Datasets:

susnato

/

PR_comments

like 0

ceph/ceph-ansible

common: stop using ceph/daemon entrypoint

major refactor in order to make ceph-ansible use the `ceph/ceph` container image instead of `ceph/daemon`

2022-10-07 12:52:53+00:00

2023-05-31 21:07:13+00:00

roles/ceph-nfs/templates/ceph-nfs.service.j2

[Unit] Description=NFS-Ganesha file server Documentation=http://github.com/nfs-ganesha/nfs-ganesha/wiki {% if container_binary == 'docker' %} After=docker.service network-online.target local-fs.target time-sync.target Requires=docker.service {% else %} After=network-online.target local-fs.target time-sync.target {% endif %} Wants=network-online.target local-fs.target time-sync.target [Service] EnvironmentFile=-/etc/environment {% if container_binary == 'podman' %} ExecStartPre=-/usr/bin/rm -f /%t/%n-pid /%t/%n-cid ExecStartPre=-/usr/bin/{{ container_binary }} rm --storage ceph-nfs-%i ExecStartPre=-/usr/bin/mkdir -p /var/log/ceph /var/log/ganesha {% endif %} ExecStartPre=-/usr/bin/{{ container_binary }} rm ceph-nfs-%i ExecStartPre={{ '/bin/mkdir' if ansible_facts['os_family'] == 'Debian' else '/usr/bin/mkdir' }} -p /etc/ceph /etc/ganesha /var/lib/nfs/ganesha /var/log/ganesha ExecStart=/usr/bin/{{ container_binary }} run --rm --net=host \ {% if container_binary == 'podman' %} -d --log-driver journald --conmon-pidfile /%t/%n-pid --cidfile /%t/%n-cid \ {% endif %} --pids-limit={{ 0 if container_binary == 'podman' else -1 }} \ -v /etc/ceph:/etc/ceph:z \ -v /var/lib/ceph/radosgw/{{ cluster }}-rgw.{{ ansible_facts['hostname'] }}/keyring:/var/lib/ceph/radosgw/{{ cluster }}-rgw.{{ ansible_facts['hostname'] }}/keyring:z \ -v /var/lib/ceph/radosgw/{{ cluster }}-nfs.{{ ansible_facts['hostname'] }}/keyring:/etc/ceph/keyring:z \ -v /etc/ganesha:/etc/ganesha:z \ -v /var/run/dbus/system_bus_socket:/var/run/dbus/system_bus_socket \ -v /var/run/ceph:/var/run/ceph:z \ -v /var/log/ceph:/var/log/ceph:z \ -v /var/log/ganesha:/var/log/ganesha:z \ -v /etc/localtime:/etc/localtime:ro \ {{ ceph_nfs_docker_extra_env }} \ --entrypoint=/usr/bin/ganesha.nfsd \ --name=ceph-nfs-{{ ceph_nfs_service_suffix | default(ansible_facts['hostname']) }} \ {{ ceph_docker_registry }}/{{ ceph_docker_image }}:{{ ceph_docker_image_tag }} \ -F -L STDOUT {% if container_binary == 'podman' %} ExecStop=-/usr/bin/sh -c "/usr/bin/{{ container_binary }} rm -f `cat /%t/%n-cid`" {% else %} ExecStopPost=-/usr/bin/{{ container_binary }} stop ceph-nfs-%i {% endif %} KillMode=none Restart=always RestartSec=10s TimeoutStartSec=120 TimeoutStopSec=15 {% if container_binary == 'podman' %} Type=forking PIDFile=/%t/%n-pid {% endif %} [Install] WantedBy=multi-user.target

[Unit] Description=NFS-Ganesha file server Documentation=http://github.com/nfs-ganesha/nfs-ganesha/wiki {% if container_binary == 'docker' %} After=docker.service network-online.target local-fs.target time-sync.target Requires=docker.service {% else %} After=network-online.target local-fs.target time-sync.target {% endif %} Wants=network-online.target local-fs.target time-sync.target [Service] EnvironmentFile=-/etc/environment {% if container_binary == 'podman' %} ExecStartPre=-/usr/bin/rm -f /%t/%n-pid /%t/%n-cid ExecStartPre=-/usr/bin/{{ container_binary }} rm --storage ceph-nfs-%i ExecStartPre=-/usr/bin/mkdir -p /var/log/ceph /var/log/ganesha {% endif %} ExecStartPre=-/usr/bin/{{ container_binary }} rm ceph-nfs-%i ExecStartPre={{ '/bin/mkdir' if ansible_facts['os_family'] == 'Debian' else '/usr/bin/mkdir' }} -p /etc/ceph /etc/ganesha /var/lib/nfs/ganesha /var/log/ganesha ExecStart=/usr/bin/{{ container_binary }} run --rm --net=host \ {% if container_binary == 'podman' %} -d --log-driver journald --conmon-pidfile /%t/%n-pid --cidfile /%t/%n-cid \ {% endif %} --pids-limit={{ 0 if container_binary == 'podman' else -1 }} \ -v /etc/ceph:/etc/ceph:z \ -v /var/lib/ceph/radosgw/{{ cluster }}-rgw.{{ ansible_facts['hostname'] }}/keyring:/var/lib/ceph/radosgw/{{ cluster }}-rgw.{{ ansible_facts['hostname'] }}/keyring:z \ -v /var/lib/ceph/radosgw/{{ cluster }}-nfs.{{ ansible_facts['hostname'] }}/keyring:/etc/ceph/keyring:z \ -v /etc/ganesha:/etc/ganesha:z \ -v /var/run/dbus/system_bus_socket:/var/run/dbus/system_bus_socket \ -v /var/run/ceph:/var/run/ceph:z \ -v /var/log/ceph:/var/log/ceph:z \ -v /var/log/ganesha:/var/log/ganesha:z \ -v /etc/localtime:/etc/localtime:ro \ {{ ceph_nfs_docker_extra_env }} \ --entrypoint=/usr/bin/ganesha.nfsd \ --name=ceph-nfs-{{ ceph_nfs_service_suffix | default(ansible_facts['hostname']) }} \ {{ ceph_docker_registry }}/{{ ceph_docker_image }}:{{ ceph_docker_image_tag }} \ -F -L STDOUT {% if container_binary == 'podman' %} ExecStop=-/usr/bin/sh -c "/usr/bin/{{ container_binary }} rm -f `cat /%t/%n-cid`" {% else %} ExecStopPost=-/usr/bin/{{ container_binary }} stop ceph-nfs-%i {% endif %} KillMode=none Restart=always RestartSec=10s TimeoutStartSec=120 TimeoutStopSec=15 {% if container_binary == 'podman' %} Type=forking PIDFile=/%t/%n-pid {% endif %} [Install] WantedBy=multi-user.target

guits

5cd692dcdc8dd55036b8c283f1bca56f79965c1d

23a8bbc6c59d78e982f343c06cfb5fe8e86cb757

hi @gouthampacha sorry for the late answer. I've no idea about nfsv3 here but this part needs to be reworked indeed.

guits

ceph/ceph-ansible

[skip ci] Refresh /etc/ceph/osd json files content before zapping the disks

If the physical disk to device path mapping has changed since the last ceph-volume simple scan (e.g. addition or removal of disks), a wrong disk could be deleted. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=2071035 Signed-off-by: Teoman ONAY <tonay@redhat.com>

2022-07-04 10:03:03+00:00

2022-07-11 07:14:41+00:00

infrastructure-playbooks/shrink-osd.yml

--- # This playbook shrinks Ceph OSDs that have been created with ceph-volume. # It can remove any number of OSD(s) from the cluster and ALL THEIR DATA # # Use it like this: # ansible-playbook shrink-osd.yml -e osd_to_kill=0,2,6 # Prompts for confirmation to shrink, defaults to no and # doesn't shrink the cluster. yes shrinks the cluster. # # ansible-playbook -e ireallymeanit=yes|no shrink-osd.yml # Overrides the prompt using -e option. Can be used in # automation scripts to avoid interactive prompt. - name: gather facts and check the init system hosts: - "{{ mon_group_name|default('mons') }}" - "{{ osd_group_name|default('osds') }}" become: True tasks: - debug: msg="gather facts on all Ceph hosts for following reference" - name: confirm whether user really meant to remove osd(s) from the cluster hosts: "{{ groups[mon_group_name][0] }}" become: true vars_prompt: - name: ireallymeanit prompt: Are you sure you want to shrink the cluster? default: 'no' private: no vars: mon_group_name: mons osd_group_name: osds pre_tasks: - name: exit playbook, if user did not mean to shrink cluster fail: msg: "Exiting shrink-osd playbook, no osd(s) was/were removed.. To shrink the cluster, either say 'yes' on the prompt or or use `-e ireallymeanit=yes` on the command line when invoking the playbook" when: ireallymeanit != 'yes' - name: exit playbook, if no osd(s) was/were given fail: msg: "osd_to_kill must be declared Exiting shrink-osd playbook, no OSD(s) was/were removed. On the command line when invoking the playbook, you can use -e osd_to_kill=0,1,2,3 argument." when: osd_to_kill is not defined - name: check the osd ids passed have the correct format fail: msg: "The id {{ item }} has wrong format, please pass the number only" with_items: "{{ osd_to_kill.split(',') }}" when: not item is regex("^\d+$") tasks: - import_role: name: ceph-defaults - import_role: name: ceph-facts tasks_from: container_binary post_tasks: - name: set_fact container_exec_cmd build docker exec command (containerized) set_fact: container_exec_cmd: "{{ container_binary }} exec ceph-mon-{{ ansible_facts['hostname'] }}" when: containerized_deployment | bool - name: exit playbook, if can not connect to the cluster command: "{{ container_exec_cmd }} timeout 5 ceph --cluster {{ cluster }} health" register: ceph_health changed_when: false until: ceph_health.stdout.find("HEALTH") > -1 retries: 5 delay: 2 - name: find the host(s) where the osd(s) is/are running on command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd find {{ item }}" changed_when: false with_items: "{{ osd_to_kill.split(',') }}" register: find_osd_hosts - name: set_fact osd_hosts set_fact: osd_hosts: "{{ osd_hosts | default([]) + [ [ (item.stdout | from_json).crush_location.host, (item.stdout | from_json).osd_fsid, item.item ] ] }}" with_items: "{{ find_osd_hosts.results }}" - name: set_fact _osd_hosts set_fact: _osd_hosts: "{{ _osd_hosts | default([]) + [ [ item.0, item.2, item.3 ] ] }}" with_nested: - "{{ groups.get(osd_group_name) }}" - "{{ osd_hosts }}" when: hostvars[item.0]['ansible_facts']['hostname'] == item.1 - name: get ceph-volume lvm list data ceph_volume: cluster: "{{ cluster }}" action: list environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" register: _lvm_list_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" - name: set_fact _lvm_list set_fact: _lvm_list: "{{ _lvm_list | default({}) | combine(item.stdout | from_json) }}" with_items: "{{ _lvm_list_data.results }}" - name: find /etc/ceph/osd files find: paths: /etc/ceph/osd pattern: "{{ item.2 }}-*" register: ceph_osd_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 not in _lvm_list.keys() - name: slurp ceph osd files content slurp: src: "{{ item['files'][0]['path'] }}" delegate_to: "{{ item.item.0 }}" register: ceph_osd_files_content loop: "{{ ceph_osd_data.results }}" when: - item.skipped is undefined - item.matched > 0 - name: set_fact ceph_osd_files_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({ item.item.item.2: item.content | b64decode | from_json}) }}" with_items: "{{ ceph_osd_files_content.results }}" when: item.skipped is undefined - name: mark osd(s) out of the cluster ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: out environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true - name: stop osd(s) service service: name: ceph-osd@{{ item.2 }} state: stopped enabled: no loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: umount osd lockbox ansible.posix.mount: path: "/var/lib/ceph/osd-lockbox/{{ ceph_osd_data_json[item.2]['data']['uuid'] }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - not containerized_deployment | bool - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2]['data']['uuid'] is defined - name: umount osd data ansible.posix.mount: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: not containerized_deployment | bool - name: get parent device for data partition command: lsblk --noheadings --output PKNAME --nodeps "{{ ceph_osd_data_json[item.2]['data']['path'] }}" register: parent_device_data_part loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['data']['path'] is defined - name: add pkname information in ceph_osd_data_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({item.item[2]: {'pkname_data': '/dev/' + item.stdout }}, recursive=True) }}" loop: "{{ parent_device_data_part.results }}" when: item.skipped is undefined - name: close dmcrypt close on devices if needed command: "cryptsetup close {{ ceph_osd_data_json[item.2][item.3]['uuid'] }}" with_nested: - "{{ _osd_hosts }}" - [ 'block_dmcrypt', 'block.db_dmcrypt', 'block.wal_dmcrypt', 'data', 'journal_dmcrypt' ] delegate_to: "{{ item.0 }}" failed_when: false register: result until: result is succeeded when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2][item.3] is defined - name: use ceph-volume lvm zap to destroy all partitions ceph_volume: cluster: "{{ cluster }}" action: zap destroy: true data: "{{ ceph_osd_data_json[item.2]['pkname_data'] if item.3 == 'data' else ceph_osd_data_json[item.2][item.3]['path'] }}" environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" with_nested: - "{{ _osd_hosts }}" - [ 'block', 'block.db', 'block.wal', 'journal', 'data' ] delegate_to: "{{ item.0 }}" failed_when: false register: result when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2][item.3] is defined - name: zap osd devices ceph_volume: action: "zap" osd_fsid: "{{ item.1 }}" environment: CEPH_VOLUME_DEBUG: "{{ ceph_volume_debug }}" CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 in _lvm_list.keys() - name: ensure osds are marked down ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: down environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true delegate_to: "{{ groups[mon_group_name][0] }}" - name: purge osd(s) from the cluster ceph_osd: ids: "{{ item }}" cluster: "{{ cluster }}" state: purge environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true with_items: "{{ osd_to_kill.split(',') }}" - name: remove osd data dir file: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: show ceph health command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} -s" changed_when: false - name: show ceph osd tree command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd tree" changed_when: false

--- # This playbook shrinks Ceph OSDs that have been created with ceph-volume. # It can remove any number of OSD(s) from the cluster and ALL THEIR DATA # # Use it like this: # ansible-playbook shrink-osd.yml -e osd_to_kill=0,2,6 # Prompts for confirmation to shrink, defaults to no and # doesn't shrink the cluster. yes shrinks the cluster. # # ansible-playbook -e ireallymeanit=yes|no shrink-osd.yml # Overrides the prompt using -e option. Can be used in # automation scripts to avoid interactive prompt. - name: gather facts and check the init system hosts: - "{{ mon_group_name|default('mons') }}" - "{{ osd_group_name|default('osds') }}" become: True tasks: - debug: msg="gather facts on all Ceph hosts for following reference" - name: confirm whether user really meant to remove osd(s) from the cluster hosts: "{{ groups[mon_group_name][0] }}" become: true vars_prompt: - name: ireallymeanit prompt: Are you sure you want to shrink the cluster? default: 'no' private: no vars: mon_group_name: mons osd_group_name: osds pre_tasks: - name: exit playbook, if user did not mean to shrink cluster fail: msg: "Exiting shrink-osd playbook, no osd(s) was/were removed.. To shrink the cluster, either say 'yes' on the prompt or or use `-e ireallymeanit=yes` on the command line when invoking the playbook" when: ireallymeanit != 'yes' - name: exit playbook, if no osd(s) was/were given fail: msg: "osd_to_kill must be declared Exiting shrink-osd playbook, no OSD(s) was/were removed. On the command line when invoking the playbook, you can use -e osd_to_kill=0,1,2,3 argument." when: osd_to_kill is not defined - name: check the osd ids passed have the correct format fail: msg: "The id {{ item }} has wrong format, please pass the number only" with_items: "{{ osd_to_kill.split(',') }}" when: not item is regex("^\d+$") tasks: - import_role: name: ceph-defaults - import_role: name: ceph-facts tasks_from: container_binary post_tasks: - name: set_fact container_exec_cmd build docker exec command (containerized) set_fact: container_exec_cmd: "{{ container_binary }} exec ceph-mon-{{ ansible_facts['hostname'] }}" when: containerized_deployment | bool - name: exit playbook, if can not connect to the cluster command: "{{ container_exec_cmd }} timeout 5 ceph --cluster {{ cluster }} health" register: ceph_health changed_when: false until: ceph_health.stdout.find("HEALTH") > -1 retries: 5 delay: 2 - name: find the host(s) where the osd(s) is/are running on command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd find {{ item }}" changed_when: false with_items: "{{ osd_to_kill.split(',') }}" register: find_osd_hosts - name: set_fact osd_hosts set_fact: osd_hosts: "{{ osd_hosts | default([]) + [ [ (item.stdout | from_json).crush_location.host, (item.stdout | from_json).osd_fsid, item.item ] ] }}" with_items: "{{ find_osd_hosts.results }}" - name: set_fact _osd_hosts set_fact: _osd_hosts: "{{ _osd_hosts | default([]) + [ [ item.0, item.2, item.3 ] ] }}" with_nested: - "{{ groups.get(osd_group_name) }}" - "{{ osd_hosts }}" when: hostvars[item.0]['ansible_facts']['hostname'] == item.1 - name: set_fact host_list set_fact: host_list: "{{ host_list | default([]) | union([item.0]) }}" loop: "{{ _osd_hosts }}" - name: get ceph-volume lvm list data ceph_volume: cluster: "{{ cluster }}" action: list environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" register: _lvm_list_data delegate_to: "{{ item }}" loop: "{{ host_list }}" - name: set_fact _lvm_list set_fact: _lvm_list: "{{ _lvm_list | default({}) | combine(item.stdout | from_json) }}" with_items: "{{ _lvm_list_data.results }}" - name: refresh /etc/ceph/osd files non containerized_deployment ceph_volume_simple_scan: cluster: "{{ cluster }}" force: true delegate_to: "{{ item }}" loop: "{{ host_list }}" when: not containerized_deployment | bool - name: refresh /etc/ceph/osd files containerized_deployment command: "{{ container_binary }} exec ceph-osd-{{ item.2 }} ceph-volume simple scan --force /var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" changed_when: false delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: containerized_deployment | bool - name: find /etc/ceph/osd files find: paths: /etc/ceph/osd pattern: "{{ item.2 }}-*" register: ceph_osd_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 not in _lvm_list.keys() - name: slurp ceph osd files content slurp: src: "{{ item['files'][0]['path'] }}" delegate_to: "{{ item.item.0 }}" register: ceph_osd_files_content loop: "{{ ceph_osd_data.results }}" when: - item.skipped is undefined - item.matched > 0 - name: set_fact ceph_osd_files_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({ item.item.item.2: item.content | b64decode | from_json}) }}" with_items: "{{ ceph_osd_files_content.results }}" when: item.skipped is undefined - name: mark osd(s) out of the cluster ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: out environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true - name: stop osd(s) service service: name: ceph-osd@{{ item.2 }} state: stopped enabled: no loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: umount osd lockbox ansible.posix.mount: path: "/var/lib/ceph/osd-lockbox/{{ ceph_osd_data_json[item.2]['data']['uuid'] }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - not containerized_deployment | bool - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2]['data']['uuid'] is defined - name: umount osd data ansible.posix.mount: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: not containerized_deployment | bool - name: get parent device for data partition command: lsblk --noheadings --output PKNAME --nodeps "{{ ceph_osd_data_json[item.2]['data']['path'] }}" register: parent_device_data_part loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['data']['path'] is defined - name: add pkname information in ceph_osd_data_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({item.item[2]: {'pkname_data': '/dev/' + item.stdout }}, recursive=True) }}" loop: "{{ parent_device_data_part.results }}" when: item.skipped is undefined - name: close dmcrypt close on devices if needed command: "cryptsetup close {{ ceph_osd_data_json[item.2][item.3]['uuid'] }}" with_nested: - "{{ _osd_hosts }}" - [ 'block_dmcrypt', 'block.db_dmcrypt', 'block.wal_dmcrypt', 'data', 'journal_dmcrypt' ] delegate_to: "{{ item.0 }}" failed_when: false register: result until: result is succeeded when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2][item.3] is defined - name: use ceph-volume lvm zap to destroy all partitions ceph_volume: cluster: "{{ cluster }}" action: zap destroy: true data: "{{ ceph_osd_data_json[item.2]['pkname_data'] if item.3 == 'data' else ceph_osd_data_json[item.2][item.3]['path'] }}" environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" with_nested: - "{{ _osd_hosts }}" - [ 'block', 'block.db', 'block.wal', 'journal', 'data' ] delegate_to: "{{ item.0 }}" failed_when: false register: result when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2][item.3] is defined - name: zap osd devices ceph_volume: action: "zap" osd_fsid: "{{ item.1 }}" environment: CEPH_VOLUME_DEBUG: "{{ ceph_volume_debug }}" CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 in _lvm_list.keys() - name: ensure osds are marked down ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: down environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true delegate_to: "{{ groups[mon_group_name][0] }}" - name: purge osd(s) from the cluster ceph_osd: ids: "{{ item }}" cluster: "{{ cluster }}" state: purge environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true with_items: "{{ osd_to_kill.split(',') }}" - name: remove osd data dir file: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: show ceph health command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} -s" changed_when: false - name: show ceph osd tree command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd tree" changed_when: false

asm0deuz

dffe7b47de70b6eeec71a3fa86f8c407adb4dd8e

64e08f2c0bdea6f4c4ad5862dc8f350c6adbe2cd

```suggestion - name: refresh /etc/ceph/osd files ```

guits

ceph/ceph-ansible

[skip ci] Refresh /etc/ceph/osd json files content before zapping the disks

If the physical disk to device path mapping has changed since the last ceph-volume simple scan (e.g. addition or removal of disks), a wrong disk could be deleted. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=2071035 Signed-off-by: Teoman ONAY <tonay@redhat.com>

2022-07-04 10:03:03+00:00

2022-07-11 07:14:41+00:00

infrastructure-playbooks/shrink-osd.yml

--- # This playbook shrinks Ceph OSDs that have been created with ceph-volume. # It can remove any number of OSD(s) from the cluster and ALL THEIR DATA # # Use it like this: # ansible-playbook shrink-osd.yml -e osd_to_kill=0,2,6 # Prompts for confirmation to shrink, defaults to no and # doesn't shrink the cluster. yes shrinks the cluster. # # ansible-playbook -e ireallymeanit=yes|no shrink-osd.yml # Overrides the prompt using -e option. Can be used in # automation scripts to avoid interactive prompt. - name: gather facts and check the init system hosts: - "{{ mon_group_name|default('mons') }}" - "{{ osd_group_name|default('osds') }}" become: True tasks: - debug: msg="gather facts on all Ceph hosts for following reference" - name: confirm whether user really meant to remove osd(s) from the cluster hosts: "{{ groups[mon_group_name][0] }}" become: true vars_prompt: - name: ireallymeanit prompt: Are you sure you want to shrink the cluster? default: 'no' private: no vars: mon_group_name: mons osd_group_name: osds pre_tasks: - name: exit playbook, if user did not mean to shrink cluster fail: msg: "Exiting shrink-osd playbook, no osd(s) was/were removed.. To shrink the cluster, either say 'yes' on the prompt or or use `-e ireallymeanit=yes` on the command line when invoking the playbook" when: ireallymeanit != 'yes' - name: exit playbook, if no osd(s) was/were given fail: msg: "osd_to_kill must be declared Exiting shrink-osd playbook, no OSD(s) was/were removed. On the command line when invoking the playbook, you can use -e osd_to_kill=0,1,2,3 argument." when: osd_to_kill is not defined - name: check the osd ids passed have the correct format fail: msg: "The id {{ item }} has wrong format, please pass the number only" with_items: "{{ osd_to_kill.split(',') }}" when: not item is regex("^\d+$") tasks: - import_role: name: ceph-defaults - import_role: name: ceph-facts tasks_from: container_binary post_tasks: - name: set_fact container_exec_cmd build docker exec command (containerized) set_fact: container_exec_cmd: "{{ container_binary }} exec ceph-mon-{{ ansible_facts['hostname'] }}" when: containerized_deployment | bool - name: exit playbook, if can not connect to the cluster command: "{{ container_exec_cmd }} timeout 5 ceph --cluster {{ cluster }} health" register: ceph_health changed_when: false until: ceph_health.stdout.find("HEALTH") > -1 retries: 5 delay: 2 - name: find the host(s) where the osd(s) is/are running on command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd find {{ item }}" changed_when: false with_items: "{{ osd_to_kill.split(',') }}" register: find_osd_hosts - name: set_fact osd_hosts set_fact: osd_hosts: "{{ osd_hosts | default([]) + [ [ (item.stdout | from_json).crush_location.host, (item.stdout | from_json).osd_fsid, item.item ] ] }}" with_items: "{{ find_osd_hosts.results }}" - name: set_fact _osd_hosts set_fact: _osd_hosts: "{{ _osd_hosts | default([]) + [ [ item.0, item.2, item.3 ] ] }}" with_nested: - "{{ groups.get(osd_group_name) }}" - "{{ osd_hosts }}" when: hostvars[item.0]['ansible_facts']['hostname'] == item.1 - name: get ceph-volume lvm list data ceph_volume: cluster: "{{ cluster }}" action: list environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" register: _lvm_list_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" - name: set_fact _lvm_list set_fact: _lvm_list: "{{ _lvm_list | default({}) | combine(item.stdout | from_json) }}" with_items: "{{ _lvm_list_data.results }}" - name: find /etc/ceph/osd files find: paths: /etc/ceph/osd pattern: "{{ item.2 }}-*" register: ceph_osd_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 not in _lvm_list.keys() - name: slurp ceph osd files content slurp: src: "{{ item['files'][0]['path'] }}" delegate_to: "{{ item.item.0 }}" register: ceph_osd_files_content loop: "{{ ceph_osd_data.results }}" when: - item.skipped is undefined - item.matched > 0 - name: set_fact ceph_osd_files_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({ item.item.item.2: item.content | b64decode | from_json}) }}" with_items: "{{ ceph_osd_files_content.results }}" when: item.skipped is undefined - name: mark osd(s) out of the cluster ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: out environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true - name: stop osd(s) service service: name: ceph-osd@{{ item.2 }} state: stopped enabled: no loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: umount osd lockbox ansible.posix.mount: path: "/var/lib/ceph/osd-lockbox/{{ ceph_osd_data_json[item.2]['data']['uuid'] }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - not containerized_deployment | bool - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2]['data']['uuid'] is defined - name: umount osd data ansible.posix.mount: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: not containerized_deployment | bool - name: get parent device for data partition command: lsblk --noheadings --output PKNAME --nodeps "{{ ceph_osd_data_json[item.2]['data']['path'] }}" register: parent_device_data_part loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['data']['path'] is defined - name: add pkname information in ceph_osd_data_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({item.item[2]: {'pkname_data': '/dev/' + item.stdout }}, recursive=True) }}" loop: "{{ parent_device_data_part.results }}" when: item.skipped is undefined - name: close dmcrypt close on devices if needed command: "cryptsetup close {{ ceph_osd_data_json[item.2][item.3]['uuid'] }}" with_nested: - "{{ _osd_hosts }}" - [ 'block_dmcrypt', 'block.db_dmcrypt', 'block.wal_dmcrypt', 'data', 'journal_dmcrypt' ] delegate_to: "{{ item.0 }}" failed_when: false register: result until: result is succeeded when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2][item.3] is defined - name: use ceph-volume lvm zap to destroy all partitions ceph_volume: cluster: "{{ cluster }}" action: zap destroy: true data: "{{ ceph_osd_data_json[item.2]['pkname_data'] if item.3 == 'data' else ceph_osd_data_json[item.2][item.3]['path'] }}" environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" with_nested: - "{{ _osd_hosts }}" - [ 'block', 'block.db', 'block.wal', 'journal', 'data' ] delegate_to: "{{ item.0 }}" failed_when: false register: result when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2][item.3] is defined - name: zap osd devices ceph_volume: action: "zap" osd_fsid: "{{ item.1 }}" environment: CEPH_VOLUME_DEBUG: "{{ ceph_volume_debug }}" CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 in _lvm_list.keys() - name: ensure osds are marked down ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: down environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true delegate_to: "{{ groups[mon_group_name][0] }}" - name: purge osd(s) from the cluster ceph_osd: ids: "{{ item }}" cluster: "{{ cluster }}" state: purge environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true with_items: "{{ osd_to_kill.split(',') }}" - name: remove osd data dir file: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: show ceph health command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} -s" changed_when: false - name: show ceph osd tree command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd tree" changed_when: false

--- # This playbook shrinks Ceph OSDs that have been created with ceph-volume. # It can remove any number of OSD(s) from the cluster and ALL THEIR DATA # # Use it like this: # ansible-playbook shrink-osd.yml -e osd_to_kill=0,2,6 # Prompts for confirmation to shrink, defaults to no and # doesn't shrink the cluster. yes shrinks the cluster. # # ansible-playbook -e ireallymeanit=yes|no shrink-osd.yml # Overrides the prompt using -e option. Can be used in # automation scripts to avoid interactive prompt. - name: gather facts and check the init system hosts: - "{{ mon_group_name|default('mons') }}" - "{{ osd_group_name|default('osds') }}" become: True tasks: - debug: msg="gather facts on all Ceph hosts for following reference" - name: confirm whether user really meant to remove osd(s) from the cluster hosts: "{{ groups[mon_group_name][0] }}" become: true vars_prompt: - name: ireallymeanit prompt: Are you sure you want to shrink the cluster? default: 'no' private: no vars: mon_group_name: mons osd_group_name: osds pre_tasks: - name: exit playbook, if user did not mean to shrink cluster fail: msg: "Exiting shrink-osd playbook, no osd(s) was/were removed.. To shrink the cluster, either say 'yes' on the prompt or or use `-e ireallymeanit=yes` on the command line when invoking the playbook" when: ireallymeanit != 'yes' - name: exit playbook, if no osd(s) was/were given fail: msg: "osd_to_kill must be declared Exiting shrink-osd playbook, no OSD(s) was/were removed. On the command line when invoking the playbook, you can use -e osd_to_kill=0,1,2,3 argument." when: osd_to_kill is not defined - name: check the osd ids passed have the correct format fail: msg: "The id {{ item }} has wrong format, please pass the number only" with_items: "{{ osd_to_kill.split(',') }}" when: not item is regex("^\d+$") tasks: - import_role: name: ceph-defaults - import_role: name: ceph-facts tasks_from: container_binary post_tasks: - name: set_fact container_exec_cmd build docker exec command (containerized) set_fact: container_exec_cmd: "{{ container_binary }} exec ceph-mon-{{ ansible_facts['hostname'] }}" when: containerized_deployment | bool - name: exit playbook, if can not connect to the cluster command: "{{ container_exec_cmd }} timeout 5 ceph --cluster {{ cluster }} health" register: ceph_health changed_when: false until: ceph_health.stdout.find("HEALTH") > -1 retries: 5 delay: 2 - name: find the host(s) where the osd(s) is/are running on command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd find {{ item }}" changed_when: false with_items: "{{ osd_to_kill.split(',') }}" register: find_osd_hosts - name: set_fact osd_hosts set_fact: osd_hosts: "{{ osd_hosts | default([]) + [ [ (item.stdout | from_json).crush_location.host, (item.stdout | from_json).osd_fsid, item.item ] ] }}" with_items: "{{ find_osd_hosts.results }}" - name: set_fact _osd_hosts set_fact: _osd_hosts: "{{ _osd_hosts | default([]) + [ [ item.0, item.2, item.3 ] ] }}" with_nested: - "{{ groups.get(osd_group_name) }}" - "{{ osd_hosts }}" when: hostvars[item.0]['ansible_facts']['hostname'] == item.1 - name: set_fact host_list set_fact: host_list: "{{ host_list | default([]) | union([item.0]) }}" loop: "{{ _osd_hosts }}" - name: get ceph-volume lvm list data ceph_volume: cluster: "{{ cluster }}" action: list environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" register: _lvm_list_data delegate_to: "{{ item }}" loop: "{{ host_list }}" - name: set_fact _lvm_list set_fact: _lvm_list: "{{ _lvm_list | default({}) | combine(item.stdout | from_json) }}" with_items: "{{ _lvm_list_data.results }}" - name: refresh /etc/ceph/osd files non containerized_deployment ceph_volume_simple_scan: cluster: "{{ cluster }}" force: true delegate_to: "{{ item }}" loop: "{{ host_list }}" when: not containerized_deployment | bool - name: refresh /etc/ceph/osd files containerized_deployment command: "{{ container_binary }} exec ceph-osd-{{ item.2 }} ceph-volume simple scan --force /var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" changed_when: false delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: containerized_deployment | bool - name: find /etc/ceph/osd files find: paths: /etc/ceph/osd pattern: "{{ item.2 }}-*" register: ceph_osd_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 not in _lvm_list.keys() - name: slurp ceph osd files content slurp: src: "{{ item['files'][0]['path'] }}" delegate_to: "{{ item.item.0 }}" register: ceph_osd_files_content loop: "{{ ceph_osd_data.results }}" when: - item.skipped is undefined - item.matched > 0 - name: set_fact ceph_osd_files_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({ item.item.item.2: item.content | b64decode | from_json}) }}" with_items: "{{ ceph_osd_files_content.results }}" when: item.skipped is undefined - name: mark osd(s) out of the cluster ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: out environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true - name: stop osd(s) service service: name: ceph-osd@{{ item.2 }} state: stopped enabled: no loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: umount osd lockbox ansible.posix.mount: path: "/var/lib/ceph/osd-lockbox/{{ ceph_osd_data_json[item.2]['data']['uuid'] }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - not containerized_deployment | bool - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2]['data']['uuid'] is defined - name: umount osd data ansible.posix.mount: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: not containerized_deployment | bool - name: get parent device for data partition command: lsblk --noheadings --output PKNAME --nodeps "{{ ceph_osd_data_json[item.2]['data']['path'] }}" register: parent_device_data_part loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['data']['path'] is defined - name: add pkname information in ceph_osd_data_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({item.item[2]: {'pkname_data': '/dev/' + item.stdout }}, recursive=True) }}" loop: "{{ parent_device_data_part.results }}" when: item.skipped is undefined - name: close dmcrypt close on devices if needed command: "cryptsetup close {{ ceph_osd_data_json[item.2][item.3]['uuid'] }}" with_nested: - "{{ _osd_hosts }}" - [ 'block_dmcrypt', 'block.db_dmcrypt', 'block.wal_dmcrypt', 'data', 'journal_dmcrypt' ] delegate_to: "{{ item.0 }}" failed_when: false register: result until: result is succeeded when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2][item.3] is defined - name: use ceph-volume lvm zap to destroy all partitions ceph_volume: cluster: "{{ cluster }}" action: zap destroy: true data: "{{ ceph_osd_data_json[item.2]['pkname_data'] if item.3 == 'data' else ceph_osd_data_json[item.2][item.3]['path'] }}" environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" with_nested: - "{{ _osd_hosts }}" - [ 'block', 'block.db', 'block.wal', 'journal', 'data' ] delegate_to: "{{ item.0 }}" failed_when: false register: result when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2][item.3] is defined - name: zap osd devices ceph_volume: action: "zap" osd_fsid: "{{ item.1 }}" environment: CEPH_VOLUME_DEBUG: "{{ ceph_volume_debug }}" CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 in _lvm_list.keys() - name: ensure osds are marked down ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: down environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true delegate_to: "{{ groups[mon_group_name][0] }}" - name: purge osd(s) from the cluster ceph_osd: ids: "{{ item }}" cluster: "{{ cluster }}" state: purge environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true with_items: "{{ osd_to_kill.split(',') }}" - name: remove osd data dir file: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: show ceph health command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} -s" changed_when: false - name: show ceph osd tree command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd tree" changed_when: false

asm0deuz

dffe7b47de70b6eeec71a3fa86f8c407adb4dd8e

64e08f2c0bdea6f4c4ad5862dc8f350c6adbe2cd

When passing a list of osd ids, (`-e osd_to_kill=1,3,5`), if the osds are on the same host, that command would be run multiple times unnecessarily on that host.

guits

ceph/ceph-ansible

[skip ci] Refresh /etc/ceph/osd json files content before zapping the disks

If the physical disk to device path mapping has changed since the last ceph-volume simple scan (e.g. addition or removal of disks), a wrong disk could be deleted. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=2071035 Signed-off-by: Teoman ONAY <tonay@redhat.com>

2022-07-04 10:03:03+00:00

2022-07-11 07:14:41+00:00

infrastructure-playbooks/shrink-osd.yml

--- # This playbook shrinks Ceph OSDs that have been created with ceph-volume. # It can remove any number of OSD(s) from the cluster and ALL THEIR DATA # # Use it like this: # ansible-playbook shrink-osd.yml -e osd_to_kill=0,2,6 # Prompts for confirmation to shrink, defaults to no and # doesn't shrink the cluster. yes shrinks the cluster. # # ansible-playbook -e ireallymeanit=yes|no shrink-osd.yml # Overrides the prompt using -e option. Can be used in # automation scripts to avoid interactive prompt. - name: gather facts and check the init system hosts: - "{{ mon_group_name|default('mons') }}" - "{{ osd_group_name|default('osds') }}" become: True tasks: - debug: msg="gather facts on all Ceph hosts for following reference" - name: confirm whether user really meant to remove osd(s) from the cluster hosts: "{{ groups[mon_group_name][0] }}" become: true vars_prompt: - name: ireallymeanit prompt: Are you sure you want to shrink the cluster? default: 'no' private: no vars: mon_group_name: mons osd_group_name: osds pre_tasks: - name: exit playbook, if user did not mean to shrink cluster fail: msg: "Exiting shrink-osd playbook, no osd(s) was/were removed.. To shrink the cluster, either say 'yes' on the prompt or or use `-e ireallymeanit=yes` on the command line when invoking the playbook" when: ireallymeanit != 'yes' - name: exit playbook, if no osd(s) was/were given fail: msg: "osd_to_kill must be declared Exiting shrink-osd playbook, no OSD(s) was/were removed. On the command line when invoking the playbook, you can use -e osd_to_kill=0,1,2,3 argument." when: osd_to_kill is not defined - name: check the osd ids passed have the correct format fail: msg: "The id {{ item }} has wrong format, please pass the number only" with_items: "{{ osd_to_kill.split(',') }}" when: not item is regex("^\d+$") tasks: - import_role: name: ceph-defaults - import_role: name: ceph-facts tasks_from: container_binary post_tasks: - name: set_fact container_exec_cmd build docker exec command (containerized) set_fact: container_exec_cmd: "{{ container_binary }} exec ceph-mon-{{ ansible_facts['hostname'] }}" when: containerized_deployment | bool - name: exit playbook, if can not connect to the cluster command: "{{ container_exec_cmd }} timeout 5 ceph --cluster {{ cluster }} health" register: ceph_health changed_when: false until: ceph_health.stdout.find("HEALTH") > -1 retries: 5 delay: 2 - name: find the host(s) where the osd(s) is/are running on command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd find {{ item }}" changed_when: false with_items: "{{ osd_to_kill.split(',') }}" register: find_osd_hosts - name: set_fact osd_hosts set_fact: osd_hosts: "{{ osd_hosts | default([]) + [ [ (item.stdout | from_json).crush_location.host, (item.stdout | from_json).osd_fsid, item.item ] ] }}" with_items: "{{ find_osd_hosts.results }}" - name: set_fact _osd_hosts set_fact: _osd_hosts: "{{ _osd_hosts | default([]) + [ [ item.0, item.2, item.3 ] ] }}" with_nested: - "{{ groups.get(osd_group_name) }}" - "{{ osd_hosts }}" when: hostvars[item.0]['ansible_facts']['hostname'] == item.1 - name: get ceph-volume lvm list data ceph_volume: cluster: "{{ cluster }}" action: list environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" register: _lvm_list_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" - name: set_fact _lvm_list set_fact: _lvm_list: "{{ _lvm_list | default({}) | combine(item.stdout | from_json) }}" with_items: "{{ _lvm_list_data.results }}" - name: find /etc/ceph/osd files find: paths: /etc/ceph/osd pattern: "{{ item.2 }}-*" register: ceph_osd_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 not in _lvm_list.keys() - name: slurp ceph osd files content slurp: src: "{{ item['files'][0]['path'] }}" delegate_to: "{{ item.item.0 }}" register: ceph_osd_files_content loop: "{{ ceph_osd_data.results }}" when: - item.skipped is undefined - item.matched > 0 - name: set_fact ceph_osd_files_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({ item.item.item.2: item.content | b64decode | from_json}) }}" with_items: "{{ ceph_osd_files_content.results }}" when: item.skipped is undefined - name: mark osd(s) out of the cluster ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: out environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true - name: stop osd(s) service service: name: ceph-osd@{{ item.2 }} state: stopped enabled: no loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: umount osd lockbox ansible.posix.mount: path: "/var/lib/ceph/osd-lockbox/{{ ceph_osd_data_json[item.2]['data']['uuid'] }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - not containerized_deployment | bool - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2]['data']['uuid'] is defined - name: umount osd data ansible.posix.mount: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: not containerized_deployment | bool - name: get parent device for data partition command: lsblk --noheadings --output PKNAME --nodeps "{{ ceph_osd_data_json[item.2]['data']['path'] }}" register: parent_device_data_part loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['data']['path'] is defined - name: add pkname information in ceph_osd_data_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({item.item[2]: {'pkname_data': '/dev/' + item.stdout }}, recursive=True) }}" loop: "{{ parent_device_data_part.results }}" when: item.skipped is undefined - name: close dmcrypt close on devices if needed command: "cryptsetup close {{ ceph_osd_data_json[item.2][item.3]['uuid'] }}" with_nested: - "{{ _osd_hosts }}" - [ 'block_dmcrypt', 'block.db_dmcrypt', 'block.wal_dmcrypt', 'data', 'journal_dmcrypt' ] delegate_to: "{{ item.0 }}" failed_when: false register: result until: result is succeeded when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2][item.3] is defined - name: use ceph-volume lvm zap to destroy all partitions ceph_volume: cluster: "{{ cluster }}" action: zap destroy: true data: "{{ ceph_osd_data_json[item.2]['pkname_data'] if item.3 == 'data' else ceph_osd_data_json[item.2][item.3]['path'] }}" environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" with_nested: - "{{ _osd_hosts }}" - [ 'block', 'block.db', 'block.wal', 'journal', 'data' ] delegate_to: "{{ item.0 }}" failed_when: false register: result when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2][item.3] is defined - name: zap osd devices ceph_volume: action: "zap" osd_fsid: "{{ item.1 }}" environment: CEPH_VOLUME_DEBUG: "{{ ceph_volume_debug }}" CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 in _lvm_list.keys() - name: ensure osds are marked down ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: down environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true delegate_to: "{{ groups[mon_group_name][0] }}" - name: purge osd(s) from the cluster ceph_osd: ids: "{{ item }}" cluster: "{{ cluster }}" state: purge environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true with_items: "{{ osd_to_kill.split(',') }}" - name: remove osd data dir file: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: show ceph health command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} -s" changed_when: false - name: show ceph osd tree command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd tree" changed_when: false

--- # This playbook shrinks Ceph OSDs that have been created with ceph-volume. # It can remove any number of OSD(s) from the cluster and ALL THEIR DATA # # Use it like this: # ansible-playbook shrink-osd.yml -e osd_to_kill=0,2,6 # Prompts for confirmation to shrink, defaults to no and # doesn't shrink the cluster. yes shrinks the cluster. # # ansible-playbook -e ireallymeanit=yes|no shrink-osd.yml # Overrides the prompt using -e option. Can be used in # automation scripts to avoid interactive prompt. - name: gather facts and check the init system hosts: - "{{ mon_group_name|default('mons') }}" - "{{ osd_group_name|default('osds') }}" become: True tasks: - debug: msg="gather facts on all Ceph hosts for following reference" - name: confirm whether user really meant to remove osd(s) from the cluster hosts: "{{ groups[mon_group_name][0] }}" become: true vars_prompt: - name: ireallymeanit prompt: Are you sure you want to shrink the cluster? default: 'no' private: no vars: mon_group_name: mons osd_group_name: osds pre_tasks: - name: exit playbook, if user did not mean to shrink cluster fail: msg: "Exiting shrink-osd playbook, no osd(s) was/were removed.. To shrink the cluster, either say 'yes' on the prompt or or use `-e ireallymeanit=yes` on the command line when invoking the playbook" when: ireallymeanit != 'yes' - name: exit playbook, if no osd(s) was/were given fail: msg: "osd_to_kill must be declared Exiting shrink-osd playbook, no OSD(s) was/were removed. On the command line when invoking the playbook, you can use -e osd_to_kill=0,1,2,3 argument." when: osd_to_kill is not defined - name: check the osd ids passed have the correct format fail: msg: "The id {{ item }} has wrong format, please pass the number only" with_items: "{{ osd_to_kill.split(',') }}" when: not item is regex("^\d+$") tasks: - import_role: name: ceph-defaults - import_role: name: ceph-facts tasks_from: container_binary post_tasks: - name: set_fact container_exec_cmd build docker exec command (containerized) set_fact: container_exec_cmd: "{{ container_binary }} exec ceph-mon-{{ ansible_facts['hostname'] }}" when: containerized_deployment | bool - name: exit playbook, if can not connect to the cluster command: "{{ container_exec_cmd }} timeout 5 ceph --cluster {{ cluster }} health" register: ceph_health changed_when: false until: ceph_health.stdout.find("HEALTH") > -1 retries: 5 delay: 2 - name: find the host(s) where the osd(s) is/are running on command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd find {{ item }}" changed_when: false with_items: "{{ osd_to_kill.split(',') }}" register: find_osd_hosts - name: set_fact osd_hosts set_fact: osd_hosts: "{{ osd_hosts | default([]) + [ [ (item.stdout | from_json).crush_location.host, (item.stdout | from_json).osd_fsid, item.item ] ] }}" with_items: "{{ find_osd_hosts.results }}" - name: set_fact _osd_hosts set_fact: _osd_hosts: "{{ _osd_hosts | default([]) + [ [ item.0, item.2, item.3 ] ] }}" with_nested: - "{{ groups.get(osd_group_name) }}" - "{{ osd_hosts }}" when: hostvars[item.0]['ansible_facts']['hostname'] == item.1 - name: set_fact host_list set_fact: host_list: "{{ host_list | default([]) | union([item.0]) }}" loop: "{{ _osd_hosts }}" - name: get ceph-volume lvm list data ceph_volume: cluster: "{{ cluster }}" action: list environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" register: _lvm_list_data delegate_to: "{{ item }}" loop: "{{ host_list }}" - name: set_fact _lvm_list set_fact: _lvm_list: "{{ _lvm_list | default({}) | combine(item.stdout | from_json) }}" with_items: "{{ _lvm_list_data.results }}" - name: refresh /etc/ceph/osd files non containerized_deployment ceph_volume_simple_scan: cluster: "{{ cluster }}" force: true delegate_to: "{{ item }}" loop: "{{ host_list }}" when: not containerized_deployment | bool - name: refresh /etc/ceph/osd files containerized_deployment command: "{{ container_binary }} exec ceph-osd-{{ item.2 }} ceph-volume simple scan --force /var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" changed_when: false delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: containerized_deployment | bool - name: find /etc/ceph/osd files find: paths: /etc/ceph/osd pattern: "{{ item.2 }}-*" register: ceph_osd_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 not in _lvm_list.keys() - name: slurp ceph osd files content slurp: src: "{{ item['files'][0]['path'] }}" delegate_to: "{{ item.item.0 }}" register: ceph_osd_files_content loop: "{{ ceph_osd_data.results }}" when: - item.skipped is undefined - item.matched > 0 - name: set_fact ceph_osd_files_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({ item.item.item.2: item.content | b64decode | from_json}) }}" with_items: "{{ ceph_osd_files_content.results }}" when: item.skipped is undefined - name: mark osd(s) out of the cluster ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: out environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true - name: stop osd(s) service service: name: ceph-osd@{{ item.2 }} state: stopped enabled: no loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: umount osd lockbox ansible.posix.mount: path: "/var/lib/ceph/osd-lockbox/{{ ceph_osd_data_json[item.2]['data']['uuid'] }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - not containerized_deployment | bool - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2]['data']['uuid'] is defined - name: umount osd data ansible.posix.mount: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: not containerized_deployment | bool - name: get parent device for data partition command: lsblk --noheadings --output PKNAME --nodeps "{{ ceph_osd_data_json[item.2]['data']['path'] }}" register: parent_device_data_part loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['data']['path'] is defined - name: add pkname information in ceph_osd_data_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({item.item[2]: {'pkname_data': '/dev/' + item.stdout }}, recursive=True) }}" loop: "{{ parent_device_data_part.results }}" when: item.skipped is undefined - name: close dmcrypt close on devices if needed command: "cryptsetup close {{ ceph_osd_data_json[item.2][item.3]['uuid'] }}" with_nested: - "{{ _osd_hosts }}" - [ 'block_dmcrypt', 'block.db_dmcrypt', 'block.wal_dmcrypt', 'data', 'journal_dmcrypt' ] delegate_to: "{{ item.0 }}" failed_when: false register: result until: result is succeeded when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2][item.3] is defined - name: use ceph-volume lvm zap to destroy all partitions ceph_volume: cluster: "{{ cluster }}" action: zap destroy: true data: "{{ ceph_osd_data_json[item.2]['pkname_data'] if item.3 == 'data' else ceph_osd_data_json[item.2][item.3]['path'] }}" environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" with_nested: - "{{ _osd_hosts }}" - [ 'block', 'block.db', 'block.wal', 'journal', 'data' ] delegate_to: "{{ item.0 }}" failed_when: false register: result when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2][item.3] is defined - name: zap osd devices ceph_volume: action: "zap" osd_fsid: "{{ item.1 }}" environment: CEPH_VOLUME_DEBUG: "{{ ceph_volume_debug }}" CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 in _lvm_list.keys() - name: ensure osds are marked down ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: down environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true delegate_to: "{{ groups[mon_group_name][0] }}" - name: purge osd(s) from the cluster ceph_osd: ids: "{{ item }}" cluster: "{{ cluster }}" state: purge environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true with_items: "{{ osd_to_kill.split(',') }}" - name: remove osd data dir file: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: show ceph health command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} -s" changed_when: false - name: show ceph osd tree command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd tree" changed_when: false

asm0deuz

dffe7b47de70b6eeec71a3fa86f8c407adb4dd8e

64e08f2c0bdea6f4c4ad5862dc8f350c6adbe2cd

```suggestion - name: refresh /etc/ceph/osd files ceph_volume_simple_scan: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" cluster: "{{ cluster }}" force: true environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" ```

guits

ceph/ceph-ansible

[skip ci] Refresh /etc/ceph/osd json files content before zapping the disks

If the physical disk to device path mapping has changed since the last ceph-volume simple scan (e.g. addition or removal of disks), a wrong disk could be deleted. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=2071035 Signed-off-by: Teoman ONAY <tonay@redhat.com>

2022-07-04 10:03:03+00:00

2022-07-11 07:14:41+00:00

infrastructure-playbooks/shrink-osd.yml

--- # This playbook shrinks Ceph OSDs that have been created with ceph-volume. # It can remove any number of OSD(s) from the cluster and ALL THEIR DATA # # Use it like this: # ansible-playbook shrink-osd.yml -e osd_to_kill=0,2,6 # Prompts for confirmation to shrink, defaults to no and # doesn't shrink the cluster. yes shrinks the cluster. # # ansible-playbook -e ireallymeanit=yes|no shrink-osd.yml # Overrides the prompt using -e option. Can be used in # automation scripts to avoid interactive prompt. - name: gather facts and check the init system hosts: - "{{ mon_group_name|default('mons') }}" - "{{ osd_group_name|default('osds') }}" become: True tasks: - debug: msg="gather facts on all Ceph hosts for following reference" - name: confirm whether user really meant to remove osd(s) from the cluster hosts: "{{ groups[mon_group_name][0] }}" become: true vars_prompt: - name: ireallymeanit prompt: Are you sure you want to shrink the cluster? default: 'no' private: no vars: mon_group_name: mons osd_group_name: osds pre_tasks: - name: exit playbook, if user did not mean to shrink cluster fail: msg: "Exiting shrink-osd playbook, no osd(s) was/were removed.. To shrink the cluster, either say 'yes' on the prompt or or use `-e ireallymeanit=yes` on the command line when invoking the playbook" when: ireallymeanit != 'yes' - name: exit playbook, if no osd(s) was/were given fail: msg: "osd_to_kill must be declared Exiting shrink-osd playbook, no OSD(s) was/were removed. On the command line when invoking the playbook, you can use -e osd_to_kill=0,1,2,3 argument." when: osd_to_kill is not defined - name: check the osd ids passed have the correct format fail: msg: "The id {{ item }} has wrong format, please pass the number only" with_items: "{{ osd_to_kill.split(',') }}" when: not item is regex("^\d+$") tasks: - import_role: name: ceph-defaults - import_role: name: ceph-facts tasks_from: container_binary post_tasks: - name: set_fact container_exec_cmd build docker exec command (containerized) set_fact: container_exec_cmd: "{{ container_binary }} exec ceph-mon-{{ ansible_facts['hostname'] }}" when: containerized_deployment | bool - name: exit playbook, if can not connect to the cluster command: "{{ container_exec_cmd }} timeout 5 ceph --cluster {{ cluster }} health" register: ceph_health changed_when: false until: ceph_health.stdout.find("HEALTH") > -1 retries: 5 delay: 2 - name: find the host(s) where the osd(s) is/are running on command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd find {{ item }}" changed_when: false with_items: "{{ osd_to_kill.split(',') }}" register: find_osd_hosts - name: set_fact osd_hosts set_fact: osd_hosts: "{{ osd_hosts | default([]) + [ [ (item.stdout | from_json).crush_location.host, (item.stdout | from_json).osd_fsid, item.item ] ] }}" with_items: "{{ find_osd_hosts.results }}" - name: set_fact _osd_hosts set_fact: _osd_hosts: "{{ _osd_hosts | default([]) + [ [ item.0, item.2, item.3 ] ] }}" with_nested: - "{{ groups.get(osd_group_name) }}" - "{{ osd_hosts }}" when: hostvars[item.0]['ansible_facts']['hostname'] == item.1 - name: get ceph-volume lvm list data ceph_volume: cluster: "{{ cluster }}" action: list environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" register: _lvm_list_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" - name: set_fact _lvm_list set_fact: _lvm_list: "{{ _lvm_list | default({}) | combine(item.stdout | from_json) }}" with_items: "{{ _lvm_list_data.results }}" - name: find /etc/ceph/osd files find: paths: /etc/ceph/osd pattern: "{{ item.2 }}-*" register: ceph_osd_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 not in _lvm_list.keys() - name: slurp ceph osd files content slurp: src: "{{ item['files'][0]['path'] }}" delegate_to: "{{ item.item.0 }}" register: ceph_osd_files_content loop: "{{ ceph_osd_data.results }}" when: - item.skipped is undefined - item.matched > 0 - name: set_fact ceph_osd_files_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({ item.item.item.2: item.content | b64decode | from_json}) }}" with_items: "{{ ceph_osd_files_content.results }}" when: item.skipped is undefined - name: mark osd(s) out of the cluster ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: out environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true - name: stop osd(s) service service: name: ceph-osd@{{ item.2 }} state: stopped enabled: no loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: umount osd lockbox ansible.posix.mount: path: "/var/lib/ceph/osd-lockbox/{{ ceph_osd_data_json[item.2]['data']['uuid'] }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - not containerized_deployment | bool - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2]['data']['uuid'] is defined - name: umount osd data ansible.posix.mount: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: not containerized_deployment | bool - name: get parent device for data partition command: lsblk --noheadings --output PKNAME --nodeps "{{ ceph_osd_data_json[item.2]['data']['path'] }}" register: parent_device_data_part loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['data']['path'] is defined - name: add pkname information in ceph_osd_data_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({item.item[2]: {'pkname_data': '/dev/' + item.stdout }}, recursive=True) }}" loop: "{{ parent_device_data_part.results }}" when: item.skipped is undefined - name: close dmcrypt close on devices if needed command: "cryptsetup close {{ ceph_osd_data_json[item.2][item.3]['uuid'] }}" with_nested: - "{{ _osd_hosts }}" - [ 'block_dmcrypt', 'block.db_dmcrypt', 'block.wal_dmcrypt', 'data', 'journal_dmcrypt' ] delegate_to: "{{ item.0 }}" failed_when: false register: result until: result is succeeded when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2][item.3] is defined - name: use ceph-volume lvm zap to destroy all partitions ceph_volume: cluster: "{{ cluster }}" action: zap destroy: true data: "{{ ceph_osd_data_json[item.2]['pkname_data'] if item.3 == 'data' else ceph_osd_data_json[item.2][item.3]['path'] }}" environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" with_nested: - "{{ _osd_hosts }}" - [ 'block', 'block.db', 'block.wal', 'journal', 'data' ] delegate_to: "{{ item.0 }}" failed_when: false register: result when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2][item.3] is defined - name: zap osd devices ceph_volume: action: "zap" osd_fsid: "{{ item.1 }}" environment: CEPH_VOLUME_DEBUG: "{{ ceph_volume_debug }}" CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 in _lvm_list.keys() - name: ensure osds are marked down ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: down environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true delegate_to: "{{ groups[mon_group_name][0] }}" - name: purge osd(s) from the cluster ceph_osd: ids: "{{ item }}" cluster: "{{ cluster }}" state: purge environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true with_items: "{{ osd_to_kill.split(',') }}" - name: remove osd data dir file: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: show ceph health command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} -s" changed_when: false - name: show ceph osd tree command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd tree" changed_when: false

--- # This playbook shrinks Ceph OSDs that have been created with ceph-volume. # It can remove any number of OSD(s) from the cluster and ALL THEIR DATA # # Use it like this: # ansible-playbook shrink-osd.yml -e osd_to_kill=0,2,6 # Prompts for confirmation to shrink, defaults to no and # doesn't shrink the cluster. yes shrinks the cluster. # # ansible-playbook -e ireallymeanit=yes|no shrink-osd.yml # Overrides the prompt using -e option. Can be used in # automation scripts to avoid interactive prompt. - name: gather facts and check the init system hosts: - "{{ mon_group_name|default('mons') }}" - "{{ osd_group_name|default('osds') }}" become: True tasks: - debug: msg="gather facts on all Ceph hosts for following reference" - name: confirm whether user really meant to remove osd(s) from the cluster hosts: "{{ groups[mon_group_name][0] }}" become: true vars_prompt: - name: ireallymeanit prompt: Are you sure you want to shrink the cluster? default: 'no' private: no vars: mon_group_name: mons osd_group_name: osds pre_tasks: - name: exit playbook, if user did not mean to shrink cluster fail: msg: "Exiting shrink-osd playbook, no osd(s) was/were removed.. To shrink the cluster, either say 'yes' on the prompt or or use `-e ireallymeanit=yes` on the command line when invoking the playbook" when: ireallymeanit != 'yes' - name: exit playbook, if no osd(s) was/were given fail: msg: "osd_to_kill must be declared Exiting shrink-osd playbook, no OSD(s) was/were removed. On the command line when invoking the playbook, you can use -e osd_to_kill=0,1,2,3 argument." when: osd_to_kill is not defined - name: check the osd ids passed have the correct format fail: msg: "The id {{ item }} has wrong format, please pass the number only" with_items: "{{ osd_to_kill.split(',') }}" when: not item is regex("^\d+$") tasks: - import_role: name: ceph-defaults - import_role: name: ceph-facts tasks_from: container_binary post_tasks: - name: set_fact container_exec_cmd build docker exec command (containerized) set_fact: container_exec_cmd: "{{ container_binary }} exec ceph-mon-{{ ansible_facts['hostname'] }}" when: containerized_deployment | bool - name: exit playbook, if can not connect to the cluster command: "{{ container_exec_cmd }} timeout 5 ceph --cluster {{ cluster }} health" register: ceph_health changed_when: false until: ceph_health.stdout.find("HEALTH") > -1 retries: 5 delay: 2 - name: find the host(s) where the osd(s) is/are running on command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd find {{ item }}" changed_when: false with_items: "{{ osd_to_kill.split(',') }}" register: find_osd_hosts - name: set_fact osd_hosts set_fact: osd_hosts: "{{ osd_hosts | default([]) + [ [ (item.stdout | from_json).crush_location.host, (item.stdout | from_json).osd_fsid, item.item ] ] }}" with_items: "{{ find_osd_hosts.results }}" - name: set_fact _osd_hosts set_fact: _osd_hosts: "{{ _osd_hosts | default([]) + [ [ item.0, item.2, item.3 ] ] }}" with_nested: - "{{ groups.get(osd_group_name) }}" - "{{ osd_hosts }}" when: hostvars[item.0]['ansible_facts']['hostname'] == item.1 - name: set_fact host_list set_fact: host_list: "{{ host_list | default([]) | union([item.0]) }}" loop: "{{ _osd_hosts }}" - name: get ceph-volume lvm list data ceph_volume: cluster: "{{ cluster }}" action: list environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" register: _lvm_list_data delegate_to: "{{ item }}" loop: "{{ host_list }}" - name: set_fact _lvm_list set_fact: _lvm_list: "{{ _lvm_list | default({}) | combine(item.stdout | from_json) }}" with_items: "{{ _lvm_list_data.results }}" - name: refresh /etc/ceph/osd files non containerized_deployment ceph_volume_simple_scan: cluster: "{{ cluster }}" force: true delegate_to: "{{ item }}" loop: "{{ host_list }}" when: not containerized_deployment | bool - name: refresh /etc/ceph/osd files containerized_deployment command: "{{ container_binary }} exec ceph-osd-{{ item.2 }} ceph-volume simple scan --force /var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" changed_when: false delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: containerized_deployment | bool - name: find /etc/ceph/osd files find: paths: /etc/ceph/osd pattern: "{{ item.2 }}-*" register: ceph_osd_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 not in _lvm_list.keys() - name: slurp ceph osd files content slurp: src: "{{ item['files'][0]['path'] }}" delegate_to: "{{ item.item.0 }}" register: ceph_osd_files_content loop: "{{ ceph_osd_data.results }}" when: - item.skipped is undefined - item.matched > 0 - name: set_fact ceph_osd_files_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({ item.item.item.2: item.content | b64decode | from_json}) }}" with_items: "{{ ceph_osd_files_content.results }}" when: item.skipped is undefined - name: mark osd(s) out of the cluster ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: out environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true - name: stop osd(s) service service: name: ceph-osd@{{ item.2 }} state: stopped enabled: no loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: umount osd lockbox ansible.posix.mount: path: "/var/lib/ceph/osd-lockbox/{{ ceph_osd_data_json[item.2]['data']['uuid'] }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - not containerized_deployment | bool - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2]['data']['uuid'] is defined - name: umount osd data ansible.posix.mount: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: not containerized_deployment | bool - name: get parent device for data partition command: lsblk --noheadings --output PKNAME --nodeps "{{ ceph_osd_data_json[item.2]['data']['path'] }}" register: parent_device_data_part loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['data']['path'] is defined - name: add pkname information in ceph_osd_data_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({item.item[2]: {'pkname_data': '/dev/' + item.stdout }}, recursive=True) }}" loop: "{{ parent_device_data_part.results }}" when: item.skipped is undefined - name: close dmcrypt close on devices if needed command: "cryptsetup close {{ ceph_osd_data_json[item.2][item.3]['uuid'] }}" with_nested: - "{{ _osd_hosts }}" - [ 'block_dmcrypt', 'block.db_dmcrypt', 'block.wal_dmcrypt', 'data', 'journal_dmcrypt' ] delegate_to: "{{ item.0 }}" failed_when: false register: result until: result is succeeded when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2][item.3] is defined - name: use ceph-volume lvm zap to destroy all partitions ceph_volume: cluster: "{{ cluster }}" action: zap destroy: true data: "{{ ceph_osd_data_json[item.2]['pkname_data'] if item.3 == 'data' else ceph_osd_data_json[item.2][item.3]['path'] }}" environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" with_nested: - "{{ _osd_hosts }}" - [ 'block', 'block.db', 'block.wal', 'journal', 'data' ] delegate_to: "{{ item.0 }}" failed_when: false register: result when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2][item.3] is defined - name: zap osd devices ceph_volume: action: "zap" osd_fsid: "{{ item.1 }}" environment: CEPH_VOLUME_DEBUG: "{{ ceph_volume_debug }}" CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 in _lvm_list.keys() - name: ensure osds are marked down ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: down environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true delegate_to: "{{ groups[mon_group_name][0] }}" - name: purge osd(s) from the cluster ceph_osd: ids: "{{ item }}" cluster: "{{ cluster }}" state: purge environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true with_items: "{{ osd_to_kill.split(',') }}" - name: remove osd data dir file: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: show ceph health command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} -s" changed_when: false - name: show ceph osd tree command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd tree" changed_when: false

asm0deuz

dffe7b47de70b6eeec71a3fa86f8c407adb4dd8e

64e08f2c0bdea6f4c4ad5862dc8f350c6adbe2cd

```suggestion ```

guits

ceph/ceph-ansible

[skip ci] Refresh /etc/ceph/osd json files content before zapping the disks

If the physical disk to device path mapping has changed since the last ceph-volume simple scan (e.g. addition or removal of disks), a wrong disk could be deleted. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=2071035 Signed-off-by: Teoman ONAY <tonay@redhat.com>

2022-07-04 10:03:03+00:00

2022-07-11 07:14:41+00:00

infrastructure-playbooks/shrink-osd.yml

--- # This playbook shrinks Ceph OSDs that have been created with ceph-volume. # It can remove any number of OSD(s) from the cluster and ALL THEIR DATA # # Use it like this: # ansible-playbook shrink-osd.yml -e osd_to_kill=0,2,6 # Prompts for confirmation to shrink, defaults to no and # doesn't shrink the cluster. yes shrinks the cluster. # # ansible-playbook -e ireallymeanit=yes|no shrink-osd.yml # Overrides the prompt using -e option. Can be used in # automation scripts to avoid interactive prompt. - name: gather facts and check the init system hosts: - "{{ mon_group_name|default('mons') }}" - "{{ osd_group_name|default('osds') }}" become: True tasks: - debug: msg="gather facts on all Ceph hosts for following reference" - name: confirm whether user really meant to remove osd(s) from the cluster hosts: "{{ groups[mon_group_name][0] }}" become: true vars_prompt: - name: ireallymeanit prompt: Are you sure you want to shrink the cluster? default: 'no' private: no vars: mon_group_name: mons osd_group_name: osds pre_tasks: - name: exit playbook, if user did not mean to shrink cluster fail: msg: "Exiting shrink-osd playbook, no osd(s) was/were removed.. To shrink the cluster, either say 'yes' on the prompt or or use `-e ireallymeanit=yes` on the command line when invoking the playbook" when: ireallymeanit != 'yes' - name: exit playbook, if no osd(s) was/were given fail: msg: "osd_to_kill must be declared Exiting shrink-osd playbook, no OSD(s) was/were removed. On the command line when invoking the playbook, you can use -e osd_to_kill=0,1,2,3 argument." when: osd_to_kill is not defined - name: check the osd ids passed have the correct format fail: msg: "The id {{ item }} has wrong format, please pass the number only" with_items: "{{ osd_to_kill.split(',') }}" when: not item is regex("^\d+$") tasks: - import_role: name: ceph-defaults - import_role: name: ceph-facts tasks_from: container_binary post_tasks: - name: set_fact container_exec_cmd build docker exec command (containerized) set_fact: container_exec_cmd: "{{ container_binary }} exec ceph-mon-{{ ansible_facts['hostname'] }}" when: containerized_deployment | bool - name: exit playbook, if can not connect to the cluster command: "{{ container_exec_cmd }} timeout 5 ceph --cluster {{ cluster }} health" register: ceph_health changed_when: false until: ceph_health.stdout.find("HEALTH") > -1 retries: 5 delay: 2 - name: find the host(s) where the osd(s) is/are running on command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd find {{ item }}" changed_when: false with_items: "{{ osd_to_kill.split(',') }}" register: find_osd_hosts - name: set_fact osd_hosts set_fact: osd_hosts: "{{ osd_hosts | default([]) + [ [ (item.stdout | from_json).crush_location.host, (item.stdout | from_json).osd_fsid, item.item ] ] }}" with_items: "{{ find_osd_hosts.results }}" - name: set_fact _osd_hosts set_fact: _osd_hosts: "{{ _osd_hosts | default([]) + [ [ item.0, item.2, item.3 ] ] }}" with_nested: - "{{ groups.get(osd_group_name) }}" - "{{ osd_hosts }}" when: hostvars[item.0]['ansible_facts']['hostname'] == item.1 - name: get ceph-volume lvm list data ceph_volume: cluster: "{{ cluster }}" action: list environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" register: _lvm_list_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" - name: set_fact _lvm_list set_fact: _lvm_list: "{{ _lvm_list | default({}) | combine(item.stdout | from_json) }}" with_items: "{{ _lvm_list_data.results }}" - name: find /etc/ceph/osd files find: paths: /etc/ceph/osd pattern: "{{ item.2 }}-*" register: ceph_osd_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 not in _lvm_list.keys() - name: slurp ceph osd files content slurp: src: "{{ item['files'][0]['path'] }}" delegate_to: "{{ item.item.0 }}" register: ceph_osd_files_content loop: "{{ ceph_osd_data.results }}" when: - item.skipped is undefined - item.matched > 0 - name: set_fact ceph_osd_files_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({ item.item.item.2: item.content | b64decode | from_json}) }}" with_items: "{{ ceph_osd_files_content.results }}" when: item.skipped is undefined - name: mark osd(s) out of the cluster ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: out environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true - name: stop osd(s) service service: name: ceph-osd@{{ item.2 }} state: stopped enabled: no loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: umount osd lockbox ansible.posix.mount: path: "/var/lib/ceph/osd-lockbox/{{ ceph_osd_data_json[item.2]['data']['uuid'] }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - not containerized_deployment | bool - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2]['data']['uuid'] is defined - name: umount osd data ansible.posix.mount: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: not containerized_deployment | bool - name: get parent device for data partition command: lsblk --noheadings --output PKNAME --nodeps "{{ ceph_osd_data_json[item.2]['data']['path'] }}" register: parent_device_data_part loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['data']['path'] is defined - name: add pkname information in ceph_osd_data_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({item.item[2]: {'pkname_data': '/dev/' + item.stdout }}, recursive=True) }}" loop: "{{ parent_device_data_part.results }}" when: item.skipped is undefined - name: close dmcrypt close on devices if needed command: "cryptsetup close {{ ceph_osd_data_json[item.2][item.3]['uuid'] }}" with_nested: - "{{ _osd_hosts }}" - [ 'block_dmcrypt', 'block.db_dmcrypt', 'block.wal_dmcrypt', 'data', 'journal_dmcrypt' ] delegate_to: "{{ item.0 }}" failed_when: false register: result until: result is succeeded when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2][item.3] is defined - name: use ceph-volume lvm zap to destroy all partitions ceph_volume: cluster: "{{ cluster }}" action: zap destroy: true data: "{{ ceph_osd_data_json[item.2]['pkname_data'] if item.3 == 'data' else ceph_osd_data_json[item.2][item.3]['path'] }}" environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" with_nested: - "{{ _osd_hosts }}" - [ 'block', 'block.db', 'block.wal', 'journal', 'data' ] delegate_to: "{{ item.0 }}" failed_when: false register: result when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2][item.3] is defined - name: zap osd devices ceph_volume: action: "zap" osd_fsid: "{{ item.1 }}" environment: CEPH_VOLUME_DEBUG: "{{ ceph_volume_debug }}" CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 in _lvm_list.keys() - name: ensure osds are marked down ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: down environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true delegate_to: "{{ groups[mon_group_name][0] }}" - name: purge osd(s) from the cluster ceph_osd: ids: "{{ item }}" cluster: "{{ cluster }}" state: purge environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true with_items: "{{ osd_to_kill.split(',') }}" - name: remove osd data dir file: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: show ceph health command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} -s" changed_when: false - name: show ceph osd tree command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd tree" changed_when: false

--- # This playbook shrinks Ceph OSDs that have been created with ceph-volume. # It can remove any number of OSD(s) from the cluster and ALL THEIR DATA # # Use it like this: # ansible-playbook shrink-osd.yml -e osd_to_kill=0,2,6 # Prompts for confirmation to shrink, defaults to no and # doesn't shrink the cluster. yes shrinks the cluster. # # ansible-playbook -e ireallymeanit=yes|no shrink-osd.yml # Overrides the prompt using -e option. Can be used in # automation scripts to avoid interactive prompt. - name: gather facts and check the init system hosts: - "{{ mon_group_name|default('mons') }}" - "{{ osd_group_name|default('osds') }}" become: True tasks: - debug: msg="gather facts on all Ceph hosts for following reference" - name: confirm whether user really meant to remove osd(s) from the cluster hosts: "{{ groups[mon_group_name][0] }}" become: true vars_prompt: - name: ireallymeanit prompt: Are you sure you want to shrink the cluster? default: 'no' private: no vars: mon_group_name: mons osd_group_name: osds pre_tasks: - name: exit playbook, if user did not mean to shrink cluster fail: msg: "Exiting shrink-osd playbook, no osd(s) was/were removed.. To shrink the cluster, either say 'yes' on the prompt or or use `-e ireallymeanit=yes` on the command line when invoking the playbook" when: ireallymeanit != 'yes' - name: exit playbook, if no osd(s) was/were given fail: msg: "osd_to_kill must be declared Exiting shrink-osd playbook, no OSD(s) was/were removed. On the command line when invoking the playbook, you can use -e osd_to_kill=0,1,2,3 argument." when: osd_to_kill is not defined - name: check the osd ids passed have the correct format fail: msg: "The id {{ item }} has wrong format, please pass the number only" with_items: "{{ osd_to_kill.split(',') }}" when: not item is regex("^\d+$") tasks: - import_role: name: ceph-defaults - import_role: name: ceph-facts tasks_from: container_binary post_tasks: - name: set_fact container_exec_cmd build docker exec command (containerized) set_fact: container_exec_cmd: "{{ container_binary }} exec ceph-mon-{{ ansible_facts['hostname'] }}" when: containerized_deployment | bool - name: exit playbook, if can not connect to the cluster command: "{{ container_exec_cmd }} timeout 5 ceph --cluster {{ cluster }} health" register: ceph_health changed_when: false until: ceph_health.stdout.find("HEALTH") > -1 retries: 5 delay: 2 - name: find the host(s) where the osd(s) is/are running on command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd find {{ item }}" changed_when: false with_items: "{{ osd_to_kill.split(',') }}" register: find_osd_hosts - name: set_fact osd_hosts set_fact: osd_hosts: "{{ osd_hosts | default([]) + [ [ (item.stdout | from_json).crush_location.host, (item.stdout | from_json).osd_fsid, item.item ] ] }}" with_items: "{{ find_osd_hosts.results }}" - name: set_fact _osd_hosts set_fact: _osd_hosts: "{{ _osd_hosts | default([]) + [ [ item.0, item.2, item.3 ] ] }}" with_nested: - "{{ groups.get(osd_group_name) }}" - "{{ osd_hosts }}" when: hostvars[item.0]['ansible_facts']['hostname'] == item.1 - name: set_fact host_list set_fact: host_list: "{{ host_list | default([]) | union([item.0]) }}" loop: "{{ _osd_hosts }}" - name: get ceph-volume lvm list data ceph_volume: cluster: "{{ cluster }}" action: list environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" register: _lvm_list_data delegate_to: "{{ item }}" loop: "{{ host_list }}" - name: set_fact _lvm_list set_fact: _lvm_list: "{{ _lvm_list | default({}) | combine(item.stdout | from_json) }}" with_items: "{{ _lvm_list_data.results }}" - name: refresh /etc/ceph/osd files non containerized_deployment ceph_volume_simple_scan: cluster: "{{ cluster }}" force: true delegate_to: "{{ item }}" loop: "{{ host_list }}" when: not containerized_deployment | bool - name: refresh /etc/ceph/osd files containerized_deployment command: "{{ container_binary }} exec ceph-osd-{{ item.2 }} ceph-volume simple scan --force /var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" changed_when: false delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: containerized_deployment | bool - name: find /etc/ceph/osd files find: paths: /etc/ceph/osd pattern: "{{ item.2 }}-*" register: ceph_osd_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 not in _lvm_list.keys() - name: slurp ceph osd files content slurp: src: "{{ item['files'][0]['path'] }}" delegate_to: "{{ item.item.0 }}" register: ceph_osd_files_content loop: "{{ ceph_osd_data.results }}" when: - item.skipped is undefined - item.matched > 0 - name: set_fact ceph_osd_files_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({ item.item.item.2: item.content | b64decode | from_json}) }}" with_items: "{{ ceph_osd_files_content.results }}" when: item.skipped is undefined - name: mark osd(s) out of the cluster ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: out environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true - name: stop osd(s) service service: name: ceph-osd@{{ item.2 }} state: stopped enabled: no loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: umount osd lockbox ansible.posix.mount: path: "/var/lib/ceph/osd-lockbox/{{ ceph_osd_data_json[item.2]['data']['uuid'] }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - not containerized_deployment | bool - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2]['data']['uuid'] is defined - name: umount osd data ansible.posix.mount: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: not containerized_deployment | bool - name: get parent device for data partition command: lsblk --noheadings --output PKNAME --nodeps "{{ ceph_osd_data_json[item.2]['data']['path'] }}" register: parent_device_data_part loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['data']['path'] is defined - name: add pkname information in ceph_osd_data_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({item.item[2]: {'pkname_data': '/dev/' + item.stdout }}, recursive=True) }}" loop: "{{ parent_device_data_part.results }}" when: item.skipped is undefined - name: close dmcrypt close on devices if needed command: "cryptsetup close {{ ceph_osd_data_json[item.2][item.3]['uuid'] }}" with_nested: - "{{ _osd_hosts }}" - [ 'block_dmcrypt', 'block.db_dmcrypt', 'block.wal_dmcrypt', 'data', 'journal_dmcrypt' ] delegate_to: "{{ item.0 }}" failed_when: false register: result until: result is succeeded when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2][item.3] is defined - name: use ceph-volume lvm zap to destroy all partitions ceph_volume: cluster: "{{ cluster }}" action: zap destroy: true data: "{{ ceph_osd_data_json[item.2]['pkname_data'] if item.3 == 'data' else ceph_osd_data_json[item.2][item.3]['path'] }}" environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" with_nested: - "{{ _osd_hosts }}" - [ 'block', 'block.db', 'block.wal', 'journal', 'data' ] delegate_to: "{{ item.0 }}" failed_when: false register: result when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2][item.3] is defined - name: zap osd devices ceph_volume: action: "zap" osd_fsid: "{{ item.1 }}" environment: CEPH_VOLUME_DEBUG: "{{ ceph_volume_debug }}" CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 in _lvm_list.keys() - name: ensure osds are marked down ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: down environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true delegate_to: "{{ groups[mon_group_name][0] }}" - name: purge osd(s) from the cluster ceph_osd: ids: "{{ item }}" cluster: "{{ cluster }}" state: purge environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true with_items: "{{ osd_to_kill.split(',') }}" - name: remove osd data dir file: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: show ceph health command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} -s" changed_when: false - name: show ceph osd tree command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd tree" changed_when: false

asm0deuz

dffe7b47de70b6eeec71a3fa86f8c407adb4dd8e

64e08f2c0bdea6f4c4ad5862dc8f350c6adbe2cd

```suggestion ``` doesn't make sense to keep this if we know it will run only when not containerized deployments.

guits

ceph/ceph-ansible

[skip ci] Refresh /etc/ceph/osd json files content before zapping the disks

If the physical disk to device path mapping has changed since the last ceph-volume simple scan (e.g. addition or removal of disks), a wrong disk could be deleted. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=2071035 Signed-off-by: Teoman ONAY <tonay@redhat.com>

2022-07-04 10:03:03+00:00

2022-07-11 07:14:41+00:00

infrastructure-playbooks/shrink-osd.yml

--- # This playbook shrinks Ceph OSDs that have been created with ceph-volume. # It can remove any number of OSD(s) from the cluster and ALL THEIR DATA # # Use it like this: # ansible-playbook shrink-osd.yml -e osd_to_kill=0,2,6 # Prompts for confirmation to shrink, defaults to no and # doesn't shrink the cluster. yes shrinks the cluster. # # ansible-playbook -e ireallymeanit=yes|no shrink-osd.yml # Overrides the prompt using -e option. Can be used in # automation scripts to avoid interactive prompt. - name: gather facts and check the init system hosts: - "{{ mon_group_name|default('mons') }}" - "{{ osd_group_name|default('osds') }}" become: True tasks: - debug: msg="gather facts on all Ceph hosts for following reference" - name: confirm whether user really meant to remove osd(s) from the cluster hosts: "{{ groups[mon_group_name][0] }}" become: true vars_prompt: - name: ireallymeanit prompt: Are you sure you want to shrink the cluster? default: 'no' private: no vars: mon_group_name: mons osd_group_name: osds pre_tasks: - name: exit playbook, if user did not mean to shrink cluster fail: msg: "Exiting shrink-osd playbook, no osd(s) was/were removed.. To shrink the cluster, either say 'yes' on the prompt or or use `-e ireallymeanit=yes` on the command line when invoking the playbook" when: ireallymeanit != 'yes' - name: exit playbook, if no osd(s) was/were given fail: msg: "osd_to_kill must be declared Exiting shrink-osd playbook, no OSD(s) was/were removed. On the command line when invoking the playbook, you can use -e osd_to_kill=0,1,2,3 argument." when: osd_to_kill is not defined - name: check the osd ids passed have the correct format fail: msg: "The id {{ item }} has wrong format, please pass the number only" with_items: "{{ osd_to_kill.split(',') }}" when: not item is regex("^\d+$") tasks: - import_role: name: ceph-defaults - import_role: name: ceph-facts tasks_from: container_binary post_tasks: - name: set_fact container_exec_cmd build docker exec command (containerized) set_fact: container_exec_cmd: "{{ container_binary }} exec ceph-mon-{{ ansible_facts['hostname'] }}" when: containerized_deployment | bool - name: exit playbook, if can not connect to the cluster command: "{{ container_exec_cmd }} timeout 5 ceph --cluster {{ cluster }} health" register: ceph_health changed_when: false until: ceph_health.stdout.find("HEALTH") > -1 retries: 5 delay: 2 - name: find the host(s) where the osd(s) is/are running on command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd find {{ item }}" changed_when: false with_items: "{{ osd_to_kill.split(',') }}" register: find_osd_hosts - name: set_fact osd_hosts set_fact: osd_hosts: "{{ osd_hosts | default([]) + [ [ (item.stdout | from_json).crush_location.host, (item.stdout | from_json).osd_fsid, item.item ] ] }}" with_items: "{{ find_osd_hosts.results }}" - name: set_fact _osd_hosts set_fact: _osd_hosts: "{{ _osd_hosts | default([]) + [ [ item.0, item.2, item.3 ] ] }}" with_nested: - "{{ groups.get(osd_group_name) }}" - "{{ osd_hosts }}" when: hostvars[item.0]['ansible_facts']['hostname'] == item.1 - name: get ceph-volume lvm list data ceph_volume: cluster: "{{ cluster }}" action: list environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" register: _lvm_list_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" - name: set_fact _lvm_list set_fact: _lvm_list: "{{ _lvm_list | default({}) | combine(item.stdout | from_json) }}" with_items: "{{ _lvm_list_data.results }}" - name: find /etc/ceph/osd files find: paths: /etc/ceph/osd pattern: "{{ item.2 }}-*" register: ceph_osd_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 not in _lvm_list.keys() - name: slurp ceph osd files content slurp: src: "{{ item['files'][0]['path'] }}" delegate_to: "{{ item.item.0 }}" register: ceph_osd_files_content loop: "{{ ceph_osd_data.results }}" when: - item.skipped is undefined - item.matched > 0 - name: set_fact ceph_osd_files_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({ item.item.item.2: item.content | b64decode | from_json}) }}" with_items: "{{ ceph_osd_files_content.results }}" when: item.skipped is undefined - name: mark osd(s) out of the cluster ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: out environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true - name: stop osd(s) service service: name: ceph-osd@{{ item.2 }} state: stopped enabled: no loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: umount osd lockbox ansible.posix.mount: path: "/var/lib/ceph/osd-lockbox/{{ ceph_osd_data_json[item.2]['data']['uuid'] }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - not containerized_deployment | bool - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2]['data']['uuid'] is defined - name: umount osd data ansible.posix.mount: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: not containerized_deployment | bool - name: get parent device for data partition command: lsblk --noheadings --output PKNAME --nodeps "{{ ceph_osd_data_json[item.2]['data']['path'] }}" register: parent_device_data_part loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['data']['path'] is defined - name: add pkname information in ceph_osd_data_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({item.item[2]: {'pkname_data': '/dev/' + item.stdout }}, recursive=True) }}" loop: "{{ parent_device_data_part.results }}" when: item.skipped is undefined - name: close dmcrypt close on devices if needed command: "cryptsetup close {{ ceph_osd_data_json[item.2][item.3]['uuid'] }}" with_nested: - "{{ _osd_hosts }}" - [ 'block_dmcrypt', 'block.db_dmcrypt', 'block.wal_dmcrypt', 'data', 'journal_dmcrypt' ] delegate_to: "{{ item.0 }}" failed_when: false register: result until: result is succeeded when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2][item.3] is defined - name: use ceph-volume lvm zap to destroy all partitions ceph_volume: cluster: "{{ cluster }}" action: zap destroy: true data: "{{ ceph_osd_data_json[item.2]['pkname_data'] if item.3 == 'data' else ceph_osd_data_json[item.2][item.3]['path'] }}" environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" with_nested: - "{{ _osd_hosts }}" - [ 'block', 'block.db', 'block.wal', 'journal', 'data' ] delegate_to: "{{ item.0 }}" failed_when: false register: result when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2][item.3] is defined - name: zap osd devices ceph_volume: action: "zap" osd_fsid: "{{ item.1 }}" environment: CEPH_VOLUME_DEBUG: "{{ ceph_volume_debug }}" CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 in _lvm_list.keys() - name: ensure osds are marked down ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: down environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true delegate_to: "{{ groups[mon_group_name][0] }}" - name: purge osd(s) from the cluster ceph_osd: ids: "{{ item }}" cluster: "{{ cluster }}" state: purge environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true with_items: "{{ osd_to_kill.split(',') }}" - name: remove osd data dir file: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: show ceph health command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} -s" changed_when: false - name: show ceph osd tree command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd tree" changed_when: false

--- # This playbook shrinks Ceph OSDs that have been created with ceph-volume. # It can remove any number of OSD(s) from the cluster and ALL THEIR DATA # # Use it like this: # ansible-playbook shrink-osd.yml -e osd_to_kill=0,2,6 # Prompts for confirmation to shrink, defaults to no and # doesn't shrink the cluster. yes shrinks the cluster. # # ansible-playbook -e ireallymeanit=yes|no shrink-osd.yml # Overrides the prompt using -e option. Can be used in # automation scripts to avoid interactive prompt. - name: gather facts and check the init system hosts: - "{{ mon_group_name|default('mons') }}" - "{{ osd_group_name|default('osds') }}" become: True tasks: - debug: msg="gather facts on all Ceph hosts for following reference" - name: confirm whether user really meant to remove osd(s) from the cluster hosts: "{{ groups[mon_group_name][0] }}" become: true vars_prompt: - name: ireallymeanit prompt: Are you sure you want to shrink the cluster? default: 'no' private: no vars: mon_group_name: mons osd_group_name: osds pre_tasks: - name: exit playbook, if user did not mean to shrink cluster fail: msg: "Exiting shrink-osd playbook, no osd(s) was/were removed.. To shrink the cluster, either say 'yes' on the prompt or or use `-e ireallymeanit=yes` on the command line when invoking the playbook" when: ireallymeanit != 'yes' - name: exit playbook, if no osd(s) was/were given fail: msg: "osd_to_kill must be declared Exiting shrink-osd playbook, no OSD(s) was/were removed. On the command line when invoking the playbook, you can use -e osd_to_kill=0,1,2,3 argument." when: osd_to_kill is not defined - name: check the osd ids passed have the correct format fail: msg: "The id {{ item }} has wrong format, please pass the number only" with_items: "{{ osd_to_kill.split(',') }}" when: not item is regex("^\d+$") tasks: - import_role: name: ceph-defaults - import_role: name: ceph-facts tasks_from: container_binary post_tasks: - name: set_fact container_exec_cmd build docker exec command (containerized) set_fact: container_exec_cmd: "{{ container_binary }} exec ceph-mon-{{ ansible_facts['hostname'] }}" when: containerized_deployment | bool - name: exit playbook, if can not connect to the cluster command: "{{ container_exec_cmd }} timeout 5 ceph --cluster {{ cluster }} health" register: ceph_health changed_when: false until: ceph_health.stdout.find("HEALTH") > -1 retries: 5 delay: 2 - name: find the host(s) where the osd(s) is/are running on command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd find {{ item }}" changed_when: false with_items: "{{ osd_to_kill.split(',') }}" register: find_osd_hosts - name: set_fact osd_hosts set_fact: osd_hosts: "{{ osd_hosts | default([]) + [ [ (item.stdout | from_json).crush_location.host, (item.stdout | from_json).osd_fsid, item.item ] ] }}" with_items: "{{ find_osd_hosts.results }}" - name: set_fact _osd_hosts set_fact: _osd_hosts: "{{ _osd_hosts | default([]) + [ [ item.0, item.2, item.3 ] ] }}" with_nested: - "{{ groups.get(osd_group_name) }}" - "{{ osd_hosts }}" when: hostvars[item.0]['ansible_facts']['hostname'] == item.1 - name: set_fact host_list set_fact: host_list: "{{ host_list | default([]) | union([item.0]) }}" loop: "{{ _osd_hosts }}" - name: get ceph-volume lvm list data ceph_volume: cluster: "{{ cluster }}" action: list environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" register: _lvm_list_data delegate_to: "{{ item }}" loop: "{{ host_list }}" - name: set_fact _lvm_list set_fact: _lvm_list: "{{ _lvm_list | default({}) | combine(item.stdout | from_json) }}" with_items: "{{ _lvm_list_data.results }}" - name: refresh /etc/ceph/osd files non containerized_deployment ceph_volume_simple_scan: cluster: "{{ cluster }}" force: true delegate_to: "{{ item }}" loop: "{{ host_list }}" when: not containerized_deployment | bool - name: refresh /etc/ceph/osd files containerized_deployment command: "{{ container_binary }} exec ceph-osd-{{ item.2 }} ceph-volume simple scan --force /var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" changed_when: false delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: containerized_deployment | bool - name: find /etc/ceph/osd files find: paths: /etc/ceph/osd pattern: "{{ item.2 }}-*" register: ceph_osd_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 not in _lvm_list.keys() - name: slurp ceph osd files content slurp: src: "{{ item['files'][0]['path'] }}" delegate_to: "{{ item.item.0 }}" register: ceph_osd_files_content loop: "{{ ceph_osd_data.results }}" when: - item.skipped is undefined - item.matched > 0 - name: set_fact ceph_osd_files_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({ item.item.item.2: item.content | b64decode | from_json}) }}" with_items: "{{ ceph_osd_files_content.results }}" when: item.skipped is undefined - name: mark osd(s) out of the cluster ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: out environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true - name: stop osd(s) service service: name: ceph-osd@{{ item.2 }} state: stopped enabled: no loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: umount osd lockbox ansible.posix.mount: path: "/var/lib/ceph/osd-lockbox/{{ ceph_osd_data_json[item.2]['data']['uuid'] }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - not containerized_deployment | bool - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2]['data']['uuid'] is defined - name: umount osd data ansible.posix.mount: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: not containerized_deployment | bool - name: get parent device for data partition command: lsblk --noheadings --output PKNAME --nodeps "{{ ceph_osd_data_json[item.2]['data']['path'] }}" register: parent_device_data_part loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['data']['path'] is defined - name: add pkname information in ceph_osd_data_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({item.item[2]: {'pkname_data': '/dev/' + item.stdout }}, recursive=True) }}" loop: "{{ parent_device_data_part.results }}" when: item.skipped is undefined - name: close dmcrypt close on devices if needed command: "cryptsetup close {{ ceph_osd_data_json[item.2][item.3]['uuid'] }}" with_nested: - "{{ _osd_hosts }}" - [ 'block_dmcrypt', 'block.db_dmcrypt', 'block.wal_dmcrypt', 'data', 'journal_dmcrypt' ] delegate_to: "{{ item.0 }}" failed_when: false register: result until: result is succeeded when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2][item.3] is defined - name: use ceph-volume lvm zap to destroy all partitions ceph_volume: cluster: "{{ cluster }}" action: zap destroy: true data: "{{ ceph_osd_data_json[item.2]['pkname_data'] if item.3 == 'data' else ceph_osd_data_json[item.2][item.3]['path'] }}" environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" with_nested: - "{{ _osd_hosts }}" - [ 'block', 'block.db', 'block.wal', 'journal', 'data' ] delegate_to: "{{ item.0 }}" failed_when: false register: result when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2][item.3] is defined - name: zap osd devices ceph_volume: action: "zap" osd_fsid: "{{ item.1 }}" environment: CEPH_VOLUME_DEBUG: "{{ ceph_volume_debug }}" CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 in _lvm_list.keys() - name: ensure osds are marked down ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: down environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true delegate_to: "{{ groups[mon_group_name][0] }}" - name: purge osd(s) from the cluster ceph_osd: ids: "{{ item }}" cluster: "{{ cluster }}" state: purge environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true with_items: "{{ osd_to_kill.split(',') }}" - name: remove osd data dir file: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: show ceph health command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} -s" changed_when: false - name: show ceph osd tree command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd tree" changed_when: false

asm0deuz

dffe7b47de70b6eeec71a3fa86f8c407adb4dd8e

64e08f2c0bdea6f4c4ad5862dc8f350c6adbe2cd

this command won't work in containerized deployment, you must run it from within a container..

guits

ceph/ceph-ansible

[skip ci] Refresh /etc/ceph/osd json files content before zapping the disks

If the physical disk to device path mapping has changed since the last ceph-volume simple scan (e.g. addition or removal of disks), a wrong disk could be deleted. Closes: https://bugzilla.redhat.com/show_bug.cgi?id=2071035 Signed-off-by: Teoman ONAY <tonay@redhat.com>

2022-07-04 10:03:03+00:00

2022-07-11 07:14:41+00:00

infrastructure-playbooks/shrink-osd.yml

--- # This playbook shrinks Ceph OSDs that have been created with ceph-volume. # It can remove any number of OSD(s) from the cluster and ALL THEIR DATA # # Use it like this: # ansible-playbook shrink-osd.yml -e osd_to_kill=0,2,6 # Prompts for confirmation to shrink, defaults to no and # doesn't shrink the cluster. yes shrinks the cluster. # # ansible-playbook -e ireallymeanit=yes|no shrink-osd.yml # Overrides the prompt using -e option. Can be used in # automation scripts to avoid interactive prompt. - name: gather facts and check the init system hosts: - "{{ mon_group_name|default('mons') }}" - "{{ osd_group_name|default('osds') }}" become: True tasks: - debug: msg="gather facts on all Ceph hosts for following reference" - name: confirm whether user really meant to remove osd(s) from the cluster hosts: "{{ groups[mon_group_name][0] }}" become: true vars_prompt: - name: ireallymeanit prompt: Are you sure you want to shrink the cluster? default: 'no' private: no vars: mon_group_name: mons osd_group_name: osds pre_tasks: - name: exit playbook, if user did not mean to shrink cluster fail: msg: "Exiting shrink-osd playbook, no osd(s) was/were removed.. To shrink the cluster, either say 'yes' on the prompt or or use `-e ireallymeanit=yes` on the command line when invoking the playbook" when: ireallymeanit != 'yes' - name: exit playbook, if no osd(s) was/were given fail: msg: "osd_to_kill must be declared Exiting shrink-osd playbook, no OSD(s) was/were removed. On the command line when invoking the playbook, you can use -e osd_to_kill=0,1,2,3 argument." when: osd_to_kill is not defined - name: check the osd ids passed have the correct format fail: msg: "The id {{ item }} has wrong format, please pass the number only" with_items: "{{ osd_to_kill.split(',') }}" when: not item is regex("^\d+$") tasks: - import_role: name: ceph-defaults - import_role: name: ceph-facts tasks_from: container_binary post_tasks: - name: set_fact container_exec_cmd build docker exec command (containerized) set_fact: container_exec_cmd: "{{ container_binary }} exec ceph-mon-{{ ansible_facts['hostname'] }}" when: containerized_deployment | bool - name: exit playbook, if can not connect to the cluster command: "{{ container_exec_cmd }} timeout 5 ceph --cluster {{ cluster }} health" register: ceph_health changed_when: false until: ceph_health.stdout.find("HEALTH") > -1 retries: 5 delay: 2 - name: find the host(s) where the osd(s) is/are running on command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd find {{ item }}" changed_when: false with_items: "{{ osd_to_kill.split(',') }}" register: find_osd_hosts - name: set_fact osd_hosts set_fact: osd_hosts: "{{ osd_hosts | default([]) + [ [ (item.stdout | from_json).crush_location.host, (item.stdout | from_json).osd_fsid, item.item ] ] }}" with_items: "{{ find_osd_hosts.results }}" - name: set_fact _osd_hosts set_fact: _osd_hosts: "{{ _osd_hosts | default([]) + [ [ item.0, item.2, item.3 ] ] }}" with_nested: - "{{ groups.get(osd_group_name) }}" - "{{ osd_hosts }}" when: hostvars[item.0]['ansible_facts']['hostname'] == item.1 - name: get ceph-volume lvm list data ceph_volume: cluster: "{{ cluster }}" action: list environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" register: _lvm_list_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" - name: set_fact _lvm_list set_fact: _lvm_list: "{{ _lvm_list | default({}) | combine(item.stdout | from_json) }}" with_items: "{{ _lvm_list_data.results }}" - name: find /etc/ceph/osd files find: paths: /etc/ceph/osd pattern: "{{ item.2 }}-*" register: ceph_osd_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 not in _lvm_list.keys() - name: slurp ceph osd files content slurp: src: "{{ item['files'][0]['path'] }}" delegate_to: "{{ item.item.0 }}" register: ceph_osd_files_content loop: "{{ ceph_osd_data.results }}" when: - item.skipped is undefined - item.matched > 0 - name: set_fact ceph_osd_files_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({ item.item.item.2: item.content | b64decode | from_json}) }}" with_items: "{{ ceph_osd_files_content.results }}" when: item.skipped is undefined - name: mark osd(s) out of the cluster ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: out environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true - name: stop osd(s) service service: name: ceph-osd@{{ item.2 }} state: stopped enabled: no loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: umount osd lockbox ansible.posix.mount: path: "/var/lib/ceph/osd-lockbox/{{ ceph_osd_data_json[item.2]['data']['uuid'] }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - not containerized_deployment | bool - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2]['data']['uuid'] is defined - name: umount osd data ansible.posix.mount: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: not containerized_deployment | bool - name: get parent device for data partition command: lsblk --noheadings --output PKNAME --nodeps "{{ ceph_osd_data_json[item.2]['data']['path'] }}" register: parent_device_data_part loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['data']['path'] is defined - name: add pkname information in ceph_osd_data_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({item.item[2]: {'pkname_data': '/dev/' + item.stdout }}, recursive=True) }}" loop: "{{ parent_device_data_part.results }}" when: item.skipped is undefined - name: close dmcrypt close on devices if needed command: "cryptsetup close {{ ceph_osd_data_json[item.2][item.3]['uuid'] }}" with_nested: - "{{ _osd_hosts }}" - [ 'block_dmcrypt', 'block.db_dmcrypt', 'block.wal_dmcrypt', 'data', 'journal_dmcrypt' ] delegate_to: "{{ item.0 }}" failed_when: false register: result until: result is succeeded when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2][item.3] is defined - name: use ceph-volume lvm zap to destroy all partitions ceph_volume: cluster: "{{ cluster }}" action: zap destroy: true data: "{{ ceph_osd_data_json[item.2]['pkname_data'] if item.3 == 'data' else ceph_osd_data_json[item.2][item.3]['path'] }}" environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" with_nested: - "{{ _osd_hosts }}" - [ 'block', 'block.db', 'block.wal', 'journal', 'data' ] delegate_to: "{{ item.0 }}" failed_when: false register: result when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2][item.3] is defined - name: zap osd devices ceph_volume: action: "zap" osd_fsid: "{{ item.1 }}" environment: CEPH_VOLUME_DEBUG: "{{ ceph_volume_debug }}" CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 in _lvm_list.keys() - name: ensure osds are marked down ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: down environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true delegate_to: "{{ groups[mon_group_name][0] }}" - name: purge osd(s) from the cluster ceph_osd: ids: "{{ item }}" cluster: "{{ cluster }}" state: purge environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true with_items: "{{ osd_to_kill.split(',') }}" - name: remove osd data dir file: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: show ceph health command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} -s" changed_when: false - name: show ceph osd tree command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd tree" changed_when: false

--- # This playbook shrinks Ceph OSDs that have been created with ceph-volume. # It can remove any number of OSD(s) from the cluster and ALL THEIR DATA # # Use it like this: # ansible-playbook shrink-osd.yml -e osd_to_kill=0,2,6 # Prompts for confirmation to shrink, defaults to no and # doesn't shrink the cluster. yes shrinks the cluster. # # ansible-playbook -e ireallymeanit=yes|no shrink-osd.yml # Overrides the prompt using -e option. Can be used in # automation scripts to avoid interactive prompt. - name: gather facts and check the init system hosts: - "{{ mon_group_name|default('mons') }}" - "{{ osd_group_name|default('osds') }}" become: True tasks: - debug: msg="gather facts on all Ceph hosts for following reference" - name: confirm whether user really meant to remove osd(s) from the cluster hosts: "{{ groups[mon_group_name][0] }}" become: true vars_prompt: - name: ireallymeanit prompt: Are you sure you want to shrink the cluster? default: 'no' private: no vars: mon_group_name: mons osd_group_name: osds pre_tasks: - name: exit playbook, if user did not mean to shrink cluster fail: msg: "Exiting shrink-osd playbook, no osd(s) was/were removed.. To shrink the cluster, either say 'yes' on the prompt or or use `-e ireallymeanit=yes` on the command line when invoking the playbook" when: ireallymeanit != 'yes' - name: exit playbook, if no osd(s) was/were given fail: msg: "osd_to_kill must be declared Exiting shrink-osd playbook, no OSD(s) was/were removed. On the command line when invoking the playbook, you can use -e osd_to_kill=0,1,2,3 argument." when: osd_to_kill is not defined - name: check the osd ids passed have the correct format fail: msg: "The id {{ item }} has wrong format, please pass the number only" with_items: "{{ osd_to_kill.split(',') }}" when: not item is regex("^\d+$") tasks: - import_role: name: ceph-defaults - import_role: name: ceph-facts tasks_from: container_binary post_tasks: - name: set_fact container_exec_cmd build docker exec command (containerized) set_fact: container_exec_cmd: "{{ container_binary }} exec ceph-mon-{{ ansible_facts['hostname'] }}" when: containerized_deployment | bool - name: exit playbook, if can not connect to the cluster command: "{{ container_exec_cmd }} timeout 5 ceph --cluster {{ cluster }} health" register: ceph_health changed_when: false until: ceph_health.stdout.find("HEALTH") > -1 retries: 5 delay: 2 - name: find the host(s) where the osd(s) is/are running on command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd find {{ item }}" changed_when: false with_items: "{{ osd_to_kill.split(',') }}" register: find_osd_hosts - name: set_fact osd_hosts set_fact: osd_hosts: "{{ osd_hosts | default([]) + [ [ (item.stdout | from_json).crush_location.host, (item.stdout | from_json).osd_fsid, item.item ] ] }}" with_items: "{{ find_osd_hosts.results }}" - name: set_fact _osd_hosts set_fact: _osd_hosts: "{{ _osd_hosts | default([]) + [ [ item.0, item.2, item.3 ] ] }}" with_nested: - "{{ groups.get(osd_group_name) }}" - "{{ osd_hosts }}" when: hostvars[item.0]['ansible_facts']['hostname'] == item.1 - name: set_fact host_list set_fact: host_list: "{{ host_list | default([]) | union([item.0]) }}" loop: "{{ _osd_hosts }}" - name: get ceph-volume lvm list data ceph_volume: cluster: "{{ cluster }}" action: list environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" register: _lvm_list_data delegate_to: "{{ item }}" loop: "{{ host_list }}" - name: set_fact _lvm_list set_fact: _lvm_list: "{{ _lvm_list | default({}) | combine(item.stdout | from_json) }}" with_items: "{{ _lvm_list_data.results }}" - name: refresh /etc/ceph/osd files non containerized_deployment ceph_volume_simple_scan: cluster: "{{ cluster }}" force: true delegate_to: "{{ item }}" loop: "{{ host_list }}" when: not containerized_deployment | bool - name: refresh /etc/ceph/osd files containerized_deployment command: "{{ container_binary }} exec ceph-osd-{{ item.2 }} ceph-volume simple scan --force /var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" changed_when: false delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: containerized_deployment | bool - name: find /etc/ceph/osd files find: paths: /etc/ceph/osd pattern: "{{ item.2 }}-*" register: ceph_osd_data delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 not in _lvm_list.keys() - name: slurp ceph osd files content slurp: src: "{{ item['files'][0]['path'] }}" delegate_to: "{{ item.item.0 }}" register: ceph_osd_files_content loop: "{{ ceph_osd_data.results }}" when: - item.skipped is undefined - item.matched > 0 - name: set_fact ceph_osd_files_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({ item.item.item.2: item.content | b64decode | from_json}) }}" with_items: "{{ ceph_osd_files_content.results }}" when: item.skipped is undefined - name: mark osd(s) out of the cluster ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: out environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true - name: stop osd(s) service service: name: ceph-osd@{{ item.2 }} state: stopped enabled: no loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: umount osd lockbox ansible.posix.mount: path: "/var/lib/ceph/osd-lockbox/{{ ceph_osd_data_json[item.2]['data']['uuid'] }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - not containerized_deployment | bool - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2]['data']['uuid'] is defined - name: umount osd data ansible.posix.mount: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: not containerized_deployment | bool - name: get parent device for data partition command: lsblk --noheadings --output PKNAME --nodeps "{{ ceph_osd_data_json[item.2]['data']['path'] }}" register: parent_device_data_part loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['data']['path'] is defined - name: add pkname information in ceph_osd_data_json set_fact: ceph_osd_data_json: "{{ ceph_osd_data_json | default({}) | combine({item.item[2]: {'pkname_data': '/dev/' + item.stdout }}, recursive=True) }}" loop: "{{ parent_device_data_part.results }}" when: item.skipped is undefined - name: close dmcrypt close on devices if needed command: "cryptsetup close {{ ceph_osd_data_json[item.2][item.3]['uuid'] }}" with_nested: - "{{ _osd_hosts }}" - [ 'block_dmcrypt', 'block.db_dmcrypt', 'block.wal_dmcrypt', 'data', 'journal_dmcrypt' ] delegate_to: "{{ item.0 }}" failed_when: false register: result until: result is succeeded when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2]['encrypted'] | default(False) | bool - ceph_osd_data_json[item.2][item.3] is defined - name: use ceph-volume lvm zap to destroy all partitions ceph_volume: cluster: "{{ cluster }}" action: zap destroy: true data: "{{ ceph_osd_data_json[item.2]['pkname_data'] if item.3 == 'data' else ceph_osd_data_json[item.2][item.3]['path'] }}" environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" with_nested: - "{{ _osd_hosts }}" - [ 'block', 'block.db', 'block.wal', 'journal', 'data' ] delegate_to: "{{ item.0 }}" failed_when: false register: result when: - item.2 not in _lvm_list.keys() - ceph_osd_data_json[item.2][item.3] is defined - name: zap osd devices ceph_volume: action: "zap" osd_fsid: "{{ item.1 }}" environment: CEPH_VOLUME_DEBUG: "{{ ceph_volume_debug }}" CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" delegate_to: "{{ item.0 }}" loop: "{{ _osd_hosts }}" when: item.2 in _lvm_list.keys() - name: ensure osds are marked down ceph_osd: ids: "{{ osd_to_kill.split(',') }}" cluster: "{{ cluster }}" state: down environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true delegate_to: "{{ groups[mon_group_name][0] }}" - name: purge osd(s) from the cluster ceph_osd: ids: "{{ item }}" cluster: "{{ cluster }}" state: purge environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" run_once: true with_items: "{{ osd_to_kill.split(',') }}" - name: remove osd data dir file: path: "/var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" state: absent loop: "{{ _osd_hosts }}" delegate_to: "{{ item.0 }}" - name: show ceph health command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} -s" changed_when: false - name: show ceph osd tree command: "{{ container_exec_cmd }} ceph --cluster {{ cluster }} osd tree" changed_when: false

asm0deuz

dffe7b47de70b6eeec71a3fa86f8c407adb4dd8e

64e08f2c0bdea6f4c4ad5862dc8f350c6adbe2cd

```suggestion command: "{{ container_binary }} exec ceph-osd-{{ item.2 }} ceph-volume simple scan --force /var/lib/ceph/osd/{{ cluster }}-{{ item.2 }}" ```

guits

ceph/ceph-ansible

fix(ceph-grafana): make dashboard download work again

This fixes the dashboard download for pacific and later. Since ceph switched to Prometheus Monitoring Mixins the path to the generated dashboards has changed. It is still working for octopus but it's broken from pacific onwards. This change fixes the issue. Currently I only added the two latest releases to the check. Also, this is only related to non containerized deployments.

2022-06-10 15:27:51+00:00

2022-06-14 12:36:24+00:00

roles/ceph-grafana/tasks/configure_grafana.yml

--- - name: install ceph-grafana-dashboards package on RedHat or SUSE package: name: ceph-grafana-dashboards state: "{{ (upgrade_ceph_packages|bool) | ternary('latest','present') }}" register: result until: result is succeeded when: - not containerized_deployment | bool - ansible_facts['os_family'] in ['RedHat', 'Suse'] tags: package-install - name: make sure grafana is down service: name: grafana-server state: stopped - name: wait for grafana to be stopped wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}' state: stopped - name: make sure grafana configuration directories exist file: path: "{{ item }}" state: directory recurse: yes owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" with_items: - "/etc/grafana/dashboards/ceph-dashboard" - "/etc/grafana/provisioning/datasources" - "/etc/grafana/provisioning/dashboards" - "/etc/grafana/provisioning/notifiers" - name: download ceph grafana dashboards get_url: url: "https://raw.githubusercontent.com/ceph/ceph/{{ grafana_dashboard_version }}/monitoring/grafana/dashboards/{{ item }}" dest: "/etc/grafana/dashboards/ceph-dashboard/{{ item }}" with_items: "{{ grafana_dashboard_files }}" when: - not containerized_deployment | bool - not ansible_facts['os_family'] in ['RedHat', 'Suse'] - name: write grafana.ini openstack.config_template.config_template: src: grafana.ini.j2 dest: /etc/grafana/grafana.ini owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 config_type: ini config_overrides: "{{ grafana_conf_overrides }}" - name: write datasources provisioning config file template: src: datasources-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/datasources/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 - name: Write dashboards provisioning config file template: src: dashboards-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/dashboards/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 when: not containerized_deployment | bool - name: copy grafana SSL certificate file copy: src: "{{ grafana_crt }}" dest: "/etc/grafana/ceph-dashboard.crt" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_crt | length > 0 - dashboard_protocol == "https" - name: copy grafana SSL certificate key copy: src: "{{ grafana_key }}" dest: "/etc/grafana/ceph-dashboard.key" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0440 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_key | length > 0 - dashboard_protocol == "https" - name: generate a Self Signed OpenSSL certificate for dashboard shell: | test -f /etc/grafana/ceph-dashboard.key -a -f /etc/grafana/ceph-dashboard.crt || \ (openssl req -new -nodes -x509 -subj '/O=IT/CN=ceph-grafana' -days 3650 -keyout /etc/grafana/ceph-dashboard.key -out /etc/grafana/ceph-dashboard.crt -extensions v3_ca && \ chown {{ grafana_uid }}:{{ grafana_uid }} /etc/grafana/ceph-dashboard.key /etc/grafana/ceph-dashboard.crt) when: - dashboard_protocol == "https" - grafana_key | length == 0 or grafana_crt | length == 0 - name: enable and start grafana service: name: grafana-server state: restarted enabled: true - name: wait for grafana to start wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}'

--- - name: install ceph-grafana-dashboards package on RedHat or SUSE package: name: ceph-grafana-dashboards state: "{{ (upgrade_ceph_packages|bool) | ternary('latest','present') }}" register: result until: result is succeeded when: - not containerized_deployment | bool - ansible_facts['os_family'] in ['RedHat', 'Suse'] tags: package-install - name: make sure grafana is down service: name: grafana-server state: stopped - name: wait for grafana to be stopped wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}' state: stopped - name: make sure grafana configuration directories exist file: path: "{{ item }}" state: directory recurse: yes owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" with_items: - "/etc/grafana/dashboards/ceph-dashboard" - "/etc/grafana/provisioning/datasources" - "/etc/grafana/provisioning/dashboards" - "/etc/grafana/provisioning/notifiers" - name: download ceph grafana dashboards get_url: url: "https://raw.githubusercontent.com/ceph/ceph/{{ grafana_dashboard_version }}/monitoring/ceph-mixin/dashboards_out/{{ item }}" dest: "/etc/grafana/dashboards/ceph-dashboard/{{ item }}" with_items: "{{ grafana_dashboard_files }}" when: - not containerized_deployment | bool - not ansible_facts['os_family'] in ['RedHat', 'Suse'] - name: write grafana.ini openstack.config_template.config_template: src: grafana.ini.j2 dest: /etc/grafana/grafana.ini owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 config_type: ini config_overrides: "{{ grafana_conf_overrides }}" - name: write datasources provisioning config file template: src: datasources-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/datasources/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 - name: Write dashboards provisioning config file template: src: dashboards-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/dashboards/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 when: not containerized_deployment | bool - name: copy grafana SSL certificate file copy: src: "{{ grafana_crt }}" dest: "/etc/grafana/ceph-dashboard.crt" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_crt | length > 0 - dashboard_protocol == "https" - name: copy grafana SSL certificate key copy: src: "{{ grafana_key }}" dest: "/etc/grafana/ceph-dashboard.key" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0440 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_key | length > 0 - dashboard_protocol == "https" - name: generate a Self Signed OpenSSL certificate for dashboard shell: | test -f /etc/grafana/ceph-dashboard.key -a -f /etc/grafana/ceph-dashboard.crt || \ (openssl req -new -nodes -x509 -subj '/O=IT/CN=ceph-grafana' -days 3650 -keyout /etc/grafana/ceph-dashboard.key -out /etc/grafana/ceph-dashboard.crt -extensions v3_ca && \ chown {{ grafana_uid }}:{{ grafana_uid }} /etc/grafana/ceph-dashboard.key /etc/grafana/ceph-dashboard.crt) when: - dashboard_protocol == "https" - grafana_key | length == 0 or grafana_crt | length == 0 - name: enable and start grafana service: name: grafana-server state: restarted enabled: true - name: wait for grafana to start wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}'

mitch000001

8a5fb702f2a3df46834baf6019285463bbfcc4fb

4edaab5f4c5445cb1fafc5d8824c49717e9f96c8

```suggestion url: "https://raw.githubusercontent.com/ceph/ceph/{{ grafana_dashboard_version }}/monitoring/ceph-mixin/dashboards_out/{{ item }}" ```

guits

ceph/ceph-ansible

fix(ceph-grafana): make dashboard download work again

This fixes the dashboard download for pacific and later. Since ceph switched to Prometheus Monitoring Mixins the path to the generated dashboards has changed. It is still working for octopus but it's broken from pacific onwards. This change fixes the issue. Currently I only added the two latest releases to the check. Also, this is only related to non containerized deployments.

2022-06-10 15:27:51+00:00

2022-06-14 12:36:24+00:00

roles/ceph-grafana/tasks/configure_grafana.yml

--- - name: install ceph-grafana-dashboards package on RedHat or SUSE package: name: ceph-grafana-dashboards state: "{{ (upgrade_ceph_packages|bool) | ternary('latest','present') }}" register: result until: result is succeeded when: - not containerized_deployment | bool - ansible_facts['os_family'] in ['RedHat', 'Suse'] tags: package-install - name: make sure grafana is down service: name: grafana-server state: stopped - name: wait for grafana to be stopped wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}' state: stopped - name: make sure grafana configuration directories exist file: path: "{{ item }}" state: directory recurse: yes owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" with_items: - "/etc/grafana/dashboards/ceph-dashboard" - "/etc/grafana/provisioning/datasources" - "/etc/grafana/provisioning/dashboards" - "/etc/grafana/provisioning/notifiers" - name: download ceph grafana dashboards get_url: url: "https://raw.githubusercontent.com/ceph/ceph/{{ grafana_dashboard_version }}/monitoring/grafana/dashboards/{{ item }}" dest: "/etc/grafana/dashboards/ceph-dashboard/{{ item }}" with_items: "{{ grafana_dashboard_files }}" when: - not containerized_deployment | bool - not ansible_facts['os_family'] in ['RedHat', 'Suse'] - name: write grafana.ini openstack.config_template.config_template: src: grafana.ini.j2 dest: /etc/grafana/grafana.ini owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 config_type: ini config_overrides: "{{ grafana_conf_overrides }}" - name: write datasources provisioning config file template: src: datasources-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/datasources/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 - name: Write dashboards provisioning config file template: src: dashboards-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/dashboards/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 when: not containerized_deployment | bool - name: copy grafana SSL certificate file copy: src: "{{ grafana_crt }}" dest: "/etc/grafana/ceph-dashboard.crt" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_crt | length > 0 - dashboard_protocol == "https" - name: copy grafana SSL certificate key copy: src: "{{ grafana_key }}" dest: "/etc/grafana/ceph-dashboard.key" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0440 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_key | length > 0 - dashboard_protocol == "https" - name: generate a Self Signed OpenSSL certificate for dashboard shell: | test -f /etc/grafana/ceph-dashboard.key -a -f /etc/grafana/ceph-dashboard.crt || \ (openssl req -new -nodes -x509 -subj '/O=IT/CN=ceph-grafana' -days 3650 -keyout /etc/grafana/ceph-dashboard.key -out /etc/grafana/ceph-dashboard.crt -extensions v3_ca && \ chown {{ grafana_uid }}:{{ grafana_uid }} /etc/grafana/ceph-dashboard.key /etc/grafana/ceph-dashboard.crt) when: - dashboard_protocol == "https" - grafana_key | length == 0 or grafana_crt | length == 0 - name: enable and start grafana service: name: grafana-server state: restarted enabled: true - name: wait for grafana to start wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}'

--- - name: install ceph-grafana-dashboards package on RedHat or SUSE package: name: ceph-grafana-dashboards state: "{{ (upgrade_ceph_packages|bool) | ternary('latest','present') }}" register: result until: result is succeeded when: - not containerized_deployment | bool - ansible_facts['os_family'] in ['RedHat', 'Suse'] tags: package-install - name: make sure grafana is down service: name: grafana-server state: stopped - name: wait for grafana to be stopped wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}' state: stopped - name: make sure grafana configuration directories exist file: path: "{{ item }}" state: directory recurse: yes owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" with_items: - "/etc/grafana/dashboards/ceph-dashboard" - "/etc/grafana/provisioning/datasources" - "/etc/grafana/provisioning/dashboards" - "/etc/grafana/provisioning/notifiers" - name: download ceph grafana dashboards get_url: url: "https://raw.githubusercontent.com/ceph/ceph/{{ grafana_dashboard_version }}/monitoring/ceph-mixin/dashboards_out/{{ item }}" dest: "/etc/grafana/dashboards/ceph-dashboard/{{ item }}" with_items: "{{ grafana_dashboard_files }}" when: - not containerized_deployment | bool - not ansible_facts['os_family'] in ['RedHat', 'Suse'] - name: write grafana.ini openstack.config_template.config_template: src: grafana.ini.j2 dest: /etc/grafana/grafana.ini owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 config_type: ini config_overrides: "{{ grafana_conf_overrides }}" - name: write datasources provisioning config file template: src: datasources-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/datasources/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 - name: Write dashboards provisioning config file template: src: dashboards-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/dashboards/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 when: not containerized_deployment | bool - name: copy grafana SSL certificate file copy: src: "{{ grafana_crt }}" dest: "/etc/grafana/ceph-dashboard.crt" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_crt | length > 0 - dashboard_protocol == "https" - name: copy grafana SSL certificate key copy: src: "{{ grafana_key }}" dest: "/etc/grafana/ceph-dashboard.key" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0440 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_key | length > 0 - dashboard_protocol == "https" - name: generate a Self Signed OpenSSL certificate for dashboard shell: | test -f /etc/grafana/ceph-dashboard.key -a -f /etc/grafana/ceph-dashboard.crt || \ (openssl req -new -nodes -x509 -subj '/O=IT/CN=ceph-grafana' -days 3650 -keyout /etc/grafana/ceph-dashboard.key -out /etc/grafana/ceph-dashboard.crt -extensions v3_ca && \ chown {{ grafana_uid }}:{{ grafana_uid }} /etc/grafana/ceph-dashboard.key /etc/grafana/ceph-dashboard.crt) when: - dashboard_protocol == "https" - grafana_key | length == 0 or grafana_crt | length == 0 - name: enable and start grafana service: name: grafana-server state: restarted enabled: true - name: wait for grafana to start wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}'

mitch000001

8a5fb702f2a3df46834baf6019285463bbfcc4fb

4edaab5f4c5445cb1fafc5d8824c49717e9f96c8

```suggestion ```

guits

ceph/ceph-ansible

fix(ceph-grafana): make dashboard download work again

This fixes the dashboard download for pacific and later. Since ceph switched to Prometheus Monitoring Mixins the path to the generated dashboards has changed. It is still working for octopus but it's broken from pacific onwards. This change fixes the issue. Currently I only added the two latest releases to the check. Also, this is only related to non containerized deployments.

2022-06-10 15:27:51+00:00

2022-06-14 12:36:24+00:00

roles/ceph-grafana/tasks/configure_grafana.yml

--- - name: install ceph-grafana-dashboards package on RedHat or SUSE package: name: ceph-grafana-dashboards state: "{{ (upgrade_ceph_packages|bool) | ternary('latest','present') }}" register: result until: result is succeeded when: - not containerized_deployment | bool - ansible_facts['os_family'] in ['RedHat', 'Suse'] tags: package-install - name: make sure grafana is down service: name: grafana-server state: stopped - name: wait for grafana to be stopped wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}' state: stopped - name: make sure grafana configuration directories exist file: path: "{{ item }}" state: directory recurse: yes owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" with_items: - "/etc/grafana/dashboards/ceph-dashboard" - "/etc/grafana/provisioning/datasources" - "/etc/grafana/provisioning/dashboards" - "/etc/grafana/provisioning/notifiers" - name: download ceph grafana dashboards get_url: url: "https://raw.githubusercontent.com/ceph/ceph/{{ grafana_dashboard_version }}/monitoring/grafana/dashboards/{{ item }}" dest: "/etc/grafana/dashboards/ceph-dashboard/{{ item }}" with_items: "{{ grafana_dashboard_files }}" when: - not containerized_deployment | bool - not ansible_facts['os_family'] in ['RedHat', 'Suse'] - name: write grafana.ini openstack.config_template.config_template: src: grafana.ini.j2 dest: /etc/grafana/grafana.ini owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 config_type: ini config_overrides: "{{ grafana_conf_overrides }}" - name: write datasources provisioning config file template: src: datasources-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/datasources/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 - name: Write dashboards provisioning config file template: src: dashboards-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/dashboards/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 when: not containerized_deployment | bool - name: copy grafana SSL certificate file copy: src: "{{ grafana_crt }}" dest: "/etc/grafana/ceph-dashboard.crt" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_crt | length > 0 - dashboard_protocol == "https" - name: copy grafana SSL certificate key copy: src: "{{ grafana_key }}" dest: "/etc/grafana/ceph-dashboard.key" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0440 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_key | length > 0 - dashboard_protocol == "https" - name: generate a Self Signed OpenSSL certificate for dashboard shell: | test -f /etc/grafana/ceph-dashboard.key -a -f /etc/grafana/ceph-dashboard.crt || \ (openssl req -new -nodes -x509 -subj '/O=IT/CN=ceph-grafana' -days 3650 -keyout /etc/grafana/ceph-dashboard.key -out /etc/grafana/ceph-dashboard.crt -extensions v3_ca && \ chown {{ grafana_uid }}:{{ grafana_uid }} /etc/grafana/ceph-dashboard.key /etc/grafana/ceph-dashboard.crt) when: - dashboard_protocol == "https" - grafana_key | length == 0 or grafana_crt | length == 0 - name: enable and start grafana service: name: grafana-server state: restarted enabled: true - name: wait for grafana to start wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}'

--- - name: install ceph-grafana-dashboards package on RedHat or SUSE package: name: ceph-grafana-dashboards state: "{{ (upgrade_ceph_packages|bool) | ternary('latest','present') }}" register: result until: result is succeeded when: - not containerized_deployment | bool - ansible_facts['os_family'] in ['RedHat', 'Suse'] tags: package-install - name: make sure grafana is down service: name: grafana-server state: stopped - name: wait for grafana to be stopped wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}' state: stopped - name: make sure grafana configuration directories exist file: path: "{{ item }}" state: directory recurse: yes owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" with_items: - "/etc/grafana/dashboards/ceph-dashboard" - "/etc/grafana/provisioning/datasources" - "/etc/grafana/provisioning/dashboards" - "/etc/grafana/provisioning/notifiers" - name: download ceph grafana dashboards get_url: url: "https://raw.githubusercontent.com/ceph/ceph/{{ grafana_dashboard_version }}/monitoring/ceph-mixin/dashboards_out/{{ item }}" dest: "/etc/grafana/dashboards/ceph-dashboard/{{ item }}" with_items: "{{ grafana_dashboard_files }}" when: - not containerized_deployment | bool - not ansible_facts['os_family'] in ['RedHat', 'Suse'] - name: write grafana.ini openstack.config_template.config_template: src: grafana.ini.j2 dest: /etc/grafana/grafana.ini owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 config_type: ini config_overrides: "{{ grafana_conf_overrides }}" - name: write datasources provisioning config file template: src: datasources-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/datasources/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 - name: Write dashboards provisioning config file template: src: dashboards-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/dashboards/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 when: not containerized_deployment | bool - name: copy grafana SSL certificate file copy: src: "{{ grafana_crt }}" dest: "/etc/grafana/ceph-dashboard.crt" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_crt | length > 0 - dashboard_protocol == "https" - name: copy grafana SSL certificate key copy: src: "{{ grafana_key }}" dest: "/etc/grafana/ceph-dashboard.key" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0440 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_key | length > 0 - dashboard_protocol == "https" - name: generate a Self Signed OpenSSL certificate for dashboard shell: | test -f /etc/grafana/ceph-dashboard.key -a -f /etc/grafana/ceph-dashboard.crt || \ (openssl req -new -nodes -x509 -subj '/O=IT/CN=ceph-grafana' -days 3650 -keyout /etc/grafana/ceph-dashboard.key -out /etc/grafana/ceph-dashboard.crt -extensions v3_ca && \ chown {{ grafana_uid }}:{{ grafana_uid }} /etc/grafana/ceph-dashboard.key /etc/grafana/ceph-dashboard.crt) when: - dashboard_protocol == "https" - grafana_key | length == 0 or grafana_crt | length == 0 - name: enable and start grafana service: name: grafana-server state: restarted enabled: true - name: wait for grafana to start wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}'

mitch000001

8a5fb702f2a3df46834baf6019285463bbfcc4fb

4edaab5f4c5445cb1fafc5d8824c49717e9f96c8

The change I proposed would be backwards compatible as the dashboards for octopus and earlier life in the old location still.

mitch000001

ceph/ceph-ansible

fix(ceph-grafana): make dashboard download work again

This fixes the dashboard download for pacific and later. Since ceph switched to Prometheus Monitoring Mixins the path to the generated dashboards has changed. It is still working for octopus but it's broken from pacific onwards. This change fixes the issue. Currently I only added the two latest releases to the check. Also, this is only related to non containerized deployments.

2022-06-10 15:27:51+00:00

2022-06-14 12:36:24+00:00

roles/ceph-grafana/tasks/configure_grafana.yml

--- - name: install ceph-grafana-dashboards package on RedHat or SUSE package: name: ceph-grafana-dashboards state: "{{ (upgrade_ceph_packages|bool) | ternary('latest','present') }}" register: result until: result is succeeded when: - not containerized_deployment | bool - ansible_facts['os_family'] in ['RedHat', 'Suse'] tags: package-install - name: make sure grafana is down service: name: grafana-server state: stopped - name: wait for grafana to be stopped wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}' state: stopped - name: make sure grafana configuration directories exist file: path: "{{ item }}" state: directory recurse: yes owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" with_items: - "/etc/grafana/dashboards/ceph-dashboard" - "/etc/grafana/provisioning/datasources" - "/etc/grafana/provisioning/dashboards" - "/etc/grafana/provisioning/notifiers" - name: download ceph grafana dashboards get_url: url: "https://raw.githubusercontent.com/ceph/ceph/{{ grafana_dashboard_version }}/monitoring/grafana/dashboards/{{ item }}" dest: "/etc/grafana/dashboards/ceph-dashboard/{{ item }}" with_items: "{{ grafana_dashboard_files }}" when: - not containerized_deployment | bool - not ansible_facts['os_family'] in ['RedHat', 'Suse'] - name: write grafana.ini openstack.config_template.config_template: src: grafana.ini.j2 dest: /etc/grafana/grafana.ini owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 config_type: ini config_overrides: "{{ grafana_conf_overrides }}" - name: write datasources provisioning config file template: src: datasources-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/datasources/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 - name: Write dashboards provisioning config file template: src: dashboards-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/dashboards/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 when: not containerized_deployment | bool - name: copy grafana SSL certificate file copy: src: "{{ grafana_crt }}" dest: "/etc/grafana/ceph-dashboard.crt" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_crt | length > 0 - dashboard_protocol == "https" - name: copy grafana SSL certificate key copy: src: "{{ grafana_key }}" dest: "/etc/grafana/ceph-dashboard.key" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0440 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_key | length > 0 - dashboard_protocol == "https" - name: generate a Self Signed OpenSSL certificate for dashboard shell: | test -f /etc/grafana/ceph-dashboard.key -a -f /etc/grafana/ceph-dashboard.crt || \ (openssl req -new -nodes -x509 -subj '/O=IT/CN=ceph-grafana' -days 3650 -keyout /etc/grafana/ceph-dashboard.key -out /etc/grafana/ceph-dashboard.crt -extensions v3_ca && \ chown {{ grafana_uid }}:{{ grafana_uid }} /etc/grafana/ceph-dashboard.key /etc/grafana/ceph-dashboard.crt) when: - dashboard_protocol == "https" - grafana_key | length == 0 or grafana_crt | length == 0 - name: enable and start grafana service: name: grafana-server state: restarted enabled: true - name: wait for grafana to start wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}'

--- - name: install ceph-grafana-dashboards package on RedHat or SUSE package: name: ceph-grafana-dashboards state: "{{ (upgrade_ceph_packages|bool) | ternary('latest','present') }}" register: result until: result is succeeded when: - not containerized_deployment | bool - ansible_facts['os_family'] in ['RedHat', 'Suse'] tags: package-install - name: make sure grafana is down service: name: grafana-server state: stopped - name: wait for grafana to be stopped wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}' state: stopped - name: make sure grafana configuration directories exist file: path: "{{ item }}" state: directory recurse: yes owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" with_items: - "/etc/grafana/dashboards/ceph-dashboard" - "/etc/grafana/provisioning/datasources" - "/etc/grafana/provisioning/dashboards" - "/etc/grafana/provisioning/notifiers" - name: download ceph grafana dashboards get_url: url: "https://raw.githubusercontent.com/ceph/ceph/{{ grafana_dashboard_version }}/monitoring/ceph-mixin/dashboards_out/{{ item }}" dest: "/etc/grafana/dashboards/ceph-dashboard/{{ item }}" with_items: "{{ grafana_dashboard_files }}" when: - not containerized_deployment | bool - not ansible_facts['os_family'] in ['RedHat', 'Suse'] - name: write grafana.ini openstack.config_template.config_template: src: grafana.ini.j2 dest: /etc/grafana/grafana.ini owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 config_type: ini config_overrides: "{{ grafana_conf_overrides }}" - name: write datasources provisioning config file template: src: datasources-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/datasources/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 - name: Write dashboards provisioning config file template: src: dashboards-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/dashboards/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 when: not containerized_deployment | bool - name: copy grafana SSL certificate file copy: src: "{{ grafana_crt }}" dest: "/etc/grafana/ceph-dashboard.crt" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_crt | length > 0 - dashboard_protocol == "https" - name: copy grafana SSL certificate key copy: src: "{{ grafana_key }}" dest: "/etc/grafana/ceph-dashboard.key" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0440 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_key | length > 0 - dashboard_protocol == "https" - name: generate a Self Signed OpenSSL certificate for dashboard shell: | test -f /etc/grafana/ceph-dashboard.key -a -f /etc/grafana/ceph-dashboard.crt || \ (openssl req -new -nodes -x509 -subj '/O=IT/CN=ceph-grafana' -days 3650 -keyout /etc/grafana/ceph-dashboard.key -out /etc/grafana/ceph-dashboard.crt -extensions v3_ca && \ chown {{ grafana_uid }}:{{ grafana_uid }} /etc/grafana/ceph-dashboard.key /etc/grafana/ceph-dashboard.crt) when: - dashboard_protocol == "https" - grafana_key | length == 0 or grafana_crt | length == 0 - name: enable and start grafana service: name: grafana-server state: restarted enabled: true - name: wait for grafana to start wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}'

mitch000001

8a5fb702f2a3df46834baf6019285463bbfcc4fb

4edaab5f4c5445cb1fafc5d8824c49717e9f96c8

I mean, we could also change this like that within master and cherry-pick into stable-6.0 onwards. Works for me. Is that the direction to go?

mitch000001

ceph/ceph-ansible

fix(ceph-grafana): make dashboard download work again

This fixes the dashboard download for pacific and later. Since ceph switched to Prometheus Monitoring Mixins the path to the generated dashboards has changed. It is still working for octopus but it's broken from pacific onwards. This change fixes the issue. Currently I only added the two latest releases to the check. Also, this is only related to non containerized deployments.

2022-06-10 15:27:51+00:00

2022-06-14 12:36:24+00:00

roles/ceph-grafana/tasks/configure_grafana.yml

--- - name: install ceph-grafana-dashboards package on RedHat or SUSE package: name: ceph-grafana-dashboards state: "{{ (upgrade_ceph_packages|bool) | ternary('latest','present') }}" register: result until: result is succeeded when: - not containerized_deployment | bool - ansible_facts['os_family'] in ['RedHat', 'Suse'] tags: package-install - name: make sure grafana is down service: name: grafana-server state: stopped - name: wait for grafana to be stopped wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}' state: stopped - name: make sure grafana configuration directories exist file: path: "{{ item }}" state: directory recurse: yes owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" with_items: - "/etc/grafana/dashboards/ceph-dashboard" - "/etc/grafana/provisioning/datasources" - "/etc/grafana/provisioning/dashboards" - "/etc/grafana/provisioning/notifiers" - name: download ceph grafana dashboards get_url: url: "https://raw.githubusercontent.com/ceph/ceph/{{ grafana_dashboard_version }}/monitoring/grafana/dashboards/{{ item }}" dest: "/etc/grafana/dashboards/ceph-dashboard/{{ item }}" with_items: "{{ grafana_dashboard_files }}" when: - not containerized_deployment | bool - not ansible_facts['os_family'] in ['RedHat', 'Suse'] - name: write grafana.ini openstack.config_template.config_template: src: grafana.ini.j2 dest: /etc/grafana/grafana.ini owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 config_type: ini config_overrides: "{{ grafana_conf_overrides }}" - name: write datasources provisioning config file template: src: datasources-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/datasources/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 - name: Write dashboards provisioning config file template: src: dashboards-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/dashboards/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 when: not containerized_deployment | bool - name: copy grafana SSL certificate file copy: src: "{{ grafana_crt }}" dest: "/etc/grafana/ceph-dashboard.crt" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_crt | length > 0 - dashboard_protocol == "https" - name: copy grafana SSL certificate key copy: src: "{{ grafana_key }}" dest: "/etc/grafana/ceph-dashboard.key" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0440 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_key | length > 0 - dashboard_protocol == "https" - name: generate a Self Signed OpenSSL certificate for dashboard shell: | test -f /etc/grafana/ceph-dashboard.key -a -f /etc/grafana/ceph-dashboard.crt || \ (openssl req -new -nodes -x509 -subj '/O=IT/CN=ceph-grafana' -days 3650 -keyout /etc/grafana/ceph-dashboard.key -out /etc/grafana/ceph-dashboard.crt -extensions v3_ca && \ chown {{ grafana_uid }}:{{ grafana_uid }} /etc/grafana/ceph-dashboard.key /etc/grafana/ceph-dashboard.crt) when: - dashboard_protocol == "https" - grafana_key | length == 0 or grafana_crt | length == 0 - name: enable and start grafana service: name: grafana-server state: restarted enabled: true - name: wait for grafana to start wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}'

--- - name: install ceph-grafana-dashboards package on RedHat or SUSE package: name: ceph-grafana-dashboards state: "{{ (upgrade_ceph_packages|bool) | ternary('latest','present') }}" register: result until: result is succeeded when: - not containerized_deployment | bool - ansible_facts['os_family'] in ['RedHat', 'Suse'] tags: package-install - name: make sure grafana is down service: name: grafana-server state: stopped - name: wait for grafana to be stopped wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}' state: stopped - name: make sure grafana configuration directories exist file: path: "{{ item }}" state: directory recurse: yes owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" with_items: - "/etc/grafana/dashboards/ceph-dashboard" - "/etc/grafana/provisioning/datasources" - "/etc/grafana/provisioning/dashboards" - "/etc/grafana/provisioning/notifiers" - name: download ceph grafana dashboards get_url: url: "https://raw.githubusercontent.com/ceph/ceph/{{ grafana_dashboard_version }}/monitoring/ceph-mixin/dashboards_out/{{ item }}" dest: "/etc/grafana/dashboards/ceph-dashboard/{{ item }}" with_items: "{{ grafana_dashboard_files }}" when: - not containerized_deployment | bool - not ansible_facts['os_family'] in ['RedHat', 'Suse'] - name: write grafana.ini openstack.config_template.config_template: src: grafana.ini.j2 dest: /etc/grafana/grafana.ini owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 config_type: ini config_overrides: "{{ grafana_conf_overrides }}" - name: write datasources provisioning config file template: src: datasources-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/datasources/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 - name: Write dashboards provisioning config file template: src: dashboards-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/dashboards/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 when: not containerized_deployment | bool - name: copy grafana SSL certificate file copy: src: "{{ grafana_crt }}" dest: "/etc/grafana/ceph-dashboard.crt" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_crt | length > 0 - dashboard_protocol == "https" - name: copy grafana SSL certificate key copy: src: "{{ grafana_key }}" dest: "/etc/grafana/ceph-dashboard.key" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0440 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_key | length > 0 - dashboard_protocol == "https" - name: generate a Self Signed OpenSSL certificate for dashboard shell: | test -f /etc/grafana/ceph-dashboard.key -a -f /etc/grafana/ceph-dashboard.crt || \ (openssl req -new -nodes -x509 -subj '/O=IT/CN=ceph-grafana' -days 3650 -keyout /etc/grafana/ceph-dashboard.key -out /etc/grafana/ceph-dashboard.crt -extensions v3_ca && \ chown {{ grafana_uid }}:{{ grafana_uid }} /etc/grafana/ceph-dashboard.key /etc/grafana/ceph-dashboard.crt) when: - dashboard_protocol == "https" - grafana_key | length == 0 or grafana_crt | length == 0 - name: enable and start grafana service: name: grafana-server state: restarted enabled: true - name: wait for grafana to start wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}'

mitch000001

8a5fb702f2a3df46834baf6019285463bbfcc4fb

4edaab5f4c5445cb1fafc5d8824c49717e9f96c8

> I mean, we could also change this like that within master and cherry-pick into stable-6.0 onwards. Works for me. Is that the direction to go? yes

guits

ceph/ceph-ansible

fix(ceph-grafana): make dashboard download work again

This fixes the dashboard download for pacific and later. Since ceph switched to Prometheus Monitoring Mixins the path to the generated dashboards has changed. It is still working for octopus but it's broken from pacific onwards. This change fixes the issue. Currently I only added the two latest releases to the check. Also, this is only related to non containerized deployments.

2022-06-10 15:27:51+00:00

2022-06-14 12:36:24+00:00

roles/ceph-grafana/tasks/configure_grafana.yml

--- - name: install ceph-grafana-dashboards package on RedHat or SUSE package: name: ceph-grafana-dashboards state: "{{ (upgrade_ceph_packages|bool) | ternary('latest','present') }}" register: result until: result is succeeded when: - not containerized_deployment | bool - ansible_facts['os_family'] in ['RedHat', 'Suse'] tags: package-install - name: make sure grafana is down service: name: grafana-server state: stopped - name: wait for grafana to be stopped wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}' state: stopped - name: make sure grafana configuration directories exist file: path: "{{ item }}" state: directory recurse: yes owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" with_items: - "/etc/grafana/dashboards/ceph-dashboard" - "/etc/grafana/provisioning/datasources" - "/etc/grafana/provisioning/dashboards" - "/etc/grafana/provisioning/notifiers" - name: download ceph grafana dashboards get_url: url: "https://raw.githubusercontent.com/ceph/ceph/{{ grafana_dashboard_version }}/monitoring/grafana/dashboards/{{ item }}" dest: "/etc/grafana/dashboards/ceph-dashboard/{{ item }}" with_items: "{{ grafana_dashboard_files }}" when: - not containerized_deployment | bool - not ansible_facts['os_family'] in ['RedHat', 'Suse'] - name: write grafana.ini openstack.config_template.config_template: src: grafana.ini.j2 dest: /etc/grafana/grafana.ini owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 config_type: ini config_overrides: "{{ grafana_conf_overrides }}" - name: write datasources provisioning config file template: src: datasources-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/datasources/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 - name: Write dashboards provisioning config file template: src: dashboards-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/dashboards/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 when: not containerized_deployment | bool - name: copy grafana SSL certificate file copy: src: "{{ grafana_crt }}" dest: "/etc/grafana/ceph-dashboard.crt" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_crt | length > 0 - dashboard_protocol == "https" - name: copy grafana SSL certificate key copy: src: "{{ grafana_key }}" dest: "/etc/grafana/ceph-dashboard.key" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0440 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_key | length > 0 - dashboard_protocol == "https" - name: generate a Self Signed OpenSSL certificate for dashboard shell: | test -f /etc/grafana/ceph-dashboard.key -a -f /etc/grafana/ceph-dashboard.crt || \ (openssl req -new -nodes -x509 -subj '/O=IT/CN=ceph-grafana' -days 3650 -keyout /etc/grafana/ceph-dashboard.key -out /etc/grafana/ceph-dashboard.crt -extensions v3_ca && \ chown {{ grafana_uid }}:{{ grafana_uid }} /etc/grafana/ceph-dashboard.key /etc/grafana/ceph-dashboard.crt) when: - dashboard_protocol == "https" - grafana_key | length == 0 or grafana_crt | length == 0 - name: enable and start grafana service: name: grafana-server state: restarted enabled: true - name: wait for grafana to start wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}'

--- - name: install ceph-grafana-dashboards package on RedHat or SUSE package: name: ceph-grafana-dashboards state: "{{ (upgrade_ceph_packages|bool) | ternary('latest','present') }}" register: result until: result is succeeded when: - not containerized_deployment | bool - ansible_facts['os_family'] in ['RedHat', 'Suse'] tags: package-install - name: make sure grafana is down service: name: grafana-server state: stopped - name: wait for grafana to be stopped wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}' state: stopped - name: make sure grafana configuration directories exist file: path: "{{ item }}" state: directory recurse: yes owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" with_items: - "/etc/grafana/dashboards/ceph-dashboard" - "/etc/grafana/provisioning/datasources" - "/etc/grafana/provisioning/dashboards" - "/etc/grafana/provisioning/notifiers" - name: download ceph grafana dashboards get_url: url: "https://raw.githubusercontent.com/ceph/ceph/{{ grafana_dashboard_version }}/monitoring/ceph-mixin/dashboards_out/{{ item }}" dest: "/etc/grafana/dashboards/ceph-dashboard/{{ item }}" with_items: "{{ grafana_dashboard_files }}" when: - not containerized_deployment | bool - not ansible_facts['os_family'] in ['RedHat', 'Suse'] - name: write grafana.ini openstack.config_template.config_template: src: grafana.ini.j2 dest: /etc/grafana/grafana.ini owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 config_type: ini config_overrides: "{{ grafana_conf_overrides }}" - name: write datasources provisioning config file template: src: datasources-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/datasources/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 - name: Write dashboards provisioning config file template: src: dashboards-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/dashboards/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 when: not containerized_deployment | bool - name: copy grafana SSL certificate file copy: src: "{{ grafana_crt }}" dest: "/etc/grafana/ceph-dashboard.crt" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_crt | length > 0 - dashboard_protocol == "https" - name: copy grafana SSL certificate key copy: src: "{{ grafana_key }}" dest: "/etc/grafana/ceph-dashboard.key" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0440 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_key | length > 0 - dashboard_protocol == "https" - name: generate a Self Signed OpenSSL certificate for dashboard shell: | test -f /etc/grafana/ceph-dashboard.key -a -f /etc/grafana/ceph-dashboard.crt || \ (openssl req -new -nodes -x509 -subj '/O=IT/CN=ceph-grafana' -days 3650 -keyout /etc/grafana/ceph-dashboard.key -out /etc/grafana/ceph-dashboard.crt -extensions v3_ca && \ chown {{ grafana_uid }}:{{ grafana_uid }} /etc/grafana/ceph-dashboard.key /etc/grafana/ceph-dashboard.crt) when: - dashboard_protocol == "https" - grafana_key | length == 0 or grafana_crt | length == 0 - name: enable and start grafana service: name: grafana-server state: restarted enabled: true - name: wait for grafana to start wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}'

mitch000001

8a5fb702f2a3df46834baf6019285463bbfcc4fb

4edaab5f4c5445cb1fafc5d8824c49717e9f96c8

> The change I proposed would be backwards compatible as the dashboards for octopus and earlier life in the old location still. the branch 'main' isn't intended to be used for deploying stables releases.

guits

ceph/ceph-ansible

fix(ceph-grafana): make dashboard download work again

This fixes the dashboard download for pacific and later. Since ceph switched to Prometheus Monitoring Mixins the path to the generated dashboards has changed. It is still working for octopus but it's broken from pacific onwards. This change fixes the issue. Currently I only added the two latest releases to the check. Also, this is only related to non containerized deployments.

2022-06-10 15:27:51+00:00

2022-06-14 12:36:24+00:00

roles/ceph-grafana/tasks/configure_grafana.yml

--- - name: install ceph-grafana-dashboards package on RedHat or SUSE package: name: ceph-grafana-dashboards state: "{{ (upgrade_ceph_packages|bool) | ternary('latest','present') }}" register: result until: result is succeeded when: - not containerized_deployment | bool - ansible_facts['os_family'] in ['RedHat', 'Suse'] tags: package-install - name: make sure grafana is down service: name: grafana-server state: stopped - name: wait for grafana to be stopped wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}' state: stopped - name: make sure grafana configuration directories exist file: path: "{{ item }}" state: directory recurse: yes owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" with_items: - "/etc/grafana/dashboards/ceph-dashboard" - "/etc/grafana/provisioning/datasources" - "/etc/grafana/provisioning/dashboards" - "/etc/grafana/provisioning/notifiers" - name: download ceph grafana dashboards get_url: url: "https://raw.githubusercontent.com/ceph/ceph/{{ grafana_dashboard_version }}/monitoring/grafana/dashboards/{{ item }}" dest: "/etc/grafana/dashboards/ceph-dashboard/{{ item }}" with_items: "{{ grafana_dashboard_files }}" when: - not containerized_deployment | bool - not ansible_facts['os_family'] in ['RedHat', 'Suse'] - name: write grafana.ini openstack.config_template.config_template: src: grafana.ini.j2 dest: /etc/grafana/grafana.ini owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 config_type: ini config_overrides: "{{ grafana_conf_overrides }}" - name: write datasources provisioning config file template: src: datasources-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/datasources/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 - name: Write dashboards provisioning config file template: src: dashboards-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/dashboards/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 when: not containerized_deployment | bool - name: copy grafana SSL certificate file copy: src: "{{ grafana_crt }}" dest: "/etc/grafana/ceph-dashboard.crt" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_crt | length > 0 - dashboard_protocol == "https" - name: copy grafana SSL certificate key copy: src: "{{ grafana_key }}" dest: "/etc/grafana/ceph-dashboard.key" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0440 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_key | length > 0 - dashboard_protocol == "https" - name: generate a Self Signed OpenSSL certificate for dashboard shell: | test -f /etc/grafana/ceph-dashboard.key -a -f /etc/grafana/ceph-dashboard.crt || \ (openssl req -new -nodes -x509 -subj '/O=IT/CN=ceph-grafana' -days 3650 -keyout /etc/grafana/ceph-dashboard.key -out /etc/grafana/ceph-dashboard.crt -extensions v3_ca && \ chown {{ grafana_uid }}:{{ grafana_uid }} /etc/grafana/ceph-dashboard.key /etc/grafana/ceph-dashboard.crt) when: - dashboard_protocol == "https" - grafana_key | length == 0 or grafana_crt | length == 0 - name: enable and start grafana service: name: grafana-server state: restarted enabled: true - name: wait for grafana to start wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}'

--- - name: install ceph-grafana-dashboards package on RedHat or SUSE package: name: ceph-grafana-dashboards state: "{{ (upgrade_ceph_packages|bool) | ternary('latest','present') }}" register: result until: result is succeeded when: - not containerized_deployment | bool - ansible_facts['os_family'] in ['RedHat', 'Suse'] tags: package-install - name: make sure grafana is down service: name: grafana-server state: stopped - name: wait for grafana to be stopped wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}' state: stopped - name: make sure grafana configuration directories exist file: path: "{{ item }}" state: directory recurse: yes owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" with_items: - "/etc/grafana/dashboards/ceph-dashboard" - "/etc/grafana/provisioning/datasources" - "/etc/grafana/provisioning/dashboards" - "/etc/grafana/provisioning/notifiers" - name: download ceph grafana dashboards get_url: url: "https://raw.githubusercontent.com/ceph/ceph/{{ grafana_dashboard_version }}/monitoring/ceph-mixin/dashboards_out/{{ item }}" dest: "/etc/grafana/dashboards/ceph-dashboard/{{ item }}" with_items: "{{ grafana_dashboard_files }}" when: - not containerized_deployment | bool - not ansible_facts['os_family'] in ['RedHat', 'Suse'] - name: write grafana.ini openstack.config_template.config_template: src: grafana.ini.j2 dest: /etc/grafana/grafana.ini owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 config_type: ini config_overrides: "{{ grafana_conf_overrides }}" - name: write datasources provisioning config file template: src: datasources-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/datasources/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 - name: Write dashboards provisioning config file template: src: dashboards-ceph-dashboard.yml.j2 dest: /etc/grafana/provisioning/dashboards/ceph-dashboard.yml owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 when: not containerized_deployment | bool - name: copy grafana SSL certificate file copy: src: "{{ grafana_crt }}" dest: "/etc/grafana/ceph-dashboard.crt" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0640 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_crt | length > 0 - dashboard_protocol == "https" - name: copy grafana SSL certificate key copy: src: "{{ grafana_key }}" dest: "/etc/grafana/ceph-dashboard.key" owner: "{{ grafana_uid }}" group: "{{ grafana_uid }}" mode: 0440 remote_src: "{{ dashboard_tls_external | bool }}" when: - grafana_key | length > 0 - dashboard_protocol == "https" - name: generate a Self Signed OpenSSL certificate for dashboard shell: | test -f /etc/grafana/ceph-dashboard.key -a -f /etc/grafana/ceph-dashboard.crt || \ (openssl req -new -nodes -x509 -subj '/O=IT/CN=ceph-grafana' -days 3650 -keyout /etc/grafana/ceph-dashboard.key -out /etc/grafana/ceph-dashboard.crt -extensions v3_ca && \ chown {{ grafana_uid }}:{{ grafana_uid }} /etc/grafana/ceph-dashboard.key /etc/grafana/ceph-dashboard.crt) when: - dashboard_protocol == "https" - grafana_key | length == 0 or grafana_crt | length == 0 - name: enable and start grafana service: name: grafana-server state: restarted enabled: true - name: wait for grafana to start wait_for: host: '{{ grafana_server_addr if ip_version == "ipv4" else grafana_server_addr[1:-1] }}' port: '{{ grafana_port }}'

mitch000001

8a5fb702f2a3df46834baf6019285463bbfcc4fb

4edaab5f4c5445cb1fafc5d8824c49717e9f96c8

Changed.

mitch000001

ceph/ceph-ansible

[skip ci] rbd-mirror: major refactor

- Use config-key store to add cluster peer. - Support multiple pools mirroring. Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>

2022-05-12 15:23:32+00:00

2022-07-29 15:33:26+00:00

roles/ceph-rbd-mirror/tasks/configure_mirroring.yml

--- - name: enable mirroring on the pool command: "{{ container_exec_cmd | default('') }} rbd --cluster {{ cluster }} --keyring /etc/ceph/{{ cluster }}.client.rbd-mirror.{{ ansible_facts['hostname'] }}.keyring --name client.rbd-mirror.{{ ansible_facts['hostname'] }} mirror pool enable {{ ceph_rbd_mirror_pool }} {{ ceph_rbd_mirror_mode }}" register: result changed_when: false retries: 90 delay: 1 until: result is succeeded - name: list mirroring peer command: "{{ container_exec_cmd | default('') }} rbd --cluster {{ cluster }} --keyring /etc/ceph/{{ cluster }}.client.rbd-mirror.{{ ansible_facts['hostname'] }}.keyring --name client.rbd-mirror.{{ ansible_facts['hostname'] }} mirror pool info {{ ceph_rbd_mirror_pool }}" changed_when: false register: mirror_peer - name: add a mirroring peer command: "{{ container_exec_cmd | default('') }} rbd --cluster {{ cluster }} --keyring /etc/ceph/{{ cluster }}.client.rbd-mirror.{{ ansible_facts['hostname'] }}.keyring --name client.rbd-mirror.{{ ansible_facts['hostname'] }} mirror pool peer add {{ ceph_rbd_mirror_pool }} {{ ceph_rbd_mirror_remote_user }}@{{ ceph_rbd_mirror_remote_cluster }}" changed_when: false when: ceph_rbd_mirror_remote_user not in mirror_peer.stdout

--- - name: cephx tasks when: - cephx | bool block: - name: get client.bootstrap-rbd-mirror from ceph monitor ceph_key: name: client.bootstrap-rbd-mirror cluster: "{{ cluster }}" output_format: plain state: info environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" register: _bootstrap_rbd_mirror_key delegate_to: "{{ groups.get(mon_group_name)[0] }}" run_once: true no_log: "{{ no_log_on_ceph_key_tasks }}" - name: copy ceph key(s) copy: dest: "/var/lib/ceph/bootstrap-rbd-mirror/{{ cluster }}.keyring" content: "{{ _bootstrap_rbd_mirror_key.stdout + '\n' }}" owner: "{{ ceph_uid if containerized_deployment | bool else 'ceph' }}" group: "{{ ceph_uid if containerized_deployment | bool else 'ceph' }}" mode: "{{ ceph_keyring_permissions }}" no_log: "{{ no_log_on_ceph_key_tasks }}" - name: create rbd-mirror keyrings ceph_key: name: "{{ item.name }}" cluster: "{{ cluster }}" user: client.admin user_key: "/etc/ceph/{{ cluster }}.client.admin.keyring" caps: mon: "profile rbd-mirror" osd: "profile rbd" dest: "{{ item.dest }}" secret: "{{ item.secret | default(omit) }}" import_key: true owner: "{{ ceph_uid if containerized_deployment | bool else 'ceph' }}" group: "{{ ceph_uid if containerized_deployment | bool else 'ceph' }}" mode: "{{ ceph_keyring_permissions }}" no_log: "{{ no_log_on_ceph_key_tasks }}" environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" delegate_to: "{{ groups.get(mon_group_name)[0] }}" loop: - { name: "client.rbd-mirror.{{ ansible_facts['hostname'] }}", dest: "/etc/ceph/{{ cluster }}.client.rbd-mirror.{{ ansible_facts['hostname'] }}.keyring" } - { name: "{{ ceph_rbd_mirror_local_user }}", dest: "/etc/ceph/{{ cluster }}.{{ ceph_rbd_mirror_local_user }}.keyring", secret: "{{ ceph_rbd_mirror_local_user_secret | default('') }}" } - name: get "client.rbd-mirror.{{ ansible_facts['hostname'] }}" from ceph monitor ceph_key: name: "client.rbd-mirror.{{ ansible_facts['hostname'] }}" cluster: "{{ cluster }}" output_format: plain state: info environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" register: _rbd_mirror_key delegate_to: "{{ groups.get(mon_group_name)[0] }}" run_once: true no_log: "{{ no_log_on_ceph_key_tasks }}" - name: copy ceph key copy: dest: "/etc/ceph/{{ cluster }}.client.rbd-mirror.{{ ansible_facts['hostname'] }}.keyring" content: "{{ _rbd_mirror_key.stdout + '\n' }}" owner: "{{ ceph_uid if containerized_deployment | bool else 'ceph' }}" group: "{{ ceph_uid if containerized_deployment | bool else 'ceph' }}" mode: "{{ ceph_keyring_permissions }}" no_log: false - name: start and add the rbd-mirror service instance service: name: "ceph-rbd-mirror@rbd-mirror.{{ ansible_facts['hostname'] }}" state: started enabled: yes masked: no changed_when: false when: - not containerized_deployment | bool - ceph_rbd_mirror_remote_user is defined - name: set_fact ceph_rbd_mirror_pools set_fact: ceph_rbd_mirror_pools: - name: "{{ ceph_rbd_mirror_pool }}" when: ceph_rbd_mirror_pools is undefined - name: create pool if it doesn't exist ceph_pool: name: "{{ item.name }}" cluster: "{{ cluster }}" pg_num: "{{ item.pg_num | default(omit) }}" pgp_num: "{{ item.pgp_num | default(omit) }}" size: "{{ item.size | default(omit) }}" min_size: "{{ item.min_size | default(omit) }}" pool_type: "{{ item.type | default('replicated') }}" rule_name: "{{ item.rule_name | default(omit) }}" erasure_profile: "{{ item.erasure_profile | default(omit) }}" pg_autoscale_mode: "{{ item.pg_autoscale_mode | default(omit) }}" target_size_ratio: "{{ item.target_size_ratio | default(omit) }}" application: "{{ item.application | default('rbd') }}" delegate_to: "{{ groups[mon_group_name][0] }}" loop: "{{ ceph_rbd_mirror_pools }}" environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" - name: enable mirroring on the pool command: "{{ rbd_cmd }} --cluster {{ cluster }} mirror pool enable {{ item.name }} {{ ceph_rbd_mirror_mode }}" register: result changed_when: false retries: 60 delay: 1 until: result is succeeded loop: "{{ ceph_rbd_mirror_pools }}" delegate_to: "{{ groups[mon_group_name][0] }}" - name: add mirroring peer when: ceph_rbd_mirror_remote_user is defined block: - name: list mirroring peer command: "{{ rbd_cmd }} --cluster {{ cluster }} mirror pool info {{ item.name }}" changed_when: false register: mirror_peer loop: "{{ ceph_rbd_mirror_pools }}" delegate_to: "{{ groups[mon_group_name][0] }}" - name: create a temporary file tempfile: path: /etc/ceph state: file suffix: _ceph-ansible register: tmp_file delegate_to: "{{ groups[mon_group_name][0] }}" - name: write secret to temporary file copy: dest: "{{ tmp_file.path }}" content: "{{ ceph_rbd_mirror_remote_key }}" delegate_to: "{{ groups[mon_group_name][0] }}" - name: add a mirroring peer command: "{{ rbd_cmd }} --cluster {{ cluster }} mirror pool peer add {{ item.item.name }} {{ ceph_rbd_mirror_remote_user }}@{{ ceph_rbd_mirror_remote_cluster }} --remote-mon-host {{ ceph_rbd_mirror_remote_mon_hosts }} --remote-key-file {{ tmp_file.path }}" changed_when: false delegate_to: "{{ groups[mon_group_name][0] }}" loop: "{{ mirror_peer.results }}" when: ceph_rbd_mirror_remote_user not in item.stdout - name: rm temporary file file: path: "{{ tmp_file.path }}" state: absent delegate_to: "{{ groups[mon_group_name][0] }}"

guits

3a8daafbe8c9023c6dcd8034adfcc98893e5c303

b74ff6e22c0d1b95e71384e4d7e2fb2ad556ac39

add a `ceph_pool` task? (create the pool if it doesn't exist already?)

guits

ceph/ceph-ansible

[skip ci] rbd-mirror: major refactor

- Use config-key store to add cluster peer. - Support multiple pools mirroring. Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>

2022-05-12 15:23:32+00:00

2022-07-29 15:33:26+00:00

roles/ceph-rbd-mirror/tasks/configure_mirroring.yml

--- - name: enable mirroring on the pool command: "{{ container_exec_cmd | default('') }} rbd --cluster {{ cluster }} --keyring /etc/ceph/{{ cluster }}.client.rbd-mirror.{{ ansible_facts['hostname'] }}.keyring --name client.rbd-mirror.{{ ansible_facts['hostname'] }} mirror pool enable {{ ceph_rbd_mirror_pool }} {{ ceph_rbd_mirror_mode }}" register: result changed_when: false retries: 90 delay: 1 until: result is succeeded - name: list mirroring peer command: "{{ container_exec_cmd | default('') }} rbd --cluster {{ cluster }} --keyring /etc/ceph/{{ cluster }}.client.rbd-mirror.{{ ansible_facts['hostname'] }}.keyring --name client.rbd-mirror.{{ ansible_facts['hostname'] }} mirror pool info {{ ceph_rbd_mirror_pool }}" changed_when: false register: mirror_peer - name: add a mirroring peer command: "{{ container_exec_cmd | default('') }} rbd --cluster {{ cluster }} --keyring /etc/ceph/{{ cluster }}.client.rbd-mirror.{{ ansible_facts['hostname'] }}.keyring --name client.rbd-mirror.{{ ansible_facts['hostname'] }} mirror pool peer add {{ ceph_rbd_mirror_pool }} {{ ceph_rbd_mirror_remote_user }}@{{ ceph_rbd_mirror_remote_cluster }}" changed_when: false when: ceph_rbd_mirror_remote_user not in mirror_peer.stdout

--- - name: cephx tasks when: - cephx | bool block: - name: get client.bootstrap-rbd-mirror from ceph monitor ceph_key: name: client.bootstrap-rbd-mirror cluster: "{{ cluster }}" output_format: plain state: info environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" register: _bootstrap_rbd_mirror_key delegate_to: "{{ groups.get(mon_group_name)[0] }}" run_once: true no_log: "{{ no_log_on_ceph_key_tasks }}" - name: copy ceph key(s) copy: dest: "/var/lib/ceph/bootstrap-rbd-mirror/{{ cluster }}.keyring" content: "{{ _bootstrap_rbd_mirror_key.stdout + '\n' }}" owner: "{{ ceph_uid if containerized_deployment | bool else 'ceph' }}" group: "{{ ceph_uid if containerized_deployment | bool else 'ceph' }}" mode: "{{ ceph_keyring_permissions }}" no_log: "{{ no_log_on_ceph_key_tasks }}" - name: create rbd-mirror keyrings ceph_key: name: "{{ item.name }}" cluster: "{{ cluster }}" user: client.admin user_key: "/etc/ceph/{{ cluster }}.client.admin.keyring" caps: mon: "profile rbd-mirror" osd: "profile rbd" dest: "{{ item.dest }}" secret: "{{ item.secret | default(omit) }}" import_key: true owner: "{{ ceph_uid if containerized_deployment | bool else 'ceph' }}" group: "{{ ceph_uid if containerized_deployment | bool else 'ceph' }}" mode: "{{ ceph_keyring_permissions }}" no_log: "{{ no_log_on_ceph_key_tasks }}" environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" delegate_to: "{{ groups.get(mon_group_name)[0] }}" loop: - { name: "client.rbd-mirror.{{ ansible_facts['hostname'] }}", dest: "/etc/ceph/{{ cluster }}.client.rbd-mirror.{{ ansible_facts['hostname'] }}.keyring" } - { name: "{{ ceph_rbd_mirror_local_user }}", dest: "/etc/ceph/{{ cluster }}.{{ ceph_rbd_mirror_local_user }}.keyring", secret: "{{ ceph_rbd_mirror_local_user_secret | default('') }}" } - name: get "client.rbd-mirror.{{ ansible_facts['hostname'] }}" from ceph monitor ceph_key: name: "client.rbd-mirror.{{ ansible_facts['hostname'] }}" cluster: "{{ cluster }}" output_format: plain state: info environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" register: _rbd_mirror_key delegate_to: "{{ groups.get(mon_group_name)[0] }}" run_once: true no_log: "{{ no_log_on_ceph_key_tasks }}" - name: copy ceph key copy: dest: "/etc/ceph/{{ cluster }}.client.rbd-mirror.{{ ansible_facts['hostname'] }}.keyring" content: "{{ _rbd_mirror_key.stdout + '\n' }}" owner: "{{ ceph_uid if containerized_deployment | bool else 'ceph' }}" group: "{{ ceph_uid if containerized_deployment | bool else 'ceph' }}" mode: "{{ ceph_keyring_permissions }}" no_log: false - name: start and add the rbd-mirror service instance service: name: "ceph-rbd-mirror@rbd-mirror.{{ ansible_facts['hostname'] }}" state: started enabled: yes masked: no changed_when: false when: - not containerized_deployment | bool - ceph_rbd_mirror_remote_user is defined - name: set_fact ceph_rbd_mirror_pools set_fact: ceph_rbd_mirror_pools: - name: "{{ ceph_rbd_mirror_pool }}" when: ceph_rbd_mirror_pools is undefined - name: create pool if it doesn't exist ceph_pool: name: "{{ item.name }}" cluster: "{{ cluster }}" pg_num: "{{ item.pg_num | default(omit) }}" pgp_num: "{{ item.pgp_num | default(omit) }}" size: "{{ item.size | default(omit) }}" min_size: "{{ item.min_size | default(omit) }}" pool_type: "{{ item.type | default('replicated') }}" rule_name: "{{ item.rule_name | default(omit) }}" erasure_profile: "{{ item.erasure_profile | default(omit) }}" pg_autoscale_mode: "{{ item.pg_autoscale_mode | default(omit) }}" target_size_ratio: "{{ item.target_size_ratio | default(omit) }}" application: "{{ item.application | default('rbd') }}" delegate_to: "{{ groups[mon_group_name][0] }}" loop: "{{ ceph_rbd_mirror_pools }}" environment: CEPH_CONTAINER_IMAGE: "{{ ceph_docker_registry + '/' + ceph_docker_image + ':' + ceph_docker_image_tag if containerized_deployment | bool else None }}" CEPH_CONTAINER_BINARY: "{{ container_binary }}" - name: enable mirroring on the pool command: "{{ rbd_cmd }} --cluster {{ cluster }} mirror pool enable {{ item.name }} {{ ceph_rbd_mirror_mode }}" register: result changed_when: false retries: 60 delay: 1 until: result is succeeded loop: "{{ ceph_rbd_mirror_pools }}" delegate_to: "{{ groups[mon_group_name][0] }}" - name: add mirroring peer when: ceph_rbd_mirror_remote_user is defined block: - name: list mirroring peer command: "{{ rbd_cmd }} --cluster {{ cluster }} mirror pool info {{ item.name }}" changed_when: false register: mirror_peer loop: "{{ ceph_rbd_mirror_pools }}" delegate_to: "{{ groups[mon_group_name][0] }}" - name: create a temporary file tempfile: path: /etc/ceph state: file suffix: _ceph-ansible register: tmp_file delegate_to: "{{ groups[mon_group_name][0] }}" - name: write secret to temporary file copy: dest: "{{ tmp_file.path }}" content: "{{ ceph_rbd_mirror_remote_key }}" delegate_to: "{{ groups[mon_group_name][0] }}" - name: add a mirroring peer command: "{{ rbd_cmd }} --cluster {{ cluster }} mirror pool peer add {{ item.item.name }} {{ ceph_rbd_mirror_remote_user }}@{{ ceph_rbd_mirror_remote_cluster }} --remote-mon-host {{ ceph_rbd_mirror_remote_mon_hosts }} --remote-key-file {{ tmp_file.path }}" changed_when: false delegate_to: "{{ groups[mon_group_name][0] }}" loop: "{{ mirror_peer.results }}" when: ceph_rbd_mirror_remote_user not in item.stdout - name: rm temporary file file: path: "{{ tmp_file.path }}" state: absent delegate_to: "{{ groups[mon_group_name][0] }}"

guits

3a8daafbe8c9023c6dcd8034adfcc98893e5c303

b74ff6e22c0d1b95e71384e4d7e2fb2ad556ac39

legacy from testing? ```suggestion retries: 60 ```

guits

yogeshojha/rengine

Fix report generation when `Ignore Informational Vulnerabilities` checked

When **Ignore Informational Vulnerabilities** is checked there are still info vulns datas. I've reworked the queries that display vulnerabilities to prevent info vulns to display in the : - **Quick summary** Info blue box - **Reconnaissance Findings** - **Vulnerabilities Discovered** Info blue box I've also fixed the **Vulnerabilities Discovered** listing by doing a correct loop through regrouped values because values withe the same path but not the same severity does not display well Tested and working on current master branch

2023-12-05 01:25:41+00:00

2023-12-08 05:48:36+00:00

web/startScan/views.py

import markdown from celery import group from weasyprint import HTML from datetime import datetime from django.contrib import messages from django.db.models import Count from django.http import HttpResponse, HttpResponseRedirect, JsonResponse from django.shortcuts import get_object_or_404, render from django.template.loader import get_template from django.urls import reverse from django.utils import timezone from django_celery_beat.models import (ClockedSchedule, IntervalSchedule, PeriodicTask) from rolepermissions.decorators import has_permission_decorator from reNgine.celery import app from reNgine.common_func import * from reNgine.definitions import ABORTED_TASK, SUCCESS_TASK from reNgine.tasks import create_scan_activity, initiate_scan, run_command from scanEngine.models import EngineType from startScan.models import * from targetApp.models import * def scan_history(request, slug): host = ScanHistory.objects.filter(domain__project__slug=slug).order_by('-start_scan_date') context = {'scan_history_active': 'active', "scan_history": host} return render(request, 'startScan/history.html', context) def subscan_history(request, slug): subscans = SubScan.objects.filter(scan_history__domain__project__slug=slug).order_by('-start_scan_date') context = {'scan_history_active': 'active', "subscans": subscans} return render(request, 'startScan/subscan_history.html', context) def detail_scan(request, id, slug): ctx = {} # Get scan objects scan = get_object_or_404(ScanHistory, id=id) domain_id = scan.domain.id scan_engines = EngineType.objects.order_by('engine_name').all() recent_scans = ScanHistory.objects.filter(domain__id=domain_id) last_scans = ( ScanHistory.objects .filter(domain__id=domain_id) .filter(tasks__overlap=['subdomain_discovery']) .filter(id__lte=id) .filter(scan_status=2) ) # Get all kind of objects associated with our ScanHistory object emails = Email.objects.filter(emails__in=[scan]) employees = Employee.objects.filter(employees__in=[scan]) subdomains = Subdomain.objects.filter(scan_history=scan) endpoints = EndPoint.objects.filter(scan_history=scan) vulns = Vulnerability.objects.filter(scan_history=scan) vulns_tags = VulnerabilityTags.objects.filter(vuln_tags__in=vulns) ip_addresses = IpAddress.objects.filter(ip_addresses__in=subdomains) geo_isos = CountryISO.objects.filter(ipaddress__in=ip_addresses) scan_activity = ScanActivity.objects.filter(scan_of__id=id).order_by('time') cves = CveId.objects.filter(cve_ids__in=vulns) cwes = CweId.objects.filter(cwe_ids__in=vulns) # HTTP statuses http_statuses = ( subdomains .exclude(http_status=0) .values('http_status') .annotate(Count('http_status')) ) # CVEs / CWes common_cves = ( cves .annotate(nused=Count('cve_ids')) .order_by('-nused') .values('name', 'nused') [:10] ) common_cwes = ( cwes .annotate(nused=Count('cwe_ids')) .order_by('-nused') .values('name', 'nused') [:10] ) # Tags common_tags = ( vulns_tags .annotate(nused=Count('vuln_tags')) .order_by('-nused') .values('name', 'nused') [:7] ) # Countries asset_countries = ( geo_isos .annotate(count=Count('iso')) .order_by('-count') ) # Subdomains subdomain_count = ( subdomains .values('name') .distinct() .count() ) alive_count = ( subdomains .values('name') .distinct() .filter(http_status__exact=200) .count() ) important_count = ( subdomains .values('name') .distinct() .filter(is_important=True) .count() ) # Endpoints endpoint_count = ( endpoints .values('http_url') .distinct() .count() ) endpoint_alive_count = ( endpoints .filter(http_status__exact=200) # TODO: use is_alive() func as it's more precise .values('http_url') .distinct() .count() ) # Vulnerabilities common_vulns = ( vulns .exclude(severity=0) .values('name', 'severity') .annotate(count=Count('name')) .order_by('-count') [:10] ) info_count = vulns.filter(severity=0).count() low_count = vulns.filter(severity=1).count() medium_count = vulns.filter(severity=2).count() high_count = vulns.filter(severity=3).count() critical_count = vulns.filter(severity=4).count() unknown_count = vulns.filter(severity=-1).count() total_count = vulns.count() total_count_ignore_info = vulns.exclude(severity=0).count() # Emails exposed_count = emails.exclude(password__isnull=True).count() # Build render context ctx = { 'scan_history_id': id, 'history': scan, 'scan_activity': scan_activity, 'subdomain_count': subdomain_count, 'alive_count': alive_count, 'important_count': important_count, 'endpoint_count': endpoint_count, 'endpoint_alive_count': endpoint_alive_count, 'info_count': info_count, 'low_count': low_count, 'medium_count': medium_count, 'high_count': high_count, 'critical_count': critical_count, 'unknown_count': unknown_count, 'total_vulnerability_count': total_count, 'total_vul_ignore_info_count': total_count_ignore_info, 'vulnerability_list': vulns.order_by('-severity').all(), 'scan_history_active': 'active', 'scan_engines': scan_engines, 'exposed_count': exposed_count, 'email_count': emails.count(), 'employees_count': employees.count(), 'most_recent_scans': recent_scans.order_by('-start_scan_date')[:1], 'http_status_breakdown': http_statuses, 'most_common_cve': common_cves, 'most_common_cwe': common_cwes, 'most_common_tags': common_tags, 'most_common_vulnerability': common_vulns, 'asset_countries': asset_countries, } # Find number of matched GF patterns if scan.used_gf_patterns: count_gf = {} for gf in scan.used_gf_patterns.split(','): count_gf[gf] = ( endpoints .filter(matched_gf_patterns__icontains=gf) .count() ) ctx['matched_gf_count'] = count_gf # Find last scan for this domain if last_scans.count() > 1: last_scan = last_scans.order_by('-start_scan_date')[1] ctx['last_scan'] = last_scan return render(request, 'startScan/detail_scan.html', ctx) def all_subdomains(request, slug): subdomains = Subdomain.objects.filter(target_domain__project__slug=slug) scan_engines = EngineType.objects.order_by('engine_name').all() alive_subdomains = subdomains.filter(http_status__exact=200) # TODO: replace this with is_alive() function important_subdomains = ( subdomains .filter(is_important=True) .values('name') .distinct() .count() ) context = { 'scan_history_id': id, 'scan_history_active': 'active', 'scan_engines': scan_engines, 'subdomain_count': subdomains.values('name').distinct().count(), 'alive_count': alive_subdomains.values('name').distinct().count(), 'important_count': important_subdomains } return render(request, 'startScan/subdomains.html', context) def detail_vuln_scan(request, slug, id=None): if id: history = get_object_or_404(ScanHistory, id=id) history.filter(domain__project__slug=slug) context = {'scan_history_id': id, 'history': history} else: context = {'vuln_scan_active': 'true'} return render(request, 'startScan/vulnerabilities.html', context) def all_endpoints(request, slug): context = { 'scan_history_active': 'active' } return render(request, 'startScan/endpoints.html', context) def start_scan_ui(request, slug, domain_id): domain = get_object_or_404(Domain, id=domain_id) if request.method == "POST": # Get imported and out-of-scope subdomains subdomains_in = request.POST['importSubdomainTextArea'].split() subdomains_in = [s.rstrip() for s in subdomains_in if s] subdomains_out = request.POST['outOfScopeSubdomainTextarea'].split() subdomains_out = [s.rstrip() for s in subdomains_out if s] paths = request.POST['filterPath'].split() filterPath = [s.rstrip() for s in paths if s] if len(filterPath) > 0: filterPath = filterPath[0] else: filterPath = '' # Get engine type engine_id = request.POST['scan_mode'] # Create ScanHistory object scan_history_id = create_scan_object(domain_id, engine_id) scan = ScanHistory.objects.get(pk=scan_history_id) # Start the celery task kwargs = { 'scan_history_id': scan.id, 'domain_id': domain.id, 'engine_id': engine_id, 'scan_type': LIVE_SCAN, 'results_dir': '/usr/src/scan_results', 'imported_subdomains': subdomains_in, 'out_of_scope_subdomains': subdomains_out, 'url_filter': filterPath } initiate_scan.apply_async(kwargs=kwargs) scan.save() # Send start notif messages.add_message( request, messages.INFO, f'Scan Started for {domain.name}') return HttpResponseRedirect(reverse('scan_history', kwargs={'slug': slug})) # GET request engine = EngineType.objects.order_by('engine_name') custom_engine_count = ( EngineType.objects .filter(default_engine=False) .count() ) context = { 'scan_history_active': 'active', 'domain': domain, 'engines': engine, 'custom_engine_count': custom_engine_count} return render(request, 'startScan/start_scan_ui.html', context) @has_permission_decorator(PERM_INITATE_SCANS_SUBSCANS, redirect_url=FOUR_OH_FOUR_URL) def start_multiple_scan(request, slug): # domain = get_object_or_404(Domain, id=host_id) if request.method == "POST": if request.POST.get('scan_mode', 0): # if scan mode is available, then start the scan # get engine type engine_id = request.POST['scan_mode'] list_of_domains = request.POST['list_of_domain_id'] grouped_scans = [] for domain_id in list_of_domains.split(","): # Start the celery task scan_history_id = create_scan_object(domain_id, engine_id) # domain = get_object_or_404(Domain, id=domain_id) kwargs = { 'scan_history_id': scan_history_id, 'domain_id': domain_id, 'engine_id': engine_id, 'scan_type': LIVE_SCAN, 'results_dir': '/usr/src/scan_results', # TODO: Add this to multiple scan view # 'imported_subdomains': subdomains_in, # 'out_of_scope_subdomains': subdomains_out } _scan_task = initiate_scan.si(**kwargs) grouped_scans.append(_scan_task) celery_group = group(grouped_scans) celery_group.apply_async() # Send start notif messages.add_message( request, messages.INFO, 'Scan Started for multiple targets') return HttpResponseRedirect(reverse('scan_history', kwargs={'slug': slug})) else: # this else condition will have post request from the scan page # containing all the targets id list_of_domain_name = [] list_of_domain_id = [] for key, value in request.POST.items(): if key != "list_target_table_length" and key != "csrfmiddlewaretoken": domain = get_object_or_404(Domain, id=value) list_of_domain_name.append(domain.name) list_of_domain_id.append(value) domain_ids = ",".join(list_of_domain_id) # GET request engines = EngineType.objects custom_engine_count = ( engines .filter(default_engine=False) .count() ) context = { 'scan_history_active': 'active', 'engines': engines, 'domain_list': list_of_domain_name, 'domain_ids': domain_ids, 'custom_engine_count': custom_engine_count } return render(request, 'startScan/start_multiple_scan_ui.html', context) def export_subdomains(request, scan_id): subdomain_list = Subdomain.objects.filter(scan_history__id=scan_id) scan = ScanHistory.objects.get(id=scan_id) response_body = "" for domain in subdomain_list: response_body += response_body + domain.name + "\n" scan_start_date_str = str(scan.start_scan_date.date()) domain_name = scan.domain.name response = HttpResponse(response_body, content_type='text/plain') response['Content-Disposition'] = ( f'attachment; filename="subdomains_{domain_name}_{scan_start_date_str}.txt"' ) return response def export_endpoints(request, scan_id): endpoint_list = EndPoint.objects.filter(scan_history__id=scan_id) scan = ScanHistory.objects.get(id=scan_id) response_body = "" for endpoint in endpoint_list: response_body += endpoint.http_url + "\n" scan_start_date_str = str(scan.start_scan_date.date()) domain_name = scan.domain.name response = HttpResponse(response_body, content_type='text/plain') response['Content-Disposition'] = ( f'attachment; filename="endpoints_{domain_name}_{scan_start_date_str}.txt"' ) return response def export_urls(request, scan_id): urls_list = Subdomain.objects.filter(scan_history__id=scan_id) scan = ScanHistory.objects.get(id=scan_id) response_body = "" for url in urls_list: if url.http_url: response_body += response_body + url.http_url + "\n" scan_start_date_str = str(scan.start_scan_date.date()) domain_name = scan.domain.name response = HttpResponse(response_body, content_type='text/plain') response['Content-Disposition'] = ( f'attachment; filename="urls_{domain_name}_{scan_start_date_str}.txt"' ) return response @has_permission_decorator(PERM_MODIFY_SCAN_RESULTS, redirect_url=FOUR_OH_FOUR_URL) def delete_scan(request, id): obj = get_object_or_404(ScanHistory, id=id) if request.method == "POST": delete_dir = obj.results_dir run_command('rm -rf ' + delete_dir) obj.delete() messageData = {'status': 'true'} messages.add_message( request, messages.INFO, 'Scan history successfully deleted!' ) else: messageData = {'status': 'false'} messages.add_message( request, messages.INFO, 'Oops! something went wrong!' ) return JsonResponse(messageData) @has_permission_decorator(PERM_INITATE_SCANS_SUBSCANS, redirect_url=FOUR_OH_FOUR_URL) def stop_scan(request, id): if request.method == "POST": scan = get_object_or_404(ScanHistory, id=id) scan.scan_status = ABORTED_TASK scan.save() try: for task_id in scan.celery_ids: app.control.revoke(task_id, terminate=True, signal='SIGKILL') tasks = ( ScanActivity.objects .filter(scan_of=scan) .filter(status=RUNNING_TASK) .order_by('-pk') ) for task in tasks: task.status = ABORTED_TASK task.time = timezone.now() task.save() create_scan_activity(scan.id, "Scan aborted", SUCCESS_TASK) response = {'status': True} messages.add_message( request, messages.INFO, 'Scan successfully stopped!' ) except Exception as e: logger.error(e) response = {'status': False} messages.add_message( request, messages.ERROR, f'Scan failed to stop ! Error: {str(e)}' ) return JsonResponse(response) return scan_history(request) @has_permission_decorator(PERM_INITATE_SCANS_SUBSCANS, redirect_url=FOUR_OH_FOUR_URL) def schedule_scan(request, host_id, slug): domain = Domain.objects.get(id=host_id) if request.method == "POST": scheduled_mode = request.POST['scheduled_mode'] engine_type = int(request.POST['scan_mode']) # Get imported and out-of-scope subdomains subdomains_in = request.POST['importSubdomainTextArea'].split() subdomains_in = [s.rstrip() for s in subdomains_in if s] subdomains_out = request.POST['outOfScopeSubdomainTextarea'].split() subdomains_out = [s.rstrip() for s in subdomains_out if s] # Get engine type engine = get_object_or_404(EngineType, id=engine_type) timestr = str(datetime.strftime(timezone.now(), '%Y_%m_%d_%H_%M_%S')) task_name = f'{engine.engine_name} for {domain.name}: {timestr}' if scheduled_mode == 'periodic': frequency_value = int(request.POST['frequency']) frequency_type = request.POST['frequency_type'] if frequency_type == 'minutes': period = IntervalSchedule.MINUTES elif frequency_type == 'hours': period = IntervalSchedule.HOURS elif frequency_type == 'days': period = IntervalSchedule.DAYS elif frequency_type == 'weeks': period = IntervalSchedule.DAYS frequency_value *= 7 elif frequency_type == 'months': period = IntervalSchedule.DAYS frequency_value *= 30 schedule, _ = IntervalSchedule.objects.get_or_create( every=frequency_value, period=period) kwargs = { 'domain_id': host_id, 'engine_id': engine.id, 'scan_history_id': 1, 'scan_type': SCHEDULED_SCAN, 'imported_subdomains': subdomains_in, 'out_of_scope_subdomains': subdomains_out } PeriodicTask.objects.create(interval=schedule, name=task_name, task='reNgine.tasks.initiate_scan', kwargs=json.dumps(kwargs)) elif scheduled_mode == 'clocked': schedule_time = request.POST['scheduled_time'] clock, _ = ClockedSchedule.objects.get_or_create( clocked_time=schedule_time) kwargs = { 'scan_history_id': 0, 'domain_id': host_id, 'engine_id': engine.id, 'scan_type': SCHEDULED_SCAN, 'imported_subdomains': subdomains_in, 'out_of_scope_subdomains': subdomains_out } PeriodicTask.objects.create(clocked=clock, one_off=True, name=task_name, task='reNgine.tasks.initiate_scan', kwargs=json.dumps(kwargs)) messages.add_message( request, messages.INFO, f'Scan Scheduled for {domain.name}' ) return HttpResponseRedirect(reverse('scheduled_scan_view', kwargs={'slug': slug})) # GET request engines = EngineType.objects custom_engine_count = ( engines .filter(default_engine=False) .count() ) context = { 'scan_history_active': 'active', 'domain': domain, 'engines': engines, 'custom_engine_count': custom_engine_count} return render(request, 'startScan/schedule_scan_ui.html', context) def scheduled_scan_view(request, slug): scheduled_tasks = ( PeriodicTask.objects .all() .exclude(name='celery.backend_cleanup') ) context = { 'scheduled_scan_active': 'active', 'scheduled_tasks': scheduled_tasks, } return render(request, 'startScan/schedule_scan_list.html', context) @has_permission_decorator(PERM_MODIFY_SCAN_RESULTS, redirect_url=FOUR_OH_FOUR_URL) def delete_scheduled_task(request, id): task_object = get_object_or_404(PeriodicTask, id=id) if request.method == "POST": task_object.delete() messageData = {'status': 'true'} messages.add_message( request, messages.INFO, 'Scheduled Scan successfully deleted!') else: messageData = {'status': 'false'} messages.add_message( request, messages.INFO, 'Oops! something went wrong!') return JsonResponse(messageData) @has_permission_decorator(PERM_MODIFY_SCAN_RESULTS, redirect_url=FOUR_OH_FOUR_URL) def change_scheduled_task_status(request, id): if request.method == 'POST': task = PeriodicTask.objects.get(id=id) task.enabled = not task.enabled task.save() return HttpResponse('') def change_vuln_status(request, id): if request.method == 'POST': vuln = Vulnerability.objects.get(id=id) vuln.open_status = not vuln.open_status vuln.save() return HttpResponse('') def create_scan_object(host_id, engine_id): ''' create task with pending status so that celery task will execute when threads are free ''' # get current time current_scan_time = timezone.now() # fetch engine and domain object engine = EngineType.objects.get(pk=engine_id) domain = Domain.objects.get(pk=host_id) scan = ScanHistory() scan.scan_status = INITIATED_TASK scan.domain = domain scan.scan_type = engine scan.start_scan_date = current_scan_time scan.save() # save last scan date for domain model domain.start_scan_date = current_scan_time domain.save() return scan.id @has_permission_decorator(PERM_MODIFY_SYSTEM_CONFIGURATIONS, redirect_url=FOUR_OH_FOUR_URL) def delete_all_scan_results(request): if request.method == 'POST': ScanHistory.objects.all().delete() messageData = {'status': 'true'} messages.add_message( request, messages.INFO, 'All Scan History successfully deleted!') return JsonResponse(messageData) @has_permission_decorator(PERM_MODIFY_SYSTEM_CONFIGURATIONS, redirect_url=FOUR_OH_FOUR_URL) def delete_all_screenshots(request): if request.method == 'POST': run_command('rm -rf /usr/src/scan_results/*') messageData = {'status': 'true'} messages.add_message( request, messages.INFO, 'Screenshots successfully deleted!') return JsonResponse(messageData) def visualise(request, id): scan = ScanHistory.objects.get(id=id) context = { 'scan_id': id, 'scan_history': scan, } return render(request, 'startScan/visualise.html', context) @has_permission_decorator(PERM_INITATE_SCANS_SUBSCANS, redirect_url=FOUR_OH_FOUR_URL) def start_organization_scan(request, id, slug): organization = get_object_or_404(Organization, id=id) if request.method == "POST": engine_id = request.POST['scan_mode'] # Start Celery task for each organization's domains for domain in organization.get_domains(): scan_history_id = create_scan_object(domain.id, engine_id) scan = ScanHistory.objects.get(pk=scan_history_id) kwargs = { 'scan_history_id': scan.id, 'domain_id': domain.id, 'engine_id': engine_id, 'scan_type': LIVE_SCAN, 'results_dir': '/usr/src/scan_results', # TODO: Add this to multiple scan view # 'imported_subdomains': subdomains_in, # 'out_of_scope_subdomains': subdomains_out } initiate_scan.apply_async(kwargs=kwargs) scan.save() # Send start notif ndomains = len(organization.get_domains()) messages.add_message( request, messages.INFO, f'Scan Started for {ndomains} domains in organization {organization.name}') return HttpResponseRedirect(reverse('scan_history', kwargs={'slug': slug})) # GET request engine = EngineType.objects.order_by('engine_name') custom_engine_count = EngineType.objects.filter(default_engine=False).count() domain_list = organization.get_domains() context = { 'organization_data_active': 'true', 'list_organization_li': 'active', 'organization': organization, 'engines': engine, 'domain_list': domain_list, 'custom_engine_count': custom_engine_count} return render(request, 'organization/start_scan.html', context) @has_permission_decorator(PERM_INITATE_SCANS_SUBSCANS, redirect_url=FOUR_OH_FOUR_URL) def schedule_organization_scan(request, slug, id): organization =Organization.objects.get(id=id) if request.method == "POST": engine_type = int(request.POST['scan_mode']) engine = get_object_or_404(EngineType, id=engine_type) scheduled_mode = request.POST['scheduled_mode'] for domain in organization.get_domains(): timestr = str(datetime.strftime(timezone.now(), '%Y_%m_%d_%H_%M_%S')) task_name = f'{engine.engine_name} for {domain.name}: {timestr}' # Period task if scheduled_mode == 'periodic': frequency_value = int(request.POST['frequency']) frequency_type = request.POST['frequency_type'] if frequency_type == 'minutes': period = IntervalSchedule.MINUTES elif frequency_type == 'hours': period = IntervalSchedule.HOURS elif frequency_type == 'days': period = IntervalSchedule.DAYS elif frequency_type == 'weeks': period = IntervalSchedule.DAYS frequency_value *= 7 elif frequency_type == 'months': period = IntervalSchedule.DAYS frequency_value *= 30 schedule, _ = IntervalSchedule.objects.get_or_create( every=frequency_value, period=period ) _kwargs = json.dumps({ 'domain_id': domain.id, 'engine_id': engine.id, 'scan_history_id': 0, 'scan_type': SCHEDULED_SCAN, 'imported_subdomains': None }) PeriodicTask.objects.create( interval=schedule, name=task_name, task='reNgine.tasks.initiate_scan', kwargs=_kwargs ) # Clocked task elif scheduled_mode == 'clocked': schedule_time = request.POST['scheduled_time'] clock, _ = ClockedSchedule.objects.get_or_create( clocked_time=schedule_time ) _kwargs = json.dumps({ 'domain_id': domain.id, 'engine_id': engine.id, 'scan_history_id': 0, 'scan_type': LIVE_SCAN, 'imported_subdomains': None }) PeriodicTask.objects.create(clocked=clock, one_off=True, name=task_name, task='reNgine.tasks.initiate_scan', kwargs=_kwargs ) # Send start notif ndomains = len(organization.get_domains()) messages.add_message( request, messages.INFO, f'Scan started for {ndomains} domains in organization {organization.name}' ) return HttpResponseRedirect(reverse('scheduled_scan_view', kwargs={'slug': slug, 'id': id})) # GET request engine = EngineType.objects custom_engine_count = EngineType.objects.filter(default_engine=False).count() context = { 'scan_history_active': 'active', 'organization': organization, 'domain_list': organization.get_domains(), 'engines': engine, 'custom_engine_count': custom_engine_count } return render(request, 'organization/schedule_scan_ui.html', context) @has_permission_decorator(PERM_MODIFY_SCAN_RESULTS, redirect_url=FOUR_OH_FOUR_URL) def delete_scans(request, slug): if request.method == "POST": for key, value in request.POST.items(): if key == 'scan_history_table_length' or key == 'csrfmiddlewaretoken': continue scan = get_object_or_404(ScanHistory, id=value) delete_dir = scan.results_dir run_command('rm -rf ' + delete_dir) scan.delete() messages.add_message( request, messages.INFO, 'All Scans deleted!') return HttpResponseRedirect(reverse('scan_history', kwargs={'slug': slug})) @has_permission_decorator(PERM_MODIFY_SCAN_REPORT, redirect_url=FOUR_OH_FOUR_URL) def customize_report(request, id): scan = ScanHistory.objects.get(id=id) context = { 'scan_id': id, 'scan_history': scan, } return render(request, 'startScan/customize_report.html', context) @has_permission_decorator(PERM_MODIFY_SCAN_REPORT, redirect_url=FOUR_OH_FOUR_URL) def create_report(request, id): primary_color = '#FFB74D' secondary_color = '#212121' # get report type report_type = request.GET['report_type'] if 'report_type' in request.GET else 'full' is_ignore_info_vuln = True if 'ignore_info_vuln' in request.GET else False if report_type == 'recon': show_recon = True show_vuln = False report_name = 'Reconnaissance Report' elif report_type == 'vulnerability': show_recon = False show_vuln = True report_name = 'Vulnerability Report' else: # default show_recon = True show_vuln = True report_name = 'Full Scan Report' scan = ScanHistory.objects.get(id=id) vulns = ( Vulnerability.objects .filter(scan_history=scan) .order_by('-severity') ) if not is_ignore_info_vuln else ( Vulnerability.objects .filter(scan_history=scan) .exclude(severity=0) .order_by('-severity') ) unique_vulns = ( Vulnerability.objects .filter(scan_history=scan) .values("name", "severity") .annotate(count=Count('name')) .order_by('-severity', '-count') ) if not is_ignore_info_vuln else ( Vulnerability.objects .filter(scan_history=scan) .exclude(severity=0) .values("name", "severity") .annotate(count=Count('name')) .order_by('-severity', '-count') ) subdomains = ( Subdomain.objects .filter(scan_history=scan) .order_by('-content_length') ) subdomain_alive_count = ( Subdomain.objects .filter(scan_history__id=id) .values('name') .distinct() .filter(http_status__exact=200) .count() ) interesting_subdomains = get_interesting_subdomains(scan_history=id) ip_addresses = ( IpAddress.objects .filter(ip_addresses__in=subdomains) .distinct() ) data = { 'scan_object': scan, 'unique_vulnerabilities': unique_vulns, 'all_vulnerabilities': vulns, 'subdomain_alive_count': subdomain_alive_count, 'interesting_subdomains': interesting_subdomains, 'subdomains': subdomains, 'ip_addresses': ip_addresses, 'show_recon': show_recon, 'show_vuln': show_vuln, 'report_name': report_name, } # Get report related config vuln_report_query = VulnerabilityReportSetting.objects.all() if vuln_report_query.exists(): report = vuln_report_query[0] data['company_name'] = report.company_name data['company_address'] = report.company_address data['company_email'] = report.company_email data['company_website'] = report.company_website data['show_rengine_banner'] = report.show_rengine_banner data['show_footer'] = report.show_footer data['footer_text'] = report.footer_text data['show_executive_summary'] = report.show_executive_summary # Replace executive_summary_description with template syntax description = report.executive_summary_description description = description.replace('{scan_date}', scan.start_scan_date.strftime('%d %B, %Y')) description = description.replace('{company_name}', report.company_name) description = description.replace('{target_name}', scan.domain.name) description = description.replace('{subdomain_count}', str(subdomains.count())) description = description.replace('{vulnerability_count}', str(vulns.count())) description = description.replace('{critical_count}', str(vulns.filter(severity=4).count())) description = description.replace('{high_count}', str(vulns.filter(severity=3).count())) description = description.replace('{medium_count}', str(vulns.filter(severity=2).count())) description = description.replace('{low_count}', str(vulns.filter(severity=1).count())) description = description.replace('{info_count}', str(vulns.filter(severity=0).count())) description = description.replace('{unknown_count}', str(vulns.filter(severity=-1).count())) if scan.domain.description: description = description.replace('{target_description}', scan.domain.description) # Convert to Markdown data['executive_summary_description'] = markdown.markdown(description) primary_color = report.primary_color secondary_color = report.secondary_color data['primary_color'] = primary_color data['secondary_color'] = secondary_color template = get_template('report/template.html') html = template.render(data) pdf = HTML(string=html).write_pdf() if 'download' in request.GET: response = HttpResponse(pdf, content_type='application/octet-stream') else: response = HttpResponse(pdf, content_type='application/pdf') return response

import markdown from celery import group from weasyprint import HTML from datetime import datetime from django.contrib import messages from django.db.models import Count from django.http import HttpResponse, HttpResponseRedirect, JsonResponse from django.shortcuts import get_object_or_404, render from django.template.loader import get_template from django.urls import reverse from django.utils import timezone from django_celery_beat.models import (ClockedSchedule, IntervalSchedule, PeriodicTask) from rolepermissions.decorators import has_permission_decorator from reNgine.celery import app from reNgine.common_func import * from reNgine.definitions import ABORTED_TASK, SUCCESS_TASK from reNgine.tasks import create_scan_activity, initiate_scan, run_command from scanEngine.models import EngineType from startScan.models import * from targetApp.models import * def scan_history(request, slug): host = ScanHistory.objects.filter(domain__project__slug=slug).order_by('-start_scan_date') context = {'scan_history_active': 'active', "scan_history": host} return render(request, 'startScan/history.html', context) def subscan_history(request, slug): subscans = SubScan.objects.filter(scan_history__domain__project__slug=slug).order_by('-start_scan_date') context = {'scan_history_active': 'active', "subscans": subscans} return render(request, 'startScan/subscan_history.html', context) def detail_scan(request, id, slug): ctx = {} # Get scan objects scan = get_object_or_404(ScanHistory, id=id) domain_id = scan.domain.id scan_engines = EngineType.objects.order_by('engine_name').all() recent_scans = ScanHistory.objects.filter(domain__id=domain_id) last_scans = ( ScanHistory.objects .filter(domain__id=domain_id) .filter(tasks__overlap=['subdomain_discovery']) .filter(id__lte=id) .filter(scan_status=2) ) # Get all kind of objects associated with our ScanHistory object emails = Email.objects.filter(emails__in=[scan]) employees = Employee.objects.filter(employees__in=[scan]) subdomains = Subdomain.objects.filter(scan_history=scan) endpoints = EndPoint.objects.filter(scan_history=scan) vulns = Vulnerability.objects.filter(scan_history=scan) vulns_tags = VulnerabilityTags.objects.filter(vuln_tags__in=vulns) ip_addresses = IpAddress.objects.filter(ip_addresses__in=subdomains) geo_isos = CountryISO.objects.filter(ipaddress__in=ip_addresses) scan_activity = ScanActivity.objects.filter(scan_of__id=id).order_by('time') cves = CveId.objects.filter(cve_ids__in=vulns) cwes = CweId.objects.filter(cwe_ids__in=vulns) # HTTP statuses http_statuses = ( subdomains .exclude(http_status=0) .values('http_status') .annotate(Count('http_status')) ) # CVEs / CWes common_cves = ( cves .annotate(nused=Count('cve_ids')) .order_by('-nused') .values('name', 'nused') [:10] ) common_cwes = ( cwes .annotate(nused=Count('cwe_ids')) .order_by('-nused') .values('name', 'nused') [:10] ) # Tags common_tags = ( vulns_tags .annotate(nused=Count('vuln_tags')) .order_by('-nused') .values('name', 'nused') [:7] ) # Countries asset_countries = ( geo_isos .annotate(count=Count('iso')) .order_by('-count') ) # Subdomains subdomain_count = ( subdomains .values('name') .distinct() .count() ) alive_count = ( subdomains .values('name') .distinct() .filter(http_status__exact=200) .count() ) important_count = ( subdomains .values('name') .distinct() .filter(is_important=True) .count() ) # Endpoints endpoint_count = ( endpoints .values('http_url') .distinct() .count() ) endpoint_alive_count = ( endpoints .filter(http_status__exact=200) # TODO: use is_alive() func as it's more precise .values('http_url') .distinct() .count() ) # Vulnerabilities common_vulns = ( vulns .exclude(severity=0) .values('name', 'severity') .annotate(count=Count('name')) .order_by('-count') [:10] ) info_count = vulns.filter(severity=0).count() low_count = vulns.filter(severity=1).count() medium_count = vulns.filter(severity=2).count() high_count = vulns.filter(severity=3).count() critical_count = vulns.filter(severity=4).count() unknown_count = vulns.filter(severity=-1).count() total_count = vulns.count() total_count_ignore_info = vulns.exclude(severity=0).count() # Emails exposed_count = emails.exclude(password__isnull=True).count() # Build render context ctx = { 'scan_history_id': id, 'history': scan, 'scan_activity': scan_activity, 'subdomain_count': subdomain_count, 'alive_count': alive_count, 'important_count': important_count, 'endpoint_count': endpoint_count, 'endpoint_alive_count': endpoint_alive_count, 'info_count': info_count, 'low_count': low_count, 'medium_count': medium_count, 'high_count': high_count, 'critical_count': critical_count, 'unknown_count': unknown_count, 'total_vulnerability_count': total_count, 'total_vul_ignore_info_count': total_count_ignore_info, 'vulnerability_list': vulns.order_by('-severity').all(), 'scan_history_active': 'active', 'scan_engines': scan_engines, 'exposed_count': exposed_count, 'email_count': emails.count(), 'employees_count': employees.count(), 'most_recent_scans': recent_scans.order_by('-start_scan_date')[:1], 'http_status_breakdown': http_statuses, 'most_common_cve': common_cves, 'most_common_cwe': common_cwes, 'most_common_tags': common_tags, 'most_common_vulnerability': common_vulns, 'asset_countries': asset_countries, } # Find number of matched GF patterns if scan.used_gf_patterns: count_gf = {} for gf in scan.used_gf_patterns.split(','): count_gf[gf] = ( endpoints .filter(matched_gf_patterns__icontains=gf) .count() ) ctx['matched_gf_count'] = count_gf # Find last scan for this domain if last_scans.count() > 1: last_scan = last_scans.order_by('-start_scan_date')[1] ctx['last_scan'] = last_scan return render(request, 'startScan/detail_scan.html', ctx) def all_subdomains(request, slug): subdomains = Subdomain.objects.filter(target_domain__project__slug=slug) scan_engines = EngineType.objects.order_by('engine_name').all() alive_subdomains = subdomains.filter(http_status__exact=200) # TODO: replace this with is_alive() function important_subdomains = ( subdomains .filter(is_important=True) .values('name') .distinct() .count() ) context = { 'scan_history_id': id, 'scan_history_active': 'active', 'scan_engines': scan_engines, 'subdomain_count': subdomains.values('name').distinct().count(), 'alive_count': alive_subdomains.values('name').distinct().count(), 'important_count': important_subdomains } return render(request, 'startScan/subdomains.html', context) def detail_vuln_scan(request, slug, id=None): if id: history = get_object_or_404(ScanHistory, id=id) history.filter(domain__project__slug=slug) context = {'scan_history_id': id, 'history': history} else: context = {'vuln_scan_active': 'true'} return render(request, 'startScan/vulnerabilities.html', context) def all_endpoints(request, slug): context = { 'scan_history_active': 'active' } return render(request, 'startScan/endpoints.html', context) def start_scan_ui(request, slug, domain_id): domain = get_object_or_404(Domain, id=domain_id) if request.method == "POST": # Get imported and out-of-scope subdomains subdomains_in = request.POST['importSubdomainTextArea'].split() subdomains_in = [s.rstrip() for s in subdomains_in if s] subdomains_out = request.POST['outOfScopeSubdomainTextarea'].split() subdomains_out = [s.rstrip() for s in subdomains_out if s] paths = request.POST['filterPath'].split() filterPath = [s.rstrip() for s in paths if s] if len(filterPath) > 0: filterPath = filterPath[0] else: filterPath = '' # Get engine type engine_id = request.POST['scan_mode'] # Create ScanHistory object scan_history_id = create_scan_object(domain_id, engine_id) scan = ScanHistory.objects.get(pk=scan_history_id) # Start the celery task kwargs = { 'scan_history_id': scan.id, 'domain_id': domain.id, 'engine_id': engine_id, 'scan_type': LIVE_SCAN, 'results_dir': '/usr/src/scan_results', 'imported_subdomains': subdomains_in, 'out_of_scope_subdomains': subdomains_out, 'url_filter': filterPath } initiate_scan.apply_async(kwargs=kwargs) scan.save() # Send start notif messages.add_message( request, messages.INFO, f'Scan Started for {domain.name}') return HttpResponseRedirect(reverse('scan_history', kwargs={'slug': slug})) # GET request engine = EngineType.objects.order_by('engine_name') custom_engine_count = ( EngineType.objects .filter(default_engine=False) .count() ) context = { 'scan_history_active': 'active', 'domain': domain, 'engines': engine, 'custom_engine_count': custom_engine_count} return render(request, 'startScan/start_scan_ui.html', context) @has_permission_decorator(PERM_INITATE_SCANS_SUBSCANS, redirect_url=FOUR_OH_FOUR_URL) def start_multiple_scan(request, slug): # domain = get_object_or_404(Domain, id=host_id) if request.method == "POST": if request.POST.get('scan_mode', 0): # if scan mode is available, then start the scan # get engine type engine_id = request.POST['scan_mode'] list_of_domains = request.POST['list_of_domain_id'] grouped_scans = [] for domain_id in list_of_domains.split(","): # Start the celery task scan_history_id = create_scan_object(domain_id, engine_id) # domain = get_object_or_404(Domain, id=domain_id) kwargs = { 'scan_history_id': scan_history_id, 'domain_id': domain_id, 'engine_id': engine_id, 'scan_type': LIVE_SCAN, 'results_dir': '/usr/src/scan_results', # TODO: Add this to multiple scan view # 'imported_subdomains': subdomains_in, # 'out_of_scope_subdomains': subdomains_out } _scan_task = initiate_scan.si(**kwargs) grouped_scans.append(_scan_task) celery_group = group(grouped_scans) celery_group.apply_async() # Send start notif messages.add_message( request, messages.INFO, 'Scan Started for multiple targets') return HttpResponseRedirect(reverse('scan_history', kwargs={'slug': slug})) else: # this else condition will have post request from the scan page # containing all the targets id list_of_domain_name = [] list_of_domain_id = [] for key, value in request.POST.items(): if key != "list_target_table_length" and key != "csrfmiddlewaretoken": domain = get_object_or_404(Domain, id=value) list_of_domain_name.append(domain.name) list_of_domain_id.append(value) domain_ids = ",".join(list_of_domain_id) # GET request engines = EngineType.objects custom_engine_count = ( engines .filter(default_engine=False) .count() ) context = { 'scan_history_active': 'active', 'engines': engines, 'domain_list': list_of_domain_name, 'domain_ids': domain_ids, 'custom_engine_count': custom_engine_count } return render(request, 'startScan/start_multiple_scan_ui.html', context) def export_subdomains(request, scan_id): subdomain_list = Subdomain.objects.filter(scan_history__id=scan_id) scan = ScanHistory.objects.get(id=scan_id) response_body = "" for domain in subdomain_list: response_body += response_body + domain.name + "\n" scan_start_date_str = str(scan.start_scan_date.date()) domain_name = scan.domain.name response = HttpResponse(response_body, content_type='text/plain') response['Content-Disposition'] = ( f'attachment; filename="subdomains_{domain_name}_{scan_start_date_str}.txt"' ) return response def export_endpoints(request, scan_id): endpoint_list = EndPoint.objects.filter(scan_history__id=scan_id) scan = ScanHistory.objects.get(id=scan_id) response_body = "" for endpoint in endpoint_list: response_body += endpoint.http_url + "\n" scan_start_date_str = str(scan.start_scan_date.date()) domain_name = scan.domain.name response = HttpResponse(response_body, content_type='text/plain') response['Content-Disposition'] = ( f'attachment; filename="endpoints_{domain_name}_{scan_start_date_str}.txt"' ) return response def export_urls(request, scan_id): urls_list = Subdomain.objects.filter(scan_history__id=scan_id) scan = ScanHistory.objects.get(id=scan_id) response_body = "" for url in urls_list: if url.http_url: response_body += response_body + url.http_url + "\n" scan_start_date_str = str(scan.start_scan_date.date()) domain_name = scan.domain.name response = HttpResponse(response_body, content_type='text/plain') response['Content-Disposition'] = ( f'attachment; filename="urls_{domain_name}_{scan_start_date_str}.txt"' ) return response @has_permission_decorator(PERM_MODIFY_SCAN_RESULTS, redirect_url=FOUR_OH_FOUR_URL) def delete_scan(request, id): obj = get_object_or_404(ScanHistory, id=id) if request.method == "POST": delete_dir = obj.results_dir run_command('rm -rf ' + delete_dir) obj.delete() messageData = {'status': 'true'} messages.add_message( request, messages.INFO, 'Scan history successfully deleted!' ) else: messageData = {'status': 'false'} messages.add_message( request, messages.INFO, 'Oops! something went wrong!' ) return JsonResponse(messageData) @has_permission_decorator(PERM_INITATE_SCANS_SUBSCANS, redirect_url=FOUR_OH_FOUR_URL) def stop_scan(request, id): if request.method == "POST": scan = get_object_or_404(ScanHistory, id=id) scan.scan_status = ABORTED_TASK scan.save() try: for task_id in scan.celery_ids: app.control.revoke(task_id, terminate=True, signal='SIGKILL') tasks = ( ScanActivity.objects .filter(scan_of=scan) .filter(status=RUNNING_TASK) .order_by('-pk') ) for task in tasks: task.status = ABORTED_TASK task.time = timezone.now() task.save() create_scan_activity(scan.id, "Scan aborted", SUCCESS_TASK) response = {'status': True} messages.add_message( request, messages.INFO, 'Scan successfully stopped!' ) except Exception as e: logger.error(e) response = {'status': False} messages.add_message( request, messages.ERROR, f'Scan failed to stop ! Error: {str(e)}' ) return JsonResponse(response) return scan_history(request) @has_permission_decorator(PERM_INITATE_SCANS_SUBSCANS, redirect_url=FOUR_OH_FOUR_URL) def schedule_scan(request, host_id, slug): domain = Domain.objects.get(id=host_id) if request.method == "POST": scheduled_mode = request.POST['scheduled_mode'] engine_type = int(request.POST['scan_mode']) # Get imported and out-of-scope subdomains subdomains_in = request.POST['importSubdomainTextArea'].split() subdomains_in = [s.rstrip() for s in subdomains_in if s] subdomains_out = request.POST['outOfScopeSubdomainTextarea'].split() subdomains_out = [s.rstrip() for s in subdomains_out if s] # Get engine type engine = get_object_or_404(EngineType, id=engine_type) timestr = str(datetime.strftime(timezone.now(), '%Y_%m_%d_%H_%M_%S')) task_name = f'{engine.engine_name} for {domain.name}: {timestr}' if scheduled_mode == 'periodic': frequency_value = int(request.POST['frequency']) frequency_type = request.POST['frequency_type'] if frequency_type == 'minutes': period = IntervalSchedule.MINUTES elif frequency_type == 'hours': period = IntervalSchedule.HOURS elif frequency_type == 'days': period = IntervalSchedule.DAYS elif frequency_type == 'weeks': period = IntervalSchedule.DAYS frequency_value *= 7 elif frequency_type == 'months': period = IntervalSchedule.DAYS frequency_value *= 30 schedule, _ = IntervalSchedule.objects.get_or_create( every=frequency_value, period=period) kwargs = { 'domain_id': host_id, 'engine_id': engine.id, 'scan_history_id': 1, 'scan_type': SCHEDULED_SCAN, 'imported_subdomains': subdomains_in, 'out_of_scope_subdomains': subdomains_out } PeriodicTask.objects.create(interval=schedule, name=task_name, task='reNgine.tasks.initiate_scan', kwargs=json.dumps(kwargs)) elif scheduled_mode == 'clocked': schedule_time = request.POST['scheduled_time'] clock, _ = ClockedSchedule.objects.get_or_create( clocked_time=schedule_time) kwargs = { 'scan_history_id': 0, 'domain_id': host_id, 'engine_id': engine.id, 'scan_type': SCHEDULED_SCAN, 'imported_subdomains': subdomains_in, 'out_of_scope_subdomains': subdomains_out } PeriodicTask.objects.create(clocked=clock, one_off=True, name=task_name, task='reNgine.tasks.initiate_scan', kwargs=json.dumps(kwargs)) messages.add_message( request, messages.INFO, f'Scan Scheduled for {domain.name}' ) return HttpResponseRedirect(reverse('scheduled_scan_view', kwargs={'slug': slug})) # GET request engines = EngineType.objects custom_engine_count = ( engines .filter(default_engine=False) .count() ) context = { 'scan_history_active': 'active', 'domain': domain, 'engines': engines, 'custom_engine_count': custom_engine_count} return render(request, 'startScan/schedule_scan_ui.html', context) def scheduled_scan_view(request, slug): scheduled_tasks = ( PeriodicTask.objects .all() .exclude(name='celery.backend_cleanup') ) context = { 'scheduled_scan_active': 'active', 'scheduled_tasks': scheduled_tasks, } return render(request, 'startScan/schedule_scan_list.html', context) @has_permission_decorator(PERM_MODIFY_SCAN_RESULTS, redirect_url=FOUR_OH_FOUR_URL) def delete_scheduled_task(request, id): task_object = get_object_or_404(PeriodicTask, id=id) if request.method == "POST": task_object.delete() messageData = {'status': 'true'} messages.add_message( request, messages.INFO, 'Scheduled Scan successfully deleted!') else: messageData = {'status': 'false'} messages.add_message( request, messages.INFO, 'Oops! something went wrong!') return JsonResponse(messageData) @has_permission_decorator(PERM_MODIFY_SCAN_RESULTS, redirect_url=FOUR_OH_FOUR_URL) def change_scheduled_task_status(request, id): if request.method == 'POST': task = PeriodicTask.objects.get(id=id) task.enabled = not task.enabled task.save() return HttpResponse('') def change_vuln_status(request, id): if request.method == 'POST': vuln = Vulnerability.objects.get(id=id) vuln.open_status = not vuln.open_status vuln.save() return HttpResponse('') def create_scan_object(host_id, engine_id): ''' create task with pending status so that celery task will execute when threads are free ''' # get current time current_scan_time = timezone.now() # fetch engine and domain object engine = EngineType.objects.get(pk=engine_id) domain = Domain.objects.get(pk=host_id) scan = ScanHistory() scan.scan_status = INITIATED_TASK scan.domain = domain scan.scan_type = engine scan.start_scan_date = current_scan_time scan.save() # save last scan date for domain model domain.start_scan_date = current_scan_time domain.save() return scan.id @has_permission_decorator(PERM_MODIFY_SYSTEM_CONFIGURATIONS, redirect_url=FOUR_OH_FOUR_URL) def delete_all_scan_results(request): if request.method == 'POST': ScanHistory.objects.all().delete() messageData = {'status': 'true'} messages.add_message( request, messages.INFO, 'All Scan History successfully deleted!') return JsonResponse(messageData) @has_permission_decorator(PERM_MODIFY_SYSTEM_CONFIGURATIONS, redirect_url=FOUR_OH_FOUR_URL) def delete_all_screenshots(request): if request.method == 'POST': run_command('rm -rf /usr/src/scan_results/*') messageData = {'status': 'true'} messages.add_message( request, messages.INFO, 'Screenshots successfully deleted!') return JsonResponse(messageData) def visualise(request, id): scan = ScanHistory.objects.get(id=id) context = { 'scan_id': id, 'scan_history': scan, } return render(request, 'startScan/visualise.html', context) @has_permission_decorator(PERM_INITATE_SCANS_SUBSCANS, redirect_url=FOUR_OH_FOUR_URL) def start_organization_scan(request, id, slug): organization = get_object_or_404(Organization, id=id) if request.method == "POST": engine_id = request.POST['scan_mode'] # Start Celery task for each organization's domains for domain in organization.get_domains(): scan_history_id = create_scan_object(domain.id, engine_id) scan = ScanHistory.objects.get(pk=scan_history_id) kwargs = { 'scan_history_id': scan.id, 'domain_id': domain.id, 'engine_id': engine_id, 'scan_type': LIVE_SCAN, 'results_dir': '/usr/src/scan_results', # TODO: Add this to multiple scan view # 'imported_subdomains': subdomains_in, # 'out_of_scope_subdomains': subdomains_out } initiate_scan.apply_async(kwargs=kwargs) scan.save() # Send start notif ndomains = len(organization.get_domains()) messages.add_message( request, messages.INFO, f'Scan Started for {ndomains} domains in organization {organization.name}') return HttpResponseRedirect(reverse('scan_history', kwargs={'slug': slug})) # GET request engine = EngineType.objects.order_by('engine_name') custom_engine_count = EngineType.objects.filter(default_engine=False).count() domain_list = organization.get_domains() context = { 'organization_data_active': 'true', 'list_organization_li': 'active', 'organization': organization, 'engines': engine, 'domain_list': domain_list, 'custom_engine_count': custom_engine_count} return render(request, 'organization/start_scan.html', context) @has_permission_decorator(PERM_INITATE_SCANS_SUBSCANS, redirect_url=FOUR_OH_FOUR_URL) def schedule_organization_scan(request, slug, id): organization =Organization.objects.get(id=id) if request.method == "POST": engine_type = int(request.POST['scan_mode']) engine = get_object_or_404(EngineType, id=engine_type) scheduled_mode = request.POST['scheduled_mode'] for domain in organization.get_domains(): timestr = str(datetime.strftime(timezone.now(), '%Y_%m_%d_%H_%M_%S')) task_name = f'{engine.engine_name} for {domain.name}: {timestr}' # Period task if scheduled_mode == 'periodic': frequency_value = int(request.POST['frequency']) frequency_type = request.POST['frequency_type'] if frequency_type == 'minutes': period = IntervalSchedule.MINUTES elif frequency_type == 'hours': period = IntervalSchedule.HOURS elif frequency_type == 'days': period = IntervalSchedule.DAYS elif frequency_type == 'weeks': period = IntervalSchedule.DAYS frequency_value *= 7 elif frequency_type == 'months': period = IntervalSchedule.DAYS frequency_value *= 30 schedule, _ = IntervalSchedule.objects.get_or_create( every=frequency_value, period=period ) _kwargs = json.dumps({ 'domain_id': domain.id, 'engine_id': engine.id, 'scan_history_id': 0, 'scan_type': SCHEDULED_SCAN, 'imported_subdomains': None }) PeriodicTask.objects.create( interval=schedule, name=task_name, task='reNgine.tasks.initiate_scan', kwargs=_kwargs ) # Clocked task elif scheduled_mode == 'clocked': schedule_time = request.POST['scheduled_time'] clock, _ = ClockedSchedule.objects.get_or_create( clocked_time=schedule_time ) _kwargs = json.dumps({ 'domain_id': domain.id, 'engine_id': engine.id, 'scan_history_id': 0, 'scan_type': LIVE_SCAN, 'imported_subdomains': None }) PeriodicTask.objects.create(clocked=clock, one_off=True, name=task_name, task='reNgine.tasks.initiate_scan', kwargs=_kwargs ) # Send start notif ndomains = len(organization.get_domains()) messages.add_message( request, messages.INFO, f'Scan started for {ndomains} domains in organization {organization.name}' ) return HttpResponseRedirect(reverse('scheduled_scan_view', kwargs={'slug': slug, 'id': id})) # GET request engine = EngineType.objects custom_engine_count = EngineType.objects.filter(default_engine=False).count() context = { 'scan_history_active': 'active', 'organization': organization, 'domain_list': organization.get_domains(), 'engines': engine, 'custom_engine_count': custom_engine_count } return render(request, 'organization/schedule_scan_ui.html', context) @has_permission_decorator(PERM_MODIFY_SCAN_RESULTS, redirect_url=FOUR_OH_FOUR_URL) def delete_scans(request, slug): if request.method == "POST": for key, value in request.POST.items(): if key == 'scan_history_table_length' or key == 'csrfmiddlewaretoken': continue scan = get_object_or_404(ScanHistory, id=value) delete_dir = scan.results_dir run_command('rm -rf ' + delete_dir) scan.delete() messages.add_message( request, messages.INFO, 'All Scans deleted!') return HttpResponseRedirect(reverse('scan_history', kwargs={'slug': slug})) @has_permission_decorator(PERM_MODIFY_SCAN_REPORT, redirect_url=FOUR_OH_FOUR_URL) def customize_report(request, id): scan = ScanHistory.objects.get(id=id) context = { 'scan_id': id, 'scan_history': scan, } return render(request, 'startScan/customize_report.html', context) @has_permission_decorator(PERM_MODIFY_SCAN_REPORT, redirect_url=FOUR_OH_FOUR_URL) def create_report(request, id): primary_color = '#FFB74D' secondary_color = '#212121' # get report type report_type = request.GET['report_type'] if 'report_type' in request.GET else 'full' is_ignore_info_vuln = True if 'ignore_info_vuln' in request.GET else False if report_type == 'recon': show_recon = True show_vuln = False report_name = 'Reconnaissance Report' elif report_type == 'vulnerability': show_recon = False show_vuln = True report_name = 'Vulnerability Report' else: # default show_recon = True show_vuln = True report_name = 'Full Scan Report' scan = ScanHistory.objects.get(id=id) vulns = ( Vulnerability.objects .filter(scan_history=scan) .order_by('-severity') ) if not is_ignore_info_vuln else ( Vulnerability.objects .filter(scan_history=scan) .exclude(severity=0) .order_by('-severity') ) unique_vulns = ( Vulnerability.objects .filter(scan_history=scan) .values("name", "severity") .annotate(count=Count('name')) .order_by('-severity', '-count') ) if not is_ignore_info_vuln else ( Vulnerability.objects .filter(scan_history=scan) .exclude(severity=0) .values("name", "severity") .annotate(count=Count('name')) .order_by('-severity', '-count') ) subdomains = ( Subdomain.objects .filter(scan_history=scan) .order_by('-content_length') ) subdomain_alive_count = ( Subdomain.objects .filter(scan_history__id=id) .values('name') .distinct() .filter(http_status__exact=200) .count() ) interesting_subdomains = get_interesting_subdomains(scan_history=id) ip_addresses = ( IpAddress.objects .filter(ip_addresses__in=subdomains) .distinct() ) data = { 'scan_object': scan, 'unique_vulnerabilities': unique_vulns, 'all_vulnerabilities': vulns, 'all_vulnerabilities_count': vulns.count(), 'subdomain_alive_count': subdomain_alive_count, 'interesting_subdomains': interesting_subdomains, 'subdomains': subdomains, 'ip_addresses': ip_addresses, 'show_recon': show_recon, 'show_vuln': show_vuln, 'report_name': report_name, 'is_ignore_info_vuln': is_ignore_info_vuln, } # Get report related config vuln_report_query = VulnerabilityReportSetting.objects.all() if vuln_report_query.exists(): report = vuln_report_query[0] data['company_name'] = report.company_name data['company_address'] = report.company_address data['company_email'] = report.company_email data['company_website'] = report.company_website data['show_rengine_banner'] = report.show_rengine_banner data['show_footer'] = report.show_footer data['footer_text'] = report.footer_text data['show_executive_summary'] = report.show_executive_summary # Replace executive_summary_description with template syntax description = report.executive_summary_description description = description.replace('{scan_date}', scan.start_scan_date.strftime('%d %B, %Y')) description = description.replace('{company_name}', report.company_name) description = description.replace('{target_name}', scan.domain.name) description = description.replace('{subdomain_count}', str(subdomains.count())) description = description.replace('{vulnerability_count}', str(vulns.count())) description = description.replace('{critical_count}', str(vulns.filter(severity=4).count())) description = description.replace('{high_count}', str(vulns.filter(severity=3).count())) description = description.replace('{medium_count}', str(vulns.filter(severity=2).count())) description = description.replace('{low_count}', str(vulns.filter(severity=1).count())) description = description.replace('{info_count}', str(vulns.filter(severity=0).count())) description = description.replace('{unknown_count}', str(vulns.filter(severity=-1).count())) if scan.domain.description: description = description.replace('{target_description}', scan.domain.description) # Convert to Markdown data['executive_summary_description'] = markdown.markdown(description) primary_color = report.primary_color secondary_color = report.secondary_color data['primary_color'] = primary_color data['secondary_color'] = secondary_color template = get_template('report/template.html') html = template.render(data) pdf = HTML(string=html).write_pdf() if 'download' in request.GET: response = HttpResponse(pdf, content_type='application/octet-stream') else: response = HttpResponse(pdf, content_type='application/pdf') return response

psyray

4341d9834865240222a8dc72c01caaec0d7bed44

69231095782663fe0fe8b0e49b8aa995aa042723

In django template you can use `|length` you can use, `{{all_vulnerabilities|length}}`

yogeshojha

yogeshojha/rengine

Fix report generation when `Ignore Informational Vulnerabilities` checked

When **Ignore Informational Vulnerabilities** is checked there are still info vulns datas. I've reworked the queries that display vulnerabilities to prevent info vulns to display in the : - **Quick summary** Info blue box - **Reconnaissance Findings** - **Vulnerabilities Discovered** Info blue box I've also fixed the **Vulnerabilities Discovered** listing by doing a correct loop through regrouped values because values withe the same path but not the same severity does not display well Tested and working on current master branch

2023-12-05 01:25:41+00:00

2023-12-08 05:48:36+00:00

web/templates/report/template.html

<html> <head> <meta charset="utf-8"> <title>Report</title> <link href="https://fonts.googleapis.com/css2?family=Inter:wght@100;200;300;400;500&display=swap" rel="stylesheet"> <style> @page { size: A4; @top-left { background: {{primary_color}}; content: counter(page); height: 1cm; text-align: center; width: 1cm; } @top-center { background: {{primary_color}}; content: ''; display: block; height: .05cm; opacity: .5; width: 100%; } @top-right { content: string(heading); font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% if show_footer %} @bottom-left { content: "{{footer_text}}"; font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% endif %} } @page :blank { @top-left { background: none; content: '' } @top-center { content: none } @top-right { content: none } } @page no-chapter { @top-left { background: none; content: none } @top-center { content: none } @top-right { content: none } } @page :first { background-color: {{secondary_color}}; background-size: cover; margin: 0; } @page chapter { background: {{primary_color}}; margin: 0; @top-left { content: none } @top-center { content: none } @top-right { content: none } } html { color: #393939; font-family: 'Inter'; font-weight: 300; font-size: 11pt; font-weight: 300; line-height: 1.5; } h1 { font-family: 'Inter'; font-weight: 200; font-size: 38pt; margin: 5cm 2cm 0 2cm; page: no-chapter; width: 100%; line-height: normal; } h2, h3, h4 { font-family: 'Inter'; font-weight: 200; color: black; font-weight: 400; line-height: normal; } #cover { align-content: space-between; display: flex; flex-wrap: wrap; height: 297mm; } #cover-subheading { font-family: 'Inter'; font-weight: 200; font-size: 22pt; width: 100%; } #cover footer { background: {{primary_color}}; flex: 1 33%; margin: 0 -2cm; padding: 1cm 0; white-space: pre-wrap; } #cover footer:first-of-type { padding-left: 3cm; } #cover-line { margin-top: 6px; border-bottom: 1px double {{primary_color}}; } #summary { display: flex; flex-wrap: wrap; justify-content: space-between; } #contents { page: no-chapter; } #contents h2 { font-size: 20pt; Intereight: 400; margin-bottom: 3cm; } #contents h3 { font-weight: 400; margin: 3em 0 1em; } #contents h3::before { background: {{primary_color}}; content: ''; display: block; height: .08cm; margin-bottom: .25cm; width: 2cm; } #contents ul { list-style: none; padding-left: 0; } #contents ul li { border-top: .25pt solid #c1c1c1; margin: .25cm 0; padding-top: .25cm; } #contents ul li::before { color: {{primary_color}}; content: '• '; font-size: 30pt; line-height: 16pt; vertical-align: bottom; } #contents ul li a { color: inherit; text-decoration-line: inherit; } #contents ul li a::before { content: target-text(attr(href)); } #contents ul li a::after { color: {{primary_color}}; content: target-counter(attr(href), page); float: right; } #columns section { columns: 2; column-gap: 1cm; padding-top: 1cm; } #columns section p { text-align: justify; } #columns section p:first-of-type { font-weight: 700; } #chapter { align-items: center; display: flex; height: 297mm; justify-content: center; page: chapter; } #boxes { display: flex; flex-wrap: wrap; justify-content: space-between; } #boxes section h4 { margin-bottom: 0; } #boxes section p { background: {{primary_color}}; display: block; font-size: 15pt; margin-bottom: 0; padding: .25cm 0; text-align: center; height: 85px; color: #37474F; } .bg-critical { background: #EF9A9A !important; } .bg-high { background: #FFAB91 !important; } .bg-medium { background: #FFCC80 !important; } .bg-low { background: #FFE082 !important; } .bg-success { background-color: #A5D6A7 !important; } .bg-grey { background-color: #B0BEC5 !important; } .bg-info { background-color: #90CAF9 !important; } .critical-color { color: #EF9A9A; } .high-color { color: #dc3545; } .medium-color { color: #FFCC80; } .low-color { color: #FFE082; } .success-color { color: #A5D6A7; } .grey-color { color: #212121; } .info-color { color: #90CAF9; } .primary-color { color: {{primary_color}}; } .text-blue{ color: #007bff!important; } .badge { display: inline-block; padding-left: 12px; padding-right: 12px; text-align: center } .critical-hr-line { border-bottom: 3px solid #EF9A9A !important; } .high-hr-line { border-bottom: 3px solid #FFAB91 !important; } .medium-hr-line { border-bottom: 3px solid #FFCC80 !important; } .low-hr-line { border-bottom: 3px solid #FFE082 !important; } .info-hr-line { border-bottom: 3px solid #90CAF9 !important; } .grey-hr-line { border-bottom: 3px solid #212121 !important; } .inside-box-counter { font-size: 28pt; } .table { margin: 0 0 40px 0; width: 100%; box-shadow: 0 1px 3px rgba(0, 0, 0, 0.2); display: table; border-spacing: 0 0.4em; } .row { display: table-row; background: #f6f6f6; } .cell { padding: 6px 6px 6px 6px; display: table-cell; } .header { Intereight: 900; color: #ffffff; } .page_title{ font-weight: 300; font-size: 20pt; } .subheading{ font-weight: 300; font-size: 14pt; } .content-heading{ font-weight: 300; font-size: 12pt; } .mini-heading{ font-weight: 400; font-size: 11pt; } .table-border{ border-style:solid; border-width: 1px; border-color: #90CAF9 !important; } a{ color: #007bff; text-decoration: none; } .ml-8{ margin-left: 8px; } </style> </head> <body> <article id="cover"> <h1 style="color:{{primary_color}}">{{report_name}} <br> {{scan_object.domain.name}} <div id="cover-line"></div> {# generated date #} <span id="cover-subheading">{% now "F j, Y" %}</span> </h1> <footer> {{company_name}} {{company_address}} </footer> <footer> {{company_email}} {{company_website}} </footer> <footer> {% if show_rengine_banner %}Generated by reNgine https://github.com/yogeshojha/rengine {% endif %} </footer> </article> <article id="contents"> <h2>&nbsp;</h2> <h3>Table of contents</h3> <ul> {% if show_executive_summary %} <li><a href="#executive-summary"></a></li> {% endif %} <li><a href="#quick-summary"></a></li> <li><a href="#assessment-timeline"></a></li> {% if interesting_subdomains and show_recon %} <li><a href="#interesting-recon-data"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerability-summary"></a></li> {% endif %} {% if show_recon %} <li><a href="#reconnaissance-results"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerabilities-discovered"></a></li> {% endif %} </ul> </article> {% if show_executive_summary %} <article id="summary" style="page-break-before: always"> <h2 id="executive-summary" class="page_title">Executive summary</h2> <br> {{executive_summary_description | safe }} </article> {% endif %} <article id="summary" style="page-break-before: always"> <h2 id="quick-summary" class="page_title">Quick Summary</h2> <p>This section contains quick summary of scan performed on <span class="primary-color">{{scan_object.domain.name}}</span></p> <br> </article> {# recon section #} {% if show_recon %} <h4 id="reconnaissance-summary" class="subheading">Reconnaissance</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-success">Subdomains <br> <span class="inside-box-counter"> {{scan_object.get_subdomain_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Endpoints <br> <span class="inside-box-counter"> {{scan_object.get_endpoint_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-critical">Vulnerabilities <br> <span class="inside-box-counter"> {{scan_object.get_vulnerability_count}} </span> </p> </section> </div> {% endif %} <!-- vulnerability section, hide if only recon report --> {% if show_vuln %} <article> <br> <h4 id="vulnerability-summary" class="subheading">Vulnerability Summary</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width:30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {{scan_object.get_info_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {% endif %} <article> <h3 id="assessment-timeline" class="page_title">Timeline of the Assessment</h3> <p> Scan started on: {{scan_object.start_scan_date|date:"F j, Y h:i"}} <br> Total time taken: {% if scan_object.scan_status == 0 %} {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% elif scan_object.scan_status == 1 %} {{ scan_object.get_elapsed_time }} {% elif scan_object.scan_status == 2 %} {% if scan_object.get_completed_time_in_sec < 60 %} Completed in < 1 minutes {% else %} Completed in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} {% elif scan_object.scan_status == 3 %} Aborted in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} <br> Report Generated on: {% now "F j, Y" %} </p> </article> {# show interesting_subdomains section only when show_recon result is there #} {% if interesting_subdomains and show_recon %} <article style="page-break-before: always" class="summary"> <h3 id="interesting-recon-data" class="page_title">Interesting Recon Data</h3> <p>Listed below are the {{interesting_subdomains.count}} interesting subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-success"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 33%"> Subdomain </div> <div class="cell grey-color" style="width: 33%"> Page title </div> <div class="cell grey-color" style="width: 15%"> HTTP Status </div> </div> {% for subdomain in interesting_subdomains %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 35%"> {{subdomain.name}} </div> <div class="cell" style="width: 35%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> <div class="cell" style="width: 15%;"> {% if subdomain.http_status %} {{subdomain.http_status}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> </div> {% endfor %} </div> </article> {% endif %} {# vulnerability_summary only when vuln_report #} {% if show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerability-summary" class="page_title">Summary of Vulnerabilities Identified</h3> {% if all_vulnerabilities.count > 0 %} <p>Listed below are the vulnerabilities identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-critical"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 50%;"> Vulnerability Name </div> <div class="cell grey-color" style="width: 19%;"> Times Identified </div> <div class="cell grey-color" style="width: 15%"> Severity </div> </div> {% for vulnerability in unique_vulnerabilities %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 50%"> <a href="#vuln_{{vulnerability.name.split|join:'_'}}">{{vulnerability.name}}</a> </div> <div class="cell" style="float: right; width: 19%;"> {{vulnerability.count}} </div> {% if vulnerability.severity == -1 %} <div class="cell bg-grey" style="width: 15%"> <span class="severity-title-box">Unknown</span> {% elif vulnerability.severity == 0 %} <div class="cell bg-info" style="width: 15%"> <span class="severity-title-box">Informational</span> {% elif vulnerability.severity == 1 %} <div class="cell bg-low" style="width: 15%"> <span class="severity-title-box">Low</span> {% elif vulnerability.severity == 2 %} <div class="cell bg-medium" style="width: 15%"> <span class="severity-title-box">Medium</span> {% elif vulnerability.severity == 3 %} <div class="cell bg-high" style="width: 15%"> <span class="severity-title-box">High</span> {% elif vulnerability.severity == 4 %} <div class="cell bg-critical" style="width: 15%"> <span class="severity-title-box">Critical</span> {% endif %} </div> </div> {% endfor %} {% else %} <h3 class='info-color'>No Vulnerabilities were Discovered.</h3> {% endif %} </div> </article> {% endif %} {# show discovered assets only for show_recon report #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 id="reconnaissance-results" class="page_title">Discovered Assets</h3> <h4 class="subheading">Subdomains</h4> <p> During the reconnaissance phase, {{scan_object.get_subdomain_count}} subdomains were discovered. Out of {{scan_object.get_subdomain_count}} subdomains, {{subdomain_alive_count}} returned HTTP status 200. {{interesting_subdomains.count}} interesting subdomains were also identified based on the interesting keywords used. </p> <h4>{{scan_object.get_subdomain_count}} subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> Subdomain </div> <div class="cell grey-color" style="width: 38%"> Page title </div> <div class="cell grey-color" style="width: 18%"> HTTP Status </div> </div> {% for subdomain in subdomains %} <div class="row"> <div class="cell" style="width: 38%"> {{subdomain.name}} </div> <div class="cell" style="width: 38%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% endif %} </div> <div class="cell" style="width: 18%"> {{subdomain.http_status}} </div> </div> {% endfor %} </div> {% if ip_addresses.count %} <h4 class="subheading" style="margin-top: 10px;">IP Addresses</h4> <h4>{{ip_addresses.count}} IP Addresses were identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> IP </div> <div class="cell grey-color" style="width: 38%"> Open Ports </div> <div class="cell grey-color" style="width: 18%"> Remarks </div> </div> {% for ip in ip_addresses %} <div class="row"> <div class="cell" style="width: 38%"> {{ip.address}} </div> <div class="cell" style="width: 38%"> {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </div> {% if ip.is_cdn %} <div class="cell medium" style="width: 18%"> CDN IP Address {% else %} <div class="cell" style="width: 18%"> {% endif %} </div> </div> {% endfor %} </div> {% endif %} </article> <br> {% endif %} {# reconnaissance finding only when show_recon #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 class="page_title">Reconnaissance Findings</h3> {% for subdomain in subdomains %} <table class="table" cellspacing="0" style="border-collapse: collapse;"> <tr> <td style="width: 2%" class="cell table-border">{{ forloop.counter }}.</td> <td style="width: 80%" class="cell table-border">{{subdomain.name}}</td> {% if subdomain.http_status == 200 %} <td style="width: 10%" class="cell table-border bg-success">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 300 and subdomain.http_status < 400 %} <td style="width: 10%" class="cell table-border bg-medium">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 400 %} <td style="width: 10%" class="cell table-border bg-high">{{subdomain.http_status}}</td> {% elif subdomain.http_status == 0 %} <td style="width: 10%" class="cell table-border">N/A</td> {% else %} <td style="width: 10%" class="cell table-border">{{subdomain.http_status}}</td> {% endif %} </tr> {% if subdomain.page_title %} <tr> <td colspan="3" class="cell table-border"><strong>Page Title: </strong>{{subdomain.page_title}}</td> </tr> {% endif %} {% if subdomain.ip_addresses.all %} <tr> <td colspan="3" class="cell table-border"> IP Address: <ul> {% for ip in subdomain.ip_addresses.all %} <li>{{ip.address}} {% if ip.ports.all %} <ul> <li>Open Ports: &nbsp; {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </li> </ul> {% endif %} </li> {% endfor %} </ul> </td> </tr> {% endif %} {% if subdomain.get_vulnerabilities %} <tr> <td colspan="3" class="cell table-border"> Vulnerabilities {% regroup subdomain.get_vulnerabilities by name as vuln_list %} <ul> {% for vulnerability in vuln_list %} <li> <a href="#vuln_{{vulnerability.list.0.name.split|join:'_'}}">{{ vulnerability.grouper }}</a> </li> {% endfor %} </ul> </td> </tr> {% endif %} </table> {% endfor %} </article> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerabilities-discovered" class="page_title">Vulnerabilities Discovered</h3> <p> This section reports the security issues found during the audit. <br> A Total of {{scan_object.get_vulnerability_count}} were discovered in {{scan_object.domain.name}}, {{scan_object.get_critical_vulnerability_count}} of them were Critical, {{scan_object.get_high_vulnerability_count}} of them were High Severity, {{scan_object.get_medium_vulnerability_count}} of them were Medium severity, {{scan_object.get_low_vulnerability_count}} of them were Low severity, and {{scan_object.get_info_vulnerability_count}} of them were Informational. {{scan_object.get_unknown_vulnerability_count}} of them were Unknown Severity. </p> <h4 class="subheading">Vulnerability Breakdown by Severity</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {{scan_object.get_info_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {# start vulnerability #} {% if show_vuln %} <article class=""> {% regroup all_vulnerabilities by get_path as grouped_vulnerabilities %} {% for vulnerability in grouped_vulnerabilities %} <div> <h4 class="content-heading" id="vuln_{{vulnerability.list.0.name.split|join:'_'}}"> <span>{{vulnerability.list.0.name}} <br>in {{vulnerability.grouper}}</span> {% if vulnerability.list.0.severity == -1 %} <span style="float: right;" class="badge bg-grey">Unknown</span> <div class="grey-hr-line" ></div> {% elif vulnerability.list.0.severity == 0 %} <span style="float: right;" class="badge bg-info">INFO</span> <div class="info-hr-line" ></div> {% elif vulnerability.list.0.severity == 1 %} <span style="float: right;" class="badge bg-low">LOW</span> <div class="low-hr-line" ></div> {% elif vulnerability.list.0.severity == 2 %} <span style="float: right;" class="badge bg-medium">MEDIUM</span> <div class="medium-hr-line" ></div> {% elif vulnerability.list.0.severity == 3 %} <span style="float: right;" class="badge bg-high">HIGH</span> <div class="high-hr-line" ></div> {% elif vulnerability.list.0.severity == 4 %} <span style="float: right;" class="badge bg-critical">CRITICAL</span> <div class="critical-hr-line" ></div> {% endif %} </h4> <!-- show vulnerability classification --> <span class="mini-heading">Vulnerability Source: {{vulnerability.list.0.source|upper}}</span><br> {% if vulnerability.list.0.cvss_metrics or vulnerability.list.0.cvss_score or vulnerability.list.0.cve_ids.all or vulnerability.list.0.cve_ids.all %} <span class="mini-heading">Vulnerability Classification</span><br> {% if vulnerability.list.0.cvss_metrics %} <span class="mini-heading ml-8">CVSS Metrics: {{vulnerability.list.0.cvss_metrics}}</span> {% endif %} {% if vulnerability.list.0.cvss_score %} <br> <span class="mini-heading ml-8">CVSS Score:</span>&nbsp;<span class="high-color">{{vulnerability.list.0.cvss_score}}</span> {% endif %} {% if vulnerability.list.0.cve_ids.all %} <br> <span class="mini-heading ml-8">CVE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cve in vulnerability.list.0.cve_ids.all %} {{cve}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} {% if vulnerability.list.0.cwe_ids.all %} <br> <span class="mini-heading ml-8">CWE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cwe in vulnerability.list.0.cwe_ids.all %} {{cwe}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} <br> {% endif %} {% if vulnerability.list.0.description %} <br> <span class="mini-heading">Description</span><br> {{vulnerability.list.0.description|linebreaks}} {% endif %} {% if vulnerability.list.0.impact %} <br> <span class="mini-heading">Impact</span><br> {{vulnerability.list.0.impact|linebreaks}} {% endif %} {% if vulnerability.list.0.remediation %} <br> <span class="mini-heading">Remediation</span><br> {{vulnerability.list.0.remediation|linebreaks}} {% endif %} <br> <span class="mini-heading">Vulnerable URL(s)</span><br> <ul> {% for vuln in vulnerability.list %} <li class="text-blue">{{vuln.http_url}}</li> {% endfor %} </ul> <!-- {% regroup vulnerability.list by http_url as vuln_http_url_list %} --> <!-- <ul> {% for vuln_urls in vuln_http_url_list %} <li>{{vuln_urls.grouper}}</li> <span class="mini-heading">Result/Findings</span><br> {% for vuln in vuln_urls.list %} {% if vuln.matcher_name %} {% if not forloop.first %} • {% endif %} {{vuln.matcher_name}} {% endif %} {% if vuln.extracted_results %} {% for res in vuln.extracted_results %} {% if not forloop.first %} • {% endif %} {{res}} {% endfor %} {% endif %} {% endfor %} {% endfor %} </ul> --> {% if vulnerability.list.0.references.all %} <span class="mini-heading">References</span><br> <ul> {% for ref in vulnerability.list.0.references.all %} <li> <span class="text-blue"> {{ref}} </span> </li> {% endfor %} </ul> {% endif %} <br> <br> </div> {% endfor %} </article> {% endif %} {% endif %} <article id="chapter"> <h2 id="chapter-title">END OF REPORT</h2> </article> </body> </html>

<html> <head> <meta charset="utf-8"> <title>Report</title> <link href="https://fonts.googleapis.com/css2?family=Inter:wght@100;200;300;400;500&display=swap" rel="stylesheet"> <style> @page { size: A4; @top-left { background: {{primary_color}}; content: counter(page); height: 1cm; text-align: center; width: 1cm; } @top-center { background: {{primary_color}}; content: ''; display: block; height: .05cm; opacity: .5; width: 100%; } @top-right { content: string(heading); font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% if show_footer %} @bottom-left { content: "{{footer_text}}"; font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% endif %} } @page :blank { @top-left { background: none; content: '' } @top-center { content: none } @top-right { content: none } } @page no-chapter { @top-left { background: none; content: none } @top-center { content: none } @top-right { content: none } } @page :first { background-color: {{secondary_color}}; background-size: cover; margin: 0; } @page chapter { background: {{primary_color}}; margin: 0; @top-left { content: none } @top-center { content: none } @top-right { content: none } } html { color: #393939; font-family: 'Inter'; font-weight: 300; font-size: 11pt; font-weight: 300; line-height: 1.5; } h1 { font-family: 'Inter'; font-weight: 200; font-size: 38pt; margin: 5cm 2cm 0 2cm; page: no-chapter; width: 100%; line-height: normal; } h2, h3, h4 { font-family: 'Inter'; font-weight: 200; color: black; font-weight: 400; line-height: normal; } #cover { align-content: space-between; display: flex; flex-wrap: wrap; height: 297mm; } #cover-subheading { font-family: 'Inter'; font-weight: 200; font-size: 22pt; width: 100%; } #cover footer { background: {{primary_color}}; flex: 1 33%; margin: 0 -2cm; padding: 1cm 0; white-space: pre-wrap; } #cover footer:first-of-type { padding-left: 3cm; } #cover-line { margin-top: 6px; border-bottom: 1px double {{primary_color}}; } #summary { display: flex; flex-wrap: wrap; justify-content: space-between; } #contents { page: no-chapter; } #contents h2 { font-size: 20pt; Intereight: 400; margin-bottom: 3cm; } #contents h3 { font-weight: 400; margin: 3em 0 1em; } #contents h3::before { background: {{primary_color}}; content: ''; display: block; height: .08cm; margin-bottom: .25cm; width: 2cm; } #contents ul { list-style: none; padding-left: 0; } #contents ul li { border-top: .25pt solid #c1c1c1; margin: .25cm 0; padding-top: .25cm; } #contents ul li::before { color: {{primary_color}}; content: '• '; font-size: 30pt; line-height: 16pt; vertical-align: bottom; } #contents ul li a { color: inherit; text-decoration-line: inherit; } #contents ul li a::before { content: target-text(attr(href)); } #contents ul li a::after { color: {{primary_color}}; content: target-counter(attr(href), page); float: right; } #columns section { columns: 2; column-gap: 1cm; padding-top: 1cm; } #columns section p { text-align: justify; } #columns section p:first-of-type { font-weight: 700; } #chapter { align-items: center; display: flex; height: 297mm; justify-content: center; page: chapter; } #boxes { display: flex; flex-wrap: wrap; justify-content: space-between; } #boxes section h4 { margin-bottom: 0; } #boxes section p { background: {{primary_color}}; display: block; font-size: 15pt; margin-bottom: 0; padding: .25cm 0; text-align: center; height: 85px; color: #37474F; } .bg-critical { background: #EF9A9A !important; } .bg-high { background: #FFAB91 !important; } .bg-medium { background: #FFCC80 !important; } .bg-low { background: #FFE082 !important; } .bg-success { background-color: #A5D6A7 !important; } .bg-grey { background-color: #B0BEC5 !important; } .bg-info { background-color: #90CAF9 !important; } .critical-color { color: #EF9A9A; } .high-color { color: #dc3545; } .medium-color { color: #FFCC80; } .low-color { color: #FFE082; } .success-color { color: #A5D6A7; } .grey-color { color: #212121; } .info-color { color: #90CAF9; } .primary-color { color: {{primary_color}}; } .text-blue{ color: #007bff!important; } .badge { display: inline-block; padding-left: 12px; padding-right: 12px; text-align: center } .critical-hr-line { border-bottom: 3px solid #EF9A9A !important; } .high-hr-line { border-bottom: 3px solid #FFAB91 !important; } .medium-hr-line { border-bottom: 3px solid #FFCC80 !important; } .low-hr-line { border-bottom: 3px solid #FFE082 !important; } .info-hr-line { border-bottom: 3px solid #90CAF9 !important; } .grey-hr-line { border-bottom: 3px solid #212121 !important; } .inside-box-counter { font-size: 28pt; } .table { margin: 0 0 40px 0; width: 100%; box-shadow: 0 1px 3px rgba(0, 0, 0, 0.2); display: table; border-spacing: 0 0.4em; } .row { display: table-row; background: #f6f6f6; } .cell { padding: 6px 6px 6px 6px; display: table-cell; } .header { Intereight: 900; color: #ffffff; } .page_title{ font-weight: 300; font-size: 20pt; } .subheading{ font-weight: 300; font-size: 14pt; } .content-heading{ font-weight: 300; font-size: 12pt; } .mini-heading{ font-weight: 400; font-size: 11pt; } .table-border{ border-style:solid; border-width: 1px; border-color: #90CAF9 !important; } a{ color: #007bff; text-decoration: none; } .ml-8{ margin-left: 8px; } </style> </head> <body> <article id="cover"> <h1 style="color:{{primary_color}}">{{report_name}} <br> {{scan_object.domain.name}} <div id="cover-line"></div> {# generated date #} <span id="cover-subheading">{% now "F j, Y" %}</span> </h1> <footer> {{company_name}} {{company_address}} </footer> <footer> {{company_email}} {{company_website}} </footer> <footer> {% if show_rengine_banner %}Generated by reNgine https://github.com/yogeshojha/rengine {% endif %} </footer> </article> <article id="contents"> <h2>&nbsp;</h2> <h3>Table of contents</h3> <ul> {% if show_executive_summary %} <li><a href="#executive-summary"></a></li> {% endif %} <li><a href="#quick-summary"></a></li> <li><a href="#assessment-timeline"></a></li> {% if interesting_subdomains and show_recon %} <li><a href="#interesting-recon-data"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerability-summary"></a></li> {% endif %} {% if show_recon %} <li><a href="#reconnaissance-results"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerabilities-discovered"></a></li> {% endif %} </ul> </article> {% if show_executive_summary %} <article id="summary" style="page-break-before: always"> <h2 id="executive-summary" class="page_title">Executive summary</h2> <br> {{executive_summary_description | safe }} </article> {% endif %} <article id="summary" style="page-break-before: always"> <h2 id="quick-summary" class="page_title">Quick Summary</h2> <p>This section contains quick summary of scan performed on <span class="primary-color">{{scan_object.domain.name}}</span></p> <br> </article> {# recon section #} {% if show_recon %} <h4 id="reconnaissance-summary" class="subheading">Reconnaissance</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-success">Subdomains <br> <span class="inside-box-counter"> {{scan_object.get_subdomain_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Endpoints <br> <span class="inside-box-counter"> {{scan_object.get_endpoint_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-critical">Vulnerabilities <br> <span class="inside-box-counter"> {{all_vulnerabilities_count}} </span> </p> </section> </div> {% endif %} <!-- vulnerability section, hide if only recon report --> {% if show_vuln %} <article> <br> <h4 id="vulnerability-summary" class="subheading">Vulnerability Summary</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width:30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {% if is_ignore_info_vuln %} 0 {% else %} {{scan_object.get_info_vulnerability_count}} {% endif %} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {% endif %} <article> <h3 id="assessment-timeline" class="page_title">Timeline of the Assessment</h3> <p> Scan started on: {{scan_object.start_scan_date|date:"F j, Y h:i"}} <br> Total time taken: {% if scan_object.scan_status == 0 %} {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% elif scan_object.scan_status == 1 %} {{ scan_object.get_elapsed_time }} {% elif scan_object.scan_status == 2 %} {% if scan_object.get_completed_time_in_sec < 60 %} Completed in < 1 minutes {% else %} Completed in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} {% elif scan_object.scan_status == 3 %} Aborted in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} <br> Report Generated on: {% now "F j, Y" %} </p> </article> {# show interesting_subdomains section only when show_recon result is there #} {% if interesting_subdomains and show_recon %} <article style="page-break-before: always" class="summary"> <h3 id="interesting-recon-data" class="page_title">Interesting Recon Data</h3> <p>Listed below are the {{interesting_subdomains.count}} interesting subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-success"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 33%"> Subdomain </div> <div class="cell grey-color" style="width: 33%"> Page title </div> <div class="cell grey-color" style="width: 15%"> HTTP Status </div> </div> {% for subdomain in interesting_subdomains %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 35%"> {{subdomain.name}} </div> <div class="cell" style="width: 35%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> <div class="cell" style="width: 15%;"> {% if subdomain.http_status %} {{subdomain.http_status}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> </div> {% endfor %} </div> </article> {% endif %} {# vulnerability_summary only when vuln_report #} {% if show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerability-summary" class="page_title">Summary of Vulnerabilities Identified</h3> {% if all_vulnerabilities.count > 0 %} <p>Listed below are the vulnerabilities identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-critical"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 50%;"> Vulnerability Name </div> <div class="cell grey-color" style="width: 19%;"> Times Identified </div> <div class="cell grey-color" style="width: 15%"> Severity </div> </div> {% for vulnerability in unique_vulnerabilities %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 50%"> <a href="#vuln_{{vulnerability.name.split|join:'_'}}">{{vulnerability.name}}</a> </div> <div class="cell" style="float: right; width: 19%;"> {{vulnerability.count}} </div> {% if vulnerability.severity == -1 %} <div class="cell bg-grey" style="width: 15%"> <span class="severity-title-box">Unknown</span> {% elif vulnerability.severity == 0 %} <div class="cell bg-info" style="width: 15%"> <span class="severity-title-box">Informational</span> {% elif vulnerability.severity == 1 %} <div class="cell bg-low" style="width: 15%"> <span class="severity-title-box">Low</span> {% elif vulnerability.severity == 2 %} <div class="cell bg-medium" style="width: 15%"> <span class="severity-title-box">Medium</span> {% elif vulnerability.severity == 3 %} <div class="cell bg-high" style="width: 15%"> <span class="severity-title-box">High</span> {% elif vulnerability.severity == 4 %} <div class="cell bg-critical" style="width: 15%"> <span class="severity-title-box">Critical</span> {% endif %} </div> </div> {% endfor %} {% else %} <h3 class='info-color'>No Vulnerabilities were Discovered.</h3> {% endif %} </div> </article> {% endif %} {# show discovered assets only for show_recon report #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 id="reconnaissance-results" class="page_title">Discovered Assets</h3> <h4 class="subheading">Subdomains</h4> <p> During the reconnaissance phase, {{scan_object.get_subdomain_count}} subdomains were discovered. Out of {{scan_object.get_subdomain_count}} subdomains, {{subdomain_alive_count}} returned HTTP status 200. {{interesting_subdomains.count}} interesting subdomains were also identified based on the interesting keywords used. </p> <h4>{{scan_object.get_subdomain_count}} subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> Subdomain </div> <div class="cell grey-color" style="width: 38%"> Page title </div> <div class="cell grey-color" style="width: 18%"> HTTP Status </div> </div> {% for subdomain in subdomains %} <div class="row"> <div class="cell" style="width: 38%"> {{subdomain.name}} </div> <div class="cell" style="width: 38%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% endif %} </div> <div class="cell" style="width: 18%"> {{subdomain.http_status}} </div> </div> {% endfor %} </div> {% if ip_addresses.count %} <h4 class="subheading" style="margin-top: 10px;">IP Addresses</h4> <h4>{{ip_addresses.count}} IP Addresses were identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> IP </div> <div class="cell grey-color" style="width: 38%"> Open Ports </div> <div class="cell grey-color" style="width: 18%"> Remarks </div> </div> {% for ip in ip_addresses %} <div class="row"> <div class="cell" style="width: 38%"> {{ip.address}} </div> <div class="cell" style="width: 38%"> {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </div> {% if ip.is_cdn %} <div class="cell medium" style="width: 18%"> CDN IP Address {% else %} <div class="cell" style="width: 18%"> {% endif %} </div> </div> {% endfor %} </div> {% endif %} </article> <br> {% endif %} {# reconnaissance finding only when show_recon #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 class="page_title">Reconnaissance Findings</h3> {% for subdomain in subdomains %} <table class="table" cellspacing="0" style="border-collapse: collapse;"> <tr> <td style="width: 2%" class="cell table-border">{{ forloop.counter }}.</td> <td style="width: 80%" class="cell table-border">{{subdomain.name}}</td> {% if subdomain.http_status == 200 %} <td style="width: 10%" class="cell table-border bg-success">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 300 and subdomain.http_status < 400 %} <td style="width: 10%" class="cell table-border bg-medium">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 400 %} <td style="width: 10%" class="cell table-border bg-high">{{subdomain.http_status}}</td> {% elif subdomain.http_status == 0 %} <td style="width: 10%" class="cell table-border">N/A</td> {% else %} <td style="width: 10%" class="cell table-border">{{subdomain.http_status}}</td> {% endif %} </tr> {% if subdomain.page_title %} <tr> <td colspan="3" class="cell table-border"><strong>Page Title: </strong>{{subdomain.page_title}}</td> </tr> {% endif %} {% if subdomain.ip_addresses.all %} <tr> <td colspan="3" class="cell table-border"> IP Address: <ul> {% for ip in subdomain.ip_addresses.all %} <li>{{ip.address}} {% if ip.ports.all %} <ul> <li>Open Ports: &nbsp; {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </li> </ul> {% endif %} </li> {% endfor %} </ul> </td> </tr> {% endif %} {% if subdomain.get_vulnerabilities_without_info %} <tr> <td colspan="3" class="cell table-border"> Vulnerabilities {% regroup subdomain.get_vulnerabilities_without_info by name as vuln_list %} <ul> {% for vulnerability in vuln_list %} <li> <a href="#vuln_{{vulnerability.list.0.name.split|join:'_'}}">{{ vulnerability.grouper }}</a> </li> {% endfor %} </ul> </td> </tr> {% endif %} </table> {% endfor %} </article> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerabilities-discovered" class="page_title">Vulnerabilities Discovered</h3> <p> This section reports the security issues found during the audit. <br> A Total of {{scan_object.get_vulnerability_count}} were discovered in {{scan_object.domain.name}}, {{scan_object.get_critical_vulnerability_count}} of them were Critical, {{scan_object.get_high_vulnerability_count}} of them were High Severity, {{scan_object.get_medium_vulnerability_count}} of them were Medium severity, {% if is_ignore_info_vuln %}0{% else %}{{scan_object.get_info_vulnerability_count}}{% endif %} of them were Low severity, and {{scan_object.get_info_vulnerability_count}} of them were Informational. {{scan_object.get_unknown_vulnerability_count}} of them were Unknown Severity. </p> <h4 class="subheading">Vulnerability Breakdown by Severity</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {% if is_ignore_info_vuln %} 0 {% else %} {{scan_object.get_info_vulnerability_count}} {% endif %} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {# start vulnerability #} {% if show_vuln %} <article class=""> {% regroup all_vulnerabilities by get_path as grouped_vulnerabilities %} {% for vulnerabilities in grouped_vulnerabilities %} {% for vulnerability in vulnerabilities.list %} <div> <h4 class="content-heading" id="vuln_{{vulnerability.name.split|join:'_'}}"> <span>{{vulnerability.name}} <br>in {{vulnerabilities.grouper}}</span> {% if vulnerability.severity == -1 %} <span style="float: right;" class="badge bg-grey">Unknown</span> <div class="grey-hr-line" ></div> {% elif vulnerability.severity == 0 %} <span style="float: right;" class="badge bg-info">INFO</span> <div class="info-hr-line" ></div> {% elif vulnerability.severity == 1 %} <span style="float: right;" class="badge bg-low">LOW</span> <div class="low-hr-line" ></div> {% elif vulnerability.severity == 2 %} <span style="float: right;" class="badge bg-medium">MEDIUM</span> <div class="medium-hr-line" ></div> {% elif vulnerability.severity == 3 %} <span style="float: right;" class="badge bg-high">HIGH</span> <div class="high-hr-line" ></div> {% elif vulnerability.severity == 4 %} <span style="float: right;" class="badge bg-critical">CRITICAL</span> <div class="critical-hr-line" ></div> {% endif %} </h4> <!-- show vulnerability classification --> <span class="mini-heading">Vulnerability Source: {{vulnerability.source|upper}}</span><br> {% if vulnerability.cvss_metrics or vulnerability.cvss_score or vulnerability.cve_ids.all or vulnerability.cve_ids.all %} <span class="mini-heading">Vulnerability Classification</span><br> {% if vulnerability.cvss_metrics %} <span class="mini-heading ml-8">CVSS Metrics: {{vulnerability.cvss_metrics}}</span> {% endif %} {% if vulnerability.cvss_score %} <br> <span class="mini-heading ml-8">CVSS Score:</span>&nbsp;<span class="high-color">{{vulnerability.cvss_score}}</span> {% endif %} {% if vulnerability.cve_ids.all %} <br> <span class="mini-heading ml-8">CVE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cve in vulnerability.cve_ids.all %} {{cve}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} {% if vulnerability.cwe_ids.all %} <br> <span class="mini-heading ml-8">CWE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cwe in vulnerability.cwe_ids.all %} {{cwe}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} <br> {% endif %} {% if vulnerability.description %} <br> <span class="mini-heading">Description</span><br> {{vulnerability.description|linebreaks}} {% endif %} {% if vulnerability.impact %} <br> <span class="mini-heading">Impact</span><br> {{vulnerability.impact|linebreaks}} {% endif %} {% if vulnerability.remediation %} <br> <span class="mini-heading">Remediation</span><br> {{vulnerability.remediation|linebreaks}} {% endif %} <br> <span class="mini-heading">Vulnerable URL(s)</span><br> <ul> <li class="text-blue"><a href="{{vulnerability.http_url}}" target="_blank" rel="noopener noreferrer">{{vulnerability.http_url}}</a></li> </ul> <!-- {% regroup vulnerability.list by http_url as vuln_http_url_list %} --> <!-- <ul> {% for vuln_urls in vuln_http_url_list %} <li>{{vuln_urls.grouper}}</li> <span class="mini-heading">Result/Findings</span><br> {% for vuln in vuln_urls.list %} {% if vuln.matcher_name %} {% if not forloop.first %} • {% endif %} {{vuln.matcher_name}} {% endif %} {% if vuln.extracted_results %} {% for res in vuln.extracted_results %} {% if not forloop.first %} • {% endif %} {{res}} {% endfor %} {% endif %} {% endfor %} {% endfor %} </ul> --> {% if vulnerability.references.all %} <span class="mini-heading">References</span><br> <ul> {% for ref in vulnerability.references.all %} <li> <span class="text-blue"><a href="{{ref}}" target="_blank" rel="noopener noreferrer">{{ref}}</a></span> </li> {% endfor %} </ul> {% endif %} <br> <br> </div> {% endfor %} {% endfor %} </article> {% endif %} {% endif %} <article id="chapter"> <h2 id="chapter-title">END OF REPORT</h2> </article> </body> </html>

psyray

4341d9834865240222a8dc72c01caaec0d7bed44

69231095782663fe0fe8b0e49b8aa995aa042723

## Potentially unsafe external link External links without noopener/noreferrer are a potential security risk. [Show more details](https://github.com/yogeshojha/rengine/security/code-scanning/171)

github-advanced-security[bot]

yogeshojha/rengine

Fix report generation when `Ignore Informational Vulnerabilities` checked

When **Ignore Informational Vulnerabilities** is checked there are still info vulns datas. I've reworked the queries that display vulnerabilities to prevent info vulns to display in the : - **Quick summary** Info blue box - **Reconnaissance Findings** - **Vulnerabilities Discovered** Info blue box I've also fixed the **Vulnerabilities Discovered** listing by doing a correct loop through regrouped values because values withe the same path but not the same severity does not display well Tested and working on current master branch

2023-12-05 01:25:41+00:00

2023-12-08 05:48:36+00:00

web/templates/report/template.html

<html> <head> <meta charset="utf-8"> <title>Report</title> <link href="https://fonts.googleapis.com/css2?family=Inter:wght@100;200;300;400;500&display=swap" rel="stylesheet"> <style> @page { size: A4; @top-left { background: {{primary_color}}; content: counter(page); height: 1cm; text-align: center; width: 1cm; } @top-center { background: {{primary_color}}; content: ''; display: block; height: .05cm; opacity: .5; width: 100%; } @top-right { content: string(heading); font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% if show_footer %} @bottom-left { content: "{{footer_text}}"; font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% endif %} } @page :blank { @top-left { background: none; content: '' } @top-center { content: none } @top-right { content: none } } @page no-chapter { @top-left { background: none; content: none } @top-center { content: none } @top-right { content: none } } @page :first { background-color: {{secondary_color}}; background-size: cover; margin: 0; } @page chapter { background: {{primary_color}}; margin: 0; @top-left { content: none } @top-center { content: none } @top-right { content: none } } html { color: #393939; font-family: 'Inter'; font-weight: 300; font-size: 11pt; font-weight: 300; line-height: 1.5; } h1 { font-family: 'Inter'; font-weight: 200; font-size: 38pt; margin: 5cm 2cm 0 2cm; page: no-chapter; width: 100%; line-height: normal; } h2, h3, h4 { font-family: 'Inter'; font-weight: 200; color: black; font-weight: 400; line-height: normal; } #cover { align-content: space-between; display: flex; flex-wrap: wrap; height: 297mm; } #cover-subheading { font-family: 'Inter'; font-weight: 200; font-size: 22pt; width: 100%; } #cover footer { background: {{primary_color}}; flex: 1 33%; margin: 0 -2cm; padding: 1cm 0; white-space: pre-wrap; } #cover footer:first-of-type { padding-left: 3cm; } #cover-line { margin-top: 6px; border-bottom: 1px double {{primary_color}}; } #summary { display: flex; flex-wrap: wrap; justify-content: space-between; } #contents { page: no-chapter; } #contents h2 { font-size: 20pt; Intereight: 400; margin-bottom: 3cm; } #contents h3 { font-weight: 400; margin: 3em 0 1em; } #contents h3::before { background: {{primary_color}}; content: ''; display: block; height: .08cm; margin-bottom: .25cm; width: 2cm; } #contents ul { list-style: none; padding-left: 0; } #contents ul li { border-top: .25pt solid #c1c1c1; margin: .25cm 0; padding-top: .25cm; } #contents ul li::before { color: {{primary_color}}; content: '• '; font-size: 30pt; line-height: 16pt; vertical-align: bottom; } #contents ul li a { color: inherit; text-decoration-line: inherit; } #contents ul li a::before { content: target-text(attr(href)); } #contents ul li a::after { color: {{primary_color}}; content: target-counter(attr(href), page); float: right; } #columns section { columns: 2; column-gap: 1cm; padding-top: 1cm; } #columns section p { text-align: justify; } #columns section p:first-of-type { font-weight: 700; } #chapter { align-items: center; display: flex; height: 297mm; justify-content: center; page: chapter; } #boxes { display: flex; flex-wrap: wrap; justify-content: space-between; } #boxes section h4 { margin-bottom: 0; } #boxes section p { background: {{primary_color}}; display: block; font-size: 15pt; margin-bottom: 0; padding: .25cm 0; text-align: center; height: 85px; color: #37474F; } .bg-critical { background: #EF9A9A !important; } .bg-high { background: #FFAB91 !important; } .bg-medium { background: #FFCC80 !important; } .bg-low { background: #FFE082 !important; } .bg-success { background-color: #A5D6A7 !important; } .bg-grey { background-color: #B0BEC5 !important; } .bg-info { background-color: #90CAF9 !important; } .critical-color { color: #EF9A9A; } .high-color { color: #dc3545; } .medium-color { color: #FFCC80; } .low-color { color: #FFE082; } .success-color { color: #A5D6A7; } .grey-color { color: #212121; } .info-color { color: #90CAF9; } .primary-color { color: {{primary_color}}; } .text-blue{ color: #007bff!important; } .badge { display: inline-block; padding-left: 12px; padding-right: 12px; text-align: center } .critical-hr-line { border-bottom: 3px solid #EF9A9A !important; } .high-hr-line { border-bottom: 3px solid #FFAB91 !important; } .medium-hr-line { border-bottom: 3px solid #FFCC80 !important; } .low-hr-line { border-bottom: 3px solid #FFE082 !important; } .info-hr-line { border-bottom: 3px solid #90CAF9 !important; } .grey-hr-line { border-bottom: 3px solid #212121 !important; } .inside-box-counter { font-size: 28pt; } .table { margin: 0 0 40px 0; width: 100%; box-shadow: 0 1px 3px rgba(0, 0, 0, 0.2); display: table; border-spacing: 0 0.4em; } .row { display: table-row; background: #f6f6f6; } .cell { padding: 6px 6px 6px 6px; display: table-cell; } .header { Intereight: 900; color: #ffffff; } .page_title{ font-weight: 300; font-size: 20pt; } .subheading{ font-weight: 300; font-size: 14pt; } .content-heading{ font-weight: 300; font-size: 12pt; } .mini-heading{ font-weight: 400; font-size: 11pt; } .table-border{ border-style:solid; border-width: 1px; border-color: #90CAF9 !important; } a{ color: #007bff; text-decoration: none; } .ml-8{ margin-left: 8px; } </style> </head> <body> <article id="cover"> <h1 style="color:{{primary_color}}">{{report_name}} <br> {{scan_object.domain.name}} <div id="cover-line"></div> {# generated date #} <span id="cover-subheading">{% now "F j, Y" %}</span> </h1> <footer> {{company_name}} {{company_address}} </footer> <footer> {{company_email}} {{company_website}} </footer> <footer> {% if show_rengine_banner %}Generated by reNgine https://github.com/yogeshojha/rengine {% endif %} </footer> </article> <article id="contents"> <h2>&nbsp;</h2> <h3>Table of contents</h3> <ul> {% if show_executive_summary %} <li><a href="#executive-summary"></a></li> {% endif %} <li><a href="#quick-summary"></a></li> <li><a href="#assessment-timeline"></a></li> {% if interesting_subdomains and show_recon %} <li><a href="#interesting-recon-data"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerability-summary"></a></li> {% endif %} {% if show_recon %} <li><a href="#reconnaissance-results"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerabilities-discovered"></a></li> {% endif %} </ul> </article> {% if show_executive_summary %} <article id="summary" style="page-break-before: always"> <h2 id="executive-summary" class="page_title">Executive summary</h2> <br> {{executive_summary_description | safe }} </article> {% endif %} <article id="summary" style="page-break-before: always"> <h2 id="quick-summary" class="page_title">Quick Summary</h2> <p>This section contains quick summary of scan performed on <span class="primary-color">{{scan_object.domain.name}}</span></p> <br> </article> {# recon section #} {% if show_recon %} <h4 id="reconnaissance-summary" class="subheading">Reconnaissance</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-success">Subdomains <br> <span class="inside-box-counter"> {{scan_object.get_subdomain_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Endpoints <br> <span class="inside-box-counter"> {{scan_object.get_endpoint_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-critical">Vulnerabilities <br> <span class="inside-box-counter"> {{scan_object.get_vulnerability_count}} </span> </p> </section> </div> {% endif %} <!-- vulnerability section, hide if only recon report --> {% if show_vuln %} <article> <br> <h4 id="vulnerability-summary" class="subheading">Vulnerability Summary</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width:30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {{scan_object.get_info_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {% endif %} <article> <h3 id="assessment-timeline" class="page_title">Timeline of the Assessment</h3> <p> Scan started on: {{scan_object.start_scan_date|date:"F j, Y h:i"}} <br> Total time taken: {% if scan_object.scan_status == 0 %} {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% elif scan_object.scan_status == 1 %} {{ scan_object.get_elapsed_time }} {% elif scan_object.scan_status == 2 %} {% if scan_object.get_completed_time_in_sec < 60 %} Completed in < 1 minutes {% else %} Completed in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} {% elif scan_object.scan_status == 3 %} Aborted in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} <br> Report Generated on: {% now "F j, Y" %} </p> </article> {# show interesting_subdomains section only when show_recon result is there #} {% if interesting_subdomains and show_recon %} <article style="page-break-before: always" class="summary"> <h3 id="interesting-recon-data" class="page_title">Interesting Recon Data</h3> <p>Listed below are the {{interesting_subdomains.count}} interesting subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-success"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 33%"> Subdomain </div> <div class="cell grey-color" style="width: 33%"> Page title </div> <div class="cell grey-color" style="width: 15%"> HTTP Status </div> </div> {% for subdomain in interesting_subdomains %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 35%"> {{subdomain.name}} </div> <div class="cell" style="width: 35%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> <div class="cell" style="width: 15%;"> {% if subdomain.http_status %} {{subdomain.http_status}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> </div> {% endfor %} </div> </article> {% endif %} {# vulnerability_summary only when vuln_report #} {% if show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerability-summary" class="page_title">Summary of Vulnerabilities Identified</h3> {% if all_vulnerabilities.count > 0 %} <p>Listed below are the vulnerabilities identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-critical"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 50%;"> Vulnerability Name </div> <div class="cell grey-color" style="width: 19%;"> Times Identified </div> <div class="cell grey-color" style="width: 15%"> Severity </div> </div> {% for vulnerability in unique_vulnerabilities %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 50%"> <a href="#vuln_{{vulnerability.name.split|join:'_'}}">{{vulnerability.name}}</a> </div> <div class="cell" style="float: right; width: 19%;"> {{vulnerability.count}} </div> {% if vulnerability.severity == -1 %} <div class="cell bg-grey" style="width: 15%"> <span class="severity-title-box">Unknown</span> {% elif vulnerability.severity == 0 %} <div class="cell bg-info" style="width: 15%"> <span class="severity-title-box">Informational</span> {% elif vulnerability.severity == 1 %} <div class="cell bg-low" style="width: 15%"> <span class="severity-title-box">Low</span> {% elif vulnerability.severity == 2 %} <div class="cell bg-medium" style="width: 15%"> <span class="severity-title-box">Medium</span> {% elif vulnerability.severity == 3 %} <div class="cell bg-high" style="width: 15%"> <span class="severity-title-box">High</span> {% elif vulnerability.severity == 4 %} <div class="cell bg-critical" style="width: 15%"> <span class="severity-title-box">Critical</span> {% endif %} </div> </div> {% endfor %} {% else %} <h3 class='info-color'>No Vulnerabilities were Discovered.</h3> {% endif %} </div> </article> {% endif %} {# show discovered assets only for show_recon report #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 id="reconnaissance-results" class="page_title">Discovered Assets</h3> <h4 class="subheading">Subdomains</h4> <p> During the reconnaissance phase, {{scan_object.get_subdomain_count}} subdomains were discovered. Out of {{scan_object.get_subdomain_count}} subdomains, {{subdomain_alive_count}} returned HTTP status 200. {{interesting_subdomains.count}} interesting subdomains were also identified based on the interesting keywords used. </p> <h4>{{scan_object.get_subdomain_count}} subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> Subdomain </div> <div class="cell grey-color" style="width: 38%"> Page title </div> <div class="cell grey-color" style="width: 18%"> HTTP Status </div> </div> {% for subdomain in subdomains %} <div class="row"> <div class="cell" style="width: 38%"> {{subdomain.name}} </div> <div class="cell" style="width: 38%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% endif %} </div> <div class="cell" style="width: 18%"> {{subdomain.http_status}} </div> </div> {% endfor %} </div> {% if ip_addresses.count %} <h4 class="subheading" style="margin-top: 10px;">IP Addresses</h4> <h4>{{ip_addresses.count}} IP Addresses were identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> IP </div> <div class="cell grey-color" style="width: 38%"> Open Ports </div> <div class="cell grey-color" style="width: 18%"> Remarks </div> </div> {% for ip in ip_addresses %} <div class="row"> <div class="cell" style="width: 38%"> {{ip.address}} </div> <div class="cell" style="width: 38%"> {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </div> {% if ip.is_cdn %} <div class="cell medium" style="width: 18%"> CDN IP Address {% else %} <div class="cell" style="width: 18%"> {% endif %} </div> </div> {% endfor %} </div> {% endif %} </article> <br> {% endif %} {# reconnaissance finding only when show_recon #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 class="page_title">Reconnaissance Findings</h3> {% for subdomain in subdomains %} <table class="table" cellspacing="0" style="border-collapse: collapse;"> <tr> <td style="width: 2%" class="cell table-border">{{ forloop.counter }}.</td> <td style="width: 80%" class="cell table-border">{{subdomain.name}}</td> {% if subdomain.http_status == 200 %} <td style="width: 10%" class="cell table-border bg-success">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 300 and subdomain.http_status < 400 %} <td style="width: 10%" class="cell table-border bg-medium">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 400 %} <td style="width: 10%" class="cell table-border bg-high">{{subdomain.http_status}}</td> {% elif subdomain.http_status == 0 %} <td style="width: 10%" class="cell table-border">N/A</td> {% else %} <td style="width: 10%" class="cell table-border">{{subdomain.http_status}}</td> {% endif %} </tr> {% if subdomain.page_title %} <tr> <td colspan="3" class="cell table-border"><strong>Page Title: </strong>{{subdomain.page_title}}</td> </tr> {% endif %} {% if subdomain.ip_addresses.all %} <tr> <td colspan="3" class="cell table-border"> IP Address: <ul> {% for ip in subdomain.ip_addresses.all %} <li>{{ip.address}} {% if ip.ports.all %} <ul> <li>Open Ports: &nbsp; {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </li> </ul> {% endif %} </li> {% endfor %} </ul> </td> </tr> {% endif %} {% if subdomain.get_vulnerabilities %} <tr> <td colspan="3" class="cell table-border"> Vulnerabilities {% regroup subdomain.get_vulnerabilities by name as vuln_list %} <ul> {% for vulnerability in vuln_list %} <li> <a href="#vuln_{{vulnerability.list.0.name.split|join:'_'}}">{{ vulnerability.grouper }}</a> </li> {% endfor %} </ul> </td> </tr> {% endif %} </table> {% endfor %} </article> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerabilities-discovered" class="page_title">Vulnerabilities Discovered</h3> <p> This section reports the security issues found during the audit. <br> A Total of {{scan_object.get_vulnerability_count}} were discovered in {{scan_object.domain.name}}, {{scan_object.get_critical_vulnerability_count}} of them were Critical, {{scan_object.get_high_vulnerability_count}} of them were High Severity, {{scan_object.get_medium_vulnerability_count}} of them were Medium severity, {{scan_object.get_low_vulnerability_count}} of them were Low severity, and {{scan_object.get_info_vulnerability_count}} of them were Informational. {{scan_object.get_unknown_vulnerability_count}} of them were Unknown Severity. </p> <h4 class="subheading">Vulnerability Breakdown by Severity</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {{scan_object.get_info_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {# start vulnerability #} {% if show_vuln %} <article class=""> {% regroup all_vulnerabilities by get_path as grouped_vulnerabilities %} {% for vulnerability in grouped_vulnerabilities %} <div> <h4 class="content-heading" id="vuln_{{vulnerability.list.0.name.split|join:'_'}}"> <span>{{vulnerability.list.0.name}} <br>in {{vulnerability.grouper}}</span> {% if vulnerability.list.0.severity == -1 %} <span style="float: right;" class="badge bg-grey">Unknown</span> <div class="grey-hr-line" ></div> {% elif vulnerability.list.0.severity == 0 %} <span style="float: right;" class="badge bg-info">INFO</span> <div class="info-hr-line" ></div> {% elif vulnerability.list.0.severity == 1 %} <span style="float: right;" class="badge bg-low">LOW</span> <div class="low-hr-line" ></div> {% elif vulnerability.list.0.severity == 2 %} <span style="float: right;" class="badge bg-medium">MEDIUM</span> <div class="medium-hr-line" ></div> {% elif vulnerability.list.0.severity == 3 %} <span style="float: right;" class="badge bg-high">HIGH</span> <div class="high-hr-line" ></div> {% elif vulnerability.list.0.severity == 4 %} <span style="float: right;" class="badge bg-critical">CRITICAL</span> <div class="critical-hr-line" ></div> {% endif %} </h4> <!-- show vulnerability classification --> <span class="mini-heading">Vulnerability Source: {{vulnerability.list.0.source|upper}}</span><br> {% if vulnerability.list.0.cvss_metrics or vulnerability.list.0.cvss_score or vulnerability.list.0.cve_ids.all or vulnerability.list.0.cve_ids.all %} <span class="mini-heading">Vulnerability Classification</span><br> {% if vulnerability.list.0.cvss_metrics %} <span class="mini-heading ml-8">CVSS Metrics: {{vulnerability.list.0.cvss_metrics}}</span> {% endif %} {% if vulnerability.list.0.cvss_score %} <br> <span class="mini-heading ml-8">CVSS Score:</span>&nbsp;<span class="high-color">{{vulnerability.list.0.cvss_score}}</span> {% endif %} {% if vulnerability.list.0.cve_ids.all %} <br> <span class="mini-heading ml-8">CVE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cve in vulnerability.list.0.cve_ids.all %} {{cve}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} {% if vulnerability.list.0.cwe_ids.all %} <br> <span class="mini-heading ml-8">CWE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cwe in vulnerability.list.0.cwe_ids.all %} {{cwe}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} <br> {% endif %} {% if vulnerability.list.0.description %} <br> <span class="mini-heading">Description</span><br> {{vulnerability.list.0.description|linebreaks}} {% endif %} {% if vulnerability.list.0.impact %} <br> <span class="mini-heading">Impact</span><br> {{vulnerability.list.0.impact|linebreaks}} {% endif %} {% if vulnerability.list.0.remediation %} <br> <span class="mini-heading">Remediation</span><br> {{vulnerability.list.0.remediation|linebreaks}} {% endif %} <br> <span class="mini-heading">Vulnerable URL(s)</span><br> <ul> {% for vuln in vulnerability.list %} <li class="text-blue">{{vuln.http_url}}</li> {% endfor %} </ul> <!-- {% regroup vulnerability.list by http_url as vuln_http_url_list %} --> <!-- <ul> {% for vuln_urls in vuln_http_url_list %} <li>{{vuln_urls.grouper}}</li> <span class="mini-heading">Result/Findings</span><br> {% for vuln in vuln_urls.list %} {% if vuln.matcher_name %} {% if not forloop.first %} • {% endif %} {{vuln.matcher_name}} {% endif %} {% if vuln.extracted_results %} {% for res in vuln.extracted_results %} {% if not forloop.first %} • {% endif %} {{res}} {% endfor %} {% endif %} {% endfor %} {% endfor %} </ul> --> {% if vulnerability.list.0.references.all %} <span class="mini-heading">References</span><br> <ul> {% for ref in vulnerability.list.0.references.all %} <li> <span class="text-blue"> {{ref}} </span> </li> {% endfor %} </ul> {% endif %} <br> <br> </div> {% endfor %} </article> {% endif %} {% endif %} <article id="chapter"> <h2 id="chapter-title">END OF REPORT</h2> </article> </body> </html>

<html> <head> <meta charset="utf-8"> <title>Report</title> <link href="https://fonts.googleapis.com/css2?family=Inter:wght@100;200;300;400;500&display=swap" rel="stylesheet"> <style> @page { size: A4; @top-left { background: {{primary_color}}; content: counter(page); height: 1cm; text-align: center; width: 1cm; } @top-center { background: {{primary_color}}; content: ''; display: block; height: .05cm; opacity: .5; width: 100%; } @top-right { content: string(heading); font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% if show_footer %} @bottom-left { content: "{{footer_text}}"; font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% endif %} } @page :blank { @top-left { background: none; content: '' } @top-center { content: none } @top-right { content: none } } @page no-chapter { @top-left { background: none; content: none } @top-center { content: none } @top-right { content: none } } @page :first { background-color: {{secondary_color}}; background-size: cover; margin: 0; } @page chapter { background: {{primary_color}}; margin: 0; @top-left { content: none } @top-center { content: none } @top-right { content: none } } html { color: #393939; font-family: 'Inter'; font-weight: 300; font-size: 11pt; font-weight: 300; line-height: 1.5; } h1 { font-family: 'Inter'; font-weight: 200; font-size: 38pt; margin: 5cm 2cm 0 2cm; page: no-chapter; width: 100%; line-height: normal; } h2, h3, h4 { font-family: 'Inter'; font-weight: 200; color: black; font-weight: 400; line-height: normal; } #cover { align-content: space-between; display: flex; flex-wrap: wrap; height: 297mm; } #cover-subheading { font-family: 'Inter'; font-weight: 200; font-size: 22pt; width: 100%; } #cover footer { background: {{primary_color}}; flex: 1 33%; margin: 0 -2cm; padding: 1cm 0; white-space: pre-wrap; } #cover footer:first-of-type { padding-left: 3cm; } #cover-line { margin-top: 6px; border-bottom: 1px double {{primary_color}}; } #summary { display: flex; flex-wrap: wrap; justify-content: space-between; } #contents { page: no-chapter; } #contents h2 { font-size: 20pt; Intereight: 400; margin-bottom: 3cm; } #contents h3 { font-weight: 400; margin: 3em 0 1em; } #contents h3::before { background: {{primary_color}}; content: ''; display: block; height: .08cm; margin-bottom: .25cm; width: 2cm; } #contents ul { list-style: none; padding-left: 0; } #contents ul li { border-top: .25pt solid #c1c1c1; margin: .25cm 0; padding-top: .25cm; } #contents ul li::before { color: {{primary_color}}; content: '• '; font-size: 30pt; line-height: 16pt; vertical-align: bottom; } #contents ul li a { color: inherit; text-decoration-line: inherit; } #contents ul li a::before { content: target-text(attr(href)); } #contents ul li a::after { color: {{primary_color}}; content: target-counter(attr(href), page); float: right; } #columns section { columns: 2; column-gap: 1cm; padding-top: 1cm; } #columns section p { text-align: justify; } #columns section p:first-of-type { font-weight: 700; } #chapter { align-items: center; display: flex; height: 297mm; justify-content: center; page: chapter; } #boxes { display: flex; flex-wrap: wrap; justify-content: space-between; } #boxes section h4 { margin-bottom: 0; } #boxes section p { background: {{primary_color}}; display: block; font-size: 15pt; margin-bottom: 0; padding: .25cm 0; text-align: center; height: 85px; color: #37474F; } .bg-critical { background: #EF9A9A !important; } .bg-high { background: #FFAB91 !important; } .bg-medium { background: #FFCC80 !important; } .bg-low { background: #FFE082 !important; } .bg-success { background-color: #A5D6A7 !important; } .bg-grey { background-color: #B0BEC5 !important; } .bg-info { background-color: #90CAF9 !important; } .critical-color { color: #EF9A9A; } .high-color { color: #dc3545; } .medium-color { color: #FFCC80; } .low-color { color: #FFE082; } .success-color { color: #A5D6A7; } .grey-color { color: #212121; } .info-color { color: #90CAF9; } .primary-color { color: {{primary_color}}; } .text-blue{ color: #007bff!important; } .badge { display: inline-block; padding-left: 12px; padding-right: 12px; text-align: center } .critical-hr-line { border-bottom: 3px solid #EF9A9A !important; } .high-hr-line { border-bottom: 3px solid #FFAB91 !important; } .medium-hr-line { border-bottom: 3px solid #FFCC80 !important; } .low-hr-line { border-bottom: 3px solid #FFE082 !important; } .info-hr-line { border-bottom: 3px solid #90CAF9 !important; } .grey-hr-line { border-bottom: 3px solid #212121 !important; } .inside-box-counter { font-size: 28pt; } .table { margin: 0 0 40px 0; width: 100%; box-shadow: 0 1px 3px rgba(0, 0, 0, 0.2); display: table; border-spacing: 0 0.4em; } .row { display: table-row; background: #f6f6f6; } .cell { padding: 6px 6px 6px 6px; display: table-cell; } .header { Intereight: 900; color: #ffffff; } .page_title{ font-weight: 300; font-size: 20pt; } .subheading{ font-weight: 300; font-size: 14pt; } .content-heading{ font-weight: 300; font-size: 12pt; } .mini-heading{ font-weight: 400; font-size: 11pt; } .table-border{ border-style:solid; border-width: 1px; border-color: #90CAF9 !important; } a{ color: #007bff; text-decoration: none; } .ml-8{ margin-left: 8px; } </style> </head> <body> <article id="cover"> <h1 style="color:{{primary_color}}">{{report_name}} <br> {{scan_object.domain.name}} <div id="cover-line"></div> {# generated date #} <span id="cover-subheading">{% now "F j, Y" %}</span> </h1> <footer> {{company_name}} {{company_address}} </footer> <footer> {{company_email}} {{company_website}} </footer> <footer> {% if show_rengine_banner %}Generated by reNgine https://github.com/yogeshojha/rengine {% endif %} </footer> </article> <article id="contents"> <h2>&nbsp;</h2> <h3>Table of contents</h3> <ul> {% if show_executive_summary %} <li><a href="#executive-summary"></a></li> {% endif %} <li><a href="#quick-summary"></a></li> <li><a href="#assessment-timeline"></a></li> {% if interesting_subdomains and show_recon %} <li><a href="#interesting-recon-data"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerability-summary"></a></li> {% endif %} {% if show_recon %} <li><a href="#reconnaissance-results"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerabilities-discovered"></a></li> {% endif %} </ul> </article> {% if show_executive_summary %} <article id="summary" style="page-break-before: always"> <h2 id="executive-summary" class="page_title">Executive summary</h2> <br> {{executive_summary_description | safe }} </article> {% endif %} <article id="summary" style="page-break-before: always"> <h2 id="quick-summary" class="page_title">Quick Summary</h2> <p>This section contains quick summary of scan performed on <span class="primary-color">{{scan_object.domain.name}}</span></p> <br> </article> {# recon section #} {% if show_recon %} <h4 id="reconnaissance-summary" class="subheading">Reconnaissance</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-success">Subdomains <br> <span class="inside-box-counter"> {{scan_object.get_subdomain_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Endpoints <br> <span class="inside-box-counter"> {{scan_object.get_endpoint_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-critical">Vulnerabilities <br> <span class="inside-box-counter"> {{all_vulnerabilities_count}} </span> </p> </section> </div> {% endif %} <!-- vulnerability section, hide if only recon report --> {% if show_vuln %} <article> <br> <h4 id="vulnerability-summary" class="subheading">Vulnerability Summary</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width:30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {% if is_ignore_info_vuln %} 0 {% else %} {{scan_object.get_info_vulnerability_count}} {% endif %} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {% endif %} <article> <h3 id="assessment-timeline" class="page_title">Timeline of the Assessment</h3> <p> Scan started on: {{scan_object.start_scan_date|date:"F j, Y h:i"}} <br> Total time taken: {% if scan_object.scan_status == 0 %} {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% elif scan_object.scan_status == 1 %} {{ scan_object.get_elapsed_time }} {% elif scan_object.scan_status == 2 %} {% if scan_object.get_completed_time_in_sec < 60 %} Completed in < 1 minutes {% else %} Completed in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} {% elif scan_object.scan_status == 3 %} Aborted in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} <br> Report Generated on: {% now "F j, Y" %} </p> </article> {# show interesting_subdomains section only when show_recon result is there #} {% if interesting_subdomains and show_recon %} <article style="page-break-before: always" class="summary"> <h3 id="interesting-recon-data" class="page_title">Interesting Recon Data</h3> <p>Listed below are the {{interesting_subdomains.count}} interesting subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-success"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 33%"> Subdomain </div> <div class="cell grey-color" style="width: 33%"> Page title </div> <div class="cell grey-color" style="width: 15%"> HTTP Status </div> </div> {% for subdomain in interesting_subdomains %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 35%"> {{subdomain.name}} </div> <div class="cell" style="width: 35%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> <div class="cell" style="width: 15%;"> {% if subdomain.http_status %} {{subdomain.http_status}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> </div> {% endfor %} </div> </article> {% endif %} {# vulnerability_summary only when vuln_report #} {% if show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerability-summary" class="page_title">Summary of Vulnerabilities Identified</h3> {% if all_vulnerabilities.count > 0 %} <p>Listed below are the vulnerabilities identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-critical"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 50%;"> Vulnerability Name </div> <div class="cell grey-color" style="width: 19%;"> Times Identified </div> <div class="cell grey-color" style="width: 15%"> Severity </div> </div> {% for vulnerability in unique_vulnerabilities %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 50%"> <a href="#vuln_{{vulnerability.name.split|join:'_'}}">{{vulnerability.name}}</a> </div> <div class="cell" style="float: right; width: 19%;"> {{vulnerability.count}} </div> {% if vulnerability.severity == -1 %} <div class="cell bg-grey" style="width: 15%"> <span class="severity-title-box">Unknown</span> {% elif vulnerability.severity == 0 %} <div class="cell bg-info" style="width: 15%"> <span class="severity-title-box">Informational</span> {% elif vulnerability.severity == 1 %} <div class="cell bg-low" style="width: 15%"> <span class="severity-title-box">Low</span> {% elif vulnerability.severity == 2 %} <div class="cell bg-medium" style="width: 15%"> <span class="severity-title-box">Medium</span> {% elif vulnerability.severity == 3 %} <div class="cell bg-high" style="width: 15%"> <span class="severity-title-box">High</span> {% elif vulnerability.severity == 4 %} <div class="cell bg-critical" style="width: 15%"> <span class="severity-title-box">Critical</span> {% endif %} </div> </div> {% endfor %} {% else %} <h3 class='info-color'>No Vulnerabilities were Discovered.</h3> {% endif %} </div> </article> {% endif %} {# show discovered assets only for show_recon report #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 id="reconnaissance-results" class="page_title">Discovered Assets</h3> <h4 class="subheading">Subdomains</h4> <p> During the reconnaissance phase, {{scan_object.get_subdomain_count}} subdomains were discovered. Out of {{scan_object.get_subdomain_count}} subdomains, {{subdomain_alive_count}} returned HTTP status 200. {{interesting_subdomains.count}} interesting subdomains were also identified based on the interesting keywords used. </p> <h4>{{scan_object.get_subdomain_count}} subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> Subdomain </div> <div class="cell grey-color" style="width: 38%"> Page title </div> <div class="cell grey-color" style="width: 18%"> HTTP Status </div> </div> {% for subdomain in subdomains %} <div class="row"> <div class="cell" style="width: 38%"> {{subdomain.name}} </div> <div class="cell" style="width: 38%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% endif %} </div> <div class="cell" style="width: 18%"> {{subdomain.http_status}} </div> </div> {% endfor %} </div> {% if ip_addresses.count %} <h4 class="subheading" style="margin-top: 10px;">IP Addresses</h4> <h4>{{ip_addresses.count}} IP Addresses were identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> IP </div> <div class="cell grey-color" style="width: 38%"> Open Ports </div> <div class="cell grey-color" style="width: 18%"> Remarks </div> </div> {% for ip in ip_addresses %} <div class="row"> <div class="cell" style="width: 38%"> {{ip.address}} </div> <div class="cell" style="width: 38%"> {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </div> {% if ip.is_cdn %} <div class="cell medium" style="width: 18%"> CDN IP Address {% else %} <div class="cell" style="width: 18%"> {% endif %} </div> </div> {% endfor %} </div> {% endif %} </article> <br> {% endif %} {# reconnaissance finding only when show_recon #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 class="page_title">Reconnaissance Findings</h3> {% for subdomain in subdomains %} <table class="table" cellspacing="0" style="border-collapse: collapse;"> <tr> <td style="width: 2%" class="cell table-border">{{ forloop.counter }}.</td> <td style="width: 80%" class="cell table-border">{{subdomain.name}}</td> {% if subdomain.http_status == 200 %} <td style="width: 10%" class="cell table-border bg-success">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 300 and subdomain.http_status < 400 %} <td style="width: 10%" class="cell table-border bg-medium">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 400 %} <td style="width: 10%" class="cell table-border bg-high">{{subdomain.http_status}}</td> {% elif subdomain.http_status == 0 %} <td style="width: 10%" class="cell table-border">N/A</td> {% else %} <td style="width: 10%" class="cell table-border">{{subdomain.http_status}}</td> {% endif %} </tr> {% if subdomain.page_title %} <tr> <td colspan="3" class="cell table-border"><strong>Page Title: </strong>{{subdomain.page_title}}</td> </tr> {% endif %} {% if subdomain.ip_addresses.all %} <tr> <td colspan="3" class="cell table-border"> IP Address: <ul> {% for ip in subdomain.ip_addresses.all %} <li>{{ip.address}} {% if ip.ports.all %} <ul> <li>Open Ports: &nbsp; {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </li> </ul> {% endif %} </li> {% endfor %} </ul> </td> </tr> {% endif %} {% if subdomain.get_vulnerabilities_without_info %} <tr> <td colspan="3" class="cell table-border"> Vulnerabilities {% regroup subdomain.get_vulnerabilities_without_info by name as vuln_list %} <ul> {% for vulnerability in vuln_list %} <li> <a href="#vuln_{{vulnerability.list.0.name.split|join:'_'}}">{{ vulnerability.grouper }}</a> </li> {% endfor %} </ul> </td> </tr> {% endif %} </table> {% endfor %} </article> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerabilities-discovered" class="page_title">Vulnerabilities Discovered</h3> <p> This section reports the security issues found during the audit. <br> A Total of {{scan_object.get_vulnerability_count}} were discovered in {{scan_object.domain.name}}, {{scan_object.get_critical_vulnerability_count}} of them were Critical, {{scan_object.get_high_vulnerability_count}} of them were High Severity, {{scan_object.get_medium_vulnerability_count}} of them were Medium severity, {% if is_ignore_info_vuln %}0{% else %}{{scan_object.get_info_vulnerability_count}}{% endif %} of them were Low severity, and {{scan_object.get_info_vulnerability_count}} of them were Informational. {{scan_object.get_unknown_vulnerability_count}} of them were Unknown Severity. </p> <h4 class="subheading">Vulnerability Breakdown by Severity</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {% if is_ignore_info_vuln %} 0 {% else %} {{scan_object.get_info_vulnerability_count}} {% endif %} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {# start vulnerability #} {% if show_vuln %} <article class=""> {% regroup all_vulnerabilities by get_path as grouped_vulnerabilities %} {% for vulnerabilities in grouped_vulnerabilities %} {% for vulnerability in vulnerabilities.list %} <div> <h4 class="content-heading" id="vuln_{{vulnerability.name.split|join:'_'}}"> <span>{{vulnerability.name}} <br>in {{vulnerabilities.grouper}}</span> {% if vulnerability.severity == -1 %} <span style="float: right;" class="badge bg-grey">Unknown</span> <div class="grey-hr-line" ></div> {% elif vulnerability.severity == 0 %} <span style="float: right;" class="badge bg-info">INFO</span> <div class="info-hr-line" ></div> {% elif vulnerability.severity == 1 %} <span style="float: right;" class="badge bg-low">LOW</span> <div class="low-hr-line" ></div> {% elif vulnerability.severity == 2 %} <span style="float: right;" class="badge bg-medium">MEDIUM</span> <div class="medium-hr-line" ></div> {% elif vulnerability.severity == 3 %} <span style="float: right;" class="badge bg-high">HIGH</span> <div class="high-hr-line" ></div> {% elif vulnerability.severity == 4 %} <span style="float: right;" class="badge bg-critical">CRITICAL</span> <div class="critical-hr-line" ></div> {% endif %} </h4> <!-- show vulnerability classification --> <span class="mini-heading">Vulnerability Source: {{vulnerability.source|upper}}</span><br> {% if vulnerability.cvss_metrics or vulnerability.cvss_score or vulnerability.cve_ids.all or vulnerability.cve_ids.all %} <span class="mini-heading">Vulnerability Classification</span><br> {% if vulnerability.cvss_metrics %} <span class="mini-heading ml-8">CVSS Metrics: {{vulnerability.cvss_metrics}}</span> {% endif %} {% if vulnerability.cvss_score %} <br> <span class="mini-heading ml-8">CVSS Score:</span>&nbsp;<span class="high-color">{{vulnerability.cvss_score}}</span> {% endif %} {% if vulnerability.cve_ids.all %} <br> <span class="mini-heading ml-8">CVE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cve in vulnerability.cve_ids.all %} {{cve}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} {% if vulnerability.cwe_ids.all %} <br> <span class="mini-heading ml-8">CWE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cwe in vulnerability.cwe_ids.all %} {{cwe}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} <br> {% endif %} {% if vulnerability.description %} <br> <span class="mini-heading">Description</span><br> {{vulnerability.description|linebreaks}} {% endif %} {% if vulnerability.impact %} <br> <span class="mini-heading">Impact</span><br> {{vulnerability.impact|linebreaks}} {% endif %} {% if vulnerability.remediation %} <br> <span class="mini-heading">Remediation</span><br> {{vulnerability.remediation|linebreaks}} {% endif %} <br> <span class="mini-heading">Vulnerable URL(s)</span><br> <ul> <li class="text-blue"><a href="{{vulnerability.http_url}}" target="_blank" rel="noopener noreferrer">{{vulnerability.http_url}}</a></li> </ul> <!-- {% regroup vulnerability.list by http_url as vuln_http_url_list %} --> <!-- <ul> {% for vuln_urls in vuln_http_url_list %} <li>{{vuln_urls.grouper}}</li> <span class="mini-heading">Result/Findings</span><br> {% for vuln in vuln_urls.list %} {% if vuln.matcher_name %} {% if not forloop.first %} • {% endif %} {{vuln.matcher_name}} {% endif %} {% if vuln.extracted_results %} {% for res in vuln.extracted_results %} {% if not forloop.first %} • {% endif %} {{res}} {% endfor %} {% endif %} {% endfor %} {% endfor %} </ul> --> {% if vulnerability.references.all %} <span class="mini-heading">References</span><br> <ul> {% for ref in vulnerability.references.all %} <li> <span class="text-blue"><a href="{{ref}}" target="_blank" rel="noopener noreferrer">{{ref}}</a></span> </li> {% endfor %} </ul> {% endif %} <br> <br> </div> {% endfor %} {% endfor %} </article> {% endif %} {% endif %} <article id="chapter"> <h2 id="chapter-title">END OF REPORT</h2> </article> </body> </html>

psyray

4341d9834865240222a8dc72c01caaec0d7bed44

69231095782663fe0fe8b0e49b8aa995aa042723

## Potentially unsafe external link External links without noopener/noreferrer are a potential security risk. [Show more details](https://github.com/yogeshojha/rengine/security/code-scanning/172)

github-advanced-security[bot]

yogeshojha/rengine

Fix report generation when `Ignore Informational Vulnerabilities` checked

When **Ignore Informational Vulnerabilities** is checked there are still info vulns datas. I've reworked the queries that display vulnerabilities to prevent info vulns to display in the : - **Quick summary** Info blue box - **Reconnaissance Findings** - **Vulnerabilities Discovered** Info blue box I've also fixed the **Vulnerabilities Discovered** listing by doing a correct loop through regrouped values because values withe the same path but not the same severity does not display well Tested and working on current master branch

2023-12-05 01:25:41+00:00

2023-12-08 05:48:36+00:00

web/templates/report/template.html

<html> <head> <meta charset="utf-8"> <title>Report</title> <link href="https://fonts.googleapis.com/css2?family=Inter:wght@100;200;300;400;500&display=swap" rel="stylesheet"> <style> @page { size: A4; @top-left { background: {{primary_color}}; content: counter(page); height: 1cm; text-align: center; width: 1cm; } @top-center { background: {{primary_color}}; content: ''; display: block; height: .05cm; opacity: .5; width: 100%; } @top-right { content: string(heading); font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% if show_footer %} @bottom-left { content: "{{footer_text}}"; font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% endif %} } @page :blank { @top-left { background: none; content: '' } @top-center { content: none } @top-right { content: none } } @page no-chapter { @top-left { background: none; content: none } @top-center { content: none } @top-right { content: none } } @page :first { background-color: {{secondary_color}}; background-size: cover; margin: 0; } @page chapter { background: {{primary_color}}; margin: 0; @top-left { content: none } @top-center { content: none } @top-right { content: none } } html { color: #393939; font-family: 'Inter'; font-weight: 300; font-size: 11pt; font-weight: 300; line-height: 1.5; } h1 { font-family: 'Inter'; font-weight: 200; font-size: 38pt; margin: 5cm 2cm 0 2cm; page: no-chapter; width: 100%; line-height: normal; } h2, h3, h4 { font-family: 'Inter'; font-weight: 200; color: black; font-weight: 400; line-height: normal; } #cover { align-content: space-between; display: flex; flex-wrap: wrap; height: 297mm; } #cover-subheading { font-family: 'Inter'; font-weight: 200; font-size: 22pt; width: 100%; } #cover footer { background: {{primary_color}}; flex: 1 33%; margin: 0 -2cm; padding: 1cm 0; white-space: pre-wrap; } #cover footer:first-of-type { padding-left: 3cm; } #cover-line { margin-top: 6px; border-bottom: 1px double {{primary_color}}; } #summary { display: flex; flex-wrap: wrap; justify-content: space-between; } #contents { page: no-chapter; } #contents h2 { font-size: 20pt; Intereight: 400; margin-bottom: 3cm; } #contents h3 { font-weight: 400; margin: 3em 0 1em; } #contents h3::before { background: {{primary_color}}; content: ''; display: block; height: .08cm; margin-bottom: .25cm; width: 2cm; } #contents ul { list-style: none; padding-left: 0; } #contents ul li { border-top: .25pt solid #c1c1c1; margin: .25cm 0; padding-top: .25cm; } #contents ul li::before { color: {{primary_color}}; content: '• '; font-size: 30pt; line-height: 16pt; vertical-align: bottom; } #contents ul li a { color: inherit; text-decoration-line: inherit; } #contents ul li a::before { content: target-text(attr(href)); } #contents ul li a::after { color: {{primary_color}}; content: target-counter(attr(href), page); float: right; } #columns section { columns: 2; column-gap: 1cm; padding-top: 1cm; } #columns section p { text-align: justify; } #columns section p:first-of-type { font-weight: 700; } #chapter { align-items: center; display: flex; height: 297mm; justify-content: center; page: chapter; } #boxes { display: flex; flex-wrap: wrap; justify-content: space-between; } #boxes section h4 { margin-bottom: 0; } #boxes section p { background: {{primary_color}}; display: block; font-size: 15pt; margin-bottom: 0; padding: .25cm 0; text-align: center; height: 85px; color: #37474F; } .bg-critical { background: #EF9A9A !important; } .bg-high { background: #FFAB91 !important; } .bg-medium { background: #FFCC80 !important; } .bg-low { background: #FFE082 !important; } .bg-success { background-color: #A5D6A7 !important; } .bg-grey { background-color: #B0BEC5 !important; } .bg-info { background-color: #90CAF9 !important; } .critical-color { color: #EF9A9A; } .high-color { color: #dc3545; } .medium-color { color: #FFCC80; } .low-color { color: #FFE082; } .success-color { color: #A5D6A7; } .grey-color { color: #212121; } .info-color { color: #90CAF9; } .primary-color { color: {{primary_color}}; } .text-blue{ color: #007bff!important; } .badge { display: inline-block; padding-left: 12px; padding-right: 12px; text-align: center } .critical-hr-line { border-bottom: 3px solid #EF9A9A !important; } .high-hr-line { border-bottom: 3px solid #FFAB91 !important; } .medium-hr-line { border-bottom: 3px solid #FFCC80 !important; } .low-hr-line { border-bottom: 3px solid #FFE082 !important; } .info-hr-line { border-bottom: 3px solid #90CAF9 !important; } .grey-hr-line { border-bottom: 3px solid #212121 !important; } .inside-box-counter { font-size: 28pt; } .table { margin: 0 0 40px 0; width: 100%; box-shadow: 0 1px 3px rgba(0, 0, 0, 0.2); display: table; border-spacing: 0 0.4em; } .row { display: table-row; background: #f6f6f6; } .cell { padding: 6px 6px 6px 6px; display: table-cell; } .header { Intereight: 900; color: #ffffff; } .page_title{ font-weight: 300; font-size: 20pt; } .subheading{ font-weight: 300; font-size: 14pt; } .content-heading{ font-weight: 300; font-size: 12pt; } .mini-heading{ font-weight: 400; font-size: 11pt; } .table-border{ border-style:solid; border-width: 1px; border-color: #90CAF9 !important; } a{ color: #007bff; text-decoration: none; } .ml-8{ margin-left: 8px; } </style> </head> <body> <article id="cover"> <h1 style="color:{{primary_color}}">{{report_name}} <br> {{scan_object.domain.name}} <div id="cover-line"></div> {# generated date #} <span id="cover-subheading">{% now "F j, Y" %}</span> </h1> <footer> {{company_name}} {{company_address}} </footer> <footer> {{company_email}} {{company_website}} </footer> <footer> {% if show_rengine_banner %}Generated by reNgine https://github.com/yogeshojha/rengine {% endif %} </footer> </article> <article id="contents"> <h2>&nbsp;</h2> <h3>Table of contents</h3> <ul> {% if show_executive_summary %} <li><a href="#executive-summary"></a></li> {% endif %} <li><a href="#quick-summary"></a></li> <li><a href="#assessment-timeline"></a></li> {% if interesting_subdomains and show_recon %} <li><a href="#interesting-recon-data"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerability-summary"></a></li> {% endif %} {% if show_recon %} <li><a href="#reconnaissance-results"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerabilities-discovered"></a></li> {% endif %} </ul> </article> {% if show_executive_summary %} <article id="summary" style="page-break-before: always"> <h2 id="executive-summary" class="page_title">Executive summary</h2> <br> {{executive_summary_description | safe }} </article> {% endif %} <article id="summary" style="page-break-before: always"> <h2 id="quick-summary" class="page_title">Quick Summary</h2> <p>This section contains quick summary of scan performed on <span class="primary-color">{{scan_object.domain.name}}</span></p> <br> </article> {# recon section #} {% if show_recon %} <h4 id="reconnaissance-summary" class="subheading">Reconnaissance</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-success">Subdomains <br> <span class="inside-box-counter"> {{scan_object.get_subdomain_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Endpoints <br> <span class="inside-box-counter"> {{scan_object.get_endpoint_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-critical">Vulnerabilities <br> <span class="inside-box-counter"> {{scan_object.get_vulnerability_count}} </span> </p> </section> </div> {% endif %} <!-- vulnerability section, hide if only recon report --> {% if show_vuln %} <article> <br> <h4 id="vulnerability-summary" class="subheading">Vulnerability Summary</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width:30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {{scan_object.get_info_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {% endif %} <article> <h3 id="assessment-timeline" class="page_title">Timeline of the Assessment</h3> <p> Scan started on: {{scan_object.start_scan_date|date:"F j, Y h:i"}} <br> Total time taken: {% if scan_object.scan_status == 0 %} {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% elif scan_object.scan_status == 1 %} {{ scan_object.get_elapsed_time }} {% elif scan_object.scan_status == 2 %} {% if scan_object.get_completed_time_in_sec < 60 %} Completed in < 1 minutes {% else %} Completed in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} {% elif scan_object.scan_status == 3 %} Aborted in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} <br> Report Generated on: {% now "F j, Y" %} </p> </article> {# show interesting_subdomains section only when show_recon result is there #} {% if interesting_subdomains and show_recon %} <article style="page-break-before: always" class="summary"> <h3 id="interesting-recon-data" class="page_title">Interesting Recon Data</h3> <p>Listed below are the {{interesting_subdomains.count}} interesting subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-success"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 33%"> Subdomain </div> <div class="cell grey-color" style="width: 33%"> Page title </div> <div class="cell grey-color" style="width: 15%"> HTTP Status </div> </div> {% for subdomain in interesting_subdomains %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 35%"> {{subdomain.name}} </div> <div class="cell" style="width: 35%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> <div class="cell" style="width: 15%;"> {% if subdomain.http_status %} {{subdomain.http_status}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> </div> {% endfor %} </div> </article> {% endif %} {# vulnerability_summary only when vuln_report #} {% if show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerability-summary" class="page_title">Summary of Vulnerabilities Identified</h3> {% if all_vulnerabilities.count > 0 %} <p>Listed below are the vulnerabilities identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-critical"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 50%;"> Vulnerability Name </div> <div class="cell grey-color" style="width: 19%;"> Times Identified </div> <div class="cell grey-color" style="width: 15%"> Severity </div> </div> {% for vulnerability in unique_vulnerabilities %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 50%"> <a href="#vuln_{{vulnerability.name.split|join:'_'}}">{{vulnerability.name}}</a> </div> <div class="cell" style="float: right; width: 19%;"> {{vulnerability.count}} </div> {% if vulnerability.severity == -1 %} <div class="cell bg-grey" style="width: 15%"> <span class="severity-title-box">Unknown</span> {% elif vulnerability.severity == 0 %} <div class="cell bg-info" style="width: 15%"> <span class="severity-title-box">Informational</span> {% elif vulnerability.severity == 1 %} <div class="cell bg-low" style="width: 15%"> <span class="severity-title-box">Low</span> {% elif vulnerability.severity == 2 %} <div class="cell bg-medium" style="width: 15%"> <span class="severity-title-box">Medium</span> {% elif vulnerability.severity == 3 %} <div class="cell bg-high" style="width: 15%"> <span class="severity-title-box">High</span> {% elif vulnerability.severity == 4 %} <div class="cell bg-critical" style="width: 15%"> <span class="severity-title-box">Critical</span> {% endif %} </div> </div> {% endfor %} {% else %} <h3 class='info-color'>No Vulnerabilities were Discovered.</h3> {% endif %} </div> </article> {% endif %} {# show discovered assets only for show_recon report #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 id="reconnaissance-results" class="page_title">Discovered Assets</h3> <h4 class="subheading">Subdomains</h4> <p> During the reconnaissance phase, {{scan_object.get_subdomain_count}} subdomains were discovered. Out of {{scan_object.get_subdomain_count}} subdomains, {{subdomain_alive_count}} returned HTTP status 200. {{interesting_subdomains.count}} interesting subdomains were also identified based on the interesting keywords used. </p> <h4>{{scan_object.get_subdomain_count}} subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> Subdomain </div> <div class="cell grey-color" style="width: 38%"> Page title </div> <div class="cell grey-color" style="width: 18%"> HTTP Status </div> </div> {% for subdomain in subdomains %} <div class="row"> <div class="cell" style="width: 38%"> {{subdomain.name}} </div> <div class="cell" style="width: 38%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% endif %} </div> <div class="cell" style="width: 18%"> {{subdomain.http_status}} </div> </div> {% endfor %} </div> {% if ip_addresses.count %} <h4 class="subheading" style="margin-top: 10px;">IP Addresses</h4> <h4>{{ip_addresses.count}} IP Addresses were identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> IP </div> <div class="cell grey-color" style="width: 38%"> Open Ports </div> <div class="cell grey-color" style="width: 18%"> Remarks </div> </div> {% for ip in ip_addresses %} <div class="row"> <div class="cell" style="width: 38%"> {{ip.address}} </div> <div class="cell" style="width: 38%"> {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </div> {% if ip.is_cdn %} <div class="cell medium" style="width: 18%"> CDN IP Address {% else %} <div class="cell" style="width: 18%"> {% endif %} </div> </div> {% endfor %} </div> {% endif %} </article> <br> {% endif %} {# reconnaissance finding only when show_recon #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 class="page_title">Reconnaissance Findings</h3> {% for subdomain in subdomains %} <table class="table" cellspacing="0" style="border-collapse: collapse;"> <tr> <td style="width: 2%" class="cell table-border">{{ forloop.counter }}.</td> <td style="width: 80%" class="cell table-border">{{subdomain.name}}</td> {% if subdomain.http_status == 200 %} <td style="width: 10%" class="cell table-border bg-success">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 300 and subdomain.http_status < 400 %} <td style="width: 10%" class="cell table-border bg-medium">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 400 %} <td style="width: 10%" class="cell table-border bg-high">{{subdomain.http_status}}</td> {% elif subdomain.http_status == 0 %} <td style="width: 10%" class="cell table-border">N/A</td> {% else %} <td style="width: 10%" class="cell table-border">{{subdomain.http_status}}</td> {% endif %} </tr> {% if subdomain.page_title %} <tr> <td colspan="3" class="cell table-border"><strong>Page Title: </strong>{{subdomain.page_title}}</td> </tr> {% endif %} {% if subdomain.ip_addresses.all %} <tr> <td colspan="3" class="cell table-border"> IP Address: <ul> {% for ip in subdomain.ip_addresses.all %} <li>{{ip.address}} {% if ip.ports.all %} <ul> <li>Open Ports: &nbsp; {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </li> </ul> {% endif %} </li> {% endfor %} </ul> </td> </tr> {% endif %} {% if subdomain.get_vulnerabilities %} <tr> <td colspan="3" class="cell table-border"> Vulnerabilities {% regroup subdomain.get_vulnerabilities by name as vuln_list %} <ul> {% for vulnerability in vuln_list %} <li> <a href="#vuln_{{vulnerability.list.0.name.split|join:'_'}}">{{ vulnerability.grouper }}</a> </li> {% endfor %} </ul> </td> </tr> {% endif %} </table> {% endfor %} </article> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerabilities-discovered" class="page_title">Vulnerabilities Discovered</h3> <p> This section reports the security issues found during the audit. <br> A Total of {{scan_object.get_vulnerability_count}} were discovered in {{scan_object.domain.name}}, {{scan_object.get_critical_vulnerability_count}} of them were Critical, {{scan_object.get_high_vulnerability_count}} of them were High Severity, {{scan_object.get_medium_vulnerability_count}} of them were Medium severity, {{scan_object.get_low_vulnerability_count}} of them were Low severity, and {{scan_object.get_info_vulnerability_count}} of them were Informational. {{scan_object.get_unknown_vulnerability_count}} of them were Unknown Severity. </p> <h4 class="subheading">Vulnerability Breakdown by Severity</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {{scan_object.get_info_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {# start vulnerability #} {% if show_vuln %} <article class=""> {% regroup all_vulnerabilities by get_path as grouped_vulnerabilities %} {% for vulnerability in grouped_vulnerabilities %} <div> <h4 class="content-heading" id="vuln_{{vulnerability.list.0.name.split|join:'_'}}"> <span>{{vulnerability.list.0.name}} <br>in {{vulnerability.grouper}}</span> {% if vulnerability.list.0.severity == -1 %} <span style="float: right;" class="badge bg-grey">Unknown</span> <div class="grey-hr-line" ></div> {% elif vulnerability.list.0.severity == 0 %} <span style="float: right;" class="badge bg-info">INFO</span> <div class="info-hr-line" ></div> {% elif vulnerability.list.0.severity == 1 %} <span style="float: right;" class="badge bg-low">LOW</span> <div class="low-hr-line" ></div> {% elif vulnerability.list.0.severity == 2 %} <span style="float: right;" class="badge bg-medium">MEDIUM</span> <div class="medium-hr-line" ></div> {% elif vulnerability.list.0.severity == 3 %} <span style="float: right;" class="badge bg-high">HIGH</span> <div class="high-hr-line" ></div> {% elif vulnerability.list.0.severity == 4 %} <span style="float: right;" class="badge bg-critical">CRITICAL</span> <div class="critical-hr-line" ></div> {% endif %} </h4> <!-- show vulnerability classification --> <span class="mini-heading">Vulnerability Source: {{vulnerability.list.0.source|upper}}</span><br> {% if vulnerability.list.0.cvss_metrics or vulnerability.list.0.cvss_score or vulnerability.list.0.cve_ids.all or vulnerability.list.0.cve_ids.all %} <span class="mini-heading">Vulnerability Classification</span><br> {% if vulnerability.list.0.cvss_metrics %} <span class="mini-heading ml-8">CVSS Metrics: {{vulnerability.list.0.cvss_metrics}}</span> {% endif %} {% if vulnerability.list.0.cvss_score %} <br> <span class="mini-heading ml-8">CVSS Score:</span>&nbsp;<span class="high-color">{{vulnerability.list.0.cvss_score}}</span> {% endif %} {% if vulnerability.list.0.cve_ids.all %} <br> <span class="mini-heading ml-8">CVE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cve in vulnerability.list.0.cve_ids.all %} {{cve}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} {% if vulnerability.list.0.cwe_ids.all %} <br> <span class="mini-heading ml-8">CWE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cwe in vulnerability.list.0.cwe_ids.all %} {{cwe}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} <br> {% endif %} {% if vulnerability.list.0.description %} <br> <span class="mini-heading">Description</span><br> {{vulnerability.list.0.description|linebreaks}} {% endif %} {% if vulnerability.list.0.impact %} <br> <span class="mini-heading">Impact</span><br> {{vulnerability.list.0.impact|linebreaks}} {% endif %} {% if vulnerability.list.0.remediation %} <br> <span class="mini-heading">Remediation</span><br> {{vulnerability.list.0.remediation|linebreaks}} {% endif %} <br> <span class="mini-heading">Vulnerable URL(s)</span><br> <ul> {% for vuln in vulnerability.list %} <li class="text-blue">{{vuln.http_url}}</li> {% endfor %} </ul> <!-- {% regroup vulnerability.list by http_url as vuln_http_url_list %} --> <!-- <ul> {% for vuln_urls in vuln_http_url_list %} <li>{{vuln_urls.grouper}}</li> <span class="mini-heading">Result/Findings</span><br> {% for vuln in vuln_urls.list %} {% if vuln.matcher_name %} {% if not forloop.first %} • {% endif %} {{vuln.matcher_name}} {% endif %} {% if vuln.extracted_results %} {% for res in vuln.extracted_results %} {% if not forloop.first %} • {% endif %} {{res}} {% endfor %} {% endif %} {% endfor %} {% endfor %} </ul> --> {% if vulnerability.list.0.references.all %} <span class="mini-heading">References</span><br> <ul> {% for ref in vulnerability.list.0.references.all %} <li> <span class="text-blue"> {{ref}} </span> </li> {% endfor %} </ul> {% endif %} <br> <br> </div> {% endfor %} </article> {% endif %} {% endif %} <article id="chapter"> <h2 id="chapter-title">END OF REPORT</h2> </article> </body> </html>

<html> <head> <meta charset="utf-8"> <title>Report</title> <link href="https://fonts.googleapis.com/css2?family=Inter:wght@100;200;300;400;500&display=swap" rel="stylesheet"> <style> @page { size: A4; @top-left { background: {{primary_color}}; content: counter(page); height: 1cm; text-align: center; width: 1cm; } @top-center { background: {{primary_color}}; content: ''; display: block; height: .05cm; opacity: .5; width: 100%; } @top-right { content: string(heading); font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% if show_footer %} @bottom-left { content: "{{footer_text}}"; font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% endif %} } @page :blank { @top-left { background: none; content: '' } @top-center { content: none } @top-right { content: none } } @page no-chapter { @top-left { background: none; content: none } @top-center { content: none } @top-right { content: none } } @page :first { background-color: {{secondary_color}}; background-size: cover; margin: 0; } @page chapter { background: {{primary_color}}; margin: 0; @top-left { content: none } @top-center { content: none } @top-right { content: none } } html { color: #393939; font-family: 'Inter'; font-weight: 300; font-size: 11pt; font-weight: 300; line-height: 1.5; } h1 { font-family: 'Inter'; font-weight: 200; font-size: 38pt; margin: 5cm 2cm 0 2cm; page: no-chapter; width: 100%; line-height: normal; } h2, h3, h4 { font-family: 'Inter'; font-weight: 200; color: black; font-weight: 400; line-height: normal; } #cover { align-content: space-between; display: flex; flex-wrap: wrap; height: 297mm; } #cover-subheading { font-family: 'Inter'; font-weight: 200; font-size: 22pt; width: 100%; } #cover footer { background: {{primary_color}}; flex: 1 33%; margin: 0 -2cm; padding: 1cm 0; white-space: pre-wrap; } #cover footer:first-of-type { padding-left: 3cm; } #cover-line { margin-top: 6px; border-bottom: 1px double {{primary_color}}; } #summary { display: flex; flex-wrap: wrap; justify-content: space-between; } #contents { page: no-chapter; } #contents h2 { font-size: 20pt; Intereight: 400; margin-bottom: 3cm; } #contents h3 { font-weight: 400; margin: 3em 0 1em; } #contents h3::before { background: {{primary_color}}; content: ''; display: block; height: .08cm; margin-bottom: .25cm; width: 2cm; } #contents ul { list-style: none; padding-left: 0; } #contents ul li { border-top: .25pt solid #c1c1c1; margin: .25cm 0; padding-top: .25cm; } #contents ul li::before { color: {{primary_color}}; content: '• '; font-size: 30pt; line-height: 16pt; vertical-align: bottom; } #contents ul li a { color: inherit; text-decoration-line: inherit; } #contents ul li a::before { content: target-text(attr(href)); } #contents ul li a::after { color: {{primary_color}}; content: target-counter(attr(href), page); float: right; } #columns section { columns: 2; column-gap: 1cm; padding-top: 1cm; } #columns section p { text-align: justify; } #columns section p:first-of-type { font-weight: 700; } #chapter { align-items: center; display: flex; height: 297mm; justify-content: center; page: chapter; } #boxes { display: flex; flex-wrap: wrap; justify-content: space-between; } #boxes section h4 { margin-bottom: 0; } #boxes section p { background: {{primary_color}}; display: block; font-size: 15pt; margin-bottom: 0; padding: .25cm 0; text-align: center; height: 85px; color: #37474F; } .bg-critical { background: #EF9A9A !important; } .bg-high { background: #FFAB91 !important; } .bg-medium { background: #FFCC80 !important; } .bg-low { background: #FFE082 !important; } .bg-success { background-color: #A5D6A7 !important; } .bg-grey { background-color: #B0BEC5 !important; } .bg-info { background-color: #90CAF9 !important; } .critical-color { color: #EF9A9A; } .high-color { color: #dc3545; } .medium-color { color: #FFCC80; } .low-color { color: #FFE082; } .success-color { color: #A5D6A7; } .grey-color { color: #212121; } .info-color { color: #90CAF9; } .primary-color { color: {{primary_color}}; } .text-blue{ color: #007bff!important; } .badge { display: inline-block; padding-left: 12px; padding-right: 12px; text-align: center } .critical-hr-line { border-bottom: 3px solid #EF9A9A !important; } .high-hr-line { border-bottom: 3px solid #FFAB91 !important; } .medium-hr-line { border-bottom: 3px solid #FFCC80 !important; } .low-hr-line { border-bottom: 3px solid #FFE082 !important; } .info-hr-line { border-bottom: 3px solid #90CAF9 !important; } .grey-hr-line { border-bottom: 3px solid #212121 !important; } .inside-box-counter { font-size: 28pt; } .table { margin: 0 0 40px 0; width: 100%; box-shadow: 0 1px 3px rgba(0, 0, 0, 0.2); display: table; border-spacing: 0 0.4em; } .row { display: table-row; background: #f6f6f6; } .cell { padding: 6px 6px 6px 6px; display: table-cell; } .header { Intereight: 900; color: #ffffff; } .page_title{ font-weight: 300; font-size: 20pt; } .subheading{ font-weight: 300; font-size: 14pt; } .content-heading{ font-weight: 300; font-size: 12pt; } .mini-heading{ font-weight: 400; font-size: 11pt; } .table-border{ border-style:solid; border-width: 1px; border-color: #90CAF9 !important; } a{ color: #007bff; text-decoration: none; } .ml-8{ margin-left: 8px; } </style> </head> <body> <article id="cover"> <h1 style="color:{{primary_color}}">{{report_name}} <br> {{scan_object.domain.name}} <div id="cover-line"></div> {# generated date #} <span id="cover-subheading">{% now "F j, Y" %}</span> </h1> <footer> {{company_name}} {{company_address}} </footer> <footer> {{company_email}} {{company_website}} </footer> <footer> {% if show_rengine_banner %}Generated by reNgine https://github.com/yogeshojha/rengine {% endif %} </footer> </article> <article id="contents"> <h2>&nbsp;</h2> <h3>Table of contents</h3> <ul> {% if show_executive_summary %} <li><a href="#executive-summary"></a></li> {% endif %} <li><a href="#quick-summary"></a></li> <li><a href="#assessment-timeline"></a></li> {% if interesting_subdomains and show_recon %} <li><a href="#interesting-recon-data"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerability-summary"></a></li> {% endif %} {% if show_recon %} <li><a href="#reconnaissance-results"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerabilities-discovered"></a></li> {% endif %} </ul> </article> {% if show_executive_summary %} <article id="summary" style="page-break-before: always"> <h2 id="executive-summary" class="page_title">Executive summary</h2> <br> {{executive_summary_description | safe }} </article> {% endif %} <article id="summary" style="page-break-before: always"> <h2 id="quick-summary" class="page_title">Quick Summary</h2> <p>This section contains quick summary of scan performed on <span class="primary-color">{{scan_object.domain.name}}</span></p> <br> </article> {# recon section #} {% if show_recon %} <h4 id="reconnaissance-summary" class="subheading">Reconnaissance</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-success">Subdomains <br> <span class="inside-box-counter"> {{scan_object.get_subdomain_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Endpoints <br> <span class="inside-box-counter"> {{scan_object.get_endpoint_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-critical">Vulnerabilities <br> <span class="inside-box-counter"> {{all_vulnerabilities_count}} </span> </p> </section> </div> {% endif %} <!-- vulnerability section, hide if only recon report --> {% if show_vuln %} <article> <br> <h4 id="vulnerability-summary" class="subheading">Vulnerability Summary</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width:30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {% if is_ignore_info_vuln %} 0 {% else %} {{scan_object.get_info_vulnerability_count}} {% endif %} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {% endif %} <article> <h3 id="assessment-timeline" class="page_title">Timeline of the Assessment</h3> <p> Scan started on: {{scan_object.start_scan_date|date:"F j, Y h:i"}} <br> Total time taken: {% if scan_object.scan_status == 0 %} {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% elif scan_object.scan_status == 1 %} {{ scan_object.get_elapsed_time }} {% elif scan_object.scan_status == 2 %} {% if scan_object.get_completed_time_in_sec < 60 %} Completed in < 1 minutes {% else %} Completed in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} {% elif scan_object.scan_status == 3 %} Aborted in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} <br> Report Generated on: {% now "F j, Y" %} </p> </article> {# show interesting_subdomains section only when show_recon result is there #} {% if interesting_subdomains and show_recon %} <article style="page-break-before: always" class="summary"> <h3 id="interesting-recon-data" class="page_title">Interesting Recon Data</h3> <p>Listed below are the {{interesting_subdomains.count}} interesting subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-success"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 33%"> Subdomain </div> <div class="cell grey-color" style="width: 33%"> Page title </div> <div class="cell grey-color" style="width: 15%"> HTTP Status </div> </div> {% for subdomain in interesting_subdomains %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 35%"> {{subdomain.name}} </div> <div class="cell" style="width: 35%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> <div class="cell" style="width: 15%;"> {% if subdomain.http_status %} {{subdomain.http_status}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> </div> {% endfor %} </div> </article> {% endif %} {# vulnerability_summary only when vuln_report #} {% if show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerability-summary" class="page_title">Summary of Vulnerabilities Identified</h3> {% if all_vulnerabilities.count > 0 %} <p>Listed below are the vulnerabilities identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-critical"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 50%;"> Vulnerability Name </div> <div class="cell grey-color" style="width: 19%;"> Times Identified </div> <div class="cell grey-color" style="width: 15%"> Severity </div> </div> {% for vulnerability in unique_vulnerabilities %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 50%"> <a href="#vuln_{{vulnerability.name.split|join:'_'}}">{{vulnerability.name}}</a> </div> <div class="cell" style="float: right; width: 19%;"> {{vulnerability.count}} </div> {% if vulnerability.severity == -1 %} <div class="cell bg-grey" style="width: 15%"> <span class="severity-title-box">Unknown</span> {% elif vulnerability.severity == 0 %} <div class="cell bg-info" style="width: 15%"> <span class="severity-title-box">Informational</span> {% elif vulnerability.severity == 1 %} <div class="cell bg-low" style="width: 15%"> <span class="severity-title-box">Low</span> {% elif vulnerability.severity == 2 %} <div class="cell bg-medium" style="width: 15%"> <span class="severity-title-box">Medium</span> {% elif vulnerability.severity == 3 %} <div class="cell bg-high" style="width: 15%"> <span class="severity-title-box">High</span> {% elif vulnerability.severity == 4 %} <div class="cell bg-critical" style="width: 15%"> <span class="severity-title-box">Critical</span> {% endif %} </div> </div> {% endfor %} {% else %} <h3 class='info-color'>No Vulnerabilities were Discovered.</h3> {% endif %} </div> </article> {% endif %} {# show discovered assets only for show_recon report #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 id="reconnaissance-results" class="page_title">Discovered Assets</h3> <h4 class="subheading">Subdomains</h4> <p> During the reconnaissance phase, {{scan_object.get_subdomain_count}} subdomains were discovered. Out of {{scan_object.get_subdomain_count}} subdomains, {{subdomain_alive_count}} returned HTTP status 200. {{interesting_subdomains.count}} interesting subdomains were also identified based on the interesting keywords used. </p> <h4>{{scan_object.get_subdomain_count}} subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> Subdomain </div> <div class="cell grey-color" style="width: 38%"> Page title </div> <div class="cell grey-color" style="width: 18%"> HTTP Status </div> </div> {% for subdomain in subdomains %} <div class="row"> <div class="cell" style="width: 38%"> {{subdomain.name}} </div> <div class="cell" style="width: 38%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% endif %} </div> <div class="cell" style="width: 18%"> {{subdomain.http_status}} </div> </div> {% endfor %} </div> {% if ip_addresses.count %} <h4 class="subheading" style="margin-top: 10px;">IP Addresses</h4> <h4>{{ip_addresses.count}} IP Addresses were identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> IP </div> <div class="cell grey-color" style="width: 38%"> Open Ports </div> <div class="cell grey-color" style="width: 18%"> Remarks </div> </div> {% for ip in ip_addresses %} <div class="row"> <div class="cell" style="width: 38%"> {{ip.address}} </div> <div class="cell" style="width: 38%"> {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </div> {% if ip.is_cdn %} <div class="cell medium" style="width: 18%"> CDN IP Address {% else %} <div class="cell" style="width: 18%"> {% endif %} </div> </div> {% endfor %} </div> {% endif %} </article> <br> {% endif %} {# reconnaissance finding only when show_recon #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 class="page_title">Reconnaissance Findings</h3> {% for subdomain in subdomains %} <table class="table" cellspacing="0" style="border-collapse: collapse;"> <tr> <td style="width: 2%" class="cell table-border">{{ forloop.counter }}.</td> <td style="width: 80%" class="cell table-border">{{subdomain.name}}</td> {% if subdomain.http_status == 200 %} <td style="width: 10%" class="cell table-border bg-success">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 300 and subdomain.http_status < 400 %} <td style="width: 10%" class="cell table-border bg-medium">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 400 %} <td style="width: 10%" class="cell table-border bg-high">{{subdomain.http_status}}</td> {% elif subdomain.http_status == 0 %} <td style="width: 10%" class="cell table-border">N/A</td> {% else %} <td style="width: 10%" class="cell table-border">{{subdomain.http_status}}</td> {% endif %} </tr> {% if subdomain.page_title %} <tr> <td colspan="3" class="cell table-border"><strong>Page Title: </strong>{{subdomain.page_title}}</td> </tr> {% endif %} {% if subdomain.ip_addresses.all %} <tr> <td colspan="3" class="cell table-border"> IP Address: <ul> {% for ip in subdomain.ip_addresses.all %} <li>{{ip.address}} {% if ip.ports.all %} <ul> <li>Open Ports: &nbsp; {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </li> </ul> {% endif %} </li> {% endfor %} </ul> </td> </tr> {% endif %} {% if subdomain.get_vulnerabilities_without_info %} <tr> <td colspan="3" class="cell table-border"> Vulnerabilities {% regroup subdomain.get_vulnerabilities_without_info by name as vuln_list %} <ul> {% for vulnerability in vuln_list %} <li> <a href="#vuln_{{vulnerability.list.0.name.split|join:'_'}}">{{ vulnerability.grouper }}</a> </li> {% endfor %} </ul> </td> </tr> {% endif %} </table> {% endfor %} </article> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerabilities-discovered" class="page_title">Vulnerabilities Discovered</h3> <p> This section reports the security issues found during the audit. <br> A Total of {{scan_object.get_vulnerability_count}} were discovered in {{scan_object.domain.name}}, {{scan_object.get_critical_vulnerability_count}} of them were Critical, {{scan_object.get_high_vulnerability_count}} of them were High Severity, {{scan_object.get_medium_vulnerability_count}} of them were Medium severity, {% if is_ignore_info_vuln %}0{% else %}{{scan_object.get_info_vulnerability_count}}{% endif %} of them were Low severity, and {{scan_object.get_info_vulnerability_count}} of them were Informational. {{scan_object.get_unknown_vulnerability_count}} of them were Unknown Severity. </p> <h4 class="subheading">Vulnerability Breakdown by Severity</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {% if is_ignore_info_vuln %} 0 {% else %} {{scan_object.get_info_vulnerability_count}} {% endif %} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {# start vulnerability #} {% if show_vuln %} <article class=""> {% regroup all_vulnerabilities by get_path as grouped_vulnerabilities %} {% for vulnerabilities in grouped_vulnerabilities %} {% for vulnerability in vulnerabilities.list %} <div> <h4 class="content-heading" id="vuln_{{vulnerability.name.split|join:'_'}}"> <span>{{vulnerability.name}} <br>in {{vulnerabilities.grouper}}</span> {% if vulnerability.severity == -1 %} <span style="float: right;" class="badge bg-grey">Unknown</span> <div class="grey-hr-line" ></div> {% elif vulnerability.severity == 0 %} <span style="float: right;" class="badge bg-info">INFO</span> <div class="info-hr-line" ></div> {% elif vulnerability.severity == 1 %} <span style="float: right;" class="badge bg-low">LOW</span> <div class="low-hr-line" ></div> {% elif vulnerability.severity == 2 %} <span style="float: right;" class="badge bg-medium">MEDIUM</span> <div class="medium-hr-line" ></div> {% elif vulnerability.severity == 3 %} <span style="float: right;" class="badge bg-high">HIGH</span> <div class="high-hr-line" ></div> {% elif vulnerability.severity == 4 %} <span style="float: right;" class="badge bg-critical">CRITICAL</span> <div class="critical-hr-line" ></div> {% endif %} </h4> <!-- show vulnerability classification --> <span class="mini-heading">Vulnerability Source: {{vulnerability.source|upper}}</span><br> {% if vulnerability.cvss_metrics or vulnerability.cvss_score or vulnerability.cve_ids.all or vulnerability.cve_ids.all %} <span class="mini-heading">Vulnerability Classification</span><br> {% if vulnerability.cvss_metrics %} <span class="mini-heading ml-8">CVSS Metrics: {{vulnerability.cvss_metrics}}</span> {% endif %} {% if vulnerability.cvss_score %} <br> <span class="mini-heading ml-8">CVSS Score:</span>&nbsp;<span class="high-color">{{vulnerability.cvss_score}}</span> {% endif %} {% if vulnerability.cve_ids.all %} <br> <span class="mini-heading ml-8">CVE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cve in vulnerability.cve_ids.all %} {{cve}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} {% if vulnerability.cwe_ids.all %} <br> <span class="mini-heading ml-8">CWE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cwe in vulnerability.cwe_ids.all %} {{cwe}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} <br> {% endif %} {% if vulnerability.description %} <br> <span class="mini-heading">Description</span><br> {{vulnerability.description|linebreaks}} {% endif %} {% if vulnerability.impact %} <br> <span class="mini-heading">Impact</span><br> {{vulnerability.impact|linebreaks}} {% endif %} {% if vulnerability.remediation %} <br> <span class="mini-heading">Remediation</span><br> {{vulnerability.remediation|linebreaks}} {% endif %} <br> <span class="mini-heading">Vulnerable URL(s)</span><br> <ul> <li class="text-blue"><a href="{{vulnerability.http_url}}" target="_blank" rel="noopener noreferrer">{{vulnerability.http_url}}</a></li> </ul> <!-- {% regroup vulnerability.list by http_url as vuln_http_url_list %} --> <!-- <ul> {% for vuln_urls in vuln_http_url_list %} <li>{{vuln_urls.grouper}}</li> <span class="mini-heading">Result/Findings</span><br> {% for vuln in vuln_urls.list %} {% if vuln.matcher_name %} {% if not forloop.first %} • {% endif %} {{vuln.matcher_name}} {% endif %} {% if vuln.extracted_results %} {% for res in vuln.extracted_results %} {% if not forloop.first %} • {% endif %} {{res}} {% endfor %} {% endif %} {% endfor %} {% endfor %} </ul> --> {% if vulnerability.references.all %} <span class="mini-heading">References</span><br> <ul> {% for ref in vulnerability.references.all %} <li> <span class="text-blue"><a href="{{ref}}" target="_blank" rel="noopener noreferrer">{{ref}}</a></span> </li> {% endfor %} </ul> {% endif %} <br> <br> </div> {% endfor %} {% endfor %} </article> {% endif %} {% endif %} <article id="chapter"> <h2 id="chapter-title">END OF REPORT</h2> </article> </body> </html>

psyray

4341d9834865240222a8dc72c01caaec0d7bed44

69231095782663fe0fe8b0e49b8aa995aa042723

Fixed

psyray

yogeshojha/rengine

Fix report generation when `Ignore Informational Vulnerabilities` checked

When **Ignore Informational Vulnerabilities** is checked there are still info vulns datas. I've reworked the queries that display vulnerabilities to prevent info vulns to display in the : - **Quick summary** Info blue box - **Reconnaissance Findings** - **Vulnerabilities Discovered** Info blue box I've also fixed the **Vulnerabilities Discovered** listing by doing a correct loop through regrouped values because values withe the same path but not the same severity does not display well Tested and working on current master branch

2023-12-05 01:25:41+00:00

2023-12-08 05:48:36+00:00

web/templates/report/template.html

<html> <head> <meta charset="utf-8"> <title>Report</title> <link href="https://fonts.googleapis.com/css2?family=Inter:wght@100;200;300;400;500&display=swap" rel="stylesheet"> <style> @page { size: A4; @top-left { background: {{primary_color}}; content: counter(page); height: 1cm; text-align: center; width: 1cm; } @top-center { background: {{primary_color}}; content: ''; display: block; height: .05cm; opacity: .5; width: 100%; } @top-right { content: string(heading); font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% if show_footer %} @bottom-left { content: "{{footer_text}}"; font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% endif %} } @page :blank { @top-left { background: none; content: '' } @top-center { content: none } @top-right { content: none } } @page no-chapter { @top-left { background: none; content: none } @top-center { content: none } @top-right { content: none } } @page :first { background-color: {{secondary_color}}; background-size: cover; margin: 0; } @page chapter { background: {{primary_color}}; margin: 0; @top-left { content: none } @top-center { content: none } @top-right { content: none } } html { color: #393939; font-family: 'Inter'; font-weight: 300; font-size: 11pt; font-weight: 300; line-height: 1.5; } h1 { font-family: 'Inter'; font-weight: 200; font-size: 38pt; margin: 5cm 2cm 0 2cm; page: no-chapter; width: 100%; line-height: normal; } h2, h3, h4 { font-family: 'Inter'; font-weight: 200; color: black; font-weight: 400; line-height: normal; } #cover { align-content: space-between; display: flex; flex-wrap: wrap; height: 297mm; } #cover-subheading { font-family: 'Inter'; font-weight: 200; font-size: 22pt; width: 100%; } #cover footer { background: {{primary_color}}; flex: 1 33%; margin: 0 -2cm; padding: 1cm 0; white-space: pre-wrap; } #cover footer:first-of-type { padding-left: 3cm; } #cover-line { margin-top: 6px; border-bottom: 1px double {{primary_color}}; } #summary { display: flex; flex-wrap: wrap; justify-content: space-between; } #contents { page: no-chapter; } #contents h2 { font-size: 20pt; Intereight: 400; margin-bottom: 3cm; } #contents h3 { font-weight: 400; margin: 3em 0 1em; } #contents h3::before { background: {{primary_color}}; content: ''; display: block; height: .08cm; margin-bottom: .25cm; width: 2cm; } #contents ul { list-style: none; padding-left: 0; } #contents ul li { border-top: .25pt solid #c1c1c1; margin: .25cm 0; padding-top: .25cm; } #contents ul li::before { color: {{primary_color}}; content: '• '; font-size: 30pt; line-height: 16pt; vertical-align: bottom; } #contents ul li a { color: inherit; text-decoration-line: inherit; } #contents ul li a::before { content: target-text(attr(href)); } #contents ul li a::after { color: {{primary_color}}; content: target-counter(attr(href), page); float: right; } #columns section { columns: 2; column-gap: 1cm; padding-top: 1cm; } #columns section p { text-align: justify; } #columns section p:first-of-type { font-weight: 700; } #chapter { align-items: center; display: flex; height: 297mm; justify-content: center; page: chapter; } #boxes { display: flex; flex-wrap: wrap; justify-content: space-between; } #boxes section h4 { margin-bottom: 0; } #boxes section p { background: {{primary_color}}; display: block; font-size: 15pt; margin-bottom: 0; padding: .25cm 0; text-align: center; height: 85px; color: #37474F; } .bg-critical { background: #EF9A9A !important; } .bg-high { background: #FFAB91 !important; } .bg-medium { background: #FFCC80 !important; } .bg-low { background: #FFE082 !important; } .bg-success { background-color: #A5D6A7 !important; } .bg-grey { background-color: #B0BEC5 !important; } .bg-info { background-color: #90CAF9 !important; } .critical-color { color: #EF9A9A; } .high-color { color: #dc3545; } .medium-color { color: #FFCC80; } .low-color { color: #FFE082; } .success-color { color: #A5D6A7; } .grey-color { color: #212121; } .info-color { color: #90CAF9; } .primary-color { color: {{primary_color}}; } .text-blue{ color: #007bff!important; } .badge { display: inline-block; padding-left: 12px; padding-right: 12px; text-align: center } .critical-hr-line { border-bottom: 3px solid #EF9A9A !important; } .high-hr-line { border-bottom: 3px solid #FFAB91 !important; } .medium-hr-line { border-bottom: 3px solid #FFCC80 !important; } .low-hr-line { border-bottom: 3px solid #FFE082 !important; } .info-hr-line { border-bottom: 3px solid #90CAF9 !important; } .grey-hr-line { border-bottom: 3px solid #212121 !important; } .inside-box-counter { font-size: 28pt; } .table { margin: 0 0 40px 0; width: 100%; box-shadow: 0 1px 3px rgba(0, 0, 0, 0.2); display: table; border-spacing: 0 0.4em; } .row { display: table-row; background: #f6f6f6; } .cell { padding: 6px 6px 6px 6px; display: table-cell; } .header { Intereight: 900; color: #ffffff; } .page_title{ font-weight: 300; font-size: 20pt; } .subheading{ font-weight: 300; font-size: 14pt; } .content-heading{ font-weight: 300; font-size: 12pt; } .mini-heading{ font-weight: 400; font-size: 11pt; } .table-border{ border-style:solid; border-width: 1px; border-color: #90CAF9 !important; } a{ color: #007bff; text-decoration: none; } .ml-8{ margin-left: 8px; } </style> </head> <body> <article id="cover"> <h1 style="color:{{primary_color}}">{{report_name}} <br> {{scan_object.domain.name}} <div id="cover-line"></div> {# generated date #} <span id="cover-subheading">{% now "F j, Y" %}</span> </h1> <footer> {{company_name}} {{company_address}} </footer> <footer> {{company_email}} {{company_website}} </footer> <footer> {% if show_rengine_banner %}Generated by reNgine https://github.com/yogeshojha/rengine {% endif %} </footer> </article> <article id="contents"> <h2>&nbsp;</h2> <h3>Table of contents</h3> <ul> {% if show_executive_summary %} <li><a href="#executive-summary"></a></li> {% endif %} <li><a href="#quick-summary"></a></li> <li><a href="#assessment-timeline"></a></li> {% if interesting_subdomains and show_recon %} <li><a href="#interesting-recon-data"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerability-summary"></a></li> {% endif %} {% if show_recon %} <li><a href="#reconnaissance-results"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerabilities-discovered"></a></li> {% endif %} </ul> </article> {% if show_executive_summary %} <article id="summary" style="page-break-before: always"> <h2 id="executive-summary" class="page_title">Executive summary</h2> <br> {{executive_summary_description | safe }} </article> {% endif %} <article id="summary" style="page-break-before: always"> <h2 id="quick-summary" class="page_title">Quick Summary</h2> <p>This section contains quick summary of scan performed on <span class="primary-color">{{scan_object.domain.name}}</span></p> <br> </article> {# recon section #} {% if show_recon %} <h4 id="reconnaissance-summary" class="subheading">Reconnaissance</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-success">Subdomains <br> <span class="inside-box-counter"> {{scan_object.get_subdomain_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Endpoints <br> <span class="inside-box-counter"> {{scan_object.get_endpoint_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-critical">Vulnerabilities <br> <span class="inside-box-counter"> {{scan_object.get_vulnerability_count}} </span> </p> </section> </div> {% endif %} <!-- vulnerability section, hide if only recon report --> {% if show_vuln %} <article> <br> <h4 id="vulnerability-summary" class="subheading">Vulnerability Summary</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width:30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {{scan_object.get_info_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {% endif %} <article> <h3 id="assessment-timeline" class="page_title">Timeline of the Assessment</h3> <p> Scan started on: {{scan_object.start_scan_date|date:"F j, Y h:i"}} <br> Total time taken: {% if scan_object.scan_status == 0 %} {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% elif scan_object.scan_status == 1 %} {{ scan_object.get_elapsed_time }} {% elif scan_object.scan_status == 2 %} {% if scan_object.get_completed_time_in_sec < 60 %} Completed in < 1 minutes {% else %} Completed in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} {% elif scan_object.scan_status == 3 %} Aborted in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} <br> Report Generated on: {% now "F j, Y" %} </p> </article> {# show interesting_subdomains section only when show_recon result is there #} {% if interesting_subdomains and show_recon %} <article style="page-break-before: always" class="summary"> <h3 id="interesting-recon-data" class="page_title">Interesting Recon Data</h3> <p>Listed below are the {{interesting_subdomains.count}} interesting subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-success"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 33%"> Subdomain </div> <div class="cell grey-color" style="width: 33%"> Page title </div> <div class="cell grey-color" style="width: 15%"> HTTP Status </div> </div> {% for subdomain in interesting_subdomains %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 35%"> {{subdomain.name}} </div> <div class="cell" style="width: 35%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> <div class="cell" style="width: 15%;"> {% if subdomain.http_status %} {{subdomain.http_status}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> </div> {% endfor %} </div> </article> {% endif %} {# vulnerability_summary only when vuln_report #} {% if show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerability-summary" class="page_title">Summary of Vulnerabilities Identified</h3> {% if all_vulnerabilities.count > 0 %} <p>Listed below are the vulnerabilities identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-critical"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 50%;"> Vulnerability Name </div> <div class="cell grey-color" style="width: 19%;"> Times Identified </div> <div class="cell grey-color" style="width: 15%"> Severity </div> </div> {% for vulnerability in unique_vulnerabilities %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 50%"> <a href="#vuln_{{vulnerability.name.split|join:'_'}}">{{vulnerability.name}}</a> </div> <div class="cell" style="float: right; width: 19%;"> {{vulnerability.count}} </div> {% if vulnerability.severity == -1 %} <div class="cell bg-grey" style="width: 15%"> <span class="severity-title-box">Unknown</span> {% elif vulnerability.severity == 0 %} <div class="cell bg-info" style="width: 15%"> <span class="severity-title-box">Informational</span> {% elif vulnerability.severity == 1 %} <div class="cell bg-low" style="width: 15%"> <span class="severity-title-box">Low</span> {% elif vulnerability.severity == 2 %} <div class="cell bg-medium" style="width: 15%"> <span class="severity-title-box">Medium</span> {% elif vulnerability.severity == 3 %} <div class="cell bg-high" style="width: 15%"> <span class="severity-title-box">High</span> {% elif vulnerability.severity == 4 %} <div class="cell bg-critical" style="width: 15%"> <span class="severity-title-box">Critical</span> {% endif %} </div> </div> {% endfor %} {% else %} <h3 class='info-color'>No Vulnerabilities were Discovered.</h3> {% endif %} </div> </article> {% endif %} {# show discovered assets only for show_recon report #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 id="reconnaissance-results" class="page_title">Discovered Assets</h3> <h4 class="subheading">Subdomains</h4> <p> During the reconnaissance phase, {{scan_object.get_subdomain_count}} subdomains were discovered. Out of {{scan_object.get_subdomain_count}} subdomains, {{subdomain_alive_count}} returned HTTP status 200. {{interesting_subdomains.count}} interesting subdomains were also identified based on the interesting keywords used. </p> <h4>{{scan_object.get_subdomain_count}} subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> Subdomain </div> <div class="cell grey-color" style="width: 38%"> Page title </div> <div class="cell grey-color" style="width: 18%"> HTTP Status </div> </div> {% for subdomain in subdomains %} <div class="row"> <div class="cell" style="width: 38%"> {{subdomain.name}} </div> <div class="cell" style="width: 38%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% endif %} </div> <div class="cell" style="width: 18%"> {{subdomain.http_status}} </div> </div> {% endfor %} </div> {% if ip_addresses.count %} <h4 class="subheading" style="margin-top: 10px;">IP Addresses</h4> <h4>{{ip_addresses.count}} IP Addresses were identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> IP </div> <div class="cell grey-color" style="width: 38%"> Open Ports </div> <div class="cell grey-color" style="width: 18%"> Remarks </div> </div> {% for ip in ip_addresses %} <div class="row"> <div class="cell" style="width: 38%"> {{ip.address}} </div> <div class="cell" style="width: 38%"> {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </div> {% if ip.is_cdn %} <div class="cell medium" style="width: 18%"> CDN IP Address {% else %} <div class="cell" style="width: 18%"> {% endif %} </div> </div> {% endfor %} </div> {% endif %} </article> <br> {% endif %} {# reconnaissance finding only when show_recon #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 class="page_title">Reconnaissance Findings</h3> {% for subdomain in subdomains %} <table class="table" cellspacing="0" style="border-collapse: collapse;"> <tr> <td style="width: 2%" class="cell table-border">{{ forloop.counter }}.</td> <td style="width: 80%" class="cell table-border">{{subdomain.name}}</td> {% if subdomain.http_status == 200 %} <td style="width: 10%" class="cell table-border bg-success">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 300 and subdomain.http_status < 400 %} <td style="width: 10%" class="cell table-border bg-medium">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 400 %} <td style="width: 10%" class="cell table-border bg-high">{{subdomain.http_status}}</td> {% elif subdomain.http_status == 0 %} <td style="width: 10%" class="cell table-border">N/A</td> {% else %} <td style="width: 10%" class="cell table-border">{{subdomain.http_status}}</td> {% endif %} </tr> {% if subdomain.page_title %} <tr> <td colspan="3" class="cell table-border"><strong>Page Title: </strong>{{subdomain.page_title}}</td> </tr> {% endif %} {% if subdomain.ip_addresses.all %} <tr> <td colspan="3" class="cell table-border"> IP Address: <ul> {% for ip in subdomain.ip_addresses.all %} <li>{{ip.address}} {% if ip.ports.all %} <ul> <li>Open Ports: &nbsp; {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </li> </ul> {% endif %} </li> {% endfor %} </ul> </td> </tr> {% endif %} {% if subdomain.get_vulnerabilities %} <tr> <td colspan="3" class="cell table-border"> Vulnerabilities {% regroup subdomain.get_vulnerabilities by name as vuln_list %} <ul> {% for vulnerability in vuln_list %} <li> <a href="#vuln_{{vulnerability.list.0.name.split|join:'_'}}">{{ vulnerability.grouper }}</a> </li> {% endfor %} </ul> </td> </tr> {% endif %} </table> {% endfor %} </article> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerabilities-discovered" class="page_title">Vulnerabilities Discovered</h3> <p> This section reports the security issues found during the audit. <br> A Total of {{scan_object.get_vulnerability_count}} were discovered in {{scan_object.domain.name}}, {{scan_object.get_critical_vulnerability_count}} of them were Critical, {{scan_object.get_high_vulnerability_count}} of them were High Severity, {{scan_object.get_medium_vulnerability_count}} of them were Medium severity, {{scan_object.get_low_vulnerability_count}} of them were Low severity, and {{scan_object.get_info_vulnerability_count}} of them were Informational. {{scan_object.get_unknown_vulnerability_count}} of them were Unknown Severity. </p> <h4 class="subheading">Vulnerability Breakdown by Severity</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {{scan_object.get_info_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {# start vulnerability #} {% if show_vuln %} <article class=""> {% regroup all_vulnerabilities by get_path as grouped_vulnerabilities %} {% for vulnerability in grouped_vulnerabilities %} <div> <h4 class="content-heading" id="vuln_{{vulnerability.list.0.name.split|join:'_'}}"> <span>{{vulnerability.list.0.name}} <br>in {{vulnerability.grouper}}</span> {% if vulnerability.list.0.severity == -1 %} <span style="float: right;" class="badge bg-grey">Unknown</span> <div class="grey-hr-line" ></div> {% elif vulnerability.list.0.severity == 0 %} <span style="float: right;" class="badge bg-info">INFO</span> <div class="info-hr-line" ></div> {% elif vulnerability.list.0.severity == 1 %} <span style="float: right;" class="badge bg-low">LOW</span> <div class="low-hr-line" ></div> {% elif vulnerability.list.0.severity == 2 %} <span style="float: right;" class="badge bg-medium">MEDIUM</span> <div class="medium-hr-line" ></div> {% elif vulnerability.list.0.severity == 3 %} <span style="float: right;" class="badge bg-high">HIGH</span> <div class="high-hr-line" ></div> {% elif vulnerability.list.0.severity == 4 %} <span style="float: right;" class="badge bg-critical">CRITICAL</span> <div class="critical-hr-line" ></div> {% endif %} </h4> <!-- show vulnerability classification --> <span class="mini-heading">Vulnerability Source: {{vulnerability.list.0.source|upper}}</span><br> {% if vulnerability.list.0.cvss_metrics or vulnerability.list.0.cvss_score or vulnerability.list.0.cve_ids.all or vulnerability.list.0.cve_ids.all %} <span class="mini-heading">Vulnerability Classification</span><br> {% if vulnerability.list.0.cvss_metrics %} <span class="mini-heading ml-8">CVSS Metrics: {{vulnerability.list.0.cvss_metrics}}</span> {% endif %} {% if vulnerability.list.0.cvss_score %} <br> <span class="mini-heading ml-8">CVSS Score:</span>&nbsp;<span class="high-color">{{vulnerability.list.0.cvss_score}}</span> {% endif %} {% if vulnerability.list.0.cve_ids.all %} <br> <span class="mini-heading ml-8">CVE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cve in vulnerability.list.0.cve_ids.all %} {{cve}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} {% if vulnerability.list.0.cwe_ids.all %} <br> <span class="mini-heading ml-8">CWE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cwe in vulnerability.list.0.cwe_ids.all %} {{cwe}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} <br> {% endif %} {% if vulnerability.list.0.description %} <br> <span class="mini-heading">Description</span><br> {{vulnerability.list.0.description|linebreaks}} {% endif %} {% if vulnerability.list.0.impact %} <br> <span class="mini-heading">Impact</span><br> {{vulnerability.list.0.impact|linebreaks}} {% endif %} {% if vulnerability.list.0.remediation %} <br> <span class="mini-heading">Remediation</span><br> {{vulnerability.list.0.remediation|linebreaks}} {% endif %} <br> <span class="mini-heading">Vulnerable URL(s)</span><br> <ul> {% for vuln in vulnerability.list %} <li class="text-blue">{{vuln.http_url}}</li> {% endfor %} </ul> <!-- {% regroup vulnerability.list by http_url as vuln_http_url_list %} --> <!-- <ul> {% for vuln_urls in vuln_http_url_list %} <li>{{vuln_urls.grouper}}</li> <span class="mini-heading">Result/Findings</span><br> {% for vuln in vuln_urls.list %} {% if vuln.matcher_name %} {% if not forloop.first %} • {% endif %} {{vuln.matcher_name}} {% endif %} {% if vuln.extracted_results %} {% for res in vuln.extracted_results %} {% if not forloop.first %} • {% endif %} {{res}} {% endfor %} {% endif %} {% endfor %} {% endfor %} </ul> --> {% if vulnerability.list.0.references.all %} <span class="mini-heading">References</span><br> <ul> {% for ref in vulnerability.list.0.references.all %} <li> <span class="text-blue"> {{ref}} </span> </li> {% endfor %} </ul> {% endif %} <br> <br> </div> {% endfor %} </article> {% endif %} {% endif %} <article id="chapter"> <h2 id="chapter-title">END OF REPORT</h2> </article> </body> </html>

<html> <head> <meta charset="utf-8"> <title>Report</title> <link href="https://fonts.googleapis.com/css2?family=Inter:wght@100;200;300;400;500&display=swap" rel="stylesheet"> <style> @page { size: A4; @top-left { background: {{primary_color}}; content: counter(page); height: 1cm; text-align: center; width: 1cm; } @top-center { background: {{primary_color}}; content: ''; display: block; height: .05cm; opacity: .5; width: 100%; } @top-right { content: string(heading); font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% if show_footer %} @bottom-left { content: "{{footer_text}}"; font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% endif %} } @page :blank { @top-left { background: none; content: '' } @top-center { content: none } @top-right { content: none } } @page no-chapter { @top-left { background: none; content: none } @top-center { content: none } @top-right { content: none } } @page :first { background-color: {{secondary_color}}; background-size: cover; margin: 0; } @page chapter { background: {{primary_color}}; margin: 0; @top-left { content: none } @top-center { content: none } @top-right { content: none } } html { color: #393939; font-family: 'Inter'; font-weight: 300; font-size: 11pt; font-weight: 300; line-height: 1.5; } h1 { font-family: 'Inter'; font-weight: 200; font-size: 38pt; margin: 5cm 2cm 0 2cm; page: no-chapter; width: 100%; line-height: normal; } h2, h3, h4 { font-family: 'Inter'; font-weight: 200; color: black; font-weight: 400; line-height: normal; } #cover { align-content: space-between; display: flex; flex-wrap: wrap; height: 297mm; } #cover-subheading { font-family: 'Inter'; font-weight: 200; font-size: 22pt; width: 100%; } #cover footer { background: {{primary_color}}; flex: 1 33%; margin: 0 -2cm; padding: 1cm 0; white-space: pre-wrap; } #cover footer:first-of-type { padding-left: 3cm; } #cover-line { margin-top: 6px; border-bottom: 1px double {{primary_color}}; } #summary { display: flex; flex-wrap: wrap; justify-content: space-between; } #contents { page: no-chapter; } #contents h2 { font-size: 20pt; Intereight: 400; margin-bottom: 3cm; } #contents h3 { font-weight: 400; margin: 3em 0 1em; } #contents h3::before { background: {{primary_color}}; content: ''; display: block; height: .08cm; margin-bottom: .25cm; width: 2cm; } #contents ul { list-style: none; padding-left: 0; } #contents ul li { border-top: .25pt solid #c1c1c1; margin: .25cm 0; padding-top: .25cm; } #contents ul li::before { color: {{primary_color}}; content: '• '; font-size: 30pt; line-height: 16pt; vertical-align: bottom; } #contents ul li a { color: inherit; text-decoration-line: inherit; } #contents ul li a::before { content: target-text(attr(href)); } #contents ul li a::after { color: {{primary_color}}; content: target-counter(attr(href), page); float: right; } #columns section { columns: 2; column-gap: 1cm; padding-top: 1cm; } #columns section p { text-align: justify; } #columns section p:first-of-type { font-weight: 700; } #chapter { align-items: center; display: flex; height: 297mm; justify-content: center; page: chapter; } #boxes { display: flex; flex-wrap: wrap; justify-content: space-between; } #boxes section h4 { margin-bottom: 0; } #boxes section p { background: {{primary_color}}; display: block; font-size: 15pt; margin-bottom: 0; padding: .25cm 0; text-align: center; height: 85px; color: #37474F; } .bg-critical { background: #EF9A9A !important; } .bg-high { background: #FFAB91 !important; } .bg-medium { background: #FFCC80 !important; } .bg-low { background: #FFE082 !important; } .bg-success { background-color: #A5D6A7 !important; } .bg-grey { background-color: #B0BEC5 !important; } .bg-info { background-color: #90CAF9 !important; } .critical-color { color: #EF9A9A; } .high-color { color: #dc3545; } .medium-color { color: #FFCC80; } .low-color { color: #FFE082; } .success-color { color: #A5D6A7; } .grey-color { color: #212121; } .info-color { color: #90CAF9; } .primary-color { color: {{primary_color}}; } .text-blue{ color: #007bff!important; } .badge { display: inline-block; padding-left: 12px; padding-right: 12px; text-align: center } .critical-hr-line { border-bottom: 3px solid #EF9A9A !important; } .high-hr-line { border-bottom: 3px solid #FFAB91 !important; } .medium-hr-line { border-bottom: 3px solid #FFCC80 !important; } .low-hr-line { border-bottom: 3px solid #FFE082 !important; } .info-hr-line { border-bottom: 3px solid #90CAF9 !important; } .grey-hr-line { border-bottom: 3px solid #212121 !important; } .inside-box-counter { font-size: 28pt; } .table { margin: 0 0 40px 0; width: 100%; box-shadow: 0 1px 3px rgba(0, 0, 0, 0.2); display: table; border-spacing: 0 0.4em; } .row { display: table-row; background: #f6f6f6; } .cell { padding: 6px 6px 6px 6px; display: table-cell; } .header { Intereight: 900; color: #ffffff; } .page_title{ font-weight: 300; font-size: 20pt; } .subheading{ font-weight: 300; font-size: 14pt; } .content-heading{ font-weight: 300; font-size: 12pt; } .mini-heading{ font-weight: 400; font-size: 11pt; } .table-border{ border-style:solid; border-width: 1px; border-color: #90CAF9 !important; } a{ color: #007bff; text-decoration: none; } .ml-8{ margin-left: 8px; } </style> </head> <body> <article id="cover"> <h1 style="color:{{primary_color}}">{{report_name}} <br> {{scan_object.domain.name}} <div id="cover-line"></div> {# generated date #} <span id="cover-subheading">{% now "F j, Y" %}</span> </h1> <footer> {{company_name}} {{company_address}} </footer> <footer> {{company_email}} {{company_website}} </footer> <footer> {% if show_rengine_banner %}Generated by reNgine https://github.com/yogeshojha/rengine {% endif %} </footer> </article> <article id="contents"> <h2>&nbsp;</h2> <h3>Table of contents</h3> <ul> {% if show_executive_summary %} <li><a href="#executive-summary"></a></li> {% endif %} <li><a href="#quick-summary"></a></li> <li><a href="#assessment-timeline"></a></li> {% if interesting_subdomains and show_recon %} <li><a href="#interesting-recon-data"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerability-summary"></a></li> {% endif %} {% if show_recon %} <li><a href="#reconnaissance-results"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerabilities-discovered"></a></li> {% endif %} </ul> </article> {% if show_executive_summary %} <article id="summary" style="page-break-before: always"> <h2 id="executive-summary" class="page_title">Executive summary</h2> <br> {{executive_summary_description | safe }} </article> {% endif %} <article id="summary" style="page-break-before: always"> <h2 id="quick-summary" class="page_title">Quick Summary</h2> <p>This section contains quick summary of scan performed on <span class="primary-color">{{scan_object.domain.name}}</span></p> <br> </article> {# recon section #} {% if show_recon %} <h4 id="reconnaissance-summary" class="subheading">Reconnaissance</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-success">Subdomains <br> <span class="inside-box-counter"> {{scan_object.get_subdomain_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Endpoints <br> <span class="inside-box-counter"> {{scan_object.get_endpoint_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-critical">Vulnerabilities <br> <span class="inside-box-counter"> {{all_vulnerabilities_count}} </span> </p> </section> </div> {% endif %} <!-- vulnerability section, hide if only recon report --> {% if show_vuln %} <article> <br> <h4 id="vulnerability-summary" class="subheading">Vulnerability Summary</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width:30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {% if is_ignore_info_vuln %} 0 {% else %} {{scan_object.get_info_vulnerability_count}} {% endif %} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {% endif %} <article> <h3 id="assessment-timeline" class="page_title">Timeline of the Assessment</h3> <p> Scan started on: {{scan_object.start_scan_date|date:"F j, Y h:i"}} <br> Total time taken: {% if scan_object.scan_status == 0 %} {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% elif scan_object.scan_status == 1 %} {{ scan_object.get_elapsed_time }} {% elif scan_object.scan_status == 2 %} {% if scan_object.get_completed_time_in_sec < 60 %} Completed in < 1 minutes {% else %} Completed in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} {% elif scan_object.scan_status == 3 %} Aborted in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} <br> Report Generated on: {% now "F j, Y" %} </p> </article> {# show interesting_subdomains section only when show_recon result is there #} {% if interesting_subdomains and show_recon %} <article style="page-break-before: always" class="summary"> <h3 id="interesting-recon-data" class="page_title">Interesting Recon Data</h3> <p>Listed below are the {{interesting_subdomains.count}} interesting subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-success"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 33%"> Subdomain </div> <div class="cell grey-color" style="width: 33%"> Page title </div> <div class="cell grey-color" style="width: 15%"> HTTP Status </div> </div> {% for subdomain in interesting_subdomains %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 35%"> {{subdomain.name}} </div> <div class="cell" style="width: 35%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> <div class="cell" style="width: 15%;"> {% if subdomain.http_status %} {{subdomain.http_status}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> </div> {% endfor %} </div> </article> {% endif %} {# vulnerability_summary only when vuln_report #} {% if show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerability-summary" class="page_title">Summary of Vulnerabilities Identified</h3> {% if all_vulnerabilities.count > 0 %} <p>Listed below are the vulnerabilities identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-critical"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 50%;"> Vulnerability Name </div> <div class="cell grey-color" style="width: 19%;"> Times Identified </div> <div class="cell grey-color" style="width: 15%"> Severity </div> </div> {% for vulnerability in unique_vulnerabilities %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 50%"> <a href="#vuln_{{vulnerability.name.split|join:'_'}}">{{vulnerability.name}}</a> </div> <div class="cell" style="float: right; width: 19%;"> {{vulnerability.count}} </div> {% if vulnerability.severity == -1 %} <div class="cell bg-grey" style="width: 15%"> <span class="severity-title-box">Unknown</span> {% elif vulnerability.severity == 0 %} <div class="cell bg-info" style="width: 15%"> <span class="severity-title-box">Informational</span> {% elif vulnerability.severity == 1 %} <div class="cell bg-low" style="width: 15%"> <span class="severity-title-box">Low</span> {% elif vulnerability.severity == 2 %} <div class="cell bg-medium" style="width: 15%"> <span class="severity-title-box">Medium</span> {% elif vulnerability.severity == 3 %} <div class="cell bg-high" style="width: 15%"> <span class="severity-title-box">High</span> {% elif vulnerability.severity == 4 %} <div class="cell bg-critical" style="width: 15%"> <span class="severity-title-box">Critical</span> {% endif %} </div> </div> {% endfor %} {% else %} <h3 class='info-color'>No Vulnerabilities were Discovered.</h3> {% endif %} </div> </article> {% endif %} {# show discovered assets only for show_recon report #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 id="reconnaissance-results" class="page_title">Discovered Assets</h3> <h4 class="subheading">Subdomains</h4> <p> During the reconnaissance phase, {{scan_object.get_subdomain_count}} subdomains were discovered. Out of {{scan_object.get_subdomain_count}} subdomains, {{subdomain_alive_count}} returned HTTP status 200. {{interesting_subdomains.count}} interesting subdomains were also identified based on the interesting keywords used. </p> <h4>{{scan_object.get_subdomain_count}} subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> Subdomain </div> <div class="cell grey-color" style="width: 38%"> Page title </div> <div class="cell grey-color" style="width: 18%"> HTTP Status </div> </div> {% for subdomain in subdomains %} <div class="row"> <div class="cell" style="width: 38%"> {{subdomain.name}} </div> <div class="cell" style="width: 38%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% endif %} </div> <div class="cell" style="width: 18%"> {{subdomain.http_status}} </div> </div> {% endfor %} </div> {% if ip_addresses.count %} <h4 class="subheading" style="margin-top: 10px;">IP Addresses</h4> <h4>{{ip_addresses.count}} IP Addresses were identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> IP </div> <div class="cell grey-color" style="width: 38%"> Open Ports </div> <div class="cell grey-color" style="width: 18%"> Remarks </div> </div> {% for ip in ip_addresses %} <div class="row"> <div class="cell" style="width: 38%"> {{ip.address}} </div> <div class="cell" style="width: 38%"> {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </div> {% if ip.is_cdn %} <div class="cell medium" style="width: 18%"> CDN IP Address {% else %} <div class="cell" style="width: 18%"> {% endif %} </div> </div> {% endfor %} </div> {% endif %} </article> <br> {% endif %} {# reconnaissance finding only when show_recon #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 class="page_title">Reconnaissance Findings</h3> {% for subdomain in subdomains %} <table class="table" cellspacing="0" style="border-collapse: collapse;"> <tr> <td style="width: 2%" class="cell table-border">{{ forloop.counter }}.</td> <td style="width: 80%" class="cell table-border">{{subdomain.name}}</td> {% if subdomain.http_status == 200 %} <td style="width: 10%" class="cell table-border bg-success">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 300 and subdomain.http_status < 400 %} <td style="width: 10%" class="cell table-border bg-medium">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 400 %} <td style="width: 10%" class="cell table-border bg-high">{{subdomain.http_status}}</td> {% elif subdomain.http_status == 0 %} <td style="width: 10%" class="cell table-border">N/A</td> {% else %} <td style="width: 10%" class="cell table-border">{{subdomain.http_status}}</td> {% endif %} </tr> {% if subdomain.page_title %} <tr> <td colspan="3" class="cell table-border"><strong>Page Title: </strong>{{subdomain.page_title}}</td> </tr> {% endif %} {% if subdomain.ip_addresses.all %} <tr> <td colspan="3" class="cell table-border"> IP Address: <ul> {% for ip in subdomain.ip_addresses.all %} <li>{{ip.address}} {% if ip.ports.all %} <ul> <li>Open Ports: &nbsp; {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </li> </ul> {% endif %} </li> {% endfor %} </ul> </td> </tr> {% endif %} {% if subdomain.get_vulnerabilities_without_info %} <tr> <td colspan="3" class="cell table-border"> Vulnerabilities {% regroup subdomain.get_vulnerabilities_without_info by name as vuln_list %} <ul> {% for vulnerability in vuln_list %} <li> <a href="#vuln_{{vulnerability.list.0.name.split|join:'_'}}">{{ vulnerability.grouper }}</a> </li> {% endfor %} </ul> </td> </tr> {% endif %} </table> {% endfor %} </article> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerabilities-discovered" class="page_title">Vulnerabilities Discovered</h3> <p> This section reports the security issues found during the audit. <br> A Total of {{scan_object.get_vulnerability_count}} were discovered in {{scan_object.domain.name}}, {{scan_object.get_critical_vulnerability_count}} of them were Critical, {{scan_object.get_high_vulnerability_count}} of them were High Severity, {{scan_object.get_medium_vulnerability_count}} of them were Medium severity, {% if is_ignore_info_vuln %}0{% else %}{{scan_object.get_info_vulnerability_count}}{% endif %} of them were Low severity, and {{scan_object.get_info_vulnerability_count}} of them were Informational. {{scan_object.get_unknown_vulnerability_count}} of them were Unknown Severity. </p> <h4 class="subheading">Vulnerability Breakdown by Severity</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {% if is_ignore_info_vuln %} 0 {% else %} {{scan_object.get_info_vulnerability_count}} {% endif %} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {# start vulnerability #} {% if show_vuln %} <article class=""> {% regroup all_vulnerabilities by get_path as grouped_vulnerabilities %} {% for vulnerabilities in grouped_vulnerabilities %} {% for vulnerability in vulnerabilities.list %} <div> <h4 class="content-heading" id="vuln_{{vulnerability.name.split|join:'_'}}"> <span>{{vulnerability.name}} <br>in {{vulnerabilities.grouper}}</span> {% if vulnerability.severity == -1 %} <span style="float: right;" class="badge bg-grey">Unknown</span> <div class="grey-hr-line" ></div> {% elif vulnerability.severity == 0 %} <span style="float: right;" class="badge bg-info">INFO</span> <div class="info-hr-line" ></div> {% elif vulnerability.severity == 1 %} <span style="float: right;" class="badge bg-low">LOW</span> <div class="low-hr-line" ></div> {% elif vulnerability.severity == 2 %} <span style="float: right;" class="badge bg-medium">MEDIUM</span> <div class="medium-hr-line" ></div> {% elif vulnerability.severity == 3 %} <span style="float: right;" class="badge bg-high">HIGH</span> <div class="high-hr-line" ></div> {% elif vulnerability.severity == 4 %} <span style="float: right;" class="badge bg-critical">CRITICAL</span> <div class="critical-hr-line" ></div> {% endif %} </h4> <!-- show vulnerability classification --> <span class="mini-heading">Vulnerability Source: {{vulnerability.source|upper}}</span><br> {% if vulnerability.cvss_metrics or vulnerability.cvss_score or vulnerability.cve_ids.all or vulnerability.cve_ids.all %} <span class="mini-heading">Vulnerability Classification</span><br> {% if vulnerability.cvss_metrics %} <span class="mini-heading ml-8">CVSS Metrics: {{vulnerability.cvss_metrics}}</span> {% endif %} {% if vulnerability.cvss_score %} <br> <span class="mini-heading ml-8">CVSS Score:</span>&nbsp;<span class="high-color">{{vulnerability.cvss_score}}</span> {% endif %} {% if vulnerability.cve_ids.all %} <br> <span class="mini-heading ml-8">CVE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cve in vulnerability.cve_ids.all %} {{cve}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} {% if vulnerability.cwe_ids.all %} <br> <span class="mini-heading ml-8">CWE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cwe in vulnerability.cwe_ids.all %} {{cwe}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} <br> {% endif %} {% if vulnerability.description %} <br> <span class="mini-heading">Description</span><br> {{vulnerability.description|linebreaks}} {% endif %} {% if vulnerability.impact %} <br> <span class="mini-heading">Impact</span><br> {{vulnerability.impact|linebreaks}} {% endif %} {% if vulnerability.remediation %} <br> <span class="mini-heading">Remediation</span><br> {{vulnerability.remediation|linebreaks}} {% endif %} <br> <span class="mini-heading">Vulnerable URL(s)</span><br> <ul> <li class="text-blue"><a href="{{vulnerability.http_url}}" target="_blank" rel="noopener noreferrer">{{vulnerability.http_url}}</a></li> </ul> <!-- {% regroup vulnerability.list by http_url as vuln_http_url_list %} --> <!-- <ul> {% for vuln_urls in vuln_http_url_list %} <li>{{vuln_urls.grouper}}</li> <span class="mini-heading">Result/Findings</span><br> {% for vuln in vuln_urls.list %} {% if vuln.matcher_name %} {% if not forloop.first %} • {% endif %} {{vuln.matcher_name}} {% endif %} {% if vuln.extracted_results %} {% for res in vuln.extracted_results %} {% if not forloop.first %} • {% endif %} {{res}} {% endfor %} {% endif %} {% endfor %} {% endfor %} </ul> --> {% if vulnerability.references.all %} <span class="mini-heading">References</span><br> <ul> {% for ref in vulnerability.references.all %} <li> <span class="text-blue"><a href="{{ref}}" target="_blank" rel="noopener noreferrer">{{ref}}</a></span> </li> {% endfor %} </ul> {% endif %} <br> <br> </div> {% endfor %} {% endfor %} </article> {% endif %} {% endif %} <article id="chapter"> <h2 id="chapter-title">END OF REPORT</h2> </article> </body> </html>

psyray

4341d9834865240222a8dc72c01caaec0d7bed44

69231095782663fe0fe8b0e49b8aa995aa042723

Fixed

psyray

yogeshojha/rengine

Fix report generation when `Ignore Informational Vulnerabilities` checked

When **Ignore Informational Vulnerabilities** is checked there are still info vulns datas. I've reworked the queries that display vulnerabilities to prevent info vulns to display in the : - **Quick summary** Info blue box - **Reconnaissance Findings** - **Vulnerabilities Discovered** Info blue box I've also fixed the **Vulnerabilities Discovered** listing by doing a correct loop through regrouped values because values withe the same path but not the same severity does not display well Tested and working on current master branch

2023-12-05 01:25:41+00:00

2023-12-08 05:48:36+00:00

web/templates/report/template.html

<html> <head> <meta charset="utf-8"> <title>Report</title> <link href="https://fonts.googleapis.com/css2?family=Inter:wght@100;200;300;400;500&display=swap" rel="stylesheet"> <style> @page { size: A4; @top-left { background: {{primary_color}}; content: counter(page); height: 1cm; text-align: center; width: 1cm; } @top-center { background: {{primary_color}}; content: ''; display: block; height: .05cm; opacity: .5; width: 100%; } @top-right { content: string(heading); font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% if show_footer %} @bottom-left { content: "{{footer_text}}"; font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% endif %} } @page :blank { @top-left { background: none; content: '' } @top-center { content: none } @top-right { content: none } } @page no-chapter { @top-left { background: none; content: none } @top-center { content: none } @top-right { content: none } } @page :first { background-color: {{secondary_color}}; background-size: cover; margin: 0; } @page chapter { background: {{primary_color}}; margin: 0; @top-left { content: none } @top-center { content: none } @top-right { content: none } } html { color: #393939; font-family: 'Inter'; font-weight: 300; font-size: 11pt; font-weight: 300; line-height: 1.5; } h1 { font-family: 'Inter'; font-weight: 200; font-size: 38pt; margin: 5cm 2cm 0 2cm; page: no-chapter; width: 100%; line-height: normal; } h2, h3, h4 { font-family: 'Inter'; font-weight: 200; color: black; font-weight: 400; line-height: normal; } #cover { align-content: space-between; display: flex; flex-wrap: wrap; height: 297mm; } #cover-subheading { font-family: 'Inter'; font-weight: 200; font-size: 22pt; width: 100%; } #cover footer { background: {{primary_color}}; flex: 1 33%; margin: 0 -2cm; padding: 1cm 0; white-space: pre-wrap; } #cover footer:first-of-type { padding-left: 3cm; } #cover-line { margin-top: 6px; border-bottom: 1px double {{primary_color}}; } #summary { display: flex; flex-wrap: wrap; justify-content: space-between; } #contents { page: no-chapter; } #contents h2 { font-size: 20pt; Intereight: 400; margin-bottom: 3cm; } #contents h3 { font-weight: 400; margin: 3em 0 1em; } #contents h3::before { background: {{primary_color}}; content: ''; display: block; height: .08cm; margin-bottom: .25cm; width: 2cm; } #contents ul { list-style: none; padding-left: 0; } #contents ul li { border-top: .25pt solid #c1c1c1; margin: .25cm 0; padding-top: .25cm; } #contents ul li::before { color: {{primary_color}}; content: '• '; font-size: 30pt; line-height: 16pt; vertical-align: bottom; } #contents ul li a { color: inherit; text-decoration-line: inherit; } #contents ul li a::before { content: target-text(attr(href)); } #contents ul li a::after { color: {{primary_color}}; content: target-counter(attr(href), page); float: right; } #columns section { columns: 2; column-gap: 1cm; padding-top: 1cm; } #columns section p { text-align: justify; } #columns section p:first-of-type { font-weight: 700; } #chapter { align-items: center; display: flex; height: 297mm; justify-content: center; page: chapter; } #boxes { display: flex; flex-wrap: wrap; justify-content: space-between; } #boxes section h4 { margin-bottom: 0; } #boxes section p { background: {{primary_color}}; display: block; font-size: 15pt; margin-bottom: 0; padding: .25cm 0; text-align: center; height: 85px; color: #37474F; } .bg-critical { background: #EF9A9A !important; } .bg-high { background: #FFAB91 !important; } .bg-medium { background: #FFCC80 !important; } .bg-low { background: #FFE082 !important; } .bg-success { background-color: #A5D6A7 !important; } .bg-grey { background-color: #B0BEC5 !important; } .bg-info { background-color: #90CAF9 !important; } .critical-color { color: #EF9A9A; } .high-color { color: #dc3545; } .medium-color { color: #FFCC80; } .low-color { color: #FFE082; } .success-color { color: #A5D6A7; } .grey-color { color: #212121; } .info-color { color: #90CAF9; } .primary-color { color: {{primary_color}}; } .text-blue{ color: #007bff!important; } .badge { display: inline-block; padding-left: 12px; padding-right: 12px; text-align: center } .critical-hr-line { border-bottom: 3px solid #EF9A9A !important; } .high-hr-line { border-bottom: 3px solid #FFAB91 !important; } .medium-hr-line { border-bottom: 3px solid #FFCC80 !important; } .low-hr-line { border-bottom: 3px solid #FFE082 !important; } .info-hr-line { border-bottom: 3px solid #90CAF9 !important; } .grey-hr-line { border-bottom: 3px solid #212121 !important; } .inside-box-counter { font-size: 28pt; } .table { margin: 0 0 40px 0; width: 100%; box-shadow: 0 1px 3px rgba(0, 0, 0, 0.2); display: table; border-spacing: 0 0.4em; } .row { display: table-row; background: #f6f6f6; } .cell { padding: 6px 6px 6px 6px; display: table-cell; } .header { Intereight: 900; color: #ffffff; } .page_title{ font-weight: 300; font-size: 20pt; } .subheading{ font-weight: 300; font-size: 14pt; } .content-heading{ font-weight: 300; font-size: 12pt; } .mini-heading{ font-weight: 400; font-size: 11pt; } .table-border{ border-style:solid; border-width: 1px; border-color: #90CAF9 !important; } a{ color: #007bff; text-decoration: none; } .ml-8{ margin-left: 8px; } </style> </head> <body> <article id="cover"> <h1 style="color:{{primary_color}}">{{report_name}} <br> {{scan_object.domain.name}} <div id="cover-line"></div> {# generated date #} <span id="cover-subheading">{% now "F j, Y" %}</span> </h1> <footer> {{company_name}} {{company_address}} </footer> <footer> {{company_email}} {{company_website}} </footer> <footer> {% if show_rengine_banner %}Generated by reNgine https://github.com/yogeshojha/rengine {% endif %} </footer> </article> <article id="contents"> <h2>&nbsp;</h2> <h3>Table of contents</h3> <ul> {% if show_executive_summary %} <li><a href="#executive-summary"></a></li> {% endif %} <li><a href="#quick-summary"></a></li> <li><a href="#assessment-timeline"></a></li> {% if interesting_subdomains and show_recon %} <li><a href="#interesting-recon-data"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerability-summary"></a></li> {% endif %} {% if show_recon %} <li><a href="#reconnaissance-results"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerabilities-discovered"></a></li> {% endif %} </ul> </article> {% if show_executive_summary %} <article id="summary" style="page-break-before: always"> <h2 id="executive-summary" class="page_title">Executive summary</h2> <br> {{executive_summary_description | safe }} </article> {% endif %} <article id="summary" style="page-break-before: always"> <h2 id="quick-summary" class="page_title">Quick Summary</h2> <p>This section contains quick summary of scan performed on <span class="primary-color">{{scan_object.domain.name}}</span></p> <br> </article> {# recon section #} {% if show_recon %} <h4 id="reconnaissance-summary" class="subheading">Reconnaissance</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-success">Subdomains <br> <span class="inside-box-counter"> {{scan_object.get_subdomain_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Endpoints <br> <span class="inside-box-counter"> {{scan_object.get_endpoint_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-critical">Vulnerabilities <br> <span class="inside-box-counter"> {{scan_object.get_vulnerability_count}} </span> </p> </section> </div> {% endif %} <!-- vulnerability section, hide if only recon report --> {% if show_vuln %} <article> <br> <h4 id="vulnerability-summary" class="subheading">Vulnerability Summary</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width:30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {{scan_object.get_info_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {% endif %} <article> <h3 id="assessment-timeline" class="page_title">Timeline of the Assessment</h3> <p> Scan started on: {{scan_object.start_scan_date|date:"F j, Y h:i"}} <br> Total time taken: {% if scan_object.scan_status == 0 %} {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% elif scan_object.scan_status == 1 %} {{ scan_object.get_elapsed_time }} {% elif scan_object.scan_status == 2 %} {% if scan_object.get_completed_time_in_sec < 60 %} Completed in < 1 minutes {% else %} Completed in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} {% elif scan_object.scan_status == 3 %} Aborted in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} <br> Report Generated on: {% now "F j, Y" %} </p> </article> {# show interesting_subdomains section only when show_recon result is there #} {% if interesting_subdomains and show_recon %} <article style="page-break-before: always" class="summary"> <h3 id="interesting-recon-data" class="page_title">Interesting Recon Data</h3> <p>Listed below are the {{interesting_subdomains.count}} interesting subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-success"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 33%"> Subdomain </div> <div class="cell grey-color" style="width: 33%"> Page title </div> <div class="cell grey-color" style="width: 15%"> HTTP Status </div> </div> {% for subdomain in interesting_subdomains %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 35%"> {{subdomain.name}} </div> <div class="cell" style="width: 35%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> <div class="cell" style="width: 15%;"> {% if subdomain.http_status %} {{subdomain.http_status}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> </div> {% endfor %} </div> </article> {% endif %} {# vulnerability_summary only when vuln_report #} {% if show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerability-summary" class="page_title">Summary of Vulnerabilities Identified</h3> {% if all_vulnerabilities.count > 0 %} <p>Listed below are the vulnerabilities identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-critical"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 50%;"> Vulnerability Name </div> <div class="cell grey-color" style="width: 19%;"> Times Identified </div> <div class="cell grey-color" style="width: 15%"> Severity </div> </div> {% for vulnerability in unique_vulnerabilities %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 50%"> <a href="#vuln_{{vulnerability.name.split|join:'_'}}">{{vulnerability.name}}</a> </div> <div class="cell" style="float: right; width: 19%;"> {{vulnerability.count}} </div> {% if vulnerability.severity == -1 %} <div class="cell bg-grey" style="width: 15%"> <span class="severity-title-box">Unknown</span> {% elif vulnerability.severity == 0 %} <div class="cell bg-info" style="width: 15%"> <span class="severity-title-box">Informational</span> {% elif vulnerability.severity == 1 %} <div class="cell bg-low" style="width: 15%"> <span class="severity-title-box">Low</span> {% elif vulnerability.severity == 2 %} <div class="cell bg-medium" style="width: 15%"> <span class="severity-title-box">Medium</span> {% elif vulnerability.severity == 3 %} <div class="cell bg-high" style="width: 15%"> <span class="severity-title-box">High</span> {% elif vulnerability.severity == 4 %} <div class="cell bg-critical" style="width: 15%"> <span class="severity-title-box">Critical</span> {% endif %} </div> </div> {% endfor %} {% else %} <h3 class='info-color'>No Vulnerabilities were Discovered.</h3> {% endif %} </div> </article> {% endif %} {# show discovered assets only for show_recon report #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 id="reconnaissance-results" class="page_title">Discovered Assets</h3> <h4 class="subheading">Subdomains</h4> <p> During the reconnaissance phase, {{scan_object.get_subdomain_count}} subdomains were discovered. Out of {{scan_object.get_subdomain_count}} subdomains, {{subdomain_alive_count}} returned HTTP status 200. {{interesting_subdomains.count}} interesting subdomains were also identified based on the interesting keywords used. </p> <h4>{{scan_object.get_subdomain_count}} subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> Subdomain </div> <div class="cell grey-color" style="width: 38%"> Page title </div> <div class="cell grey-color" style="width: 18%"> HTTP Status </div> </div> {% for subdomain in subdomains %} <div class="row"> <div class="cell" style="width: 38%"> {{subdomain.name}} </div> <div class="cell" style="width: 38%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% endif %} </div> <div class="cell" style="width: 18%"> {{subdomain.http_status}} </div> </div> {% endfor %} </div> {% if ip_addresses.count %} <h4 class="subheading" style="margin-top: 10px;">IP Addresses</h4> <h4>{{ip_addresses.count}} IP Addresses were identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> IP </div> <div class="cell grey-color" style="width: 38%"> Open Ports </div> <div class="cell grey-color" style="width: 18%"> Remarks </div> </div> {% for ip in ip_addresses %} <div class="row"> <div class="cell" style="width: 38%"> {{ip.address}} </div> <div class="cell" style="width: 38%"> {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </div> {% if ip.is_cdn %} <div class="cell medium" style="width: 18%"> CDN IP Address {% else %} <div class="cell" style="width: 18%"> {% endif %} </div> </div> {% endfor %} </div> {% endif %} </article> <br> {% endif %} {# reconnaissance finding only when show_recon #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 class="page_title">Reconnaissance Findings</h3> {% for subdomain in subdomains %} <table class="table" cellspacing="0" style="border-collapse: collapse;"> <tr> <td style="width: 2%" class="cell table-border">{{ forloop.counter }}.</td> <td style="width: 80%" class="cell table-border">{{subdomain.name}}</td> {% if subdomain.http_status == 200 %} <td style="width: 10%" class="cell table-border bg-success">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 300 and subdomain.http_status < 400 %} <td style="width: 10%" class="cell table-border bg-medium">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 400 %} <td style="width: 10%" class="cell table-border bg-high">{{subdomain.http_status}}</td> {% elif subdomain.http_status == 0 %} <td style="width: 10%" class="cell table-border">N/A</td> {% else %} <td style="width: 10%" class="cell table-border">{{subdomain.http_status}}</td> {% endif %} </tr> {% if subdomain.page_title %} <tr> <td colspan="3" class="cell table-border"><strong>Page Title: </strong>{{subdomain.page_title}}</td> </tr> {% endif %} {% if subdomain.ip_addresses.all %} <tr> <td colspan="3" class="cell table-border"> IP Address: <ul> {% for ip in subdomain.ip_addresses.all %} <li>{{ip.address}} {% if ip.ports.all %} <ul> <li>Open Ports: &nbsp; {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </li> </ul> {% endif %} </li> {% endfor %} </ul> </td> </tr> {% endif %} {% if subdomain.get_vulnerabilities %} <tr> <td colspan="3" class="cell table-border"> Vulnerabilities {% regroup subdomain.get_vulnerabilities by name as vuln_list %} <ul> {% for vulnerability in vuln_list %} <li> <a href="#vuln_{{vulnerability.list.0.name.split|join:'_'}}">{{ vulnerability.grouper }}</a> </li> {% endfor %} </ul> </td> </tr> {% endif %} </table> {% endfor %} </article> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerabilities-discovered" class="page_title">Vulnerabilities Discovered</h3> <p> This section reports the security issues found during the audit. <br> A Total of {{scan_object.get_vulnerability_count}} were discovered in {{scan_object.domain.name}}, {{scan_object.get_critical_vulnerability_count}} of them were Critical, {{scan_object.get_high_vulnerability_count}} of them were High Severity, {{scan_object.get_medium_vulnerability_count}} of them were Medium severity, {{scan_object.get_low_vulnerability_count}} of them were Low severity, and {{scan_object.get_info_vulnerability_count}} of them were Informational. {{scan_object.get_unknown_vulnerability_count}} of them were Unknown Severity. </p> <h4 class="subheading">Vulnerability Breakdown by Severity</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {{scan_object.get_info_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {# start vulnerability #} {% if show_vuln %} <article class=""> {% regroup all_vulnerabilities by get_path as grouped_vulnerabilities %} {% for vulnerability in grouped_vulnerabilities %} <div> <h4 class="content-heading" id="vuln_{{vulnerability.list.0.name.split|join:'_'}}"> <span>{{vulnerability.list.0.name}} <br>in {{vulnerability.grouper}}</span> {% if vulnerability.list.0.severity == -1 %} <span style="float: right;" class="badge bg-grey">Unknown</span> <div class="grey-hr-line" ></div> {% elif vulnerability.list.0.severity == 0 %} <span style="float: right;" class="badge bg-info">INFO</span> <div class="info-hr-line" ></div> {% elif vulnerability.list.0.severity == 1 %} <span style="float: right;" class="badge bg-low">LOW</span> <div class="low-hr-line" ></div> {% elif vulnerability.list.0.severity == 2 %} <span style="float: right;" class="badge bg-medium">MEDIUM</span> <div class="medium-hr-line" ></div> {% elif vulnerability.list.0.severity == 3 %} <span style="float: right;" class="badge bg-high">HIGH</span> <div class="high-hr-line" ></div> {% elif vulnerability.list.0.severity == 4 %} <span style="float: right;" class="badge bg-critical">CRITICAL</span> <div class="critical-hr-line" ></div> {% endif %} </h4> <!-- show vulnerability classification --> <span class="mini-heading">Vulnerability Source: {{vulnerability.list.0.source|upper}}</span><br> {% if vulnerability.list.0.cvss_metrics or vulnerability.list.0.cvss_score or vulnerability.list.0.cve_ids.all or vulnerability.list.0.cve_ids.all %} <span class="mini-heading">Vulnerability Classification</span><br> {% if vulnerability.list.0.cvss_metrics %} <span class="mini-heading ml-8">CVSS Metrics: {{vulnerability.list.0.cvss_metrics}}</span> {% endif %} {% if vulnerability.list.0.cvss_score %} <br> <span class="mini-heading ml-8">CVSS Score:</span>&nbsp;<span class="high-color">{{vulnerability.list.0.cvss_score}}</span> {% endif %} {% if vulnerability.list.0.cve_ids.all %} <br> <span class="mini-heading ml-8">CVE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cve in vulnerability.list.0.cve_ids.all %} {{cve}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} {% if vulnerability.list.0.cwe_ids.all %} <br> <span class="mini-heading ml-8">CWE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cwe in vulnerability.list.0.cwe_ids.all %} {{cwe}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} <br> {% endif %} {% if vulnerability.list.0.description %} <br> <span class="mini-heading">Description</span><br> {{vulnerability.list.0.description|linebreaks}} {% endif %} {% if vulnerability.list.0.impact %} <br> <span class="mini-heading">Impact</span><br> {{vulnerability.list.0.impact|linebreaks}} {% endif %} {% if vulnerability.list.0.remediation %} <br> <span class="mini-heading">Remediation</span><br> {{vulnerability.list.0.remediation|linebreaks}} {% endif %} <br> <span class="mini-heading">Vulnerable URL(s)</span><br> <ul> {% for vuln in vulnerability.list %} <li class="text-blue">{{vuln.http_url}}</li> {% endfor %} </ul> <!-- {% regroup vulnerability.list by http_url as vuln_http_url_list %} --> <!-- <ul> {% for vuln_urls in vuln_http_url_list %} <li>{{vuln_urls.grouper}}</li> <span class="mini-heading">Result/Findings</span><br> {% for vuln in vuln_urls.list %} {% if vuln.matcher_name %} {% if not forloop.first %} • {% endif %} {{vuln.matcher_name}} {% endif %} {% if vuln.extracted_results %} {% for res in vuln.extracted_results %} {% if not forloop.first %} • {% endif %} {{res}} {% endfor %} {% endif %} {% endfor %} {% endfor %} </ul> --> {% if vulnerability.list.0.references.all %} <span class="mini-heading">References</span><br> <ul> {% for ref in vulnerability.list.0.references.all %} <li> <span class="text-blue"> {{ref}} </span> </li> {% endfor %} </ul> {% endif %} <br> <br> </div> {% endfor %} </article> {% endif %} {% endif %} <article id="chapter"> <h2 id="chapter-title">END OF REPORT</h2> </article> </body> </html>

<html> <head> <meta charset="utf-8"> <title>Report</title> <link href="https://fonts.googleapis.com/css2?family=Inter:wght@100;200;300;400;500&display=swap" rel="stylesheet"> <style> @page { size: A4; @top-left { background: {{primary_color}}; content: counter(page); height: 1cm; text-align: center; width: 1cm; } @top-center { background: {{primary_color}}; content: ''; display: block; height: .05cm; opacity: .5; width: 100%; } @top-right { content: string(heading); font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% if show_footer %} @bottom-left { content: "{{footer_text}}"; font-size: 9pt; height: 1cm; vertical-align: middle; width: 100%; } {% endif %} } @page :blank { @top-left { background: none; content: '' } @top-center { content: none } @top-right { content: none } } @page no-chapter { @top-left { background: none; content: none } @top-center { content: none } @top-right { content: none } } @page :first { background-color: {{secondary_color}}; background-size: cover; margin: 0; } @page chapter { background: {{primary_color}}; margin: 0; @top-left { content: none } @top-center { content: none } @top-right { content: none } } html { color: #393939; font-family: 'Inter'; font-weight: 300; font-size: 11pt; font-weight: 300; line-height: 1.5; } h1 { font-family: 'Inter'; font-weight: 200; font-size: 38pt; margin: 5cm 2cm 0 2cm; page: no-chapter; width: 100%; line-height: normal; } h2, h3, h4 { font-family: 'Inter'; font-weight: 200; color: black; font-weight: 400; line-height: normal; } #cover { align-content: space-between; display: flex; flex-wrap: wrap; height: 297mm; } #cover-subheading { font-family: 'Inter'; font-weight: 200; font-size: 22pt; width: 100%; } #cover footer { background: {{primary_color}}; flex: 1 33%; margin: 0 -2cm; padding: 1cm 0; white-space: pre-wrap; } #cover footer:first-of-type { padding-left: 3cm; } #cover-line { margin-top: 6px; border-bottom: 1px double {{primary_color}}; } #summary { display: flex; flex-wrap: wrap; justify-content: space-between; } #contents { page: no-chapter; } #contents h2 { font-size: 20pt; Intereight: 400; margin-bottom: 3cm; } #contents h3 { font-weight: 400; margin: 3em 0 1em; } #contents h3::before { background: {{primary_color}}; content: ''; display: block; height: .08cm; margin-bottom: .25cm; width: 2cm; } #contents ul { list-style: none; padding-left: 0; } #contents ul li { border-top: .25pt solid #c1c1c1; margin: .25cm 0; padding-top: .25cm; } #contents ul li::before { color: {{primary_color}}; content: '• '; font-size: 30pt; line-height: 16pt; vertical-align: bottom; } #contents ul li a { color: inherit; text-decoration-line: inherit; } #contents ul li a::before { content: target-text(attr(href)); } #contents ul li a::after { color: {{primary_color}}; content: target-counter(attr(href), page); float: right; } #columns section { columns: 2; column-gap: 1cm; padding-top: 1cm; } #columns section p { text-align: justify; } #columns section p:first-of-type { font-weight: 700; } #chapter { align-items: center; display: flex; height: 297mm; justify-content: center; page: chapter; } #boxes { display: flex; flex-wrap: wrap; justify-content: space-between; } #boxes section h4 { margin-bottom: 0; } #boxes section p { background: {{primary_color}}; display: block; font-size: 15pt; margin-bottom: 0; padding: .25cm 0; text-align: center; height: 85px; color: #37474F; } .bg-critical { background: #EF9A9A !important; } .bg-high { background: #FFAB91 !important; } .bg-medium { background: #FFCC80 !important; } .bg-low { background: #FFE082 !important; } .bg-success { background-color: #A5D6A7 !important; } .bg-grey { background-color: #B0BEC5 !important; } .bg-info { background-color: #90CAF9 !important; } .critical-color { color: #EF9A9A; } .high-color { color: #dc3545; } .medium-color { color: #FFCC80; } .low-color { color: #FFE082; } .success-color { color: #A5D6A7; } .grey-color { color: #212121; } .info-color { color: #90CAF9; } .primary-color { color: {{primary_color}}; } .text-blue{ color: #007bff!important; } .badge { display: inline-block; padding-left: 12px; padding-right: 12px; text-align: center } .critical-hr-line { border-bottom: 3px solid #EF9A9A !important; } .high-hr-line { border-bottom: 3px solid #FFAB91 !important; } .medium-hr-line { border-bottom: 3px solid #FFCC80 !important; } .low-hr-line { border-bottom: 3px solid #FFE082 !important; } .info-hr-line { border-bottom: 3px solid #90CAF9 !important; } .grey-hr-line { border-bottom: 3px solid #212121 !important; } .inside-box-counter { font-size: 28pt; } .table { margin: 0 0 40px 0; width: 100%; box-shadow: 0 1px 3px rgba(0, 0, 0, 0.2); display: table; border-spacing: 0 0.4em; } .row { display: table-row; background: #f6f6f6; } .cell { padding: 6px 6px 6px 6px; display: table-cell; } .header { Intereight: 900; color: #ffffff; } .page_title{ font-weight: 300; font-size: 20pt; } .subheading{ font-weight: 300; font-size: 14pt; } .content-heading{ font-weight: 300; font-size: 12pt; } .mini-heading{ font-weight: 400; font-size: 11pt; } .table-border{ border-style:solid; border-width: 1px; border-color: #90CAF9 !important; } a{ color: #007bff; text-decoration: none; } .ml-8{ margin-left: 8px; } </style> </head> <body> <article id="cover"> <h1 style="color:{{primary_color}}">{{report_name}} <br> {{scan_object.domain.name}} <div id="cover-line"></div> {# generated date #} <span id="cover-subheading">{% now "F j, Y" %}</span> </h1> <footer> {{company_name}} {{company_address}} </footer> <footer> {{company_email}} {{company_website}} </footer> <footer> {% if show_rengine_banner %}Generated by reNgine https://github.com/yogeshojha/rengine {% endif %} </footer> </article> <article id="contents"> <h2>&nbsp;</h2> <h3>Table of contents</h3> <ul> {% if show_executive_summary %} <li><a href="#executive-summary"></a></li> {% endif %} <li><a href="#quick-summary"></a></li> <li><a href="#assessment-timeline"></a></li> {% if interesting_subdomains and show_recon %} <li><a href="#interesting-recon-data"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerability-summary"></a></li> {% endif %} {% if show_recon %} <li><a href="#reconnaissance-results"></a></li> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <li><a href="#vulnerabilities-discovered"></a></li> {% endif %} </ul> </article> {% if show_executive_summary %} <article id="summary" style="page-break-before: always"> <h2 id="executive-summary" class="page_title">Executive summary</h2> <br> {{executive_summary_description | safe }} </article> {% endif %} <article id="summary" style="page-break-before: always"> <h2 id="quick-summary" class="page_title">Quick Summary</h2> <p>This section contains quick summary of scan performed on <span class="primary-color">{{scan_object.domain.name}}</span></p> <br> </article> {# recon section #} {% if show_recon %} <h4 id="reconnaissance-summary" class="subheading">Reconnaissance</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-success">Subdomains <br> <span class="inside-box-counter"> {{scan_object.get_subdomain_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Endpoints <br> <span class="inside-box-counter"> {{scan_object.get_endpoint_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-critical">Vulnerabilities <br> <span class="inside-box-counter"> {{all_vulnerabilities_count}} </span> </p> </section> </div> {% endif %} <!-- vulnerability section, hide if only recon report --> {% if show_vuln %} <article> <br> <h4 id="vulnerability-summary" class="subheading">Vulnerability Summary</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width:30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {% if is_ignore_info_vuln %} 0 {% else %} {{scan_object.get_info_vulnerability_count}} {% endif %} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {% endif %} <article> <h3 id="assessment-timeline" class="page_title">Timeline of the Assessment</h3> <p> Scan started on: {{scan_object.start_scan_date|date:"F j, Y h:i"}} <br> Total time taken: {% if scan_object.scan_status == 0 %} {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% elif scan_object.scan_status == 1 %} {{ scan_object.get_elapsed_time }} {% elif scan_object.scan_status == 2 %} {% if scan_object.get_completed_time_in_sec < 60 %} Completed in < 1 minutes {% else %} Completed in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} {% elif scan_object.scan_status == 3 %} Aborted in {{ scan_object.start_scan_date|timesince:scan_object.stop_scan_date }} {% endif %} <br> Report Generated on: {% now "F j, Y" %} </p> </article> {# show interesting_subdomains section only when show_recon result is there #} {% if interesting_subdomains and show_recon %} <article style="page-break-before: always" class="summary"> <h3 id="interesting-recon-data" class="page_title">Interesting Recon Data</h3> <p>Listed below are the {{interesting_subdomains.count}} interesting subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-success"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 33%"> Subdomain </div> <div class="cell grey-color" style="width: 33%"> Page title </div> <div class="cell grey-color" style="width: 15%"> HTTP Status </div> </div> {% for subdomain in interesting_subdomains %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 35%"> {{subdomain.name}} </div> <div class="cell" style="width: 35%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> <div class="cell" style="width: 15%;"> {% if subdomain.http_status %} {{subdomain.http_status}} {% else %} &nbsp;&nbsp;&nbsp; {% endif %} </div> </div> {% endfor %} </div> </article> {% endif %} {# vulnerability_summary only when vuln_report #} {% if show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerability-summary" class="page_title">Summary of Vulnerabilities Identified</h3> {% if all_vulnerabilities.count > 0 %} <p>Listed below are the vulnerabilities identified on <span class="primary-color">{{scan_object.domain.name}}</span></p> <div class="table"> <div class="row header bg-critical"> <div class="cell grey-color" style="width: 5%"> # </div> <div class="cell grey-color" style="width: 50%;"> Vulnerability Name </div> <div class="cell grey-color" style="width: 19%;"> Times Identified </div> <div class="cell grey-color" style="width: 15%"> Severity </div> </div> {% for vulnerability in unique_vulnerabilities %} <div class="row"> <div class="cell" style="width: 5%"> {{ forloop.counter }} </div> <div class="cell" style="width: 50%"> <a href="#vuln_{{vulnerability.name.split|join:'_'}}">{{vulnerability.name}}</a> </div> <div class="cell" style="float: right; width: 19%;"> {{vulnerability.count}} </div> {% if vulnerability.severity == -1 %} <div class="cell bg-grey" style="width: 15%"> <span class="severity-title-box">Unknown</span> {% elif vulnerability.severity == 0 %} <div class="cell bg-info" style="width: 15%"> <span class="severity-title-box">Informational</span> {% elif vulnerability.severity == 1 %} <div class="cell bg-low" style="width: 15%"> <span class="severity-title-box">Low</span> {% elif vulnerability.severity == 2 %} <div class="cell bg-medium" style="width: 15%"> <span class="severity-title-box">Medium</span> {% elif vulnerability.severity == 3 %} <div class="cell bg-high" style="width: 15%"> <span class="severity-title-box">High</span> {% elif vulnerability.severity == 4 %} <div class="cell bg-critical" style="width: 15%"> <span class="severity-title-box">Critical</span> {% endif %} </div> </div> {% endfor %} {% else %} <h3 class='info-color'>No Vulnerabilities were Discovered.</h3> {% endif %} </div> </article> {% endif %} {# show discovered assets only for show_recon report #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 id="reconnaissance-results" class="page_title">Discovered Assets</h3> <h4 class="subheading">Subdomains</h4> <p> During the reconnaissance phase, {{scan_object.get_subdomain_count}} subdomains were discovered. Out of {{scan_object.get_subdomain_count}} subdomains, {{subdomain_alive_count}} returned HTTP status 200. {{interesting_subdomains.count}} interesting subdomains were also identified based on the interesting keywords used. </p> <h4>{{scan_object.get_subdomain_count}} subdomains identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> Subdomain </div> <div class="cell grey-color" style="width: 38%"> Page title </div> <div class="cell grey-color" style="width: 18%"> HTTP Status </div> </div> {% for subdomain in subdomains %} <div class="row"> <div class="cell" style="width: 38%"> {{subdomain.name}} </div> <div class="cell" style="width: 38%"> {% if subdomain.page_title %} {{subdomain.page_title}} {% endif %} </div> <div class="cell" style="width: 18%"> {{subdomain.http_status}} </div> </div> {% endfor %} </div> {% if ip_addresses.count %} <h4 class="subheading" style="margin-top: 10px;">IP Addresses</h4> <h4>{{ip_addresses.count}} IP Addresses were identified on <span class="primary-color">{{scan_object.domain.name}}</span></h4> <div class="table"> <div class="row header bg-info"> <div class="cell grey-color" style="width: 38%"> IP </div> <div class="cell grey-color" style="width: 38%"> Open Ports </div> <div class="cell grey-color" style="width: 18%"> Remarks </div> </div> {% for ip in ip_addresses %} <div class="row"> <div class="cell" style="width: 38%"> {{ip.address}} </div> <div class="cell" style="width: 38%"> {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </div> {% if ip.is_cdn %} <div class="cell medium" style="width: 18%"> CDN IP Address {% else %} <div class="cell" style="width: 18%"> {% endif %} </div> </div> {% endfor %} </div> {% endif %} </article> <br> {% endif %} {# reconnaissance finding only when show_recon #} {% if show_recon %} <article class="summary" style="page-break-before: always"> <h3 class="page_title">Reconnaissance Findings</h3> {% for subdomain in subdomains %} <table class="table" cellspacing="0" style="border-collapse: collapse;"> <tr> <td style="width: 2%" class="cell table-border">{{ forloop.counter }}.</td> <td style="width: 80%" class="cell table-border">{{subdomain.name}}</td> {% if subdomain.http_status == 200 %} <td style="width: 10%" class="cell table-border bg-success">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 300 and subdomain.http_status < 400 %} <td style="width: 10%" class="cell table-border bg-medium">{{subdomain.http_status}}</td> {% elif subdomain.http_status >= 400 %} <td style="width: 10%" class="cell table-border bg-high">{{subdomain.http_status}}</td> {% elif subdomain.http_status == 0 %} <td style="width: 10%" class="cell table-border">N/A</td> {% else %} <td style="width: 10%" class="cell table-border">{{subdomain.http_status}}</td> {% endif %} </tr> {% if subdomain.page_title %} <tr> <td colspan="3" class="cell table-border"><strong>Page Title: </strong>{{subdomain.page_title}}</td> </tr> {% endif %} {% if subdomain.ip_addresses.all %} <tr> <td colspan="3" class="cell table-border"> IP Address: <ul> {% for ip in subdomain.ip_addresses.all %} <li>{{ip.address}} {% if ip.ports.all %} <ul> <li>Open Ports: &nbsp; {% for port in ip.ports.all %} {{port.number}}/{{port.service_name}}{% if not forloop.last %},{% endif %} {% endfor %} </li> </ul> {% endif %} </li> {% endfor %} </ul> </td> </tr> {% endif %} {% if subdomain.get_vulnerabilities_without_info %} <tr> <td colspan="3" class="cell table-border"> Vulnerabilities {% regroup subdomain.get_vulnerabilities_without_info by name as vuln_list %} <ul> {% for vulnerability in vuln_list %} <li> <a href="#vuln_{{vulnerability.list.0.name.split|join:'_'}}">{{ vulnerability.grouper }}</a> </li> {% endfor %} </ul> </td> </tr> {% endif %} </table> {% endfor %} </article> {% endif %} {% if all_vulnerabilities.count > 0 and show_vuln %} <article style="page-break-before: always" class="summary"> <h3 id="vulnerabilities-discovered" class="page_title">Vulnerabilities Discovered</h3> <p> This section reports the security issues found during the audit. <br> A Total of {{scan_object.get_vulnerability_count}} were discovered in {{scan_object.domain.name}}, {{scan_object.get_critical_vulnerability_count}} of them were Critical, {{scan_object.get_high_vulnerability_count}} of them were High Severity, {{scan_object.get_medium_vulnerability_count}} of them were Medium severity, {% if is_ignore_info_vuln %}0{% else %}{{scan_object.get_info_vulnerability_count}}{% endif %} of them were Low severity, and {{scan_object.get_info_vulnerability_count}} of them were Informational. {{scan_object.get_unknown_vulnerability_count}} of them were Unknown Severity. </p> <h4 class="subheading">Vulnerability Breakdown by Severity</h4> <div id="boxes"> <section style="width: 30%"> <p class="bg-critical">Critical <br> <span class="inside-box-counter"> {{scan_object.get_critical_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-high">High <br> <span class="inside-box-counter"> {{scan_object.get_high_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-medium">Medium <br> <span class="inside-box-counter"> {{scan_object.get_medium_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-low">Low <br> <span class="inside-box-counter"> {{scan_object.get_low_vulnerability_count}} </span> </p> </section> <section style="width: 30%"> <p class="bg-info">Info <br> <span class="inside-box-counter"> {% if is_ignore_info_vuln %} 0 {% else %} {{scan_object.get_info_vulnerability_count}} {% endif %} </span> </p> </section> <section style="width: 30%"> <p class="bg-grey">Unknown <br> <span class="inside-box-counter"> {{scan_object.get_unknown_vulnerability_count}} </span> </p> </section> </div> </article> {# start vulnerability #} {% if show_vuln %} <article class=""> {% regroup all_vulnerabilities by get_path as grouped_vulnerabilities %} {% for vulnerabilities in grouped_vulnerabilities %} {% for vulnerability in vulnerabilities.list %} <div> <h4 class="content-heading" id="vuln_{{vulnerability.name.split|join:'_'}}"> <span>{{vulnerability.name}} <br>in {{vulnerabilities.grouper}}</span> {% if vulnerability.severity == -1 %} <span style="float: right;" class="badge bg-grey">Unknown</span> <div class="grey-hr-line" ></div> {% elif vulnerability.severity == 0 %} <span style="float: right;" class="badge bg-info">INFO</span> <div class="info-hr-line" ></div> {% elif vulnerability.severity == 1 %} <span style="float: right;" class="badge bg-low">LOW</span> <div class="low-hr-line" ></div> {% elif vulnerability.severity == 2 %} <span style="float: right;" class="badge bg-medium">MEDIUM</span> <div class="medium-hr-line" ></div> {% elif vulnerability.severity == 3 %} <span style="float: right;" class="badge bg-high">HIGH</span> <div class="high-hr-line" ></div> {% elif vulnerability.severity == 4 %} <span style="float: right;" class="badge bg-critical">CRITICAL</span> <div class="critical-hr-line" ></div> {% endif %} </h4> <!-- show vulnerability classification --> <span class="mini-heading">Vulnerability Source: {{vulnerability.source|upper}}</span><br> {% if vulnerability.cvss_metrics or vulnerability.cvss_score or vulnerability.cve_ids.all or vulnerability.cve_ids.all %} <span class="mini-heading">Vulnerability Classification</span><br> {% if vulnerability.cvss_metrics %} <span class="mini-heading ml-8">CVSS Metrics: {{vulnerability.cvss_metrics}}</span> {% endif %} {% if vulnerability.cvss_score %} <br> <span class="mini-heading ml-8">CVSS Score:</span>&nbsp;<span class="high-color">{{vulnerability.cvss_score}}</span> {% endif %} {% if vulnerability.cve_ids.all %} <br> <span class="mini-heading ml-8">CVE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cve in vulnerability.cve_ids.all %} {{cve}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} {% if vulnerability.cwe_ids.all %} <br> <span class="mini-heading ml-8">CWE IDs</span><br> &nbsp;&nbsp;&nbsp;&nbsp;{% for cwe in vulnerability.cwe_ids.all %} {{cwe}}{% if not forloop.last %}, {% endif %} {% endfor %} {% endif %} <br> {% endif %} {% if vulnerability.description %} <br> <span class="mini-heading">Description</span><br> {{vulnerability.description|linebreaks}} {% endif %} {% if vulnerability.impact %} <br> <span class="mini-heading">Impact</span><br> {{vulnerability.impact|linebreaks}} {% endif %} {% if vulnerability.remediation %} <br> <span class="mini-heading">Remediation</span><br> {{vulnerability.remediation|linebreaks}} {% endif %} <br> <span class="mini-heading">Vulnerable URL(s)</span><br> <ul> <li class="text-blue"><a href="{{vulnerability.http_url}}" target="_blank" rel="noopener noreferrer">{{vulnerability.http_url}}</a></li> </ul> <!-- {% regroup vulnerability.list by http_url as vuln_http_url_list %} --> <!-- <ul> {% for vuln_urls in vuln_http_url_list %} <li>{{vuln_urls.grouper}}</li> <span class="mini-heading">Result/Findings</span><br> {% for vuln in vuln_urls.list %} {% if vuln.matcher_name %} {% if not forloop.first %} • {% endif %} {{vuln.matcher_name}} {% endif %} {% if vuln.extracted_results %} {% for res in vuln.extracted_results %} {% if not forloop.first %} • {% endif %} {{res}} {% endfor %} {% endif %} {% endfor %} {% endfor %} </ul> --> {% if vulnerability.references.all %} <span class="mini-heading">References</span><br> <ul> {% for ref in vulnerability.references.all %} <li> <span class="text-blue"><a href="{{ref}}" target="_blank" rel="noopener noreferrer">{{ref}}</a></span> </li> {% endfor %} </ul> {% endif %} <br> <br> </div> {% endfor %} {% endfor %} </article> {% endif %} {% endif %} <article id="chapter"> <h2 id="chapter-title">END OF REPORT</h2> </article> </body> </html>

psyray

4341d9834865240222a8dc72c01caaec0d7bed44

69231095782663fe0fe8b0e49b8aa995aa042723

same here as well `{{all_vulnerabilities|length}}`

yogeshojha

yogeshojha/rengine

Fixes for #1033, #1026, #1027

Fixes - Fix Dashboard redirection error (fixes #1026) - Fix Message color (red created confusing message as error) (fixes #1027) - Update nuclei for v3, and nuclei v3 requires go 1.21. (fixes #1033) https://github.com/projectdiscovery/nuclei#install-nuclei

2023-11-23 10:33:54+00:00

2023-11-23 13:04:18+00:00

web/scanEngine/templates/scanEngine/lookup.html

{% extends 'base/base.html' %} {% load static %} {% load custom_tags %} {% block title %} Interesting entries Lookup {% endblock title %} {% block custom_js_css_link %} {% endblock custom_js_css_link %} {% block breadcrumb_title %} <li class="breadcrumb-item"><a href="{% url 'scan_engine_index' current_project.slug %}">Engines</a></li> <li class="breadcrumb-item active">Interesting Lookup</li> {% endblock breadcrumb_title %} {% block page_title %} Interesting Lookup {% endblock page_title %} {% block main_content %} <div class="row"> <div class="col-12"> <div class="card"> <div class="card-body"> <h4 class="header-title">Interesting Lookup</h4> <p> reNgine supports lookup for interesting keyword in recon data. This could be either looking up in subdomains, URLs or in page title. You can enter the keywords to lookup and reNgine will highlight the matched entries.<br> </p> <div class="alert alert-primary border-0 mb-4" role="alert"> Keywords are case insensitive. </div> <h4 class="header-title">Default Keywords</h4> <p>reNgine will use these default keywords to find the interesting subdomains or URLs from recon data.</p> <span class="lead"> {% for keyword in default_lookup %} {% for key in keyword.keywords|split:"," %} <span class="badge bg-primary"> {{key}}</span> {% endfor %} {% endfor %} </span> <h4 class="header-title mt-3">Custom Keywords</h4> <form method="POST"> {% csrf_token %} <label for="keywords" class="form-label">Interesting Keywords to look for</label> {{ form.keywords }} {# hidden value #} {{ form.custom_type }} <span class="text-danger">Press comma , to separate the keywords.</span> <h4 class=" header-title mt-3">Lookup in</h4> <div class="form-check mb-2 form-check-primary"> {{form.url_lookup}} <label class="form-check-label" for="url_lookup">Subdomains/URLs</label> </div> <div class="form-check mb-2 form-check-primary"> {{form.title_lookup}} <label class="form-check-label" for="title_lookup">Page Title</label> </div> <h4 class="header-title mt-3">Lookup Conditions</h6> <span class="text-primary">reNgine will lookup the keywords only when below conditions are met.</span> <br> <b>Lookup only when</b> <div class="form-check mt-2 mb-2 form-check-primary"> {{form.condition_200_http_lookup}} <label class="form-check-label" for="condition_200_http_lookup">HTTP Status is 200</label> </div> <button class="btn btn-primary submit-fn mt-2 float-end" type="submit">Update Lookup</button> </form> </div> </div> </div> </div> {% endblock main_content %} {% block page_level_script %} {% endblock page_level_script %}

{% extends 'base/base.html' %} {% load static %} {% load custom_tags %} {% block title %} Interesting entries Lookup {% endblock title %} {% block custom_js_css_link %} {% endblock custom_js_css_link %} {% block breadcrumb_title %} <li class="breadcrumb-item"><a href="{% url 'scan_engine_index' current_project.slug %}">Engines</a></li> <li class="breadcrumb-item active">Interesting Lookup</li> {% endblock breadcrumb_title %} {% block page_title %} Interesting Lookup {% endblock page_title %} {% block main_content %} <div class="row"> <div class="col-12"> <div class="card"> <div class="card-body"> <h4 class="header-title">Interesting Lookup</h4> <p> reNgine supports lookup for interesting keyword in recon data. This could be either looking up in subdomains, URLs or in page title. You can enter the keywords to lookup and reNgine will highlight the matched entries.<br> </p> <div class="alert alert-primary border-0 mb-4" role="alert"> Keywords are case insensitive. </div> <h4 class="header-title">Default Keywords</h4> <p>reNgine will use these default keywords to find the interesting subdomains or URLs from recon data.</p> <span class="lead"> {% for keyword in default_lookup %} {% for key in keyword.keywords|split:"," %} <span class="badge bg-primary"> {{key}}</span> {% endfor %} {% endfor %} </span> <h4 class="header-title mt-3">Custom Keywords</h4> <form method="POST"> {% csrf_token %} <label for="keywords" class="form-label">Interesting Keywords to look for</label> {{ form.keywords }} {# hidden value #} {{ form.custom_type }} <span class="text-muted">Please use a comma (,) to separate the keywords.</span> <h4 class=" header-title mt-3">Lookup in</h4> <div class="form-check mb-2 form-check-primary"> {{form.url_lookup}} <label class="form-check-label" for="url_lookup">Subdomains/URLs</label> </div> <div class="form-check mb-2 form-check-primary"> {{form.title_lookup}} <label class="form-check-label" for="title_lookup">Page Title</label> </div> <h4 class="header-title mt-3">Lookup Conditions</h6> <span class="text-primary">reNgine will lookup the keywords only when below conditions are met.</span> <br> <b>Lookup only when</b> <div class="form-check mt-2 mb-2 form-check-primary"> {{form.condition_200_http_lookup}} <label class="form-check-label" for="condition_200_http_lookup">HTTP Status is 200</label> </div> <button class="btn btn-primary submit-fn mt-2 float-end" type="submit">Update Lookup</button> </form> </div> </div> </div> </div> {% endblock main_content %} {% block page_level_script %} {% endblock page_level_script %}

yogeshojha

6c1ec3124b55404eae84c8ac721ad067563b9243

b190060d07e6ed4d6bfd969481ab6d54779c09a0

```suggestion <span class="text-muted">Please use a comma (,) to separate the keywords.</span> ```

AnonymousWP

yogeshojha/rengine

Fix crash on saving endpoint (FFUF related only)

Fix #1006 I've added : - a **try except** block to catch error on duplicate record returned by **get_or_create** in **saving_endpoint** method - a **check** on endpoint existence in **dir_file_fuzz** method Errors are logged to the console with the URL.  Tested and working Now we need to find why there are duplicates endpoints in the db But it's another issue

2023-11-22 02:57:45+00:00

2023-11-27 12:37:27+00:00

web/reNgine/tasks.py

import csv import json import os import pprint import subprocess import time import validators import whatportis import xmltodict import yaml import tldextract import concurrent.futures from datetime import datetime from urllib.parse import urlparse from api.serializers import SubdomainSerializer from celery import chain, chord, group from celery.result import allow_join_result from celery.utils.log import get_task_logger from django.db.models import Count from dotted_dict import DottedDict from django.utils import timezone from pycvesearch import CVESearch from metafinder.extractor import extract_metadata_from_google_search from reNgine.celery import app from reNgine.gpt import GPTVulnerabilityReportGenerator from reNgine.celery_custom_task import RengineTask from reNgine.common_func import * from reNgine.definitions import * from reNgine.settings import * from reNgine.gpt import * from reNgine.utilities import * from scanEngine.models import (EngineType, InstalledExternalTool, Notification, Proxy) from startScan.models import * from startScan.models import EndPoint, Subdomain, Vulnerability from targetApp.models import Domain """ Celery tasks. """ logger = get_task_logger(__name__) #----------------------# # Scan / Subscan tasks # #----------------------# @app.task(name='initiate_scan', bind=False, queue='initiate_scan_queue') def initiate_scan( scan_history_id, domain_id, engine_id=None, scan_type=LIVE_SCAN, results_dir=RENGINE_RESULTS, imported_subdomains=[], out_of_scope_subdomains=[], url_filter=''): """Initiate a new scan. Args: scan_history_id (int): ScanHistory id. domain_id (int): Domain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. imported_subdomains (list): Imported subdomains. out_of_scope_subdomains (list): Out-of-scope subdomains. url_filter (str): URL path. Default: '' """ # Get scan history scan = ScanHistory.objects.get(pk=scan_history_id) # Get scan engine engine_id = engine_id or scan.scan_type.id # scan history engine_id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, []) # Get domain and set last_scan_date domain = Domain.objects.get(pk=domain_id) domain.last_scan_date = timezone.now() domain.save() # Get path filter url_filter = url_filter.rstrip('/') # Get or create ScanHistory() object if scan_type == LIVE_SCAN: # immediate scan = ScanHistory.objects.get(pk=scan_history_id) scan.scan_status = RUNNING_TASK elif scan_type == SCHEDULED_SCAN: # scheduled scan = ScanHistory() scan.scan_status = INITIATED_TASK scan.scan_type = engine scan.celery_ids = [initiate_scan.request.id] scan.domain = domain scan.start_scan_date = timezone.now() scan.tasks = engine.tasks scan.results_dir = f'{results_dir}/{domain.name}_{scan.id}' add_gf_patterns = gf_patterns and 'fetch_url' in engine.tasks if add_gf_patterns: scan.used_gf_patterns = ','.join(gf_patterns) scan.save() # Create scan results dir os.makedirs(scan.results_dir) # Build task context ctx = { 'scan_history_id': scan_history_id, 'engine_id': engine_id, 'domain_id': domain.id, 'results_dir': scan.results_dir, 'url_filter': url_filter, 'yaml_configuration': config, 'out_of_scope_subdomains': out_of_scope_subdomains } ctx_str = json.dumps(ctx, indent=2) # Send start notif logger.warning(f'Starting scan {scan_history_id} with context:\n{ctx_str}') send_scan_notif.delay( scan_history_id, subscan_id=None, engine_id=engine_id, status=CELERY_TASK_STATUS_MAP[scan.scan_status]) # Save imported subdomains in DB save_imported_subdomains(imported_subdomains, ctx=ctx) # Create initial subdomain in DB: make a copy of domain as a subdomain so # that other tasks using subdomains can use it. subdomain_name = domain.name subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) # If enable_http_crawl is set, create an initial root HTTP endpoint so that # HTTP crawling can start somewhere http_url = f'{domain.name}{url_filter}' if url_filter else domain.name endpoint, _ = save_endpoint( http_url, ctx=ctx, crawl=enable_http_crawl, is_default=True, subdomain=subdomain ) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build Celery tasks, crafted according to the dependency graph below: # subdomain_discovery --> port_scan --> fetch_url --> dir_file_fuzz # osint vulnerability_scan # osint dalfox xss scan # screenshot # waf_detection workflow = chain( group( subdomain_discovery.si(ctx=ctx, description='Subdomain discovery'), osint.si(ctx=ctx, description='OS Intelligence') ), port_scan.si(ctx=ctx, description='Port scan'), fetch_url.si(ctx=ctx, description='Fetch URL'), group( dir_file_fuzz.si(ctx=ctx, description='Directories & files fuzz'), vulnerability_scan.si(ctx=ctx, description='Vulnerability scan'), screenshot.si(ctx=ctx, description='Screenshot'), waf_detection.si(ctx=ctx, description='WAF detection') ) ) # Build callback callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery chord logger.info(f'Running Celery workflow with {len(workflow.tasks) + 1} tasks') task = chain(workflow, callback).on_error(callback).delay() scan.celery_ids.append(task.id) scan.save() return { 'success': True, 'task_id': task.id } @app.task(name='initiate_subscan', bind=False, queue='subscan_queue') def initiate_subscan( scan_history_id, subdomain_id, engine_id=None, scan_type=None, results_dir=RENGINE_RESULTS, url_filter=''): """Initiate a new subscan. Args: scan_history_id (int): ScanHistory id. subdomain_id (int): Subdomain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. url_filter (str): URL path. Default: '' """ # Get Subdomain, Domain and ScanHistory subdomain = Subdomain.objects.get(pk=subdomain_id) scan = ScanHistory.objects.get(pk=subdomain.scan_history.id) domain = Domain.objects.get(pk=subdomain.target_domain.id) # Get EngineType engine_id = engine_id or scan.scan_type.id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Create scan activity of SubScan Model subscan = SubScan( start_scan_date=timezone.now(), celery_ids=[initiate_subscan.request.id], scan_history=scan, subdomain=subdomain, type=scan_type, status=RUNNING_TASK, engine=engine) subscan.save() # Get YAML configuration config = yaml.safe_load(engine.yaml_configuration) # Create results directory results_dir = f'{scan.results_dir}/subscans/{subscan.id}' os.makedirs(results_dir, exist_ok=True) # Run task method = globals().get(scan_type) if not method: logger.warning(f'Task {scan_type} is not supported by reNgine. Skipping') return scan.tasks.append(scan_type) scan.save() # Send start notif send_scan_notif.delay( scan.id, subscan_id=subscan.id, engine_id=engine_id, status='RUNNING') # Build context ctx = { 'scan_history_id': scan.id, 'subscan_id': subscan.id, 'engine_id': engine_id, 'domain_id': domain.id, 'subdomain_id': subdomain.id, 'yaml_configuration': config, 'results_dir': results_dir, 'url_filter': url_filter } # Create initial endpoints in DB: find domain HTTP endpoint so that HTTP # crawling can start somewhere base_url = f'{subdomain.name}{url_filter}' if url_filter else subdomain.name endpoint, _ = save_endpoint( base_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build header + callback workflow = method.si(ctx=ctx) callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery tasks task = chain(workflow, callback).on_error(callback).delay() subscan.celery_ids.append(task.id) subscan.save() return { 'success': True, 'task_id': task.id } @app.task(name='report', bind=False, queue='report_queue') def report(ctx={}, description=None): """Report task running after all other tasks. Mark ScanHistory or SubScan object as completed and update with final status, log run details and send notification. Args: description (str, optional): Task description shown in UI. """ # Get objects subscan_id = ctx.get('subscan_id') scan_id = ctx.get('scan_history_id') engine_id = ctx.get('engine_id') scan = ScanHistory.objects.filter(pk=scan_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() # Get failed tasks tasks = ScanActivity.objects.filter(scan_of=scan).all() if subscan: tasks = tasks.filter(celery_id__in=subscan.celery_ids) failed_tasks = tasks.filter(status=FAILED_TASK) # Get task status failed_count = failed_tasks.count() status = SUCCESS_TASK if failed_count == 0 else FAILED_TASK status_h = 'SUCCESS' if failed_count == 0 else 'FAILED' # Update scan / subscan status if subscan: subscan.stop_scan_date = timezone.now() subscan.status = status subscan.save() else: scan.scan_status = status scan.stop_scan_date = timezone.now() scan.save() # Send scan status notif send_scan_notif.delay( scan_history_id=scan_id, subscan_id=subscan_id, engine_id=engine_id, status=status_h) #------------------------- # # Tracked reNgine tasks # #--------------------------# @app.task(name='subdomain_discovery', queue='main_scan_queue', base=RengineTask, bind=True) def subdomain_discovery( self, host=None, ctx=None, description=None): """Uses a set of tools (see SUBDOMAIN_SCAN_DEFAULT_TOOLS) to scan all subdomains associated with a domain. Args: host (str): Hostname to scan. Returns: subdomains (list): List of subdomain names. """ if not host: host = self.subdomain.name if self.subdomain else self.domain.name if self.url_filter: logger.warning(f'Ignoring subdomains scan as an URL path filter was passed ({self.url_filter}).') return # Config config = self.yaml_configuration.get(SUBDOMAIN_DISCOVERY) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL) or self.yaml_configuration.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) tools = config.get(USES_TOOLS, SUBDOMAIN_SCAN_DEFAULT_TOOLS) default_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=True).filter(is_subdomain_gathering=True)] custom_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=False).filter(is_subdomain_gathering=True)] send_subdomain_changes, send_interesting = False, False notif = Notification.objects.first() if notif: send_subdomain_changes = notif.send_subdomain_changes_notif send_interesting = notif.send_interesting_notif # Gather tools to run for subdomain scan if ALL in tools: tools = SUBDOMAIN_SCAN_DEFAULT_TOOLS + custom_subdomain_tools tools = [t.lower() for t in tools] # Make exception for amass since tool name is amass, but command is amass-active/passive default_subdomain_tools.append('amass-passive') default_subdomain_tools.append('amass-active') # Run tools for tool in tools: cmd = None logger.info(f'Scanning subdomains for {host} with {tool}') proxy = get_random_proxy() if tool in default_subdomain_tools: if tool == 'amass-passive': cmd = f'amass enum -passive -d {host} -o {self.results_dir}/subdomains_amass.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' elif tool == 'amass-active': use_amass_config = config.get(USE_AMASS_CONFIG, False) amass_wordlist_name = config.get(AMASS_WORDLIST, 'deepmagic.com-prefixes-top50000') wordlist_path = f'/usr/src/wordlist/{amass_wordlist_name}.txt' cmd = f'amass enum -active -d {host} -o {self.results_dir}/subdomains_amass_active.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' cmd += f' -brute -w {wordlist_path}' elif tool == 'sublist3r': cmd = f'python3 /usr/src/github/Sublist3r/sublist3r.py -d {host} -t {threads} -o {self.results_dir}/subdomains_sublister.txt' elif tool == 'subfinder': cmd = f'subfinder -d {host} -o {self.results_dir}/subdomains_subfinder.txt' use_subfinder_config = config.get(USE_SUBFINDER_CONFIG, False) cmd += ' -config /root/.config/subfinder/config.yaml' if use_subfinder_config else '' cmd += f' -proxy {proxy}' if proxy else '' cmd += f' -timeout {timeout}' if timeout else '' cmd += f' -t {threads}' if threads else '' cmd += f' -silent' elif tool == 'oneforall': cmd = f'python3 /usr/src/github/OneForAll/oneforall.py --target {host} run' cmd_extract = f'cut -d\',\' -f6 /usr/src/github/OneForAll/results/{host}.csv > {self.results_dir}/subdomains_oneforall.txt' cmd_rm = f'rm -rf /usr/src/github/OneForAll/results/{host}.csv' cmd += f' && {cmd_extract} && {cmd_rm}' elif tool == 'ctfr': results_file = self.results_dir + '/subdomains_ctfr.txt' cmd = f'python3 /usr/src/github/ctfr/ctfr.py -d {host} -o {results_file}' cmd_extract = f"cat {results_file} | sed 's/\*.//g' | tail -n +12 | uniq | sort > {results_file}" cmd += f' && {cmd_extract}' elif tool == 'tlsx': results_file = self.results_dir + '/subdomains_tlsx.txt' cmd = f'tlsx -san -cn -silent -ro -host {host}' cmd += f" | sed -n '/^\([a-zA-Z0-9]\([-a-zA-Z0-9]*[a-zA-Z0-9]\)\?\.\)\+{host}$/p' | uniq | sort" cmd += f' > {results_file}' elif tool == 'netlas': results_file = self.results_dir + '/subdomains_netlas.txt' cmd = f'netlas search -d domain -i domain domain:"*.{host}" -f json' netlas_key = get_netlas_key() cmd += f' -a {netlas_key}' if netlas_key else '' cmd_extract = f"grep -oE '([a-zA-Z0-9]([-a-zA-Z0-9]*[a-zA-Z0-9])?\.)+{host}'" cmd += f' | {cmd_extract} > {results_file}' elif tool in custom_subdomain_tools: tool_query = InstalledExternalTool.objects.filter(name__icontains=tool.lower()) if not tool_query.exists(): logger.error(f'Missing {{TARGET}} and {{OUTPUT}} placeholders in {tool} configuration. Skipping.') continue custom_tool = tool_query.first() cmd = custom_tool.subdomain_gathering_command if '{TARGET}' in cmd and '{OUTPUT}' in cmd: cmd = cmd.replace('{TARGET}', host) cmd = cmd.replace('{OUTPUT}', f'{self.results_dir}/subdomains_{tool}.txt') cmd = cmd.replace('{PATH}', custom_tool.github_clone_path) if '{PATH}' in cmd else cmd else: logger.warning( f'Subdomain discovery tool "{tool}" is not supported by reNgine. Skipping.') continue # Run tool try: run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) except Exception as e: logger.error( f'Subdomain discovery tool "{tool}" raised an exception') logger.exception(e) # Gather all the tools' results in one single file. Write subdomains into # separate files, and sort all subdomains. run_command( f'cat {self.results_dir}/subdomains_*.txt > {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {self.output_path} -o {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) with open(self.output_path) as f: lines = f.readlines() # Parse the output_file file and store Subdomain and EndPoint objects found # in db. subdomain_count = 0 subdomains = [] urls = [] for line in lines: subdomain_name = line.strip() valid_url = bool(validators.url(subdomain_name)) valid_domain = ( bool(validators.domain(subdomain_name)) or bool(validators.ipv4(subdomain_name)) or bool(validators.ipv6(subdomain_name)) or valid_url ) if not valid_domain: logger.error(f'Subdomain {subdomain_name} is not a valid domain, IP or URL. Skipping.') continue if valid_url: subdomain_name = urlparse(subdomain_name).netloc if subdomain_name in self.out_of_scope_subdomains: logger.error(f'Subdomain {subdomain_name} is out of scope. Skipping.') continue # Add subdomain subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain_count += 1 subdomains.append(subdomain) urls.append(subdomain.name) # Bulk crawl subdomains if enable_http_crawl: ctx['track'] = True http_crawl(urls, ctx=ctx, is_ran_from_subdomain_scan=True) # Find root subdomain endpoints for subdomain in subdomains: pass # Send notifications subdomains_str = '\n'.join([f'• `{subdomain.name}`' for subdomain in subdomains]) self.notify(fields={ 'Subdomain count': len(subdomains), 'Subdomains': subdomains_str, }) if send_subdomain_changes and self.scan_id and self.domain_id: added = get_new_added_subdomain(self.scan_id, self.domain_id) removed = get_removed_subdomain(self.scan_id, self.domain_id) if added: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in added]) self.notify(fields={'Added subdomains': subdomains_str}) if removed: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in removed]) self.notify(fields={'Removed subdomains': subdomains_str}) if send_interesting and self.scan_id and self.domain_id: interesting_subdomains = get_interesting_subdomains(self.scan_id, self.domain_id) if interesting_subdomains: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in interesting_subdomains]) self.notify(fields={'Interesting subdomains': subdomains_str}) return SubdomainSerializer(subdomains, many=True).data @app.task(name='osint', queue='main_scan_queue', base=RengineTask, bind=True) def osint(self, host=None, ctx={}, description=None): """Run Open-Source Intelligence tools on selected domain. Args: host (str): Hostname to scan. Returns: dict: Results from osint discovery and dorking. """ config = self.yaml_configuration.get(OSINT) or OSINT_DEFAULT_CONFIG results = {} grouped_tasks = [] if 'discover' in config: ctx['track'] = False # results = osint_discovery(host=host, ctx=ctx) _task = osint_discovery.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, activity_id=self.activity_id, results_dir=self.results_dir, ctx=ctx ) grouped_tasks.append(_task) if OSINT_DORK in config or OSINT_CUSTOM_DORK in config: _task = dorking.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, results_dir=self.results_dir ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('OSINT Tasks finished...') # with open(self.output_path, 'w') as f: # json.dump(results, f, indent=4) # # return results @app.task(name='osint_discovery', queue='osint_discovery_queue', bind=False) def osint_discovery(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run OSINT discovery. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: dict: osint metadat and theHarvester and h8mail results. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) osint_lookup = config.get(OSINT_DISCOVER, []) osint_intensity = config.get(INTENSITY, 'normal') documents_limit = config.get(OSINT_DOCUMENTS_LIMIT, 50) results = {} meta_info = [] emails = [] creds = [] # Get and save meta info if 'metainfo' in osint_lookup: if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': host, 'domain': host, 'scan_id': scan_history_id, 'documents_limit': documents_limit }) meta_info.append(save_metadata_info(meta_dict)) # TODO: disabled for now # elif osint_intensity == 'deep': # subdomains = Subdomain.objects # if self.scan: # subdomains = subdomains.filter(scan_history=self.scan) # for subdomain in subdomains: # meta_dict = DottedDict({ # 'osint_target': subdomain.name, # 'domain': self.domain, # 'scan_id': self.scan_id, # 'documents_limit': documents_limit # }) # meta_info.append(save_metadata_info(meta_dict)) grouped_tasks = [] if 'emails' in osint_lookup: emails = get_and_save_emails(scan_history, activity_id, results_dir) emails_str = '\n'.join([f'• `{email}`' for email in emails]) # self.notify(fields={'Emails': emails_str}) # ctx['track'] = False _task = h8mail.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) if 'employees' in osint_lookup: ctx['track'] = False _task = theHarvester.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) # results['emails'] = results.get('emails', []) + emails # results['creds'] = creds # results['meta_info'] = meta_info return results @app.task(name='dorking', bind=False, queue='dorking_queue') def dorking(config, host, scan_history_id, results_dir): """Run Google dorks. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: list: Dorking results for each dork ran. """ # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh scan_history = ScanHistory.objects.get(pk=scan_history_id) dorks = config.get(OSINT_DORK, []) custom_dorks = config.get(OSINT_CUSTOM_DORK, []) results = [] # custom dorking has higher priority try: for custom_dork in custom_dorks: lookup_target = custom_dork.get('lookup_site') # replace with original host if _target_ lookup_target = host if lookup_target == '_target_' else lookup_target if 'lookup_extensions' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_extensions=custom_dork.get('lookup_extensions'), scan_history=scan_history ) elif 'lookup_keywords' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_keywords=custom_dork.get('lookup_keywords'), scan_history=scan_history ) except Exception as e: logger.exception(e) # default dorking try: for dork in dorks: logger.info(f'Getting dork information for {dork}') if dork == 'stackoverflow': results = get_and_save_dork_results( lookup_target='stackoverflow.com', results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'login_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/login/,login.html', page_count=5, scan_history=scan_history ) elif dork == 'admin_panels': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/admin/,admin.html', page_count=5, scan_history=scan_history ) elif dork == 'dashboard_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/dashboard/,dashboard.html', page_count=5, scan_history=scan_history ) elif dork == 'social_media' : social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'reddit.com' ] for site in social_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'project_management' : project_websites = [ 'trello.com', 'atlassian.net' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'code_sharing' : project_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'config_files' : config_file_exts = [ 'env', 'xml', 'conf', 'toml', 'yml', 'yaml', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(config_file_exts), page_count=4, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keyword = 'Jenkins' results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=lookup_keyword, page_count=1, scan_history=scan_history ) elif dork == 'wordpress_files' : lookup_keywords = [ '/wp-content/', '/wp-includes/' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'php_error' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'exposed_documents' : docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(docs_file_ext), page_count=7, scan_history=scan_history ) elif dork == 'db_files' : file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) elif dork == 'git_exposed' : file_ext = [ 'git', ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) except Exception as e: logger.exception(e) return results @app.task(name='theHarvester', queue='theHarvester_queue', bind=False) def theHarvester(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run theHarvester to get save emails, hosts, employees found in domain. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: dict: Dict of emails, employees, hosts and ips found during crawling. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) output_path_json = f'{results_dir}/theHarvester.json' theHarvester_dir = '/usr/src/github/theHarvester' history_file = f'{results_dir}/commands.txt' cmd = f'python3 {theHarvester_dir}/theHarvester.py -d {host} -b all -f {output_path_json}' # Update proxies.yaml proxy_query = Proxy.objects.all() if proxy_query.exists(): proxy = proxy_query.first() if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(f'{theHarvester_dir}/proxies.yaml', 'w') as file: yaml.dump(yaml_data, file) # Run cmd run_command( cmd, shell=False, cwd=theHarvester_dir, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) # Get file location if not os.path.isfile(output_path_json): logger.error(f'Could not open {output_path_json}') return {} # Load theHarvester results with open(output_path_json, 'r') as f: data = json.load(f) # Re-indent theHarvester JSON with open(output_path_json, 'w') as f: json.dump(data, f, indent=4) emails = data.get('emails', []) for email_address in emails: email, _ = save_email(email_address, scan_history=scan_history) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) linkedin_people = data.get('linkedin_people', []) for people in linkedin_people: employee, _ = save_employee( people, designation='linkedin', scan_history=scan_history) # if employee: # self.notify(fields={'LinkedIn people': f'• {employee.name}'}) twitter_people = data.get('twitter_people', []) for people in twitter_people: employee, _ = save_employee( people, designation='twitter', scan_history=scan_history) # if employee: # self.notify(fields={'Twitter people': f'• {employee.name}'}) hosts = data.get('hosts', []) urls = [] for host in hosts: split = tuple(host.split(':')) http_url = split[0] subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) endpoint, _ = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain) # if endpoint: # urls.append(endpoint.http_url) # self.notify(fields={'Hosts': f'• {endpoint.http_url}'}) # if enable_http_crawl: # ctx['track'] = False # http_crawl(urls, ctx=ctx) # TODO: Lots of ips unrelated with our domain are found, disabling # this for now. # ips = data.get('ips', []) # for ip_address in ips: # ip, created = save_ip_address( # ip_address, # subscan=subscan) # if ip: # send_task_notif.delay( # 'osint', # scan_history_id=scan_history_id, # subscan_id=subscan_id, # severity='success', # update_fields={'IPs': f'{ip.address}'}) return data @app.task(name='h8mail', queue='h8mail_queue', bind=False) def h8mail(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run h8mail. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: list[dict]: List of credentials info. """ logger.warning('Getting leaked credentials') scan_history = ScanHistory.objects.get(pk=scan_history_id) input_path = f'{results_dir}/emails.txt' output_file = f'{results_dir}/h8mail.json' cmd = f'h8mail -t {input_path} --json {output_file}' history_file = f'{results_dir}/commands.txt' run_command( cmd, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) with open(output_file) as f: data = json.load(f) creds = data.get('targets', []) # TODO: go through h8mail output and save emails to DB for cred in creds: logger.warning(cred) email_address = cred['target'] pwn_num = cred['pwn_num'] pwn_data = cred.get('data', []) email, created = save_email(email_address, scan_history=scan) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) return creds @app.task(name='screenshot', queue='main_scan_queue', base=RengineTask, bind=True) def screenshot(self, ctx={}, description=None): """Uses EyeWitness to gather screenshot of a domain and/or url. Args: description (str, optional): Task description shown in UI. """ # Config screenshots_path = f'{self.results_dir}/screenshots' output_path = f'{self.results_dir}/screenshots/{self.filename}' alive_endpoints_file = f'{self.results_dir}/endpoints_alive.txt' config = self.yaml_configuration.get(SCREENSHOT) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT + 5) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) # If intensity is normal, grab only the root endpoints of each subdomain strict = True if intensity == 'normal' else False # Get URLs to take screenshot of get_http_urls( is_alive=enable_http_crawl, strict=strict, write_filepath=alive_endpoints_file, get_only_default_urls=True, ctx=ctx ) # Send start notif notification = Notification.objects.first() send_output_file = notification.send_scan_output_file if notification else False # Run cmd cmd = f'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py -f {alive_endpoints_file} -d {screenshots_path} --no-prompt' cmd += f' --timeout {timeout}' if timeout > 0 else '' cmd += f' --threads {threads}' if threads > 0 else '' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(output_path): logger.error(f'Could not load EyeWitness results at {output_path} for {self.domain.name}.') return # Loop through results and save objects in DB screenshot_paths = [] with open(output_path, 'r') as file: reader = csv.reader(file) for row in reader: "Protocol,Port,Domain,Request Status,Screenshot Path, Source Path" protocol, port, subdomain_name, status, screenshot_path, source_path = tuple(row) logger.info(f'{protocol}:{port}:{subdomain_name}:{status}') subdomain_query = Subdomain.objects.filter(name=subdomain_name) if self.scan: subdomain_query = subdomain_query.filter(scan_history=self.scan) if status == 'Successful' and subdomain_query.exists(): subdomain = subdomain_query.first() screenshot_paths.append(screenshot_path) subdomain.screenshot_path = screenshot_path.replace('/usr/src/scan_results/', '') subdomain.save() logger.warning(f'Added screenshot for {subdomain.name} to DB') # Remove all db, html extra files in screenshot results run_command( 'rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format(screenshots_path), shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'rm -rf {screenshots_path}/source', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Send finish notifs screenshots_str = '• ' + '\n• '.join([f'`{path}`' for path in screenshot_paths]) self.notify(fields={'Screenshots': screenshots_str}) if send_output_file: for path in screenshot_paths: title = get_output_file_name( self.scan_id, self.subscan_id, self.filename) send_file_to_discord.delay(path, title) @app.task(name='port_scan', queue='main_scan_queue', base=RengineTask, bind=True) def port_scan(self, hosts=[], ctx={}, description=None): """Run port scan. Args: hosts (list, optional): Hosts to run port scan on. description (str, optional): Task description shown in UI. Returns: list: List of open ports (dict). """ input_file = f'{self.results_dir}/input_subdomains_port_scan.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(PORT_SCAN) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) exclude_ports = config.get(NAABU_EXCLUDE_PORTS, []) exclude_subdomains = config.get(NAABU_EXCLUDE_SUBDOMAINS, False) ports = config.get(PORTS, NAABU_DEFAULT_PORTS) ports = [str(port) for port in ports] rate_limit = config.get(NAABU_RATE) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) passive = config.get(NAABU_PASSIVE, False) use_naabu_config = config.get(USE_NAABU_CONFIG, False) exclude_ports_str = ','.join(return_iterable(exclude_ports)) # nmap args nmap_enabled = config.get(ENABLE_NMAP, False) nmap_cmd = config.get(NMAP_COMMAND, '') nmap_script = config.get(NMAP_SCRIPT, '') nmap_script = ','.join(return_iterable(nmap_script)) nmap_script_args = config.get(NMAP_SCRIPT_ARGS) if hosts: with open(input_file, 'w') as f: f.write('\n'.join(hosts)) else: hosts = get_subdomains( write_filepath=input_file, exclude_subdomains=exclude_subdomains, ctx=ctx) # Build cmd cmd = 'naabu -json -exclude-cdn' cmd += f' -list {input_file}' if len(hosts) > 0 else f' -host {hosts[0]}' if 'full' in ports or 'all' in ports: ports_str = ' -p "-"' elif 'top-100' in ports: ports_str = ' -top-ports 100' elif 'top-1000' in ports: ports_str = ' -top-ports 1000' else: ports_str = ','.join(ports) ports_str = f' -p {ports_str}' cmd += ports_str cmd += ' -config /root/.config/naabu/config.yaml' if use_naabu_config else '' cmd += f' -proxy "{proxy}"' if proxy else '' cmd += f' -c {threads}' if threads else '' cmd += f' -rate {rate_limit}' if rate_limit > 0 else '' cmd += f' -timeout {timeout*1000}' if timeout > 0 else '' cmd += f' -passive' if passive else '' cmd += f' -exclude-ports {exclude_ports_str}' if exclude_ports else '' cmd += f' -silent' # Execute cmd and gather results results = [] urls = [] ports_data = {} for line in stream_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) port_number = line['port'] ip_address = line['ip'] host = line.get('host') or ip_address if port_number == 0: continue # Grab subdomain subdomain = Subdomain.objects.filter( name=host, target_domain=self.domain, scan_history=self.scan ).first() # Add IP DB ip, _ = save_ip_address(ip_address, subdomain, subscan=self.subscan) if self.subscan: ip.ip_subscan_ids.add(self.subscan) ip.save() # Add endpoint to DB # port 80 and 443 not needed as http crawl already does that. if port_number not in [80, 443]: http_url = f'{host}:{port_number}' endpoint, _ = save_endpoint( http_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint: http_url = endpoint.http_url urls.append(http_url) # Add Port in DB port_details = whatportis.get_ports(str(port_number)) service_name = port_details[0].name if len(port_details) > 0 else 'unknown' description = port_details[0].description if len(port_details) > 0 else '' # get or create port port, created = Port.objects.get_or_create( number=port_number, service_name=service_name, description=description ) if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port.save() ip.ports.add(port) ip.save() if host in ports_data: ports_data[host].append(port_number) else: ports_data[host] = [port_number] # Send notification logger.warning(f'Found opened port {port_number} on {ip_address} ({host})') if len(ports_data) == 0: logger.info('Finished running naabu port scan - No open ports found.') if nmap_enabled: logger.info('Nmap scans skipped') return ports_data # Send notification fields_str = '' for host, ports in ports_data.items(): ports_str = ', '.join([f'`{port}`' for port in ports]) fields_str += f'• `{host}`: {ports_str}\n' self.notify(fields={'Ports discovered': fields_str}) # Save output to file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) logger.info('Finished running naabu port scan.') # Process nmap results: 1 process per host sigs = [] if nmap_enabled: logger.warning(f'Starting nmap scans ...') logger.warning(ports_data) for host, port_list in ports_data.items(): ports_str = '_'.join([str(p) for p in port_list]) ctx_nmap = ctx.copy() ctx_nmap['description'] = get_task_title(f'nmap_{host}', self.scan_id, self.subscan_id) ctx_nmap['track'] = False sig = nmap.si( cmd=nmap_cmd, ports=port_list, host=host, script=nmap_script, script_args=nmap_script_args, max_rate=rate_limit, ctx=ctx_nmap) sigs.append(sig) task = group(sigs).apply_async() with allow_join_result(): results = task.get() return ports_data @app.task(name='nmap', queue='main_scan_queue', base=RengineTask, bind=True) def nmap( self, cmd=None, ports=[], host=None, input_file=None, script=None, script_args=None, max_rate=None, ctx={}, description=None): """Run nmap on a host. Args: cmd (str, optional): Existing nmap command to complete. ports (list, optional): List of ports to scan. host (str, optional): Host to scan. input_file (str, optional): Input hosts file. script (str, optional): NSE script to run. script_args (str, optional): NSE script args. max_rate (int): Max rate. description (str, optional): Task description shown in UI. """ notif = Notification.objects.first() ports_str = ','.join(str(port) for port in ports) self.filename = self.filename.replace('.txt', '.xml') filename_vulns = self.filename.replace('.xml', '_vulns.json') output_file = self.output_path output_file_xml = f'{self.results_dir}/{host}_{self.filename}' vulns_file = f'{self.results_dir}/{host}_{filename_vulns}' logger.warning(f'Running nmap on {host}:{ports}') # Build cmd nmap_cmd = get_nmap_cmd( cmd=cmd, ports=ports_str, script=script, script_args=script_args, max_rate=max_rate, host=host, input_file=input_file, output_file=output_file_xml) # Run cmd run_command( nmap_cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Get nmap XML results and convert to JSON vulns = parse_nmap_results(output_file_xml, output_file) with open(vulns_file, 'w') as f: json.dump(vulns, f, indent=4) # Save vulnerabilities found by nmap vulns_str = '' for vuln_data in vulns: # URL is not necessarily an HTTP URL when running nmap (can be any # other vulnerable protocols). Look for existing endpoint and use its # URL as vulnerability.http_url if it exists. url = vuln_data['http_url'] endpoint = EndPoint.objects.filter(http_url__contains=url).first() if endpoint: vuln_data['http_url'] = endpoint.http_url vuln, created = save_vulnerability( target_domain=self.domain, subdomain=self.subdomain, scan_history=self.scan, subscan=self.subscan, endpoint=endpoint, **vuln_data) vulns_str += f'• {str(vuln)}\n' if created: logger.warning(str(vuln)) # Send only 1 notif for all vulns to reduce number of notifs if notif and notif.send_vuln_notif and vulns_str: logger.warning(vulns_str) self.notify(fields={'CVEs': vulns_str}) return vulns @app.task(name='waf_detection', queue='main_scan_queue', base=RengineTask, bind=True) def waf_detection(self, ctx={}, description=None): """ Uses wafw00f to check for the presence of a WAF. Args: description (str, optional): Task description shown in UI. Returns: list: List of startScan.models.Waf objects. """ input_path = f'{self.results_dir}/input_endpoints_waf_detection.txt' config = self.yaml_configuration.get(WAF_DETECTION) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Get alive endpoints from DB get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) cmd = f'wafw00f -i {input_path} -o {self.output_path}' run_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(self.output_path): logger.error(f'Could not find {self.output_path}') return with open(self.output_path) as file: wafs = file.readlines() for line in wafs: line = " ".join(line.split()) splitted = line.split(' ', 1) waf_info = splitted[1].strip() waf_name = waf_info[:waf_info.find('(')].strip() waf_manufacturer = waf_info[waf_info.find('(')+1:waf_info.find(')')].strip().replace('.', '') http_url = sanitize_url(splitted[0].strip()) if not waf_name or waf_name == 'None': continue # Add waf to db waf, _ = Waf.objects.get_or_create( name=waf_name, manufacturer=waf_manufacturer ) # Add waf info to Subdomain in DB subdomain = get_subdomain_from_url(http_url) logger.info(f'Wafw00f Subdomain : {subdomain}') subdomain_query, _ = Subdomain.objects.get_or_create(scan_history=self.scan, name=subdomain) subdomain_query.waf.add(waf) subdomain_query.save() return wafs @app.task(name='dir_file_fuzz', queue='main_scan_queue', base=RengineTask, bind=True) def dir_file_fuzz(self, ctx={}, description=None): """Perform directory scan, and currently uses `ffuf` as a default tool. Args: description (str, optional): Task description shown in UI. Returns: list: List of URLs discovered. """ # Config cmd = 'ffuf' config = self.yaml_configuration.get(DIR_FILE_FUZZ) or {} custom_header = self.yaml_configuration.get(CUSTOM_HEADER) auto_calibration = config.get(AUTO_CALIBRATION, True) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) extensions = config.get(EXTENSIONS, DEFAULT_DIR_FILE_FUZZ_EXTENSIONS) # prepend . on extensions extensions = [ext if ext.startswith('.') else '.' + ext for ext in extensions] extensions_str = ','.join(map(str, extensions)) follow_redirect = config.get(FOLLOW_REDIRECT, FFUF_DEFAULT_FOLLOW_REDIRECT) max_time = config.get(MAX_TIME, 0) match_http_status = config.get(MATCH_HTTP_STATUS, FFUF_DEFAULT_MATCH_HTTP_STATUS) mc = ','.join([str(c) for c in match_http_status]) recursive_level = config.get(RECURSIVE_LEVEL, FFUF_DEFAULT_RECURSIVE_LEVEL) stop_on_error = config.get(STOP_ON_ERROR, False) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) wordlist_name = config.get(WORDLIST, 'dicc') delay = rate_limit / (threads * 100) # calculate request pause delay from rate_limit and number of threads input_path = f'{self.results_dir}/input_dir_file_fuzz.txt' # Get wordlist wordlist_name = 'dicc' if wordlist_name == 'default' else wordlist_name wordlist_path = f'/usr/src/wordlist/{wordlist_name}.txt' # Build command cmd += f' -w {wordlist_path}' cmd += f' -e {extensions_str}' if extensions else '' cmd += f' -maxtime {max_time}' if max_time > 0 else '' cmd += f' -p {delay}' if delay > 0 else '' cmd += f' -recursion -recursion-depth {recursive_level} ' if recursive_level > 0 else '' cmd += f' -t {threads}' if threads and threads > 0 else '' cmd += f' -timeout {timeout}' if timeout and timeout > 0 else '' cmd += ' -se' if stop_on_error else '' cmd += ' -fr' if follow_redirect else '' cmd += ' -ac' if auto_calibration else '' cmd += f' -mc {mc}' if mc else '' cmd += f' -H "{custom_header}"' if custom_header else '' # Grab URLs to fuzz urls = get_http_urls( is_alive=True, ignore_files=False, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) logger.warning(urls) # Loop through URLs and run command results = [] for url in urls: ''' Above while fetching urls, we are not ignoring files, because some default urls may redirect to https://example.com/login.php so, ignore_files is set to False but, during fuzzing, we will only need part of the path, in above example it is still a good idea to ffuf base url https://example.com so files from base url ''' url_parse = urlparse(url) url = url_parse.scheme + '://' + url_parse.netloc url += '/FUZZ' # TODO: fuzz not only URL but also POST / PUT / headers proxy = get_random_proxy() # Build final cmd fcmd = cmd fcmd += f' -x {proxy}' if proxy else '' fcmd += f' -u {url} -json' # Initialize DirectoryScan object dirscan = DirectoryScan() dirscan.scanned_date = timezone.now() dirscan.command_line = fcmd dirscan.save() # Loop through results and populate EndPoint and DirectoryFile in DB results = [] for line in stream_command( fcmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) name = line['input'].get('FUZZ') length = line['length'] status = line['status'] words = line['words'] url = line['url'] lines = line['lines'] content_type = line['content-type'] duration = line['duration'] if not name: logger.error(f'FUZZ not found for "{url}"') continue endpoint, created = save_endpoint(url, crawl=False, ctx=ctx) # endpoint.is_default = False endpoint.http_status = status endpoint.content_length = length endpoint.response_time = duration / 1000000000 endpoint.save() if created: urls.append(endpoint.http_url) endpoint.status = status endpoint.content_type = content_type endpoint.content_length = length dfile, created = DirectoryFile.objects.get_or_create( name=name, length=length, words=words, lines=lines, content_type=content_type, url=url) dfile.http_status = status dfile.save() # if created: # logger.warning(f'Found new directory or file {url}') dirscan.directory_files.add(dfile) dirscan.save() if self.subscan: dirscan.dir_subscan_ids.add(self.subscan) subdomain_name = get_subdomain_from_url(endpoint.http_url) subdomain = Subdomain.objects.get(name=subdomain_name, scan_history=self.scan) subdomain.directories.add(dirscan) subdomain.save() # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl(urls, ctx=ctx) return results @app.task(name='fetch_url', queue='main_scan_queue', base=RengineTask, bind=True) def fetch_url(self, urls=[], ctx={}, description=None): """Fetch URLs using different tools like gauplus, gau, gospider, waybackurls ... Args: urls (list): List of URLs to start from. description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/input_endpoints_fetch_url.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(FETCH_URL) or {} should_remove_duplicate_endpoints = config.get(REMOVE_DUPLICATE_ENDPOINTS, True) duplicate_removal_fields = config.get(DUPLICATE_REMOVAL_FIELDS, ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, DEFAULT_GF_PATTERNS) ignore_file_extension = config.get(IGNORE_FILE_EXTENSION, DEFAULT_IGNORE_FILE_EXTENSIONS) tools = config.get(USES_TOOLS, ENDPOINT_SCAN_DEFAULT_TOOLS) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) domain_request_headers = self.domain.request_headers if self.domain else None custom_header = domain_request_headers or self.yaml_configuration.get(CUSTOM_HEADER) exclude_subdomains = config.get(EXCLUDED_SUBDOMAINS, False) # Get URLs to scan and save to input file if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, exclude_subdomains=exclude_subdomains, get_only_default_urls=True, ctx=ctx ) # Domain regex host = self.domain.name if self.domain else urlparse(urls[0]).netloc host_regex = f"\'https?://([a-z0-9]+[.])*{host}.*\'" # Tools cmds cmd_map = { 'gau': f'gau', 'gauplus': f'gauplus -random-agent', 'hakrawler': 'hakrawler -subs -u', 'waybackurls': 'waybackurls', 'gospider': f'gospider -S {input_path} --js -d 2 --sitemap --robots -w -r', 'katana': f'katana -list {input_path} -silent -jc -kf all -d 3 -fs rdn', } if proxy: cmd_map['gau'] += f' --proxy "{proxy}"' cmd_map['gauplus'] += f' -p "{proxy}"' cmd_map['gospider'] += f' -p {proxy}' cmd_map['hakrawler'] += f' -proxy {proxy}' cmd_map['katana'] += f' -proxy {proxy}' if threads > 0: cmd_map['gau'] += f' --threads {threads}' cmd_map['gauplus'] += f' -t {threads}' cmd_map['gospider'] += f' -t {threads}' cmd_map['katana'] += f' -c {threads}' if custom_header: header_string = ';;'.join([ f'{key}: {value}' for key, value in custom_header.items() ]) cmd_map['hakrawler'] += f' -h {header_string}' cmd_map['katana'] += f' -H {header_string}' header_flags = [':'.join(h) for h in header_string.split(';;')] for flag in header_flags: cmd_map['gospider'] += f' -H {flag}' cat_input = f'cat {input_path}' grep_output = f'grep -Eo {host_regex}' cmd_map = { tool: f'{cat_input} | {cmd} | {grep_output} > {self.results_dir}/urls_{tool}.txt' for tool, cmd in cmd_map.items() } tasks = group( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for tool, cmd in cmd_map.items() if tool in tools ) # Cleanup task sort_output = [ f'cat {self.results_dir}/urls_* > {self.output_path}', f'cat {input_path} >> {self.output_path}', f'sort -u {self.output_path} -o {self.output_path}', ] if ignore_file_extension: ignore_exts = '|'.join(ignore_file_extension) grep_ext_filtered_output = [ f'cat {self.output_path} | grep -Eiv "\\.({ignore_exts}).*" > {self.results_dir}/urls_filtered.txt', f'mv {self.results_dir}/urls_filtered.txt {self.output_path}' ] sort_output.extend(grep_ext_filtered_output) cleanup = chain( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for cmd in sort_output ) # Run all commands task = chord(tasks)(cleanup) with allow_join_result(): task.get() # Store all the endpoints and run httpx with open(self.output_path) as f: discovered_urls = f.readlines() self.notify(fields={'Discovered URLs': len(discovered_urls)}) # Some tools can have an URL in the format <URL>] - <PATH> or <URL> - <PATH>, add them # to the final URL list all_urls = [] for url in discovered_urls: url = url.strip() urlpath = None base_url = None if '] ' in url: # found JS scraped endpoint e.g from gospider split = tuple(url.split('] ')) if not len(split) == 2: logger.warning(f'URL format not recognized for "{url}". Skipping.') continue base_url, urlpath = split urlpath = urlpath.lstrip('- ') elif ' - ' in url: # found JS scraped endpoint e.g from gospider base_url, urlpath = tuple(url.split(' - ')) if base_url and urlpath: subdomain = urlparse(base_url) url = f'{subdomain.scheme}://{subdomain.netloc}{self.url_filter}' if not validators.url(url): logger.warning(f'Invalid URL "{url}". Skipping.') if url not in all_urls: all_urls.append(url) # Filter out URLs if a path filter was passed if self.url_filter: all_urls = [url for url in all_urls if self.url_filter in url] # Write result to output path with open(self.output_path, 'w') as f: f.write('\n'.join(all_urls)) logger.warning(f'Found {len(all_urls)} usable URLs') # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl( all_urls, ctx=ctx, should_remove_duplicate_endpoints=should_remove_duplicate_endpoints, duplicate_removal_fields=duplicate_removal_fields ) #-------------------# # GF PATTERNS MATCH # #-------------------# # Combine old gf patterns with new ones if gf_patterns: self.scan.used_gf_patterns = ','.join(gf_patterns) self.scan.save() # Run gf patterns on saved endpoints # TODO: refactor to Celery task for gf_pattern in gf_patterns: # TODO: js var is causing issues, removing for now if gf_pattern == 'jsvar': logger.info('Ignoring jsvar as it is causing issues.') continue # Run gf on current pattern logger.warning(f'Running gf on pattern "{gf_pattern}"') gf_output_file = f'{self.results_dir}/gf_patterns_{gf_pattern}.txt' cmd = f'cat {self.output_path} | gf {gf_pattern} | grep -Eo {host_regex} >> {gf_output_file}' run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Check output file if not os.path.exists(gf_output_file): logger.error(f'Could not find GF output file {gf_output_file}. Skipping GF pattern "{gf_pattern}"') continue # Read output file line by line and with open(gf_output_file, 'r') as f: lines = f.readlines() # Add endpoints / subdomains to DB for url in lines: http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue endpoint, created = save_endpoint( http_url, crawl=False, subdomain=subdomain, ctx=ctx) if not endpoint: continue earlier_pattern = None if not created: earlier_pattern = endpoint.matched_gf_patterns pattern = f'{earlier_pattern},{gf_pattern}' if earlier_pattern else gf_pattern endpoint.matched_gf_patterns = pattern endpoint.save() return all_urls def parse_curl_output(response): # TODO: Enrich from other cURL fields. CURL_REGEX_HTTP_STATUS = f'HTTP\/(?:(?:\d\.?)+)\s(\d+)\s(?:\w+)' http_status = 0 if response: failed = False regex = re.compile(CURL_REGEX_HTTP_STATUS, re.MULTILINE) try: http_status = int(regex.findall(response)[0]) except (KeyError, TypeError, IndexError): pass return { 'http_status': http_status, } @app.task(name='vulnerability_scan', queue='main_scan_queue', bind=True, base=RengineTask) def vulnerability_scan(self, urls=[], ctx={}, description=None): """ This function will serve as an entrypoint to vulnerability scan. All other vulnerability scan will be run from here including nuclei, crlfuzz, etc """ logger.info('Running Vulnerability Scan Queue') config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_run_nuclei = config.get(RUN_NUCLEI, True) should_run_crlfuzz = config.get(RUN_CRLFUZZ, False) should_run_dalfox = config.get(RUN_DALFOX, False) should_run_s3scanner = config.get(RUN_S3SCANNER, True) grouped_tasks = [] if should_run_nuclei: _task = nuclei_scan.si( urls=urls, ctx=ctx, description=f'Nuclei Scan' ) grouped_tasks.append(_task) if should_run_crlfuzz: _task = crlfuzz_scan.si( urls=urls, ctx=ctx, description=f'CRLFuzz Scan' ) grouped_tasks.append(_task) if should_run_dalfox: _task = dalfox_xss_scan.si( urls=urls, ctx=ctx, description=f'Dalfox XSS Scan' ) grouped_tasks.append(_task) if should_run_s3scanner: _task = s3scanner.si( ctx=ctx, description=f'Misconfigured S3 Buckets Scanner' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan completed...') # return results return None @app.task(name='nuclei_individual_severity_module', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_individual_severity_module(self, cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx={}, description=None): ''' This celery task will run vulnerability scan in parallel. All severities supplied should run in parallel as grouped tasks. ''' results = [] logger.info(f'Running vulnerability scan with severity: {severity}') cmd += f' -severity {severity}' # Send start notification notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) # Gather nuclei results vuln_data = parse_nuclei_result(line) # Get corresponding subdomain http_url = sanitize_url(line.get('matched-at')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) # Look for duplicate vulnerabilities by excluding records that might change but are irrelevant. object_comparison_exclude = ['response', 'curl_command', 'tags', 'references', 'cve_ids', 'cwe_ids'] # Add subdomain and target domain to the duplicate check vuln_data_copy = vuln_data.copy() vuln_data_copy['subdomain'] = subdomain vuln_data_copy['target_domain'] = self.domain # Check if record exists, if exists do not save it if record_exists(Vulnerability, data=vuln_data_copy, exclude_keys=object_comparison_exclude): logger.warning(f'Nuclei vulnerability of severity {severity} : {vuln_data_copy["name"]} for {subdomain_name} already exists') continue # Get or create EndPoint object response = line.get('response') httpx_crawl = False if response else enable_http_crawl # avoid yet another httpx crawl endpoint, _ = save_endpoint( http_url, crawl=httpx_crawl, subdomain=subdomain, ctx=ctx) if endpoint: http_url = endpoint.http_url if not httpx_crawl: output = parse_curl_output(response) endpoint.http_status = output['http_status'] endpoint.save() # Get or create Vulnerability object vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, subdomain=subdomain, **vuln_data) if not vuln: continue # Print vuln severity = line['info'].get('severity', 'unknown') logger.warning(str(vuln)) # Send notification for all vulnerabilities except info url = vuln.http_url or vuln.subdomain send_vuln = ( notif and notif.send_vuln_notif and vuln and severity in ['low', 'medium', 'high', 'critical']) if send_vuln: fields = { 'Severity': f'**{severity.upper()}**', 'URL': http_url, 'Subdomain': subdomain_name, 'Name': vuln.name, 'Type': vuln.type, 'Description': vuln.description, 'Template': vuln.template_url, 'Tags': vuln.get_tags_str(), 'CVEs': vuln.get_cve_str(), 'CWEs': vuln.get_cwe_str(), 'References': vuln.get_refs_str() } severity_map = { 'low': 'info', 'medium': 'warning', 'high': 'error', 'critical': 'error' } self.notify( f'vulnerability_scan_#{vuln.id}', severity_map[severity], fields, add_meta_info=False) # Send report to hackerone hackerone_query = Hackerone.objects.all() send_report = ( hackerone_query.exists() and severity not in ('info', 'low') and vuln.target_domain.h1_team_handle ) if send_report: hackerone = hackerone_query.first() if hackerone.send_critical and severity == 'critical': send_hackerone_report.delay(vuln.id) elif hackerone.send_high and severity == 'high': send_hackerone_report.delay(vuln.id) elif hackerone.send_medium and severity == 'medium': send_hackerone_report.delay(vuln.id) # Write results to JSON file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) # Send finish notif if send_status: vulns = Vulnerability.objects.filter(scan_history__id=self.scan_id) info_count = vulns.filter(severity=0).count() low_count = vulns.filter(severity=1).count() medium_count = vulns.filter(severity=2).count() high_count = vulns.filter(severity=3).count() critical_count = vulns.filter(severity=4).count() unknown_count = vulns.filter(severity=-1).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count + unknown_count fields = { 'Total': vulnerability_count, 'Critical': critical_count, 'High': high_count, 'Medium': medium_count, 'Low': low_count, 'Info': info_count, 'Unknown': unknown_count } self.notify(fields=fields) # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=NUCLEI ).exclude( severity=0 ) # find all unique vulnerabilities based on path and title # all unique vulnerability will go thru gpt function and get report # once report is got, it will be matched with other vulnerabilities and saved unique_vulns = set() for vuln in vulns: unique_vulns.add((vuln.name, vuln.get_path())) unique_vulns = list(unique_vulns) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in unique_vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return None def get_vulnerability_gpt_report(vuln): title = vuln[0] path = vuln[1] logger.info(f'Getting GPT Report for {title}, PATH: {path}') # check if in db already exists stored = GPTVulnerabilityReport.objects.filter( url_path=path ).filter( title=title ).first() if stored: response = { 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: report = GPTVulnerabilityReportGenerator() vulnerability_description = get_gpt_vuln_input_description( title, path ) response = report.get_vulnerability_description(vulnerability_description) add_gpt_description_db( title, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) for vuln in Vulnerability.objects.filter(name=title, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() def add_gpt_description_db(title, path, description, impact, remediation, references): gpt_report = GPTVulnerabilityReport() gpt_report.url_path = path gpt_report.title = title gpt_report.description = description gpt_report.impact = impact gpt_report.remediation = remediation gpt_report.save() for url in references: ref, created = VulnerabilityReference.objects.get_or_create(url=url) gpt_report.references.add(ref) gpt_report.save() @app.task(name='nuclei_scan', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_scan(self, urls=[], ctx={}, description=None): """HTTP vulnerability scan using Nuclei Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. Notes: Unfurl the urls to keep only domain and path, will be sent to vuln scan and ignore certain file extensions. Thanks: https://github.com/six2dez/reconftw """ # Config config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} input_path = f'{self.results_dir}/input_endpoints_vulnerability_scan.txt' enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) concurrency = config.get(NUCLEI_CONCURRENCY) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) retries = config.get(RETRIES) or self.yaml_configuration.get(RETRIES, DEFAULT_RETRIES) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) custom_header = config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) should_fetch_gpt_report = config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) proxy = get_random_proxy() nuclei_specific_config = config.get('nuclei', {}) use_nuclei_conf = nuclei_specific_config.get(USE_CONFIG, False) severities = nuclei_specific_config.get(NUCLEI_SEVERITY, NUCLEI_DEFAULT_SEVERITIES) tags = nuclei_specific_config.get(NUCLEI_TAGS, []) tags = ','.join(tags) nuclei_templates = nuclei_specific_config.get(NUCLEI_TEMPLATE) custom_nuclei_templates = nuclei_specific_config.get(NUCLEI_CUSTOM_TEMPLATE) # severities_str = ','.join(severities) # Get alive endpoints if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=enable_http_crawl, ignore_files=True, write_filepath=input_path, ctx=ctx ) if intensity == 'normal': # reduce number of endpoints to scan unfurl_filter = f'{self.results_dir}/urls_unfurled.txt' run_command( f"cat {input_path} | unfurl -u format %s://%d%p |uro > {unfurl_filter}", shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {unfurl_filter} -o {unfurl_filter}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) input_path = unfurl_filter # Build templates # logger.info('Updating Nuclei templates ...') run_command( 'nuclei -update-templates', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) templates = [] if not (nuclei_templates or custom_nuclei_templates): templates.append(NUCLEI_DEFAULT_TEMPLATES_PATH) if nuclei_templates: if ALL in nuclei_templates: template = NUCLEI_DEFAULT_TEMPLATES_PATH templates.append(template) else: templates.extend(nuclei_templates) if custom_nuclei_templates: custom_nuclei_template_paths = [f'{str(elem)}.yaml' for elem in custom_nuclei_templates] template = templates.extend(custom_nuclei_template_paths) # Build CMD cmd = 'nuclei -j' cmd += ' -config /root/.config/nuclei/config.yaml' if use_nuclei_conf else '' cmd += f' -irr' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -l {input_path}' cmd += f' -c {str(concurrency)}' if concurrency > 0 else '' cmd += f' -proxy {proxy} ' if proxy else '' cmd += f' -retries {retries}' if retries > 0 else '' cmd += f' -rl {rate_limit}' if rate_limit > 0 else '' # cmd += f' -severity {severities_str}' cmd += f' -timeout {str(timeout)}' if timeout and timeout > 0 else '' cmd += f' -tags {tags}' if tags else '' cmd += f' -silent' for tpl in templates: cmd += f' -t {tpl}' grouped_tasks = [] custom_ctx = ctx for severity in severities: custom_ctx['track'] = True _task = nuclei_individual_severity_module.si( cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx=custom_ctx, description=f'Nuclei Scan with severity {severity}' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan with all severities completed...') return None @app.task(name='dalfox_xss_scan', queue='main_scan_queue', base=RengineTask, bind=True) def dalfox_xss_scan(self, urls=[], ctx={}, description=None): """XSS Scan using dalfox Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) dalfox_config = vuln_config.get(DALFOX) or {} custom_header = dalfox_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() is_waf_evasion = dalfox_config.get(WAF_EVASION, False) blind_xss_server = dalfox_config.get(BLIND_XSS_SERVER) user_agent = dalfox_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) timeout = dalfox_config.get(TIMEOUT) delay = dalfox_config.get(DELAY) threads = dalfox_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_dalfox_xss.txt' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=False, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'dalfox --silence --no-color --no-spinner' cmd += f' --only-poc r ' cmd += f' --ignore-return 302,404,403' cmd += f' --skip-bav' cmd += f' file {input_path}' cmd += f' --proxy {proxy}' if proxy else '' cmd += f' --waf-evasion' if is_waf_evasion else '' cmd += f' -b {blind_xss_server}' if blind_xss_server else '' cmd += f' --delay {delay}' if delay else '' cmd += f' --timeout {timeout}' if timeout else '' cmd += f' --user-agent {user_agent}' if user_agent else '' cmd += f' --header {custom_header}' if custom_header else '' cmd += f' --worker {threads}' if threads else '' cmd += f' --format json' results = [] for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id, trunc_char=',' ): if not isinstance(line, dict): continue results.append(line) vuln_data = parse_dalfox_result(line) http_url = sanitize_url(line.get('data')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Dalfox Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=DALFOX ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='crlfuzz_scan', queue='main_scan_queue', base=RengineTask, bind=True) def crlfuzz_scan(self, urls=[], ctx={}, description=None): """CRLF Fuzzing with CRLFuzz Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) custom_header = vuln_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() user_agent = vuln_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) threads = vuln_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_crlf.txt' output_path = f'{self.results_dir}/{self.filename}' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=True, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'crlfuzz -s' cmd += f' -l {input_path}' cmd += f' -x {proxy}' if proxy else '' cmd += f' --H {custom_header}' if custom_header else '' cmd += f' -o {output_path}' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id ) if not os.path.isfile(output_path): logger.info('No Results from CRLFuzz') return crlfs = [] results = [] with open(output_path, 'r') as file: crlfs = file.readlines() for crlf in crlfs: url = crlf.strip() vuln_data = parse_crlfuzz_result(url) http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting CRLFuzz Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=CRLFUZZ ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='s3scanner', queue='main_scan_queue', base=RengineTask, bind=True) def s3scanner(self, ctx={}, description=None): """Bucket Scanner Args: ctx (dict): Context description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/#{self.scan_id}_subdomain_discovery.txt' vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} s3_config = vuln_config.get(S3SCANNER) or {} threads = s3_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) providers = s3_config.get(PROVIDERS, S3SCANNER_DEFAULT_PROVIDERS) scan_history = ScanHistory.objects.filter(pk=self.scan_id).first() for provider in providers: cmd = f's3scanner -bucket-file {input_path} -enumerate -provider {provider} -threads {threads} -json' for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue if line.get('bucket', {}).get('exists', 0) == 1: result = parse_s3scanner_result(line) s3bucket, created = S3Bucket.objects.get_or_create(**result) scan_history.buckets.add(s3bucket) logger.info(f"s3 bucket added {result['provider']}-{result['name']}-{result['region']}") @app.task(name='http_crawl', queue='main_scan_queue', base=RengineTask, bind=True) def http_crawl( self, urls=[], method=None, recrawl=False, ctx={}, track=True, description=None, is_ran_from_subdomain_scan=False, should_remove_duplicate_endpoints=True, duplicate_removal_fields=[]): """Use httpx to query HTTP URLs for important info like page titles, http status, etc... Args: urls (list, optional): A set of URLs to check. Overrides default behavior which queries all endpoints related to this scan. method (str): HTTP method to use (GET, HEAD, POST, PUT, DELETE). recrawl (bool, optional): If False, filter out URLs that have already been crawled. should_remove_duplicate_endpoints (bool): Whether to remove duplicate endpoints duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates Returns: list: httpx results. """ logger.info('Initiating HTTP Crawl') if is_ran_from_subdomain_scan: logger.info('Running From Subdomain Scan...') cmd = '/go/bin/httpx' cfg = self.yaml_configuration.get(HTTP_CRAWL) or {} custom_header = cfg.get(CUSTOM_HEADER, '') threads = cfg.get(THREADS, DEFAULT_THREADS) follow_redirect = cfg.get(FOLLOW_REDIRECT, True) self.output_path = None input_path = f'{self.results_dir}/httpx_input.txt' history_file = f'{self.results_dir}/commands.txt' if urls: # direct passing URLs to check if self.url_filter: urls = [u for u in urls if self.url_filter in u] with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_uncrawled=not recrawl, write_filepath=input_path, ctx=ctx ) # logger.debug(urls) # If no URLs found, skip it if not urls: return # Re-adjust thread number if few URLs to avoid spinning up a monster to # kill a fly. if len(urls) < threads: threads = len(urls) # Get random proxy proxy = get_random_proxy() # Run command cmd += f' -cl -ct -rt -location -td -websocket -cname -asn -cdn -probe -random-agent' cmd += f' -t {threads}' if threads > 0 else '' cmd += f' --http-proxy {proxy}' if proxy else '' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -json' cmd += f' -u {urls[0]}' if len(urls) == 1 else f' -l {input_path}' cmd += f' -x {method}' if method else '' cmd += f' -silent' if follow_redirect: cmd += ' -fr' results = [] endpoint_ids = [] for line in stream_command( cmd, history_file=history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not line or not isinstance(line, dict): continue logger.debug(line) # No response from endpoint if line.get('failed', False): continue # Parse httpx output host = line.get('host', '') content_length = line.get('content_length', 0) http_status = line.get('status_code') http_url, is_redirect = extract_httpx_url(line) page_title = line.get('title') webserver = line.get('webserver') cdn = line.get('cdn', False) rt = line.get('time') techs = line.get('tech', []) cname = line.get('cname', '') content_type = line.get('content_type', '') response_time = -1 if rt: response_time = float(''.join(ch for ch in rt if not ch.isalpha())) if rt[-2:] == 'ms': response_time = response_time / 1000 # Create Subdomain object in DB subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue # Save default HTTP URL to endpoint object in DB endpoint, created = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain, is_default=is_ran_from_subdomain_scan ) if not endpoint: continue endpoint.http_status = http_status endpoint.page_title = page_title endpoint.content_length = content_length endpoint.webserver = webserver endpoint.response_time = response_time endpoint.content_type = content_type endpoint.save() endpoint_str = f'{http_url} [{http_status}] `{content_length}B` `{webserver}` `{rt}`' logger.warning(endpoint_str) if endpoint and endpoint.is_alive and endpoint.http_status != 403: self.notify( fields={'Alive endpoint': f'• {endpoint_str}'}, add_meta_info=False) # Add endpoint to results line['_cmd'] = cmd line['final_url'] = http_url line['endpoint_id'] = endpoint.id line['endpoint_created'] = created line['is_redirect'] = is_redirect results.append(line) # Add technology objects to DB for technology in techs: tech, _ = Technology.objects.get_or_create(name=technology) endpoint.techs.add(tech) if is_ran_from_subdomain_scan: subdomain.technologies.add(tech) subdomain.save() endpoint.save() techs_str = ', '.join([f'`{tech}`' for tech in techs]) self.notify( fields={'Technologies': techs_str}, add_meta_info=False) # Add IP objects for 'a' records to DB a_records = line.get('a', []) for ip_address in a_records: ip, created = save_ip_address( ip_address, subdomain, subscan=self.subscan, cdn=cdn) ips_str = '• ' + '\n• '.join([f'`{ip}`' for ip in a_records]) self.notify( fields={'IPs': ips_str}, add_meta_info=False) # Add IP object for host in DB if host: ip, created = save_ip_address( host, subdomain, subscan=self.subscan, cdn=cdn) self.notify( fields={'IPs': f'• `{ip.address}`'}, add_meta_info=False) # Save subdomain and endpoint if is_ran_from_subdomain_scan: # save subdomain stuffs subdomain.http_url = http_url subdomain.http_status = http_status subdomain.page_title = page_title subdomain.content_length = content_length subdomain.webserver = webserver subdomain.response_time = response_time subdomain.content_type = content_type subdomain.cname = ','.join(cname) subdomain.is_cdn = cdn if cdn: subdomain.cdn_name = line.get('cdn_name') subdomain.save() endpoint.save() endpoint_ids.append(endpoint.id) if should_remove_duplicate_endpoints: # Remove 'fake' alive endpoints that are just redirects to the same page remove_duplicate_endpoints( self.scan_id, self.domain_id, self.subdomain_id, filter_ids=endpoint_ids ) # Remove input file run_command( f'rm {input_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) return results #---------------------# # Notifications tasks # #---------------------# @app.task(name='send_notif', bind=False, queue='send_notif_queue') def send_notif( message, scan_history_id=None, subscan_id=None, **options): if not 'title' in options: message = enrich_notification(message, scan_history_id, subscan_id) send_discord_message(message, **options) send_slack_message(message) send_telegram_message(message) @app.task(name='send_scan_notif', bind=False, queue='send_scan_notif_queue') def send_scan_notif( scan_history_id, subscan_id=None, engine_id=None, status='RUNNING'): """Send scan status notification. Works for scan or a subscan if subscan_id is passed. Args: scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Get domain, engine, scan_history objects engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() tasks = ScanActivity.objects.filter(scan_of=scan) if scan else 0 # Build notif options url = get_scan_url(scan_history_id, subscan_id) title = get_scan_title(scan_history_id, subscan_id) fields = get_scan_fields(engine, scan, subscan, status, tasks) severity = None msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) if status: severity = STATUS_TO_SEVERITIES.get(status) opts = { 'title': title, 'url': url, 'fields': fields, 'severity': severity } logger.warning(f'Sending notification "{title}" [{severity}]') # Send notification send_notif( msg, scan_history_id, subscan_id, **opts) @app.task(name='send_task_notif', bind=False, queue='send_task_notif_queue') def send_task_notif( task_name, status=None, result=None, output_path=None, traceback=None, scan_history_id=None, engine_id=None, subscan_id=None, severity=None, add_meta_info=True, update_fields={}): """Send task status notification. Args: task_name (str): Task name. status (str, optional): Task status. result (str, optional): Task result. output_path (str, optional): Task output path. traceback (str, optional): Task traceback. scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. severity (str, optional): Severity (will be mapped to notif colors) add_meta_info (bool, optional): Wheter to add scan / subscan info to notif. update_fields (dict, optional): Fields key / value to update. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Build fields url = None fields = {} if add_meta_info: engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() url = get_scan_url(scan_history_id) if status: fields['Status'] = f'**{status}**' if engine: fields['Engine'] = engine.engine_name if scan: fields['Scan ID'] = f'[#{scan.id}]({url})' if subscan: url = get_scan_url(scan_history_id, subscan_id) fields['Subscan ID'] = f'[#{subscan.id}]({url})' title = get_task_title(task_name, scan_history_id, subscan_id) if status: severity = STATUS_TO_SEVERITIES.get(status) msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) # Add fields to update for k, v in update_fields.items(): fields[k] = v # Add traceback to notif if traceback and notif.send_scan_tracebacks: fields['Traceback'] = f'```\n{traceback}\n```' # Add files to notif files = [] attach_file = ( notif.send_scan_output_file and output_path and result and not traceback ) if attach_file: output_title = output_path.split('/')[-1] files = [(output_path, output_title)] # Send notif opts = { 'title': title, 'url': url, 'files': files, 'severity': severity, 'fields': fields, 'fields_append': update_fields.keys() } send_notif( msg, scan_history_id=scan_history_id, subscan_id=subscan_id, **opts) @app.task(name='send_file_to_discord', bind=False, queue='send_file_to_discord_queue') def send_file_to_discord(file_path, title=None): notif = Notification.objects.first() do_send = notif and notif.send_to_discord and notif.discord_hook_url if not do_send: return False webhook = DiscordWebhook( url=notif.discord_hook_url, rate_limit_retry=True, username=title or "reNgine Discord Plugin" ) with open(file_path, "rb") as f: head, tail = os.path.split(file_path) webhook.add_file(file=f.read(), filename=tail) webhook.execute() @app.task(name='send_hackerone_report', bind=False, queue='send_hackerone_report_queue') def send_hackerone_report(vulnerability_id): """Send HackerOne vulnerability report. Args: vulnerability_id (int): Vulnerability id. Returns: int: HTTP response status code. """ vulnerability = Vulnerability.objects.get(id=vulnerability_id) severities = {v: k for k,v in NUCLEI_SEVERITY_MAP.items()} headers = { 'Content-Type': 'application/json', 'Accept': 'application/json' } # can only send vulnerability report if team_handle exists if len(vulnerability.target_domain.h1_team_handle) !=0: hackerone_query = Hackerone.objects.all() if hackerone_query.exists(): hackerone = Hackerone.objects.first() severity_value = severities[vulnerability.severity] tpl = hackerone.report_template # Replace syntax of report template with actual content tpl = tpl.replace('{vulnerability_name}', vulnerability.name) tpl = tpl.replace('{vulnerable_url}', vulnerability.http_url) tpl = tpl.replace('{vulnerability_severity}', severity_value) tpl = tpl.replace('{vulnerability_description}', vulnerability.description if vulnerability.description else '') tpl = tpl.replace('{vulnerability_extracted_results}', vulnerability.extracted_results if vulnerability.extracted_results else '') tpl = tpl.replace('{vulnerability_reference}', vulnerability.reference if vulnerability.reference else '') data = { "data": { "type": "report", "attributes": { "team_handle": vulnerability.target_domain.h1_team_handle, "title": '{} found in {}'.format(vulnerability.name, vulnerability.http_url), "vulnerability_information": tpl, "severity_rating": severity_value, "impact": "More information about the impact and vulnerability can be found here: \n" + vulnerability.reference if vulnerability.reference else "NA", } } } r = requests.post( 'https://api.hackerone.com/v1/hackers/reports', auth=(hackerone.username, hackerone.api_key), json=data, headers=headers ) response = r.json() status_code = r.status_code if status_code == 201: vulnerability.hackerone_report_id = response['data']["id"] vulnerability.open_status = False vulnerability.save() return status_code else: logger.error('No team handle found.') status_code = 111 return status_code #-------------# # Utils tasks # #-------------# @app.task(name='parse_nmap_results', bind=False, queue='parse_nmap_results_queue') def parse_nmap_results(xml_file, output_file=None): """Parse results from nmap output file. Args: xml_file (str): nmap XML report file path. Returns: list: List of vulnerabilities found from nmap results. """ with open(xml_file, encoding='utf8') as f: content = f.read() try: nmap_results = xmltodict.parse(content) # parse XML to dict except Exception as e: logger.exception(e) logger.error(f'Cannot parse {xml_file} to valid JSON. Skipping.') return [] # Write JSON to output file if output_file: with open(output_file, 'w') as f: json.dump(nmap_results, f, indent=4) logger.warning(json.dumps(nmap_results, indent=4)) hosts = ( nmap_results .get('nmaprun', {}) .get('host', {}) ) all_vulns = [] if isinstance(hosts, dict): hosts = [hosts] for host in hosts: # Grab hostname / IP from output hostnames_dict = host.get('hostnames', {}) if hostnames_dict: # Ensure that hostnames['hostname'] is a list for consistency hostnames_list = hostnames_dict['hostname'] if isinstance(hostnames_dict['hostname'], list) else [hostnames_dict['hostname']] # Extract all the @name values from the list of dictionaries hostnames = [entry.get('@name') for entry in hostnames_list] else: hostnames = [host.get('address')['@addr']] # Iterate over each hostname for each port for hostname in hostnames: # Grab ports from output ports = host.get('ports', {}).get('port', []) if isinstance(ports, dict): ports = [ports] for port in ports: url_vulns = [] port_number = port['@portid'] url = sanitize_url(f'{hostname}:{port_number}') logger.info(f'Parsing nmap results for {hostname}:{port_number} ...') if not port_number or not port_number.isdigit(): continue port_protocol = port['@protocol'] scripts = port.get('script', []) if isinstance(scripts, dict): scripts = [scripts] for script in scripts: script_id = script['@id'] script_output = script['@output'] script_output_table = script.get('table', []) logger.debug(f'Ran nmap script "{script_id}" on {port_number}/{port_protocol}:\n{script_output}\n') if script_id == 'vulscan': vulns = parse_nmap_vulscan_output(script_output) url_vulns.extend(vulns) elif script_id == 'vulners': vulns = parse_nmap_vulners_output(script_output) url_vulns.extend(vulns) # elif script_id == 'http-server-header': # TODO: nmap can help find technologies as well using the http-server-header script # regex = r'(\w+)/([\d.]+)\s?(?:\((\w+)\))?' # tech_name, tech_version, tech_os = re.match(regex, test_string).groups() # Technology.objects.get_or_create(...) # elif script_id == 'http_csrf': # vulns = parse_nmap_http_csrf_output(script_output) # url_vulns.extend(vulns) else: logger.warning(f'Script output parsing for script "{script_id}" is not supported yet.') # Add URL to vuln for vuln in url_vulns: # TODO: This should extend to any URL, not just HTTP vuln['http_url'] = url if 'http_path' in vuln: vuln['http_url'] += vuln['http_path'] all_vulns.append(vuln) return all_vulns def parse_nmap_http_csrf_output(script_output): pass def parse_nmap_vulscan_output(script_output): """Parse nmap vulscan script output. Args: script_output (str): Vulscan script output. Returns: list: List of Vulnerability dicts. """ data = {} vulns = [] provider_name = '' # Sort all vulns found by provider so that we can match each provider with # a function that pulls from its API to get more info about the # vulnerability. for line in script_output.splitlines(): if not line: continue if not line.startswith('['): # provider line if "No findings" in line: logger.info(f"No findings: {line}") continue elif ' - ' in line: provider_name, provider_url = tuple(line.split(' - ')) data[provider_name] = {'url': provider_url.rstrip(':'), 'entries': []} continue else: # Log a warning logger.warning(f"Unexpected line format: {line}") continue reg = r'\[(.*)\] (.*)' matches = re.match(reg, line) id, title = matches.groups() entry = {'id': id, 'title': title} data[provider_name]['entries'].append(entry) logger.warning('Vulscan parsed output:') logger.warning(pprint.pformat(data)) for provider_name in data: if provider_name == 'Exploit-DB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'IBM X-Force': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'MITRE CVE': logger.error(f'Provider {provider_name} is not supported YET.') for entry in data[provider_name]['entries']: cve_id = entry['id'] vuln = cve_to_vuln(cve_id) vulns.append(vuln) elif provider_name == 'OSVDB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'OpenVAS (Nessus)': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'SecurityFocus': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'VulDB': logger.error(f'Provider {provider_name} is not supported YET.') pass else: logger.error(f'Provider {provider_name} is not supported.') return vulns def parse_nmap_vulners_output(script_output, url=''): """Parse nmap vulners script output. TODO: Rework this as it's currently matching all CVEs no matter the confidence. Args: script_output (str): Script output. Returns: list: List of found vulnerabilities. """ vulns = [] # Check for CVE in script output CVE_REGEX = re.compile(r'.*(CVE-\d\d\d\d-\d+).*') matches = CVE_REGEX.findall(script_output) matches = list(dict.fromkeys(matches)) for cve_id in matches: # get CVE info vuln = cve_to_vuln(cve_id, vuln_type='nmap-vulners-nse') if vuln: vulns.append(vuln) return vulns def cve_to_vuln(cve_id, vuln_type=''): """Search for a CVE using CVESearch and return Vulnerability data. Args: cve_id (str): CVE ID in the form CVE-* Returns: dict: Vulnerability dict. """ cve_info = CVESearch('https://cve.circl.lu').id(cve_id) if not cve_info: logger.error(f'Could not fetch CVE info for cve {cve_id}. Skipping.') return None vuln_cve_id = cve_info['id'] vuln_name = vuln_cve_id vuln_description = cve_info.get('summary', 'none').replace(vuln_cve_id, '').strip() try: vuln_cvss = float(cve_info.get('cvss', -1)) except (ValueError, TypeError): vuln_cvss = -1 vuln_cwe_id = cve_info.get('cwe', '') exploit_ids = cve_info.get('refmap', {}).get('exploit-db', []) osvdb_ids = cve_info.get('refmap', {}).get('osvdb', []) references = cve_info.get('references', []) capec_objects = cve_info.get('capec', []) # Parse ovals for a better vuln name / type ovals = cve_info.get('oval', []) if ovals: vuln_name = ovals[0]['title'] vuln_type = ovals[0]['family'] # Set vulnerability severity based on CVSS score vuln_severity = 'info' if vuln_cvss < 4: vuln_severity = 'low' elif vuln_cvss < 7: vuln_severity = 'medium' elif vuln_cvss < 9: vuln_severity = 'high' else: vuln_severity = 'critical' # Build console warning message msg = f'{vuln_name} | {vuln_severity.upper()} | {vuln_cve_id} | {vuln_cwe_id} | {vuln_cvss}' for id in osvdb_ids: msg += f'\n\tOSVDB: {id}' for exploit_id in exploit_ids: msg += f'\n\tEXPLOITDB: {exploit_id}' logger.warning(msg) vuln = { 'name': vuln_name, 'type': vuln_type, 'severity': NUCLEI_SEVERITY_MAP[vuln_severity], 'description': vuln_description, 'cvss_score': vuln_cvss, 'references': references, 'cve_ids': [vuln_cve_id], 'cwe_ids': [vuln_cwe_id] } return vuln def parse_s3scanner_result(line): ''' Parses and returns s3Scanner Data ''' bucket = line['bucket'] return { 'name': bucket['name'], 'region': bucket['region'], 'provider': bucket['provider'], 'owner_display_name': bucket['owner_display_name'], 'owner_id': bucket['owner_id'], 'perm_auth_users_read': bucket['perm_auth_users_read'], 'perm_auth_users_write': bucket['perm_auth_users_write'], 'perm_auth_users_read_acl': bucket['perm_auth_users_read_acl'], 'perm_auth_users_write_acl': bucket['perm_auth_users_write_acl'], 'perm_auth_users_full_control': bucket['perm_auth_users_full_control'], 'perm_all_users_read': bucket['perm_all_users_read'], 'perm_all_users_write': bucket['perm_all_users_write'], 'perm_all_users_read_acl': bucket['perm_all_users_read_acl'], 'perm_all_users_write_acl': bucket['perm_all_users_write_acl'], 'perm_all_users_full_control': bucket['perm_all_users_full_control'], 'num_objects': bucket['num_objects'], 'size': bucket['bucket_size'] } def parse_nuclei_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ return { 'name': line['info'].get('name', ''), 'type': line['type'], 'severity': NUCLEI_SEVERITY_MAP[line['info'].get('severity', 'unknown')], 'template': line['template'], 'template_url': line['template-url'], 'template_id': line['template-id'], 'description': line['info'].get('description', ''), 'matcher_name': line.get('matcher-name', ''), 'curl_command': line.get('curl-command'), 'request': line.get('request'), 'response': line.get('response'), 'extracted_results': line.get('extracted-results', []), 'cvss_metrics': line['info'].get('classification', {}).get('cvss-metrics', ''), 'cvss_score': line['info'].get('classification', {}).get('cvss-score'), 'cve_ids': line['info'].get('classification', {}).get('cve_id', []) or [], 'cwe_ids': line['info'].get('classification', {}).get('cwe_id', []) or [], 'references': line['info'].get('reference', []) or [], 'tags': line['info'].get('tags', []), 'source': NUCLEI, } def parse_dalfox_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ description = '' description += f" Evidence: {line.get('evidence')} <br>" if line.get('evidence') else '' description += f" Message: {line.get('message')} <br>" if line.get('message') else '' description += f" Payload: {line.get('message_str')} <br>" if line.get('message_str') else '' description += f" Vulnerable Parameter: {line.get('param')} <br>" if line.get('param') else '' return { 'name': 'XSS (Cross Site Scripting)', 'type': 'XSS', 'severity': DALFOX_SEVERITY_MAP[line.get('severity', 'unknown')], 'description': description, 'source': DALFOX, 'cwe_ids': [line.get('cwe')] } def parse_crlfuzz_result(url): """Parse CRLF results Args: url (str): CRLF Vulnerable URL Returns: dict: Vulnerability data. """ return { 'name': 'CRLF (HTTP Response Splitting)', 'type': 'CRLF', 'severity': 2, 'description': 'A CRLF (HTTP Response Splitting) vulnerability has been discovered.', 'source': CRLFUZZ, } def record_exists(model, data, exclude_keys=[]): """ Check if a record already exists in the database based on the given data. Args: model (django.db.models.Model): The Django model to check against. data (dict): Data dictionary containing fields and values. exclude_keys (list): List of keys to exclude from the lookup. Returns: bool: True if the record exists, False otherwise. """ # Extract the keys that will be used for the lookup lookup_fields = {key: data[key] for key in data if key not in exclude_keys} # Return True if a record exists based on the lookup fields, False otherwise return model.objects.filter(**lookup_fields).exists() @app.task(name='geo_localize', bind=False, queue='geo_localize_queue') def geo_localize(host, ip_id=None): """Uses geoiplookup to find location associated with host. Args: host (str): Hostname. ip_id (int): IpAddress object id. Returns: startScan.models.CountryISO: CountryISO object from DB or None. """ if validators.ipv6(host): logger.info(f'Ipv6 "{host}" is not supported by geoiplookup. Skipping.') return None cmd = f'geoiplookup {host}' _, out = run_command(cmd) if 'IP Address not found' not in out and "can't resolve hostname" not in out: country_iso = out.split(':')[1].strip().split(',')[0] country_name = out.split(':')[1].strip().split(',')[1].strip() geo_object, _ = CountryISO.objects.get_or_create( iso=country_iso, name=country_name ) geo_json = { 'iso': country_iso, 'name': country_name } if ip_id: ip = IpAddress.objects.get(pk=ip_id) ip.geo_iso = geo_object ip.save() return geo_json logger.info(f'Geo IP lookup failed for host "{host}"') return None @app.task(name='query_whois', bind=False, queue='query_whois_queue') def query_whois(ip_domain, force_reload_whois=False): """Query WHOIS information for an IP or a domain name. Args: ip_domain (str): IP address or domain name. save_domain (bool): Whether to save domain or not, default False Returns: dict: WHOIS information. """ if not force_reload_whois and Domain.objects.filter(name=ip_domain).exists() and Domain.objects.get(name=ip_domain).domain_info: domain = Domain.objects.get(name=ip_domain) if not domain.insert_date: domain.insert_date = timezone.now() domain.save() domain_info_db = domain.domain_info domain_info = DottedDict( dnssec=domain_info_db.dnssec, created=domain_info_db.created, updated=domain_info_db.updated, expires=domain_info_db.expires, geolocation_iso=domain_info_db.geolocation_iso, status=[status['name'] for status in DomainWhoisStatusSerializer(domain_info_db.status, many=True).data], whois_server=domain_info_db.whois_server, ns_records=[ns['name'] for ns in NameServersSerializer(domain_info_db.name_servers, many=True).data], registrar_name=domain_info_db.registrar.name, registrar_phone=domain_info_db.registrar.phone, registrar_email=domain_info_db.registrar.email, registrar_url=domain_info_db.registrar.url, registrant_name=domain_info_db.registrant.name, registrant_id=domain_info_db.registrant.id_str, registrant_organization=domain_info_db.registrant.organization, registrant_city=domain_info_db.registrant.city, registrant_state=domain_info_db.registrant.state, registrant_zip_code=domain_info_db.registrant.zip_code, registrant_country=domain_info_db.registrant.country, registrant_phone=domain_info_db.registrant.phone, registrant_fax=domain_info_db.registrant.fax, registrant_email=domain_info_db.registrant.email, registrant_address=domain_info_db.registrant.address, admin_name=domain_info_db.admin.name, admin_id=domain_info_db.admin.id_str, admin_organization=domain_info_db.admin.organization, admin_city=domain_info_db.admin.city, admin_state=domain_info_db.admin.state, admin_zip_code=domain_info_db.admin.zip_code, admin_country=domain_info_db.admin.country, admin_phone=domain_info_db.admin.phone, admin_fax=domain_info_db.admin.fax, admin_email=domain_info_db.admin.email, admin_address=domain_info_db.admin.address, tech_name=domain_info_db.tech.name, tech_id=domain_info_db.tech.id_str, tech_organization=domain_info_db.tech.organization, tech_city=domain_info_db.tech.city, tech_state=domain_info_db.tech.state, tech_zip_code=domain_info_db.tech.zip_code, tech_country=domain_info_db.tech.country, tech_phone=domain_info_db.tech.phone, tech_fax=domain_info_db.tech.fax, tech_email=domain_info_db.tech.email, tech_address=domain_info_db.tech.address, related_tlds=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_tlds, many=True).data], related_domains=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_domains, many=True).data], historical_ips=[ip for ip in HistoricalIPSerializer(domain_info_db.historical_ips, many=True).data], ) if domain_info_db.dns_records: a_records = [] txt_records = [] mx_records = [] dns_records = [{'name': dns['name'], 'type': dns['type']} for dns in DomainDNSRecordSerializer(domain_info_db.dns_records, many=True).data] for dns in dns_records: if dns['type'] == 'a': a_records.append(dns['name']) elif dns['type'] == 'txt': txt_records.append(dns['name']) elif dns['type'] == 'mx': mx_records.append(dns['name']) domain_info.a_records = a_records domain_info.txt_records = txt_records domain_info.mx_records = mx_records else: logger.info(f'Domain info for "{ip_domain}" not found in DB, querying whois') domain_info = DottedDict() # find domain historical ip try: historical_ips = get_domain_historical_ip_address(ip_domain) domain_info.historical_ips = historical_ips except Exception as e: logger.error(f'HistoricalIP for {ip_domain} not found!\nError: {str(e)}') historical_ips = [] # find associated domains using ip_domain try: related_domains = reverse_whois(ip_domain.split('.')[0]) except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] # find related tlds using TLSx try: related_tlds = [] output_path = '/tmp/ip_domain_tlsx.txt' tlsx_command = f'tlsx -san -cn -silent -ro -host {ip_domain} -o {output_path}' run_command( tlsx_command, shell=True, ) tlsx_output = [] with open(output_path) as f: tlsx_output = f.readlines() tldextract_target = tldextract.extract(ip_domain) for doms in tlsx_output: doms = doms.strip() tldextract_res = tldextract.extract(doms) if ip_domain != doms and tldextract_res.domain == tldextract_target.domain and tldextract_res.subdomain == '': related_tlds.append(doms) related_tlds = list(set(related_tlds)) domain_info.related_tlds = related_tlds except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] related_domains_list = [] if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_related = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_related) related_domains_list.append(_domain['name']) for _domain in related_tlds: domain_related = RelatedDomain.objects.get_or_create( name=_domain, )[0] db_domain_info.related_tlds.add(domain_related) for _ip in historical_ips: historical_ip = HistoricalIP.objects.get_or_create( ip=_ip['ip'], owner=_ip['owner'], location=_ip['location'], last_seen=_ip['last_seen'], )[0] db_domain_info.historical_ips.add(historical_ip) domain.domain_info = db_domain_info domain.save() command = f'netlas host {ip_domain} -f json' # check if netlas key is provided netlas_key = get_netlas_key() command += f' -a {netlas_key}' if netlas_key else '' result = subprocess.check_output(command.split()).decode('utf-8') if 'Failed to parse response data' in result: # do fallback return { 'status': False, 'ip_domain': ip_domain, 'result': "Netlas limit exceeded.", 'message': 'Netlas limit exceeded.' } try: result = json.loads(result) logger.info(result) whois = result.get('whois') if result.get('whois') else {} domain_info.created = whois.get('created_date') domain_info.expires = whois.get('expiration_date') domain_info.updated = whois.get('updated_date') domain_info.whois_server = whois.get('whois_server') if 'registrant' in whois: registrant = whois.get('registrant') domain_info.registrant_name = registrant.get('name') domain_info.registrant_country = registrant.get('country') domain_info.registrant_id = registrant.get('id') domain_info.registrant_state = registrant.get('province') domain_info.registrant_city = registrant.get('city') domain_info.registrant_phone = registrant.get('phone') domain_info.registrant_address = registrant.get('street') domain_info.registrant_organization = registrant.get('organization') domain_info.registrant_fax = registrant.get('fax') domain_info.registrant_zip_code = registrant.get('postal_code') email_search = EMAIL_REGEX.search(str(registrant.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.registrant_email = field_content if 'administrative' in whois: administrative = whois.get('administrative') domain_info.admin_name = administrative.get('name') domain_info.admin_country = administrative.get('country') domain_info.admin_id = administrative.get('id') domain_info.admin_state = administrative.get('province') domain_info.admin_city = administrative.get('city') domain_info.admin_phone = administrative.get('phone') domain_info.admin_address = administrative.get('street') domain_info.admin_organization = administrative.get('organization') domain_info.admin_fax = administrative.get('fax') domain_info.admin_zip_code = administrative.get('postal_code') mail_search = EMAIL_REGEX.search(str(administrative.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.admin_email = field_content if 'technical' in whois: technical = whois.get('technical') domain_info.tech_name = technical.get('name') domain_info.tech_country = technical.get('country') domain_info.tech_state = technical.get('province') domain_info.tech_id = technical.get('id') domain_info.tech_city = technical.get('city') domain_info.tech_phone = technical.get('phone') domain_info.tech_address = technical.get('street') domain_info.tech_organization = technical.get('organization') domain_info.tech_fax = technical.get('fax') domain_info.tech_zip_code = technical.get('postal_code') mail_search = EMAIL_REGEX.search(str(technical.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.tech_email = field_content if 'dns' in result: dns = result.get('dns') domain_info.mx_records = dns.get('mx') domain_info.txt_records = dns.get('txt') domain_info.a_records = dns.get('a') domain_info.ns_records = whois.get('name_servers') domain_info.dnssec = True if whois.get('dnssec') else False domain_info.status = whois.get('status') if 'registrar' in whois: registrar = whois.get('registrar') domain_info.registrar_name = registrar.get('name') domain_info.registrar_email = registrar.get('email') domain_info.registrar_phone = registrar.get('phone') domain_info.registrar_url = registrar.get('url') # find associated domains if registrant email is found related_domains = reverse_whois(domain_info.get('registrant_email')) if domain_info.get('registrant_email') else [] for _domain in related_domains: related_domains_list.append(_domain['name']) # remove duplicate domains from related domains list related_domains_list = list(set(related_domains_list)) domain_info.related_domains = related_domains_list # save to db if domain exists if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_rel = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_rel) db_domain_info.dnssec = domain_info.get('dnssec') #dates db_domain_info.created = domain_info.get('created') db_domain_info.updated = domain_info.get('updated') db_domain_info.expires = domain_info.get('expires') #registrar db_domain_info.registrar = Registrar.objects.get_or_create( name=domain_info.get('registrar_name'), email=domain_info.get('registrar_email'), phone=domain_info.get('registrar_phone'), url=domain_info.get('registrar_url'), )[0] db_domain_info.registrant = DomainRegistration.objects.get_or_create( name=domain_info.get('registrant_name'), organization=domain_info.get('registrant_organization'), address=domain_info.get('registrant_address'), city=domain_info.get('registrant_city'), state=domain_info.get('registrant_state'), zip_code=domain_info.get('registrant_zip_code'), country=domain_info.get('registrant_country'), email=domain_info.get('registrant_email'), phone=domain_info.get('registrant_phone'), fax=domain_info.get('registrant_fax'), id_str=domain_info.get('registrant_id'), )[0] db_domain_info.admin = DomainRegistration.objects.get_or_create( name=domain_info.get('admin_name'), organization=domain_info.get('admin_organization'), address=domain_info.get('admin_address'), city=domain_info.get('admin_city'), state=domain_info.get('admin_state'), zip_code=domain_info.get('admin_zip_code'), country=domain_info.get('admin_country'), email=domain_info.get('admin_email'), phone=domain_info.get('admin_phone'), fax=domain_info.get('admin_fax'), id_str=domain_info.get('admin_id'), )[0] db_domain_info.tech = DomainRegistration.objects.get_or_create( name=domain_info.get('tech_name'), organization=domain_info.get('tech_organization'), address=domain_info.get('tech_address'), city=domain_info.get('tech_city'), state=domain_info.get('tech_state'), zip_code=domain_info.get('tech_zip_code'), country=domain_info.get('tech_country'), email=domain_info.get('tech_email'), phone=domain_info.get('tech_phone'), fax=domain_info.get('tech_fax'), id_str=domain_info.get('tech_id'), )[0] for status in domain_info.get('status') or []: _status = WhoisStatus.objects.get_or_create( name=status )[0] _status.save() db_domain_info.status.add(_status) for ns in domain_info.get('ns_records') or []: _ns = NameServer.objects.get_or_create( name=ns )[0] _ns.save() db_domain_info.name_servers.add(_ns) for a in domain_info.get('a_records') or []: _a = DNSRecord.objects.get_or_create( name=a, type='a' )[0] _a.save() db_domain_info.dns_records.add(_a) for mx in domain_info.get('mx_records') or []: _mx = DNSRecord.objects.get_or_create( name=mx, type='mx' )[0] _mx.save() db_domain_info.dns_records.add(_mx) for txt in domain_info.get('txt_records') or []: _txt = DNSRecord.objects.get_or_create( name=txt, type='txt' )[0] _txt.save() db_domain_info.dns_records.add(_txt) db_domain_info.geolocation_iso = domain_info.get('registrant_country') db_domain_info.whois_server = domain_info.get('whois_server') db_domain_info.save() domain.domain_info = db_domain_info domain.save() except Exception as e: return { 'status': False, 'ip_domain': ip_domain, 'result': "unable to fetch records from WHOIS database.", 'message': str(e) } return { 'status': True, 'ip_domain': ip_domain, 'dnssec': domain_info.get('dnssec'), 'created': domain_info.get('created'), 'updated': domain_info.get('updated'), 'expires': domain_info.get('expires'), 'geolocation_iso': domain_info.get('registrant_country'), 'domain_statuses': domain_info.get('status'), 'whois_server': domain_info.get('whois_server'), 'dns': { 'a': domain_info.get('a_records'), 'mx': domain_info.get('mx_records'), 'txt': domain_info.get('txt_records'), }, 'registrar': { 'name': domain_info.get('registrar_name'), 'phone': domain_info.get('registrar_phone'), 'email': domain_info.get('registrar_email'), 'url': domain_info.get('registrar_url'), }, 'registrant': { 'name': domain_info.get('registrant_name'), 'id': domain_info.get('registrant_id'), 'organization': domain_info.get('registrant_organization'), 'address': domain_info.get('registrant_address'), 'city': domain_info.get('registrant_city'), 'state': domain_info.get('registrant_state'), 'zipcode': domain_info.get('registrant_zip_code'), 'country': domain_info.get('registrant_country'), 'phone': domain_info.get('registrant_phone'), 'fax': domain_info.get('registrant_fax'), 'email': domain_info.get('registrant_email'), }, 'admin': { 'name': domain_info.get('admin_name'), 'id': domain_info.get('admin_id'), 'organization': domain_info.get('admin_organization'), 'address':domain_info.get('admin_address'), 'city': domain_info.get('admin_city'), 'state': domain_info.get('admin_state'), 'zipcode': domain_info.get('admin_zip_code'), 'country': domain_info.get('admin_country'), 'phone': domain_info.get('admin_phone'), 'fax': domain_info.get('admin_fax'), 'email': domain_info.get('admin_email'), }, 'technical_contact': { 'name': domain_info.get('tech_name'), 'id': domain_info.get('tech_id'), 'organization': domain_info.get('tech_organization'), 'address': domain_info.get('tech_address'), 'city': domain_info.get('tech_city'), 'state': domain_info.get('tech_state'), 'zipcode': domain_info.get('tech_zip_code'), 'country': domain_info.get('tech_country'), 'phone': domain_info.get('tech_phone'), 'fax': domain_info.get('tech_fax'), 'email': domain_info.get('tech_email'), }, 'nameservers': domain_info.get('ns_records'), # 'similar_domains': domain_info.get('similar_domains'), 'related_domains': domain_info.get('related_domains'), 'related_tlds': domain_info.get('related_tlds'), 'historical_ips': domain_info.get('historical_ips'), } @app.task(name='remove_duplicate_endpoints', bind=False, queue='remove_duplicate_endpoints_queue') def remove_duplicate_endpoints( scan_history_id, domain_id, subdomain_id=None, filter_ids=[], filter_status=[200, 301, 404], duplicate_removal_fields=ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS ): """Remove duplicate endpoints. Check for implicit redirections by comparing endpoints: - [x] `content_length` similarities indicating redirections - [x] `page_title` (check for same page title) - [ ] Sign-in / login page (check for endpoints with the same words) Args: scan_history_id: ScanHistory id. domain_id (int): Domain id. subdomain_id (int, optional): Subdomain id. filter_ids (list): List of endpoint ids to filter on. filter_status (list): List of HTTP status codes to filter on. duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates """ logger.info(f'Removing duplicate endpoints based on {duplicate_removal_fields}') endpoints = ( EndPoint.objects .filter(scan_history__id=scan_history_id) .filter(target_domain__id=domain_id) ) if filter_status: endpoints = endpoints.filter(http_status__in=filter_status) if subdomain_id: endpoints = endpoints.filter(subdomain__id=subdomain_id) if filter_ids: endpoints = endpoints.filter(id__in=filter_ids) for field_name in duplicate_removal_fields: cl_query = ( endpoints .values_list(field_name) .annotate(mc=Count(field_name)) .order_by('-mc') ) for (field_value, count) in cl_query: if count > DELETE_DUPLICATES_THRESHOLD: eps_to_delete = ( endpoints .filter(**{field_name: field_value}) .order_by('discovered_date') .all()[1:] ) msg = f'Deleting {len(eps_to_delete)} endpoints [reason: same {field_name} {field_value}]' for ep in eps_to_delete: url = urlparse(ep.http_url) if url.path in ['', '/', '/login']: # try do not delete the original page that other pages redirect to continue msg += f'\n\t {ep.http_url} [{ep.http_status}] [{field_name}={field_value}]' ep.delete() logger.warning(msg) @app.task(name='run_command', bind=False, queue='run_command_queue') def run_command(cmd, cwd=None, shell=False, history_file=None, scan_id=None, activity_id=None): """Run a given command using subprocess module. Args: cmd (str): Command to run. cwd (str): Current working directory. echo (bool): Log command. shell (bool): Run within separate shell if True. history_file (str): Write command + output to history file. Returns: tuple: Tuple with return_code, output. """ logger.info(cmd) logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Run the command using subprocess popen = subprocess.Popen( cmd if shell else cmd.split(), shell=shell, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, cwd=cwd, universal_newlines=True) output = '' for stdout_line in iter(popen.stdout.readline, ""): item = stdout_line.strip() output += '\n' + item logger.debug(item) popen.stdout.close() popen.wait() return_code = popen.returncode command_obj.output = output command_obj.return_code = return_code command_obj.save() if history_file: mode = 'a' if not os.path.exists(history_file): mode = 'w' with open(history_file, mode) as f: f.write(f'\n{cmd}\n{return_code}\n{output}\n------------------\n') return return_code, output #-------------# # Other utils # #-------------# def stream_command(cmd, cwd=None, shell=False, history_file=None, encoding='utf-8', scan_id=None, activity_id=None, trunc_char=None): # Log cmd logger.info(cmd) # logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Sanitize the cmd command = cmd if shell else cmd.split() # Run the command using subprocess process = subprocess.Popen( command, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, universal_newlines=True, shell=shell) # Log the output in real-time to the database output = "" # Process the output for line in iter(lambda: process.stdout.readline(), b''): if not line: break line = line.strip() ansi_escape = re.compile(r'\x1B(?:[@-Z\\-_]|\[[0-?]*[ -/]*[@-~])') line = ansi_escape.sub('', line) line = line.replace('\\x0d\\x0a', '\n') if trunc_char and line.endswith(trunc_char): line = line[:-1] item = line # Try to parse the line as JSON try: item = json.loads(line) except json.JSONDecodeError: pass # Yield the line #logger.debug(item) yield item # Add the log line to the output output += line + "\n" # Update the command record in the database command_obj.output = output command_obj.save() # Retrieve the return code and output process.wait() return_code = process.returncode # Update the return code and final output in the database command_obj.return_code = return_code command_obj.save() # Append the command, return code and output to the history file if history_file is not None: with open(history_file, "a") as f: f.write(f"{cmd}\n{return_code}\n{output}\n") def process_httpx_response(line): """TODO: implement this""" def extract_httpx_url(line): """Extract final URL from httpx results. Always follow redirects to find the last URL. Args: line (dict): URL data output by httpx. Returns: tuple: (final_url, redirect_bool) tuple. """ status_code = line.get('status_code', 0) final_url = line.get('final_url') location = line.get('location') chain_status_codes = line.get('chain_status_codes', []) # Final URL is already looking nice, if it exists return it if final_url: return final_url, False http_url = line['url'] # fallback to url field # Handle redirects manually REDIRECT_STATUS_CODES = [301, 302] is_redirect = ( status_code in REDIRECT_STATUS_CODES or any(x in REDIRECT_STATUS_CODES for x in chain_status_codes) ) if is_redirect and location: if location.startswith(('http', 'https')): http_url = location else: http_url = f'{http_url}/{location.lstrip("/")}' # Sanitize URL http_url = sanitize_url(http_url) return http_url, is_redirect #-------------# # OSInt utils # #-------------# def get_and_save_dork_results(lookup_target, results_dir, type, lookup_keywords=None, lookup_extensions=None, delay=3, page_count=2, scan_history=None): """ Uses gofuzz to dork and store information Args: lookup_target (str): target to look into such as stackoverflow or even the target itself results_dir (str): Results directory type (str): Dork Type Title lookup_keywords (str): comma separated keywords or paths to look for lookup_extensions (str): comma separated extensions to look for delay (int): delay between each requests page_count (int): pages in google to extract information scan_history (startScan.ScanHistory): Scan History Object """ results = [] gofuzz_command = f'{GOFUZZ_EXEC_PATH} -t {lookup_target} -d {delay} -p {page_count}' if lookup_extensions: gofuzz_command += f' -e {lookup_extensions}' elif lookup_keywords: gofuzz_command += f' -w {lookup_keywords}' output_file = f'{results_dir}/gofuzz.txt' gofuzz_command += f' -o {output_file}' history_file = f'{results_dir}/commands.txt' try: run_command( gofuzz_command, shell=False, history_file=history_file, scan_id=scan_history.id, ) if not os.path.isfile(output_file): return with open(output_file) as f: for line in f.readlines(): url = line.strip() if url: results.append(url) dork, created = Dork.objects.get_or_create( type=type, url=url ) if scan_history: scan_history.dorks.add(dork) # remove output file os.remove(output_file) except Exception as e: logger.exception(e) return results def get_and_save_emails(scan_history, activity_id, results_dir): """Get and save emails from Google, Bing and Baidu. Args: scan_history (startScan.ScanHistory): Scan history object. activity_id: ScanActivity Object results_dir (str): Results directory. Returns: list: List of emails found. """ emails = [] # Proxy settings # get_random_proxy() # Gather emails from Google, Bing and Baidu output_file = f'{results_dir}/emails_tmp.txt' history_file = f'{results_dir}/commands.txt' command = f'python3 /usr/src/github/Infoga/infoga.py --domain {scan_history.domain.name} --source all --report {output_file}' try: run_command( command, shell=False, history_file=history_file, scan_id=scan_history.id, activity_id=activity_id) if not os.path.isfile(output_file): logger.info('No Email results') return [] with open(output_file) as f: for line in f.readlines(): if 'Email' in line: split_email = line.split(' ')[2] emails.append(split_email) output_path = f'{results_dir}/emails.txt' with open(output_path, 'w') as output_file: for email_address in emails: save_email(email_address, scan_history) output_file.write(f'{email_address}\n') except Exception as e: logger.exception(e) return emails def save_metadata_info(meta_dict): """Extract metadata from Google Search. Args: meta_dict (dict): Info dict. Returns: list: List of startScan.MetaFinderDocument objects. """ logger.warning(f'Getting metadata for {meta_dict.osint_target}') scan_history = ScanHistory.objects.get(id=meta_dict.scan_id) # Proxy settings get_random_proxy() # Get metadata result = extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if not result: logger.error(f'No metadata result from Google Search for {meta_dict.osint_target}.') return [] # Add metadata info to DB results = [] for metadata_name, data in result.get_metadata().items(): subdomain = Subdomain.objects.get( scan_history=meta_dict.scan_id, name=meta_dict.osint_target) metadata = DottedDict({k: v for k, v in data.items()}) meta_finder_document = MetaFinderDocument( subdomain=subdomain, target_domain=meta_dict.domain, scan_history=scan_history, url=metadata.url, doc_name=metadata_name, http_status=metadata.status_code, producer=metadata.metadata.get('Producer'), creator=metadata.metadata.get('Creator'), creation_date=metadata.metadata.get('CreationDate'), modified_date=metadata.metadata.get('ModDate'), author=metadata.metadata.get('Author'), title=metadata.metadata.get('Title'), os=metadata.metadata.get('OSInfo')) meta_finder_document.save() results.append(data) return results #-----------------# # Utils functions # #-----------------# def create_scan_activity(scan_history_id, message, status): scan_activity = ScanActivity() scan_activity.scan_of = ScanHistory.objects.get(pk=scan_history_id) scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id #--------------------# # Database functions # #--------------------# def save_vulnerability(**vuln_data): references = vuln_data.pop('references', []) cve_ids = vuln_data.pop('cve_ids', []) cwe_ids = vuln_data.pop('cwe_ids', []) tags = vuln_data.pop('tags', []) subscan = vuln_data.pop('subscan', None) # remove nulls vuln_data = replace_nulls(vuln_data) # Create vulnerability vuln, created = Vulnerability.objects.get_or_create(**vuln_data) if created: vuln.discovered_date = timezone.now() vuln.open_status = True vuln.save() # Save vuln tags for tag_name in tags or []: tag, created = VulnerabilityTags.objects.get_or_create(name=tag_name) if tag: vuln.tags.add(tag) vuln.save() # Save CVEs for cve_id in cve_ids or []: cve, created = CveId.objects.get_or_create(name=cve_id) if cve: vuln.cve_ids.add(cve) vuln.save() # Save CWEs for cve_id in cwe_ids or []: cwe, created = CweId.objects.get_or_create(name=cve_id) if cwe: vuln.cwe_ids.add(cwe) vuln.save() # Save vuln reference for url in references or []: ref, created = VulnerabilityReference.objects.get_or_create(url=url) if created: vuln.references.add(ref) vuln.save() # Save subscan id in vuln object if subscan: vuln.vuln_subscan_ids.add(subscan) vuln.save() return vuln, created def save_endpoint( http_url, ctx={}, crawl=False, is_default=False, **endpoint_data): """Get or create EndPoint object. If crawl is True, also crawl the endpoint HTTP URL with httpx. Args: http_url (str): Input HTTP URL. is_default (bool): If the url is a default url for SubDomains. scan_history (startScan.models.ScanHistory): ScanHistory object. domain (startScan.models.Domain): Domain object. subdomain (starScan.models.Subdomain): Subdomain object. results_dir (str, optional): Results directory. crawl (bool, optional): Run httpx on endpoint if True. Default: False. force (bool, optional): Force crawl even if ENABLE_HTTP_CRAWL mode is on. subscan (startScan.models.SubScan, optional): SubScan object. Returns: tuple: (startScan.models.EndPoint, created) where `created` is a boolean indicating if the object is new or already existed. """ # remove nulls endpoint_data = replace_nulls(endpoint_data) scheme = urlparse(http_url).scheme endpoint = None created = False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in http_url: logger.error(f"{http_url} is not a URL of domain {domain.name}. Skipping.") return None, False if crawl: ctx['track'] = False results = http_crawl( urls=[http_url], method='HEAD', ctx=ctx) if results: endpoint_data = results[0] endpoint_id = endpoint_data['endpoint_id'] created = endpoint_data['endpoint_created'] endpoint = EndPoint.objects.get(pk=endpoint_id) elif not scheme: return None, False else: # add dumb endpoint without probing it scan = ScanHistory.objects.filter(pk=ctx.get('scan_history_id')).first() domain = Domain.objects.filter(pk=ctx.get('domain_id')).first() if not validators.url(http_url): return None, False http_url = sanitize_url(http_url) endpoint, created = EndPoint.objects.get_or_create( scan_history=scan, target_domain=domain, http_url=http_url, **endpoint_data) if created: endpoint.is_default = is_default endpoint.discovered_date = timezone.now() endpoint.save() subscan_id = ctx.get('subscan_id') if subscan_id: endpoint.endpoint_subscan_ids.add(subscan_id) endpoint.save() return endpoint, created def save_subdomain(subdomain_name, ctx={}): """Get or create Subdomain object. Args: subdomain_name (str): Subdomain name. scan_history (startScan.models.ScanHistory): ScanHistory object. Returns: tuple: (startScan.models.Subdomain, created) where `created` is a boolean indicating if the object has been created in DB. """ scan_id = ctx.get('scan_history_id') subscan_id = ctx.get('subscan_id') out_of_scope_subdomains = ctx.get('out_of_scope_subdomains', []) valid_domain = ( validators.domain(subdomain_name) or validators.ipv4(subdomain_name) or validators.ipv6(subdomain_name) ) if not valid_domain: logger.error(f'{subdomain_name} is not an invalid domain. Skipping.') return None, False if subdomain_name in out_of_scope_subdomains: logger.error(f'{subdomain_name} is out-of-scope. Skipping.') return None, False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in subdomain_name: logger.error(f"{subdomain_name} is not a subdomain of domain {domain.name}. Skipping.") return None, False scan = ScanHistory.objects.filter(pk=scan_id).first() domain = scan.domain if scan else None subdomain, created = Subdomain.objects.get_or_create( scan_history=scan, target_domain=domain, name=subdomain_name) if created: # logger.warning(f'Found new subdomain {subdomain_name}') subdomain.discovered_date = timezone.now() if subscan_id: subdomain.subdomain_subscan_ids.add(subscan_id) subdomain.save() return subdomain, created def save_email(email_address, scan_history=None): if not validators.email(email_address): logger.info(f'Email {email_address} is invalid. Skipping.') return None, False email, created = Email.objects.get_or_create(address=email_address) # if created: # logger.warning(f'Found new email address {email_address}') # Add email to ScanHistory if scan_history: scan_history.emails.add(email) scan_history.save() return email, created def save_employee(name, designation, scan_history=None): employee, created = Employee.objects.get_or_create( name=name, designation=designation) # if created: # logger.warning(f'Found new employee {name}') # Add employee to ScanHistory if scan_history: scan_history.employees.add(employee) scan_history.save() return employee, created def save_ip_address(ip_address, subdomain=None, subscan=None, **kwargs): if not (validators.ipv4(ip_address) or validators.ipv6(ip_address)): logger.info(f'IP {ip_address} is not a valid IP. Skipping.') return None, False ip, created = IpAddress.objects.get_or_create(address=ip_address) # if created: # logger.warning(f'Found new IP {ip_address}') # Set extra attributes for key, value in kwargs.items(): setattr(ip, key, value) ip.save() # Add IP to subdomain if subdomain: subdomain.ip_addresses.add(ip) subdomain.save() # Add subscan to IP if subscan: ip.ip_subscan_ids.add(subscan) # Geo-localize IP asynchronously if created: geo_localize.delay(ip_address, ip.id) return ip, created def save_imported_subdomains(subdomains, ctx={}): """Take a list of subdomains imported and write them to from_imported.txt. Args: subdomains (list): List of subdomain names. scan_history (startScan.models.ScanHistory): ScanHistory instance. domain (startScan.models.Domain): Domain instance. results_dir (str): Results directory. """ domain_id = ctx['domain_id'] domain = Domain.objects.get(pk=domain_id) results_dir = ctx.get('results_dir', RENGINE_RESULTS) # Validate each subdomain and de-duplicate entries subdomains = list(set([ subdomain for subdomain in subdomains if validators.domain(subdomain) and domain.name == get_domain_from_subdomain(subdomain) ])) if not subdomains: return logger.warning(f'Found {len(subdomains)} imported subdomains.') with open(f'{results_dir}/from_imported.txt', 'w+') as output_file: for name in subdomains: subdomain_name = name.strip() subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain.is_imported_subdomain = True subdomain.save() output_file.write(f'{subdomain}\n') @app.task(name='query_reverse_whois', bind=False, queue='query_reverse_whois_queue') def query_reverse_whois(lookup_keyword): """Queries Reverse WHOIS information for an organization or email address. Args: lookup_keyword (str): Registrar Name or email Returns: dict: Reverse WHOIS information. """ return get_associated_domains(lookup_keyword) @app.task(name='query_ip_history', bind=False, queue='query_ip_history_queue') def query_ip_history(domain): """Queries the IP history for a domain Args: domain (str): domain_name Returns: list: list of historical ip addresses """ return get_domain_historical_ip_address(domain) @app.task(name='gpt_vulnerability_description', bind=False, queue='gpt_queue') def gpt_vulnerability_description(vulnerability_id): """Generate and store Vulnerability Description using GPT. Args: vulnerability_id (Vulnerability Model ID): Vulnerability ID to fetch Description. """ logger.info('Getting GPT Vulnerability Description') try: lookup_vulnerability = Vulnerability.objects.get(id=vulnerability_id) lookup_url = urlparse(lookup_vulnerability.http_url) path = lookup_url.path except Exception as e: return { 'status': False, 'error': str(e) } # check in db GPTVulnerabilityReport model if vulnerability description and path matches stored = GPTVulnerabilityReport.objects.filter(url_path=path).filter(title=lookup_vulnerability.name).first() if stored: response = { 'status': True, 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: vulnerability_description = get_gpt_vuln_input_description( lookup_vulnerability.name, path ) # one can add more description here later gpt_generator = GPTVulnerabilityReportGenerator() response = gpt_generator.get_vulnerability_description(vulnerability_description) add_gpt_description_db( lookup_vulnerability.name, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) # for all vulnerabilities with the same vulnerability name this description has to be stored. # also the consition is that the url must contain a part of this. for vuln in Vulnerability.objects.filter(name=lookup_vulnerability.name, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() return response

import csv import json import os import pprint import subprocess import time import validators import whatportis import xmltodict import yaml import tldextract import concurrent.futures from datetime import datetime from urllib.parse import urlparse from api.serializers import SubdomainSerializer from celery import chain, chord, group from celery.result import allow_join_result from celery.utils.log import get_task_logger from django.db.models import Count from dotted_dict import DottedDict from django.utils import timezone from pycvesearch import CVESearch from metafinder.extractor import extract_metadata_from_google_search from reNgine.celery import app from reNgine.gpt import GPTVulnerabilityReportGenerator from reNgine.celery_custom_task import RengineTask from reNgine.common_func import * from reNgine.definitions import * from reNgine.settings import * from reNgine.gpt import * from reNgine.utilities import * from scanEngine.models import (EngineType, InstalledExternalTool, Notification, Proxy) from startScan.models import * from startScan.models import EndPoint, Subdomain, Vulnerability from targetApp.models import Domain """ Celery tasks. """ logger = get_task_logger(__name__) #----------------------# # Scan / Subscan tasks # #----------------------# @app.task(name='initiate_scan', bind=False, queue='initiate_scan_queue') def initiate_scan( scan_history_id, domain_id, engine_id=None, scan_type=LIVE_SCAN, results_dir=RENGINE_RESULTS, imported_subdomains=[], out_of_scope_subdomains=[], url_filter=''): """Initiate a new scan. Args: scan_history_id (int): ScanHistory id. domain_id (int): Domain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. imported_subdomains (list): Imported subdomains. out_of_scope_subdomains (list): Out-of-scope subdomains. url_filter (str): URL path. Default: '' """ # Get scan history scan = ScanHistory.objects.get(pk=scan_history_id) # Get scan engine engine_id = engine_id or scan.scan_type.id # scan history engine_id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, []) # Get domain and set last_scan_date domain = Domain.objects.get(pk=domain_id) domain.last_scan_date = timezone.now() domain.save() # Get path filter url_filter = url_filter.rstrip('/') # Get or create ScanHistory() object if scan_type == LIVE_SCAN: # immediate scan = ScanHistory.objects.get(pk=scan_history_id) scan.scan_status = RUNNING_TASK elif scan_type == SCHEDULED_SCAN: # scheduled scan = ScanHistory() scan.scan_status = INITIATED_TASK scan.scan_type = engine scan.celery_ids = [initiate_scan.request.id] scan.domain = domain scan.start_scan_date = timezone.now() scan.tasks = engine.tasks scan.results_dir = f'{results_dir}/{domain.name}_{scan.id}' add_gf_patterns = gf_patterns and 'fetch_url' in engine.tasks if add_gf_patterns: scan.used_gf_patterns = ','.join(gf_patterns) scan.save() # Create scan results dir os.makedirs(scan.results_dir) # Build task context ctx = { 'scan_history_id': scan_history_id, 'engine_id': engine_id, 'domain_id': domain.id, 'results_dir': scan.results_dir, 'url_filter': url_filter, 'yaml_configuration': config, 'out_of_scope_subdomains': out_of_scope_subdomains } ctx_str = json.dumps(ctx, indent=2) # Send start notif logger.warning(f'Starting scan {scan_history_id} with context:\n{ctx_str}') send_scan_notif.delay( scan_history_id, subscan_id=None, engine_id=engine_id, status=CELERY_TASK_STATUS_MAP[scan.scan_status]) # Save imported subdomains in DB save_imported_subdomains(imported_subdomains, ctx=ctx) # Create initial subdomain in DB: make a copy of domain as a subdomain so # that other tasks using subdomains can use it. subdomain_name = domain.name subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) # If enable_http_crawl is set, create an initial root HTTP endpoint so that # HTTP crawling can start somewhere http_url = f'{domain.name}{url_filter}' if url_filter else domain.name endpoint, _ = save_endpoint( http_url, ctx=ctx, crawl=enable_http_crawl, is_default=True, subdomain=subdomain ) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build Celery tasks, crafted according to the dependency graph below: # subdomain_discovery --> port_scan --> fetch_url --> dir_file_fuzz # osint vulnerability_scan # osint dalfox xss scan # screenshot # waf_detection workflow = chain( group( subdomain_discovery.si(ctx=ctx, description='Subdomain discovery'), osint.si(ctx=ctx, description='OS Intelligence') ), port_scan.si(ctx=ctx, description='Port scan'), fetch_url.si(ctx=ctx, description='Fetch URL'), group( dir_file_fuzz.si(ctx=ctx, description='Directories & files fuzz'), vulnerability_scan.si(ctx=ctx, description='Vulnerability scan'), screenshot.si(ctx=ctx, description='Screenshot'), waf_detection.si(ctx=ctx, description='WAF detection') ) ) # Build callback callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery chord logger.info(f'Running Celery workflow with {len(workflow.tasks) + 1} tasks') task = chain(workflow, callback).on_error(callback).delay() scan.celery_ids.append(task.id) scan.save() return { 'success': True, 'task_id': task.id } @app.task(name='initiate_subscan', bind=False, queue='subscan_queue') def initiate_subscan( scan_history_id, subdomain_id, engine_id=None, scan_type=None, results_dir=RENGINE_RESULTS, url_filter=''): """Initiate a new subscan. Args: scan_history_id (int): ScanHistory id. subdomain_id (int): Subdomain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. url_filter (str): URL path. Default: '' """ # Get Subdomain, Domain and ScanHistory subdomain = Subdomain.objects.get(pk=subdomain_id) scan = ScanHistory.objects.get(pk=subdomain.scan_history.id) domain = Domain.objects.get(pk=subdomain.target_domain.id) # Get EngineType engine_id = engine_id or scan.scan_type.id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Create scan activity of SubScan Model subscan = SubScan( start_scan_date=timezone.now(), celery_ids=[initiate_subscan.request.id], scan_history=scan, subdomain=subdomain, type=scan_type, status=RUNNING_TASK, engine=engine) subscan.save() # Get YAML configuration config = yaml.safe_load(engine.yaml_configuration) # Create results directory results_dir = f'{scan.results_dir}/subscans/{subscan.id}' os.makedirs(results_dir, exist_ok=True) # Run task method = globals().get(scan_type) if not method: logger.warning(f'Task {scan_type} is not supported by reNgine. Skipping') return scan.tasks.append(scan_type) scan.save() # Send start notif send_scan_notif.delay( scan.id, subscan_id=subscan.id, engine_id=engine_id, status='RUNNING') # Build context ctx = { 'scan_history_id': scan.id, 'subscan_id': subscan.id, 'engine_id': engine_id, 'domain_id': domain.id, 'subdomain_id': subdomain.id, 'yaml_configuration': config, 'results_dir': results_dir, 'url_filter': url_filter } # Create initial endpoints in DB: find domain HTTP endpoint so that HTTP # crawling can start somewhere base_url = f'{subdomain.name}{url_filter}' if url_filter else subdomain.name endpoint, _ = save_endpoint( base_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build header + callback workflow = method.si(ctx=ctx) callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery tasks task = chain(workflow, callback).on_error(callback).delay() subscan.celery_ids.append(task.id) subscan.save() return { 'success': True, 'task_id': task.id } @app.task(name='report', bind=False, queue='report_queue') def report(ctx={}, description=None): """Report task running after all other tasks. Mark ScanHistory or SubScan object as completed and update with final status, log run details and send notification. Args: description (str, optional): Task description shown in UI. """ # Get objects subscan_id = ctx.get('subscan_id') scan_id = ctx.get('scan_history_id') engine_id = ctx.get('engine_id') scan = ScanHistory.objects.filter(pk=scan_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() # Get failed tasks tasks = ScanActivity.objects.filter(scan_of=scan).all() if subscan: tasks = tasks.filter(celery_id__in=subscan.celery_ids) failed_tasks = tasks.filter(status=FAILED_TASK) # Get task status failed_count = failed_tasks.count() status = SUCCESS_TASK if failed_count == 0 else FAILED_TASK status_h = 'SUCCESS' if failed_count == 0 else 'FAILED' # Update scan / subscan status if subscan: subscan.stop_scan_date = timezone.now() subscan.status = status subscan.save() else: scan.scan_status = status scan.stop_scan_date = timezone.now() scan.save() # Send scan status notif send_scan_notif.delay( scan_history_id=scan_id, subscan_id=subscan_id, engine_id=engine_id, status=status_h) #------------------------- # # Tracked reNgine tasks # #--------------------------# @app.task(name='subdomain_discovery', queue='main_scan_queue', base=RengineTask, bind=True) def subdomain_discovery( self, host=None, ctx=None, description=None): """Uses a set of tools (see SUBDOMAIN_SCAN_DEFAULT_TOOLS) to scan all subdomains associated with a domain. Args: host (str): Hostname to scan. Returns: subdomains (list): List of subdomain names. """ if not host: host = self.subdomain.name if self.subdomain else self.domain.name if self.url_filter: logger.warning(f'Ignoring subdomains scan as an URL path filter was passed ({self.url_filter}).') return # Config config = self.yaml_configuration.get(SUBDOMAIN_DISCOVERY) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL) or self.yaml_configuration.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) tools = config.get(USES_TOOLS, SUBDOMAIN_SCAN_DEFAULT_TOOLS) default_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=True).filter(is_subdomain_gathering=True)] custom_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=False).filter(is_subdomain_gathering=True)] send_subdomain_changes, send_interesting = False, False notif = Notification.objects.first() if notif: send_subdomain_changes = notif.send_subdomain_changes_notif send_interesting = notif.send_interesting_notif # Gather tools to run for subdomain scan if ALL in tools: tools = SUBDOMAIN_SCAN_DEFAULT_TOOLS + custom_subdomain_tools tools = [t.lower() for t in tools] # Make exception for amass since tool name is amass, but command is amass-active/passive default_subdomain_tools.append('amass-passive') default_subdomain_tools.append('amass-active') # Run tools for tool in tools: cmd = None logger.info(f'Scanning subdomains for {host} with {tool}') proxy = get_random_proxy() if tool in default_subdomain_tools: if tool == 'amass-passive': cmd = f'amass enum -passive -d {host} -o {self.results_dir}/subdomains_amass.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' elif tool == 'amass-active': use_amass_config = config.get(USE_AMASS_CONFIG, False) amass_wordlist_name = config.get(AMASS_WORDLIST, 'deepmagic.com-prefixes-top50000') wordlist_path = f'/usr/src/wordlist/{amass_wordlist_name}.txt' cmd = f'amass enum -active -d {host} -o {self.results_dir}/subdomains_amass_active.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' cmd += f' -brute -w {wordlist_path}' elif tool == 'sublist3r': cmd = f'python3 /usr/src/github/Sublist3r/sublist3r.py -d {host} -t {threads} -o {self.results_dir}/subdomains_sublister.txt' elif tool == 'subfinder': cmd = f'subfinder -d {host} -o {self.results_dir}/subdomains_subfinder.txt' use_subfinder_config = config.get(USE_SUBFINDER_CONFIG, False) cmd += ' -config /root/.config/subfinder/config.yaml' if use_subfinder_config else '' cmd += f' -proxy {proxy}' if proxy else '' cmd += f' -timeout {timeout}' if timeout else '' cmd += f' -t {threads}' if threads else '' cmd += f' -silent' elif tool == 'oneforall': cmd = f'python3 /usr/src/github/OneForAll/oneforall.py --target {host} run' cmd_extract = f'cut -d\',\' -f6 /usr/src/github/OneForAll/results/{host}.csv > {self.results_dir}/subdomains_oneforall.txt' cmd_rm = f'rm -rf /usr/src/github/OneForAll/results/{host}.csv' cmd += f' && {cmd_extract} && {cmd_rm}' elif tool == 'ctfr': results_file = self.results_dir + '/subdomains_ctfr.txt' cmd = f'python3 /usr/src/github/ctfr/ctfr.py -d {host} -o {results_file}' cmd_extract = f"cat {results_file} | sed 's/\*.//g' | tail -n +12 | uniq | sort > {results_file}" cmd += f' && {cmd_extract}' elif tool == 'tlsx': results_file = self.results_dir + '/subdomains_tlsx.txt' cmd = f'tlsx -san -cn -silent -ro -host {host}' cmd += f" | sed -n '/^\([a-zA-Z0-9]\([-a-zA-Z0-9]*[a-zA-Z0-9]\)\?\.\)\+{host}$/p' | uniq | sort" cmd += f' > {results_file}' elif tool == 'netlas': results_file = self.results_dir + '/subdomains_netlas.txt' cmd = f'netlas search -d domain -i domain domain:"*.{host}" -f json' netlas_key = get_netlas_key() cmd += f' -a {netlas_key}' if netlas_key else '' cmd_extract = f"grep -oE '([a-zA-Z0-9]([-a-zA-Z0-9]*[a-zA-Z0-9])?\.)+{host}'" cmd += f' | {cmd_extract} > {results_file}' elif tool in custom_subdomain_tools: tool_query = InstalledExternalTool.objects.filter(name__icontains=tool.lower()) if not tool_query.exists(): logger.error(f'Missing {{TARGET}} and {{OUTPUT}} placeholders in {tool} configuration. Skipping.') continue custom_tool = tool_query.first() cmd = custom_tool.subdomain_gathering_command if '{TARGET}' in cmd and '{OUTPUT}' in cmd: cmd = cmd.replace('{TARGET}', host) cmd = cmd.replace('{OUTPUT}', f'{self.results_dir}/subdomains_{tool}.txt') cmd = cmd.replace('{PATH}', custom_tool.github_clone_path) if '{PATH}' in cmd else cmd else: logger.warning( f'Subdomain discovery tool "{tool}" is not supported by reNgine. Skipping.') continue # Run tool try: run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) except Exception as e: logger.error( f'Subdomain discovery tool "{tool}" raised an exception') logger.exception(e) # Gather all the tools' results in one single file. Write subdomains into # separate files, and sort all subdomains. run_command( f'cat {self.results_dir}/subdomains_*.txt > {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {self.output_path} -o {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) with open(self.output_path) as f: lines = f.readlines() # Parse the output_file file and store Subdomain and EndPoint objects found # in db. subdomain_count = 0 subdomains = [] urls = [] for line in lines: subdomain_name = line.strip() valid_url = bool(validators.url(subdomain_name)) valid_domain = ( bool(validators.domain(subdomain_name)) or bool(validators.ipv4(subdomain_name)) or bool(validators.ipv6(subdomain_name)) or valid_url ) if not valid_domain: logger.error(f'Subdomain {subdomain_name} is not a valid domain, IP or URL. Skipping.') continue if valid_url: subdomain_name = urlparse(subdomain_name).netloc if subdomain_name in self.out_of_scope_subdomains: logger.error(f'Subdomain {subdomain_name} is out of scope. Skipping.') continue # Add subdomain subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain_count += 1 subdomains.append(subdomain) urls.append(subdomain.name) # Bulk crawl subdomains if enable_http_crawl: ctx['track'] = True http_crawl(urls, ctx=ctx, is_ran_from_subdomain_scan=True) # Find root subdomain endpoints for subdomain in subdomains: pass # Send notifications subdomains_str = '\n'.join([f'• `{subdomain.name}`' for subdomain in subdomains]) self.notify(fields={ 'Subdomain count': len(subdomains), 'Subdomains': subdomains_str, }) if send_subdomain_changes and self.scan_id and self.domain_id: added = get_new_added_subdomain(self.scan_id, self.domain_id) removed = get_removed_subdomain(self.scan_id, self.domain_id) if added: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in added]) self.notify(fields={'Added subdomains': subdomains_str}) if removed: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in removed]) self.notify(fields={'Removed subdomains': subdomains_str}) if send_interesting and self.scan_id and self.domain_id: interesting_subdomains = get_interesting_subdomains(self.scan_id, self.domain_id) if interesting_subdomains: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in interesting_subdomains]) self.notify(fields={'Interesting subdomains': subdomains_str}) return SubdomainSerializer(subdomains, many=True).data @app.task(name='osint', queue='main_scan_queue', base=RengineTask, bind=True) def osint(self, host=None, ctx={}, description=None): """Run Open-Source Intelligence tools on selected domain. Args: host (str): Hostname to scan. Returns: dict: Results from osint discovery and dorking. """ config = self.yaml_configuration.get(OSINT) or OSINT_DEFAULT_CONFIG results = {} grouped_tasks = [] if 'discover' in config: ctx['track'] = False # results = osint_discovery(host=host, ctx=ctx) _task = osint_discovery.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, activity_id=self.activity_id, results_dir=self.results_dir, ctx=ctx ) grouped_tasks.append(_task) if OSINT_DORK in config or OSINT_CUSTOM_DORK in config: _task = dorking.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, results_dir=self.results_dir ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('OSINT Tasks finished...') # with open(self.output_path, 'w') as f: # json.dump(results, f, indent=4) # # return results @app.task(name='osint_discovery', queue='osint_discovery_queue', bind=False) def osint_discovery(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run OSINT discovery. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: dict: osint metadat and theHarvester and h8mail results. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) osint_lookup = config.get(OSINT_DISCOVER, []) osint_intensity = config.get(INTENSITY, 'normal') documents_limit = config.get(OSINT_DOCUMENTS_LIMIT, 50) results = {} meta_info = [] emails = [] creds = [] # Get and save meta info if 'metainfo' in osint_lookup: if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': host, 'domain': host, 'scan_id': scan_history_id, 'documents_limit': documents_limit }) meta_info.append(save_metadata_info(meta_dict)) # TODO: disabled for now # elif osint_intensity == 'deep': # subdomains = Subdomain.objects # if self.scan: # subdomains = subdomains.filter(scan_history=self.scan) # for subdomain in subdomains: # meta_dict = DottedDict({ # 'osint_target': subdomain.name, # 'domain': self.domain, # 'scan_id': self.scan_id, # 'documents_limit': documents_limit # }) # meta_info.append(save_metadata_info(meta_dict)) grouped_tasks = [] if 'emails' in osint_lookup: emails = get_and_save_emails(scan_history, activity_id, results_dir) emails_str = '\n'.join([f'• `{email}`' for email in emails]) # self.notify(fields={'Emails': emails_str}) # ctx['track'] = False _task = h8mail.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) if 'employees' in osint_lookup: ctx['track'] = False _task = theHarvester.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) # results['emails'] = results.get('emails', []) + emails # results['creds'] = creds # results['meta_info'] = meta_info return results @app.task(name='dorking', bind=False, queue='dorking_queue') def dorking(config, host, scan_history_id, results_dir): """Run Google dorks. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: list: Dorking results for each dork ran. """ # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh scan_history = ScanHistory.objects.get(pk=scan_history_id) dorks = config.get(OSINT_DORK, []) custom_dorks = config.get(OSINT_CUSTOM_DORK, []) results = [] # custom dorking has higher priority try: for custom_dork in custom_dorks: lookup_target = custom_dork.get('lookup_site') # replace with original host if _target_ lookup_target = host if lookup_target == '_target_' else lookup_target if 'lookup_extensions' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_extensions=custom_dork.get('lookup_extensions'), scan_history=scan_history ) elif 'lookup_keywords' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_keywords=custom_dork.get('lookup_keywords'), scan_history=scan_history ) except Exception as e: logger.exception(e) # default dorking try: for dork in dorks: logger.info(f'Getting dork information for {dork}') if dork == 'stackoverflow': results = get_and_save_dork_results( lookup_target='stackoverflow.com', results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'login_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/login/,login.html', page_count=5, scan_history=scan_history ) elif dork == 'admin_panels': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/admin/,admin.html', page_count=5, scan_history=scan_history ) elif dork == 'dashboard_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/dashboard/,dashboard.html', page_count=5, scan_history=scan_history ) elif dork == 'social_media' : social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'reddit.com' ] for site in social_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'project_management' : project_websites = [ 'trello.com', 'atlassian.net' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'code_sharing' : project_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'config_files' : config_file_exts = [ 'env', 'xml', 'conf', 'toml', 'yml', 'yaml', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(config_file_exts), page_count=4, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keyword = 'Jenkins' results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=lookup_keyword, page_count=1, scan_history=scan_history ) elif dork == 'wordpress_files' : lookup_keywords = [ '/wp-content/', '/wp-includes/' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'php_error' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'exposed_documents' : docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(docs_file_ext), page_count=7, scan_history=scan_history ) elif dork == 'db_files' : file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) elif dork == 'git_exposed' : file_ext = [ 'git', ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) except Exception as e: logger.exception(e) return results @app.task(name='theHarvester', queue='theHarvester_queue', bind=False) def theHarvester(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run theHarvester to get save emails, hosts, employees found in domain. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: dict: Dict of emails, employees, hosts and ips found during crawling. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) output_path_json = f'{results_dir}/theHarvester.json' theHarvester_dir = '/usr/src/github/theHarvester' history_file = f'{results_dir}/commands.txt' cmd = f'python3 {theHarvester_dir}/theHarvester.py -d {host} -b all -f {output_path_json}' # Update proxies.yaml proxy_query = Proxy.objects.all() if proxy_query.exists(): proxy = proxy_query.first() if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(f'{theHarvester_dir}/proxies.yaml', 'w') as file: yaml.dump(yaml_data, file) # Run cmd run_command( cmd, shell=False, cwd=theHarvester_dir, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) # Get file location if not os.path.isfile(output_path_json): logger.error(f'Could not open {output_path_json}') return {} # Load theHarvester results with open(output_path_json, 'r') as f: data = json.load(f) # Re-indent theHarvester JSON with open(output_path_json, 'w') as f: json.dump(data, f, indent=4) emails = data.get('emails', []) for email_address in emails: email, _ = save_email(email_address, scan_history=scan_history) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) linkedin_people = data.get('linkedin_people', []) for people in linkedin_people: employee, _ = save_employee( people, designation='linkedin', scan_history=scan_history) # if employee: # self.notify(fields={'LinkedIn people': f'• {employee.name}'}) twitter_people = data.get('twitter_people', []) for people in twitter_people: employee, _ = save_employee( people, designation='twitter', scan_history=scan_history) # if employee: # self.notify(fields={'Twitter people': f'• {employee.name}'}) hosts = data.get('hosts', []) urls = [] for host in hosts: split = tuple(host.split(':')) http_url = split[0] subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) endpoint, _ = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain) # if endpoint: # urls.append(endpoint.http_url) # self.notify(fields={'Hosts': f'• {endpoint.http_url}'}) # if enable_http_crawl: # ctx['track'] = False # http_crawl(urls, ctx=ctx) # TODO: Lots of ips unrelated with our domain are found, disabling # this for now. # ips = data.get('ips', []) # for ip_address in ips: # ip, created = save_ip_address( # ip_address, # subscan=subscan) # if ip: # send_task_notif.delay( # 'osint', # scan_history_id=scan_history_id, # subscan_id=subscan_id, # severity='success', # update_fields={'IPs': f'{ip.address}'}) return data @app.task(name='h8mail', queue='h8mail_queue', bind=False) def h8mail(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run h8mail. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: list[dict]: List of credentials info. """ logger.warning('Getting leaked credentials') scan_history = ScanHistory.objects.get(pk=scan_history_id) input_path = f'{results_dir}/emails.txt' output_file = f'{results_dir}/h8mail.json' cmd = f'h8mail -t {input_path} --json {output_file}' history_file = f'{results_dir}/commands.txt' run_command( cmd, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) with open(output_file) as f: data = json.load(f) creds = data.get('targets', []) # TODO: go through h8mail output and save emails to DB for cred in creds: logger.warning(cred) email_address = cred['target'] pwn_num = cred['pwn_num'] pwn_data = cred.get('data', []) email, created = save_email(email_address, scan_history=scan) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) return creds @app.task(name='screenshot', queue='main_scan_queue', base=RengineTask, bind=True) def screenshot(self, ctx={}, description=None): """Uses EyeWitness to gather screenshot of a domain and/or url. Args: description (str, optional): Task description shown in UI. """ # Config screenshots_path = f'{self.results_dir}/screenshots' output_path = f'{self.results_dir}/screenshots/{self.filename}' alive_endpoints_file = f'{self.results_dir}/endpoints_alive.txt' config = self.yaml_configuration.get(SCREENSHOT) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT + 5) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) # If intensity is normal, grab only the root endpoints of each subdomain strict = True if intensity == 'normal' else False # Get URLs to take screenshot of get_http_urls( is_alive=enable_http_crawl, strict=strict, write_filepath=alive_endpoints_file, get_only_default_urls=True, ctx=ctx ) # Send start notif notification = Notification.objects.first() send_output_file = notification.send_scan_output_file if notification else False # Run cmd cmd = f'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py -f {alive_endpoints_file} -d {screenshots_path} --no-prompt' cmd += f' --timeout {timeout}' if timeout > 0 else '' cmd += f' --threads {threads}' if threads > 0 else '' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(output_path): logger.error(f'Could not load EyeWitness results at {output_path} for {self.domain.name}.') return # Loop through results and save objects in DB screenshot_paths = [] with open(output_path, 'r') as file: reader = csv.reader(file) for row in reader: "Protocol,Port,Domain,Request Status,Screenshot Path, Source Path" protocol, port, subdomain_name, status, screenshot_path, source_path = tuple(row) logger.info(f'{protocol}:{port}:{subdomain_name}:{status}') subdomain_query = Subdomain.objects.filter(name=subdomain_name) if self.scan: subdomain_query = subdomain_query.filter(scan_history=self.scan) if status == 'Successful' and subdomain_query.exists(): subdomain = subdomain_query.first() screenshot_paths.append(screenshot_path) subdomain.screenshot_path = screenshot_path.replace('/usr/src/scan_results/', '') subdomain.save() logger.warning(f'Added screenshot for {subdomain.name} to DB') # Remove all db, html extra files in screenshot results run_command( 'rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format(screenshots_path), shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'rm -rf {screenshots_path}/source', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Send finish notifs screenshots_str = '• ' + '\n• '.join([f'`{path}`' for path in screenshot_paths]) self.notify(fields={'Screenshots': screenshots_str}) if send_output_file: for path in screenshot_paths: title = get_output_file_name( self.scan_id, self.subscan_id, self.filename) send_file_to_discord.delay(path, title) @app.task(name='port_scan', queue='main_scan_queue', base=RengineTask, bind=True) def port_scan(self, hosts=[], ctx={}, description=None): """Run port scan. Args: hosts (list, optional): Hosts to run port scan on. description (str, optional): Task description shown in UI. Returns: list: List of open ports (dict). """ input_file = f'{self.results_dir}/input_subdomains_port_scan.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(PORT_SCAN) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) exclude_ports = config.get(NAABU_EXCLUDE_PORTS, []) exclude_subdomains = config.get(NAABU_EXCLUDE_SUBDOMAINS, False) ports = config.get(PORTS, NAABU_DEFAULT_PORTS) ports = [str(port) for port in ports] rate_limit = config.get(NAABU_RATE) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) passive = config.get(NAABU_PASSIVE, False) use_naabu_config = config.get(USE_NAABU_CONFIG, False) exclude_ports_str = ','.join(return_iterable(exclude_ports)) # nmap args nmap_enabled = config.get(ENABLE_NMAP, False) nmap_cmd = config.get(NMAP_COMMAND, '') nmap_script = config.get(NMAP_SCRIPT, '') nmap_script = ','.join(return_iterable(nmap_script)) nmap_script_args = config.get(NMAP_SCRIPT_ARGS) if hosts: with open(input_file, 'w') as f: f.write('\n'.join(hosts)) else: hosts = get_subdomains( write_filepath=input_file, exclude_subdomains=exclude_subdomains, ctx=ctx) # Build cmd cmd = 'naabu -json -exclude-cdn' cmd += f' -list {input_file}' if len(hosts) > 0 else f' -host {hosts[0]}' if 'full' in ports or 'all' in ports: ports_str = ' -p "-"' elif 'top-100' in ports: ports_str = ' -top-ports 100' elif 'top-1000' in ports: ports_str = ' -top-ports 1000' else: ports_str = ','.join(ports) ports_str = f' -p {ports_str}' cmd += ports_str cmd += ' -config /root/.config/naabu/config.yaml' if use_naabu_config else '' cmd += f' -proxy "{proxy}"' if proxy else '' cmd += f' -c {threads}' if threads else '' cmd += f' -rate {rate_limit}' if rate_limit > 0 else '' cmd += f' -timeout {timeout*1000}' if timeout > 0 else '' cmd += f' -passive' if passive else '' cmd += f' -exclude-ports {exclude_ports_str}' if exclude_ports else '' cmd += f' -silent' # Execute cmd and gather results results = [] urls = [] ports_data = {} for line in stream_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) port_number = line['port'] ip_address = line['ip'] host = line.get('host') or ip_address if port_number == 0: continue # Grab subdomain subdomain = Subdomain.objects.filter( name=host, target_domain=self.domain, scan_history=self.scan ).first() # Add IP DB ip, _ = save_ip_address(ip_address, subdomain, subscan=self.subscan) if self.subscan: ip.ip_subscan_ids.add(self.subscan) ip.save() # Add endpoint to DB # port 80 and 443 not needed as http crawl already does that. if port_number not in [80, 443]: http_url = f'{host}:{port_number}' endpoint, _ = save_endpoint( http_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint: http_url = endpoint.http_url urls.append(http_url) # Add Port in DB port_details = whatportis.get_ports(str(port_number)) service_name = port_details[0].name if len(port_details) > 0 else 'unknown' description = port_details[0].description if len(port_details) > 0 else '' # get or create port port, created = Port.objects.get_or_create( number=port_number, service_name=service_name, description=description ) if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port.save() ip.ports.add(port) ip.save() if host in ports_data: ports_data[host].append(port_number) else: ports_data[host] = [port_number] # Send notification logger.warning(f'Found opened port {port_number} on {ip_address} ({host})') if len(ports_data) == 0: logger.info('Finished running naabu port scan - No open ports found.') if nmap_enabled: logger.info('Nmap scans skipped') return ports_data # Send notification fields_str = '' for host, ports in ports_data.items(): ports_str = ', '.join([f'`{port}`' for port in ports]) fields_str += f'• `{host}`: {ports_str}\n' self.notify(fields={'Ports discovered': fields_str}) # Save output to file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) logger.info('Finished running naabu port scan.') # Process nmap results: 1 process per host sigs = [] if nmap_enabled: logger.warning(f'Starting nmap scans ...') logger.warning(ports_data) for host, port_list in ports_data.items(): ports_str = '_'.join([str(p) for p in port_list]) ctx_nmap = ctx.copy() ctx_nmap['description'] = get_task_title(f'nmap_{host}', self.scan_id, self.subscan_id) ctx_nmap['track'] = False sig = nmap.si( cmd=nmap_cmd, ports=port_list, host=host, script=nmap_script, script_args=nmap_script_args, max_rate=rate_limit, ctx=ctx_nmap) sigs.append(sig) task = group(sigs).apply_async() with allow_join_result(): results = task.get() return ports_data @app.task(name='nmap', queue='main_scan_queue', base=RengineTask, bind=True) def nmap( self, cmd=None, ports=[], host=None, input_file=None, script=None, script_args=None, max_rate=None, ctx={}, description=None): """Run nmap on a host. Args: cmd (str, optional): Existing nmap command to complete. ports (list, optional): List of ports to scan. host (str, optional): Host to scan. input_file (str, optional): Input hosts file. script (str, optional): NSE script to run. script_args (str, optional): NSE script args. max_rate (int): Max rate. description (str, optional): Task description shown in UI. """ notif = Notification.objects.first() ports_str = ','.join(str(port) for port in ports) self.filename = self.filename.replace('.txt', '.xml') filename_vulns = self.filename.replace('.xml', '_vulns.json') output_file = self.output_path output_file_xml = f'{self.results_dir}/{host}_{self.filename}' vulns_file = f'{self.results_dir}/{host}_{filename_vulns}' logger.warning(f'Running nmap on {host}:{ports}') # Build cmd nmap_cmd = get_nmap_cmd( cmd=cmd, ports=ports_str, script=script, script_args=script_args, max_rate=max_rate, host=host, input_file=input_file, output_file=output_file_xml) # Run cmd run_command( nmap_cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Get nmap XML results and convert to JSON vulns = parse_nmap_results(output_file_xml, output_file) with open(vulns_file, 'w') as f: json.dump(vulns, f, indent=4) # Save vulnerabilities found by nmap vulns_str = '' for vuln_data in vulns: # URL is not necessarily an HTTP URL when running nmap (can be any # other vulnerable protocols). Look for existing endpoint and use its # URL as vulnerability.http_url if it exists. url = vuln_data['http_url'] endpoint = EndPoint.objects.filter(http_url__contains=url).first() if endpoint: vuln_data['http_url'] = endpoint.http_url vuln, created = save_vulnerability( target_domain=self.domain, subdomain=self.subdomain, scan_history=self.scan, subscan=self.subscan, endpoint=endpoint, **vuln_data) vulns_str += f'• {str(vuln)}\n' if created: logger.warning(str(vuln)) # Send only 1 notif for all vulns to reduce number of notifs if notif and notif.send_vuln_notif and vulns_str: logger.warning(vulns_str) self.notify(fields={'CVEs': vulns_str}) return vulns @app.task(name='waf_detection', queue='main_scan_queue', base=RengineTask, bind=True) def waf_detection(self, ctx={}, description=None): """ Uses wafw00f to check for the presence of a WAF. Args: description (str, optional): Task description shown in UI. Returns: list: List of startScan.models.Waf objects. """ input_path = f'{self.results_dir}/input_endpoints_waf_detection.txt' config = self.yaml_configuration.get(WAF_DETECTION) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Get alive endpoints from DB get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) cmd = f'wafw00f -i {input_path} -o {self.output_path}' run_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(self.output_path): logger.error(f'Could not find {self.output_path}') return with open(self.output_path) as file: wafs = file.readlines() for line in wafs: line = " ".join(line.split()) splitted = line.split(' ', 1) waf_info = splitted[1].strip() waf_name = waf_info[:waf_info.find('(')].strip() waf_manufacturer = waf_info[waf_info.find('(')+1:waf_info.find(')')].strip().replace('.', '') http_url = sanitize_url(splitted[0].strip()) if not waf_name or waf_name == 'None': continue # Add waf to db waf, _ = Waf.objects.get_or_create( name=waf_name, manufacturer=waf_manufacturer ) # Add waf info to Subdomain in DB subdomain = get_subdomain_from_url(http_url) logger.info(f'Wafw00f Subdomain : {subdomain}') subdomain_query, _ = Subdomain.objects.get_or_create(scan_history=self.scan, name=subdomain) subdomain_query.waf.add(waf) subdomain_query.save() return wafs @app.task(name='dir_file_fuzz', queue='main_scan_queue', base=RengineTask, bind=True) def dir_file_fuzz(self, ctx={}, description=None): """Perform directory scan, and currently uses `ffuf` as a default tool. Args: description (str, optional): Task description shown in UI. Returns: list: List of URLs discovered. """ # Config cmd = 'ffuf' config = self.yaml_configuration.get(DIR_FILE_FUZZ) or {} custom_header = self.yaml_configuration.get(CUSTOM_HEADER) auto_calibration = config.get(AUTO_CALIBRATION, True) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) extensions = config.get(EXTENSIONS, DEFAULT_DIR_FILE_FUZZ_EXTENSIONS) # prepend . on extensions extensions = [ext if ext.startswith('.') else '.' + ext for ext in extensions] extensions_str = ','.join(map(str, extensions)) follow_redirect = config.get(FOLLOW_REDIRECT, FFUF_DEFAULT_FOLLOW_REDIRECT) max_time = config.get(MAX_TIME, 0) match_http_status = config.get(MATCH_HTTP_STATUS, FFUF_DEFAULT_MATCH_HTTP_STATUS) mc = ','.join([str(c) for c in match_http_status]) recursive_level = config.get(RECURSIVE_LEVEL, FFUF_DEFAULT_RECURSIVE_LEVEL) stop_on_error = config.get(STOP_ON_ERROR, False) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) wordlist_name = config.get(WORDLIST, 'dicc') delay = rate_limit / (threads * 100) # calculate request pause delay from rate_limit and number of threads input_path = f'{self.results_dir}/input_dir_file_fuzz.txt' # Get wordlist wordlist_name = 'dicc' if wordlist_name == 'default' else wordlist_name wordlist_path = f'/usr/src/wordlist/{wordlist_name}.txt' # Build command cmd += f' -w {wordlist_path}' cmd += f' -e {extensions_str}' if extensions else '' cmd += f' -maxtime {max_time}' if max_time > 0 else '' cmd += f' -p {delay}' if delay > 0 else '' cmd += f' -recursion -recursion-depth {recursive_level} ' if recursive_level > 0 else '' cmd += f' -t {threads}' if threads and threads > 0 else '' cmd += f' -timeout {timeout}' if timeout and timeout > 0 else '' cmd += ' -se' if stop_on_error else '' cmd += ' -fr' if follow_redirect else '' cmd += ' -ac' if auto_calibration else '' cmd += f' -mc {mc}' if mc else '' cmd += f' -H "{custom_header}"' if custom_header else '' # Grab URLs to fuzz urls = get_http_urls( is_alive=True, ignore_files=False, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) logger.warning(urls) # Loop through URLs and run command results = [] for url in urls: ''' Above while fetching urls, we are not ignoring files, because some default urls may redirect to https://example.com/login.php so, ignore_files is set to False but, during fuzzing, we will only need part of the path, in above example it is still a good idea to ffuf base url https://example.com so files from base url ''' url_parse = urlparse(url) url = url_parse.scheme + '://' + url_parse.netloc url += '/FUZZ' # TODO: fuzz not only URL but also POST / PUT / headers proxy = get_random_proxy() # Build final cmd fcmd = cmd fcmd += f' -x {proxy}' if proxy else '' fcmd += f' -u {url} -json' # Initialize DirectoryScan object dirscan = DirectoryScan() dirscan.scanned_date = timezone.now() dirscan.command_line = fcmd dirscan.save() # Loop through results and populate EndPoint and DirectoryFile in DB results = [] for line in stream_command( fcmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) name = line['input'].get('FUZZ') length = line['length'] status = line['status'] words = line['words'] url = line['url'] lines = line['lines'] content_type = line['content-type'] duration = line['duration'] if not name: logger.error(f'FUZZ not found for "{url}"') continue endpoint, created = save_endpoint(url, crawl=False, ctx=ctx) endpoint.http_status = status endpoint.content_length = length endpoint.response_time = duration / 1000000000 endpoint.save() if created: urls.append(endpoint.http_url) endpoint.status = status endpoint.content_type = content_type endpoint.content_length = length dfile, created = DirectoryFile.objects.get_or_create( name=name, length=length, words=words, lines=lines, content_type=content_type, url=url) dfile.http_status = status dfile.save() # if created: # logger.warning(f'Found new directory or file {url}') dirscan.directory_files.add(dfile) dirscan.save() if self.subscan: dirscan.dir_subscan_ids.add(self.subscan) subdomain_name = get_subdomain_from_url(endpoint.http_url) subdomain = Subdomain.objects.get(name=subdomain_name, scan_history=self.scan) subdomain.directories.add(dirscan) subdomain.save() # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl(urls, ctx=ctx) return results @app.task(name='fetch_url', queue='main_scan_queue', base=RengineTask, bind=True) def fetch_url(self, urls=[], ctx={}, description=None): """Fetch URLs using different tools like gauplus, gau, gospider, waybackurls ... Args: urls (list): List of URLs to start from. description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/input_endpoints_fetch_url.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(FETCH_URL) or {} should_remove_duplicate_endpoints = config.get(REMOVE_DUPLICATE_ENDPOINTS, True) duplicate_removal_fields = config.get(DUPLICATE_REMOVAL_FIELDS, ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, DEFAULT_GF_PATTERNS) ignore_file_extension = config.get(IGNORE_FILE_EXTENSION, DEFAULT_IGNORE_FILE_EXTENSIONS) tools = config.get(USES_TOOLS, ENDPOINT_SCAN_DEFAULT_TOOLS) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) domain_request_headers = self.domain.request_headers if self.domain else None custom_header = domain_request_headers or self.yaml_configuration.get(CUSTOM_HEADER) exclude_subdomains = config.get(EXCLUDED_SUBDOMAINS, False) # Get URLs to scan and save to input file if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, exclude_subdomains=exclude_subdomains, get_only_default_urls=True, ctx=ctx ) # Domain regex host = self.domain.name if self.domain else urlparse(urls[0]).netloc host_regex = f"\'https?://([a-z0-9]+[.])*{host}.*\'" # Tools cmds cmd_map = { 'gau': f'gau', 'gauplus': f'gauplus -random-agent', 'hakrawler': 'hakrawler -subs -u', 'waybackurls': 'waybackurls', 'gospider': f'gospider -S {input_path} --js -d 2 --sitemap --robots -w -r', 'katana': f'katana -list {input_path} -silent -jc -kf all -d 3 -fs rdn', } if proxy: cmd_map['gau'] += f' --proxy "{proxy}"' cmd_map['gauplus'] += f' -p "{proxy}"' cmd_map['gospider'] += f' -p {proxy}' cmd_map['hakrawler'] += f' -proxy {proxy}' cmd_map['katana'] += f' -proxy {proxy}' if threads > 0: cmd_map['gau'] += f' --threads {threads}' cmd_map['gauplus'] += f' -t {threads}' cmd_map['gospider'] += f' -t {threads}' cmd_map['katana'] += f' -c {threads}' if custom_header: header_string = ';;'.join([ f'{key}: {value}' for key, value in custom_header.items() ]) cmd_map['hakrawler'] += f' -h {header_string}' cmd_map['katana'] += f' -H {header_string}' header_flags = [':'.join(h) for h in header_string.split(';;')] for flag in header_flags: cmd_map['gospider'] += f' -H {flag}' cat_input = f'cat {input_path}' grep_output = f'grep -Eo {host_regex}' cmd_map = { tool: f'{cat_input} | {cmd} | {grep_output} > {self.results_dir}/urls_{tool}.txt' for tool, cmd in cmd_map.items() } tasks = group( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for tool, cmd in cmd_map.items() if tool in tools ) # Cleanup task sort_output = [ f'cat {self.results_dir}/urls_* > {self.output_path}', f'cat {input_path} >> {self.output_path}', f'sort -u {self.output_path} -o {self.output_path}', ] if ignore_file_extension: ignore_exts = '|'.join(ignore_file_extension) grep_ext_filtered_output = [ f'cat {self.output_path} | grep -Eiv "\\.({ignore_exts}).*" > {self.results_dir}/urls_filtered.txt', f'mv {self.results_dir}/urls_filtered.txt {self.output_path}' ] sort_output.extend(grep_ext_filtered_output) cleanup = chain( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for cmd in sort_output ) # Run all commands task = chord(tasks)(cleanup) with allow_join_result(): task.get() # Store all the endpoints and run httpx with open(self.output_path) as f: discovered_urls = f.readlines() self.notify(fields={'Discovered URLs': len(discovered_urls)}) # Some tools can have an URL in the format <URL>] - <PATH> or <URL> - <PATH>, add them # to the final URL list all_urls = [] for url in discovered_urls: url = url.strip() urlpath = None base_url = None if '] ' in url: # found JS scraped endpoint e.g from gospider split = tuple(url.split('] ')) if not len(split) == 2: logger.warning(f'URL format not recognized for "{url}". Skipping.') continue base_url, urlpath = split urlpath = urlpath.lstrip('- ') elif ' - ' in url: # found JS scraped endpoint e.g from gospider base_url, urlpath = tuple(url.split(' - ')) if base_url and urlpath: subdomain = urlparse(base_url) url = f'{subdomain.scheme}://{subdomain.netloc}{self.url_filter}' if not validators.url(url): logger.warning(f'Invalid URL "{url}". Skipping.') if url not in all_urls: all_urls.append(url) # Filter out URLs if a path filter was passed if self.url_filter: all_urls = [url for url in all_urls if self.url_filter in url] # Write result to output path with open(self.output_path, 'w') as f: f.write('\n'.join(all_urls)) logger.warning(f'Found {len(all_urls)} usable URLs') # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl( all_urls, ctx=ctx, should_remove_duplicate_endpoints=should_remove_duplicate_endpoints, duplicate_removal_fields=duplicate_removal_fields ) #-------------------# # GF PATTERNS MATCH # #-------------------# # Combine old gf patterns with new ones if gf_patterns: self.scan.used_gf_patterns = ','.join(gf_patterns) self.scan.save() # Run gf patterns on saved endpoints # TODO: refactor to Celery task for gf_pattern in gf_patterns: # TODO: js var is causing issues, removing for now if gf_pattern == 'jsvar': logger.info('Ignoring jsvar as it is causing issues.') continue # Run gf on current pattern logger.warning(f'Running gf on pattern "{gf_pattern}"') gf_output_file = f'{self.results_dir}/gf_patterns_{gf_pattern}.txt' cmd = f'cat {self.output_path} | gf {gf_pattern} | grep -Eo {host_regex} >> {gf_output_file}' run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Check output file if not os.path.exists(gf_output_file): logger.error(f'Could not find GF output file {gf_output_file}. Skipping GF pattern "{gf_pattern}"') continue # Read output file line by line and with open(gf_output_file, 'r') as f: lines = f.readlines() # Add endpoints / subdomains to DB for url in lines: http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue endpoint, created = save_endpoint( http_url, crawl=False, subdomain=subdomain, ctx=ctx) if not endpoint: continue earlier_pattern = None if not created: earlier_pattern = endpoint.matched_gf_patterns pattern = f'{earlier_pattern},{gf_pattern}' if earlier_pattern else gf_pattern endpoint.matched_gf_patterns = pattern endpoint.save() return all_urls def parse_curl_output(response): # TODO: Enrich from other cURL fields. CURL_REGEX_HTTP_STATUS = f'HTTP\/(?:(?:\d\.?)+)\s(\d+)\s(?:\w+)' http_status = 0 if response: failed = False regex = re.compile(CURL_REGEX_HTTP_STATUS, re.MULTILINE) try: http_status = int(regex.findall(response)[0]) except (KeyError, TypeError, IndexError): pass return { 'http_status': http_status, } @app.task(name='vulnerability_scan', queue='main_scan_queue', bind=True, base=RengineTask) def vulnerability_scan(self, urls=[], ctx={}, description=None): """ This function will serve as an entrypoint to vulnerability scan. All other vulnerability scan will be run from here including nuclei, crlfuzz, etc """ logger.info('Running Vulnerability Scan Queue') config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_run_nuclei = config.get(RUN_NUCLEI, True) should_run_crlfuzz = config.get(RUN_CRLFUZZ, False) should_run_dalfox = config.get(RUN_DALFOX, False) should_run_s3scanner = config.get(RUN_S3SCANNER, True) grouped_tasks = [] if should_run_nuclei: _task = nuclei_scan.si( urls=urls, ctx=ctx, description=f'Nuclei Scan' ) grouped_tasks.append(_task) if should_run_crlfuzz: _task = crlfuzz_scan.si( urls=urls, ctx=ctx, description=f'CRLFuzz Scan' ) grouped_tasks.append(_task) if should_run_dalfox: _task = dalfox_xss_scan.si( urls=urls, ctx=ctx, description=f'Dalfox XSS Scan' ) grouped_tasks.append(_task) if should_run_s3scanner: _task = s3scanner.si( ctx=ctx, description=f'Misconfigured S3 Buckets Scanner' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan completed...') # return results return None @app.task(name='nuclei_individual_severity_module', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_individual_severity_module(self, cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx={}, description=None): ''' This celery task will run vulnerability scan in parallel. All severities supplied should run in parallel as grouped tasks. ''' results = [] logger.info(f'Running vulnerability scan with severity: {severity}') cmd += f' -severity {severity}' # Send start notification notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) # Gather nuclei results vuln_data = parse_nuclei_result(line) # Get corresponding subdomain http_url = sanitize_url(line.get('matched-at')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) # Look for duplicate vulnerabilities by excluding records that might change but are irrelevant. object_comparison_exclude = ['response', 'curl_command', 'tags', 'references', 'cve_ids', 'cwe_ids'] # Add subdomain and target domain to the duplicate check vuln_data_copy = vuln_data.copy() vuln_data_copy['subdomain'] = subdomain vuln_data_copy['target_domain'] = self.domain # Check if record exists, if exists do not save it if record_exists(Vulnerability, data=vuln_data_copy, exclude_keys=object_comparison_exclude): logger.warning(f'Nuclei vulnerability of severity {severity} : {vuln_data_copy["name"]} for {subdomain_name} already exists') continue # Get or create EndPoint object response = line.get('response') httpx_crawl = False if response else enable_http_crawl # avoid yet another httpx crawl endpoint, _ = save_endpoint( http_url, crawl=httpx_crawl, subdomain=subdomain, ctx=ctx) if endpoint: http_url = endpoint.http_url if not httpx_crawl: output = parse_curl_output(response) endpoint.http_status = output['http_status'] endpoint.save() # Get or create Vulnerability object vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, subdomain=subdomain, **vuln_data) if not vuln: continue # Print vuln severity = line['info'].get('severity', 'unknown') logger.warning(str(vuln)) # Send notification for all vulnerabilities except info url = vuln.http_url or vuln.subdomain send_vuln = ( notif and notif.send_vuln_notif and vuln and severity in ['low', 'medium', 'high', 'critical']) if send_vuln: fields = { 'Severity': f'**{severity.upper()}**', 'URL': http_url, 'Subdomain': subdomain_name, 'Name': vuln.name, 'Type': vuln.type, 'Description': vuln.description, 'Template': vuln.template_url, 'Tags': vuln.get_tags_str(), 'CVEs': vuln.get_cve_str(), 'CWEs': vuln.get_cwe_str(), 'References': vuln.get_refs_str() } severity_map = { 'low': 'info', 'medium': 'warning', 'high': 'error', 'critical': 'error' } self.notify( f'vulnerability_scan_#{vuln.id}', severity_map[severity], fields, add_meta_info=False) # Send report to hackerone hackerone_query = Hackerone.objects.all() send_report = ( hackerone_query.exists() and severity not in ('info', 'low') and vuln.target_domain.h1_team_handle ) if send_report: hackerone = hackerone_query.first() if hackerone.send_critical and severity == 'critical': send_hackerone_report.delay(vuln.id) elif hackerone.send_high and severity == 'high': send_hackerone_report.delay(vuln.id) elif hackerone.send_medium and severity == 'medium': send_hackerone_report.delay(vuln.id) # Write results to JSON file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) # Send finish notif if send_status: vulns = Vulnerability.objects.filter(scan_history__id=self.scan_id) info_count = vulns.filter(severity=0).count() low_count = vulns.filter(severity=1).count() medium_count = vulns.filter(severity=2).count() high_count = vulns.filter(severity=3).count() critical_count = vulns.filter(severity=4).count() unknown_count = vulns.filter(severity=-1).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count + unknown_count fields = { 'Total': vulnerability_count, 'Critical': critical_count, 'High': high_count, 'Medium': medium_count, 'Low': low_count, 'Info': info_count, 'Unknown': unknown_count } self.notify(fields=fields) # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=NUCLEI ).exclude( severity=0 ) # find all unique vulnerabilities based on path and title # all unique vulnerability will go thru gpt function and get report # once report is got, it will be matched with other vulnerabilities and saved unique_vulns = set() for vuln in vulns: unique_vulns.add((vuln.name, vuln.get_path())) unique_vulns = list(unique_vulns) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in unique_vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return None def get_vulnerability_gpt_report(vuln): title = vuln[0] path = vuln[1] logger.info(f'Getting GPT Report for {title}, PATH: {path}') # check if in db already exists stored = GPTVulnerabilityReport.objects.filter( url_path=path ).filter( title=title ).first() if stored: response = { 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: report = GPTVulnerabilityReportGenerator() vulnerability_description = get_gpt_vuln_input_description( title, path ) response = report.get_vulnerability_description(vulnerability_description) add_gpt_description_db( title, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) for vuln in Vulnerability.objects.filter(name=title, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() def add_gpt_description_db(title, path, description, impact, remediation, references): gpt_report = GPTVulnerabilityReport() gpt_report.url_path = path gpt_report.title = title gpt_report.description = description gpt_report.impact = impact gpt_report.remediation = remediation gpt_report.save() for url in references: ref, created = VulnerabilityReference.objects.get_or_create(url=url) gpt_report.references.add(ref) gpt_report.save() @app.task(name='nuclei_scan', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_scan(self, urls=[], ctx={}, description=None): """HTTP vulnerability scan using Nuclei Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. Notes: Unfurl the urls to keep only domain and path, will be sent to vuln scan and ignore certain file extensions. Thanks: https://github.com/six2dez/reconftw """ # Config config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} input_path = f'{self.results_dir}/input_endpoints_vulnerability_scan.txt' enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) concurrency = config.get(NUCLEI_CONCURRENCY) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) retries = config.get(RETRIES) or self.yaml_configuration.get(RETRIES, DEFAULT_RETRIES) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) custom_header = config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) should_fetch_gpt_report = config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) proxy = get_random_proxy() nuclei_specific_config = config.get('nuclei', {}) use_nuclei_conf = nuclei_specific_config.get(USE_CONFIG, False) severities = nuclei_specific_config.get(NUCLEI_SEVERITY, NUCLEI_DEFAULT_SEVERITIES) tags = nuclei_specific_config.get(NUCLEI_TAGS, []) tags = ','.join(tags) nuclei_templates = nuclei_specific_config.get(NUCLEI_TEMPLATE) custom_nuclei_templates = nuclei_specific_config.get(NUCLEI_CUSTOM_TEMPLATE) # severities_str = ','.join(severities) # Get alive endpoints if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=enable_http_crawl, ignore_files=True, write_filepath=input_path, ctx=ctx ) if intensity == 'normal': # reduce number of endpoints to scan unfurl_filter = f'{self.results_dir}/urls_unfurled.txt' run_command( f"cat {input_path} | unfurl -u format %s://%d%p |uro > {unfurl_filter}", shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {unfurl_filter} -o {unfurl_filter}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) input_path = unfurl_filter # Build templates # logger.info('Updating Nuclei templates ...') run_command( 'nuclei -update-templates', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) templates = [] if not (nuclei_templates or custom_nuclei_templates): templates.append(NUCLEI_DEFAULT_TEMPLATES_PATH) if nuclei_templates: if ALL in nuclei_templates: template = NUCLEI_DEFAULT_TEMPLATES_PATH templates.append(template) else: templates.extend(nuclei_templates) if custom_nuclei_templates: custom_nuclei_template_paths = [f'{str(elem)}.yaml' for elem in custom_nuclei_templates] template = templates.extend(custom_nuclei_template_paths) # Build CMD cmd = 'nuclei -j' cmd += ' -config /root/.config/nuclei/config.yaml' if use_nuclei_conf else '' cmd += f' -irr' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -l {input_path}' cmd += f' -c {str(concurrency)}' if concurrency > 0 else '' cmd += f' -proxy {proxy} ' if proxy else '' cmd += f' -retries {retries}' if retries > 0 else '' cmd += f' -rl {rate_limit}' if rate_limit > 0 else '' # cmd += f' -severity {severities_str}' cmd += f' -timeout {str(timeout)}' if timeout and timeout > 0 else '' cmd += f' -tags {tags}' if tags else '' cmd += f' -silent' for tpl in templates: cmd += f' -t {tpl}' grouped_tasks = [] custom_ctx = ctx for severity in severities: custom_ctx['track'] = True _task = nuclei_individual_severity_module.si( cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx=custom_ctx, description=f'Nuclei Scan with severity {severity}' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan with all severities completed...') return None @app.task(name='dalfox_xss_scan', queue='main_scan_queue', base=RengineTask, bind=True) def dalfox_xss_scan(self, urls=[], ctx={}, description=None): """XSS Scan using dalfox Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) dalfox_config = vuln_config.get(DALFOX) or {} custom_header = dalfox_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() is_waf_evasion = dalfox_config.get(WAF_EVASION, False) blind_xss_server = dalfox_config.get(BLIND_XSS_SERVER) user_agent = dalfox_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) timeout = dalfox_config.get(TIMEOUT) delay = dalfox_config.get(DELAY) threads = dalfox_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_dalfox_xss.txt' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=False, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'dalfox --silence --no-color --no-spinner' cmd += f' --only-poc r ' cmd += f' --ignore-return 302,404,403' cmd += f' --skip-bav' cmd += f' file {input_path}' cmd += f' --proxy {proxy}' if proxy else '' cmd += f' --waf-evasion' if is_waf_evasion else '' cmd += f' -b {blind_xss_server}' if blind_xss_server else '' cmd += f' --delay {delay}' if delay else '' cmd += f' --timeout {timeout}' if timeout else '' cmd += f' --user-agent {user_agent}' if user_agent else '' cmd += f' --header {custom_header}' if custom_header else '' cmd += f' --worker {threads}' if threads else '' cmd += f' --format json' results = [] for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id, trunc_char=',' ): if not isinstance(line, dict): continue results.append(line) vuln_data = parse_dalfox_result(line) http_url = sanitize_url(line.get('data')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Dalfox Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=DALFOX ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='crlfuzz_scan', queue='main_scan_queue', base=RengineTask, bind=True) def crlfuzz_scan(self, urls=[], ctx={}, description=None): """CRLF Fuzzing with CRLFuzz Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) custom_header = vuln_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() user_agent = vuln_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) threads = vuln_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_crlf.txt' output_path = f'{self.results_dir}/{self.filename}' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=True, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'crlfuzz -s' cmd += f' -l {input_path}' cmd += f' -x {proxy}' if proxy else '' cmd += f' --H {custom_header}' if custom_header else '' cmd += f' -o {output_path}' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id ) if not os.path.isfile(output_path): logger.info('No Results from CRLFuzz') return crlfs = [] results = [] with open(output_path, 'r') as file: crlfs = file.readlines() for crlf in crlfs: url = crlf.strip() vuln_data = parse_crlfuzz_result(url) http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting CRLFuzz Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=CRLFUZZ ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='s3scanner', queue='main_scan_queue', base=RengineTask, bind=True) def s3scanner(self, ctx={}, description=None): """Bucket Scanner Args: ctx (dict): Context description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/#{self.scan_id}_subdomain_discovery.txt' vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} s3_config = vuln_config.get(S3SCANNER) or {} threads = s3_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) providers = s3_config.get(PROVIDERS, S3SCANNER_DEFAULT_PROVIDERS) scan_history = ScanHistory.objects.filter(pk=self.scan_id).first() for provider in providers: cmd = f's3scanner -bucket-file {input_path} -enumerate -provider {provider} -threads {threads} -json' for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue if line.get('bucket', {}).get('exists', 0) == 1: result = parse_s3scanner_result(line) s3bucket, created = S3Bucket.objects.get_or_create(**result) scan_history.buckets.add(s3bucket) logger.info(f"s3 bucket added {result['provider']}-{result['name']}-{result['region']}") @app.task(name='http_crawl', queue='main_scan_queue', base=RengineTask, bind=True) def http_crawl( self, urls=[], method=None, recrawl=False, ctx={}, track=True, description=None, is_ran_from_subdomain_scan=False, should_remove_duplicate_endpoints=True, duplicate_removal_fields=[]): """Use httpx to query HTTP URLs for important info like page titles, http status, etc... Args: urls (list, optional): A set of URLs to check. Overrides default behavior which queries all endpoints related to this scan. method (str): HTTP method to use (GET, HEAD, POST, PUT, DELETE). recrawl (bool, optional): If False, filter out URLs that have already been crawled. should_remove_duplicate_endpoints (bool): Whether to remove duplicate endpoints duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates Returns: list: httpx results. """ logger.info('Initiating HTTP Crawl') if is_ran_from_subdomain_scan: logger.info('Running From Subdomain Scan...') cmd = '/go/bin/httpx' cfg = self.yaml_configuration.get(HTTP_CRAWL) or {} custom_header = cfg.get(CUSTOM_HEADER, '') threads = cfg.get(THREADS, DEFAULT_THREADS) follow_redirect = cfg.get(FOLLOW_REDIRECT, True) self.output_path = None input_path = f'{self.results_dir}/httpx_input.txt' history_file = f'{self.results_dir}/commands.txt' if urls: # direct passing URLs to check if self.url_filter: urls = [u for u in urls if self.url_filter in u] with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_uncrawled=not recrawl, write_filepath=input_path, ctx=ctx ) # logger.debug(urls) # If no URLs found, skip it if not urls: return # Re-adjust thread number if few URLs to avoid spinning up a monster to # kill a fly. if len(urls) < threads: threads = len(urls) # Get random proxy proxy = get_random_proxy() # Run command cmd += f' -cl -ct -rt -location -td -websocket -cname -asn -cdn -probe -random-agent' cmd += f' -t {threads}' if threads > 0 else '' cmd += f' --http-proxy {proxy}' if proxy else '' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -json' cmd += f' -u {urls[0]}' if len(urls) == 1 else f' -l {input_path}' cmd += f' -x {method}' if method else '' cmd += f' -silent' if follow_redirect: cmd += ' -fr' results = [] endpoint_ids = [] for line in stream_command( cmd, history_file=history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not line or not isinstance(line, dict): continue logger.debug(line) # No response from endpoint if line.get('failed', False): continue # Parse httpx output host = line.get('host', '') content_length = line.get('content_length', 0) http_status = line.get('status_code') http_url, is_redirect = extract_httpx_url(line) page_title = line.get('title') webserver = line.get('webserver') cdn = line.get('cdn', False) rt = line.get('time') techs = line.get('tech', []) cname = line.get('cname', '') content_type = line.get('content_type', '') response_time = -1 if rt: response_time = float(''.join(ch for ch in rt if not ch.isalpha())) if rt[-2:] == 'ms': response_time = response_time / 1000 # Create Subdomain object in DB subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue # Save default HTTP URL to endpoint object in DB endpoint, created = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain, is_default=is_ran_from_subdomain_scan ) if not endpoint: continue endpoint.http_status = http_status endpoint.page_title = page_title endpoint.content_length = content_length endpoint.webserver = webserver endpoint.response_time = response_time endpoint.content_type = content_type endpoint.save() endpoint_str = f'{http_url} [{http_status}] `{content_length}B` `{webserver}` `{rt}`' logger.warning(endpoint_str) if endpoint and endpoint.is_alive and endpoint.http_status != 403: self.notify( fields={'Alive endpoint': f'• {endpoint_str}'}, add_meta_info=False) # Add endpoint to results line['_cmd'] = cmd line['final_url'] = http_url line['endpoint_id'] = endpoint.id line['endpoint_created'] = created line['is_redirect'] = is_redirect results.append(line) # Add technology objects to DB for technology in techs: tech, _ = Technology.objects.get_or_create(name=technology) endpoint.techs.add(tech) if is_ran_from_subdomain_scan: subdomain.technologies.add(tech) subdomain.save() endpoint.save() techs_str = ', '.join([f'`{tech}`' for tech in techs]) self.notify( fields={'Technologies': techs_str}, add_meta_info=False) # Add IP objects for 'a' records to DB a_records = line.get('a', []) for ip_address in a_records: ip, created = save_ip_address( ip_address, subdomain, subscan=self.subscan, cdn=cdn) ips_str = '• ' + '\n• '.join([f'`{ip}`' for ip in a_records]) self.notify( fields={'IPs': ips_str}, add_meta_info=False) # Add IP object for host in DB if host: ip, created = save_ip_address( host, subdomain, subscan=self.subscan, cdn=cdn) self.notify( fields={'IPs': f'• `{ip.address}`'}, add_meta_info=False) # Save subdomain and endpoint if is_ran_from_subdomain_scan: # save subdomain stuffs subdomain.http_url = http_url subdomain.http_status = http_status subdomain.page_title = page_title subdomain.content_length = content_length subdomain.webserver = webserver subdomain.response_time = response_time subdomain.content_type = content_type subdomain.cname = ','.join(cname) subdomain.is_cdn = cdn if cdn: subdomain.cdn_name = line.get('cdn_name') subdomain.save() endpoint.save() endpoint_ids.append(endpoint.id) if should_remove_duplicate_endpoints: # Remove 'fake' alive endpoints that are just redirects to the same page remove_duplicate_endpoints( self.scan_id, self.domain_id, self.subdomain_id, filter_ids=endpoint_ids ) # Remove input file run_command( f'rm {input_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) return results #---------------------# # Notifications tasks # #---------------------# @app.task(name='send_notif', bind=False, queue='send_notif_queue') def send_notif( message, scan_history_id=None, subscan_id=None, **options): if not 'title' in options: message = enrich_notification(message, scan_history_id, subscan_id) send_discord_message(message, **options) send_slack_message(message) send_telegram_message(message) @app.task(name='send_scan_notif', bind=False, queue='send_scan_notif_queue') def send_scan_notif( scan_history_id, subscan_id=None, engine_id=None, status='RUNNING'): """Send scan status notification. Works for scan or a subscan if subscan_id is passed. Args: scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Get domain, engine, scan_history objects engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() tasks = ScanActivity.objects.filter(scan_of=scan) if scan else 0 # Build notif options url = get_scan_url(scan_history_id, subscan_id) title = get_scan_title(scan_history_id, subscan_id) fields = get_scan_fields(engine, scan, subscan, status, tasks) severity = None msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) if status: severity = STATUS_TO_SEVERITIES.get(status) opts = { 'title': title, 'url': url, 'fields': fields, 'severity': severity } logger.warning(f'Sending notification "{title}" [{severity}]') # Send notification send_notif( msg, scan_history_id, subscan_id, **opts) @app.task(name='send_task_notif', bind=False, queue='send_task_notif_queue') def send_task_notif( task_name, status=None, result=None, output_path=None, traceback=None, scan_history_id=None, engine_id=None, subscan_id=None, severity=None, add_meta_info=True, update_fields={}): """Send task status notification. Args: task_name (str): Task name. status (str, optional): Task status. result (str, optional): Task result. output_path (str, optional): Task output path. traceback (str, optional): Task traceback. scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. severity (str, optional): Severity (will be mapped to notif colors) add_meta_info (bool, optional): Wheter to add scan / subscan info to notif. update_fields (dict, optional): Fields key / value to update. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Build fields url = None fields = {} if add_meta_info: engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() url = get_scan_url(scan_history_id) if status: fields['Status'] = f'**{status}**' if engine: fields['Engine'] = engine.engine_name if scan: fields['Scan ID'] = f'[#{scan.id}]({url})' if subscan: url = get_scan_url(scan_history_id, subscan_id) fields['Subscan ID'] = f'[#{subscan.id}]({url})' title = get_task_title(task_name, scan_history_id, subscan_id) if status: severity = STATUS_TO_SEVERITIES.get(status) msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) # Add fields to update for k, v in update_fields.items(): fields[k] = v # Add traceback to notif if traceback and notif.send_scan_tracebacks: fields['Traceback'] = f'```\n{traceback}\n```' # Add files to notif files = [] attach_file = ( notif.send_scan_output_file and output_path and result and not traceback ) if attach_file: output_title = output_path.split('/')[-1] files = [(output_path, output_title)] # Send notif opts = { 'title': title, 'url': url, 'files': files, 'severity': severity, 'fields': fields, 'fields_append': update_fields.keys() } send_notif( msg, scan_history_id=scan_history_id, subscan_id=subscan_id, **opts) @app.task(name='send_file_to_discord', bind=False, queue='send_file_to_discord_queue') def send_file_to_discord(file_path, title=None): notif = Notification.objects.first() do_send = notif and notif.send_to_discord and notif.discord_hook_url if not do_send: return False webhook = DiscordWebhook( url=notif.discord_hook_url, rate_limit_retry=True, username=title or "reNgine Discord Plugin" ) with open(file_path, "rb") as f: head, tail = os.path.split(file_path) webhook.add_file(file=f.read(), filename=tail) webhook.execute() @app.task(name='send_hackerone_report', bind=False, queue='send_hackerone_report_queue') def send_hackerone_report(vulnerability_id): """Send HackerOne vulnerability report. Args: vulnerability_id (int): Vulnerability id. Returns: int: HTTP response status code. """ vulnerability = Vulnerability.objects.get(id=vulnerability_id) severities = {v: k for k,v in NUCLEI_SEVERITY_MAP.items()} headers = { 'Content-Type': 'application/json', 'Accept': 'application/json' } # can only send vulnerability report if team_handle exists if len(vulnerability.target_domain.h1_team_handle) !=0: hackerone_query = Hackerone.objects.all() if hackerone_query.exists(): hackerone = Hackerone.objects.first() severity_value = severities[vulnerability.severity] tpl = hackerone.report_template # Replace syntax of report template with actual content tpl = tpl.replace('{vulnerability_name}', vulnerability.name) tpl = tpl.replace('{vulnerable_url}', vulnerability.http_url) tpl = tpl.replace('{vulnerability_severity}', severity_value) tpl = tpl.replace('{vulnerability_description}', vulnerability.description if vulnerability.description else '') tpl = tpl.replace('{vulnerability_extracted_results}', vulnerability.extracted_results if vulnerability.extracted_results else '') tpl = tpl.replace('{vulnerability_reference}', vulnerability.reference if vulnerability.reference else '') data = { "data": { "type": "report", "attributes": { "team_handle": vulnerability.target_domain.h1_team_handle, "title": '{} found in {}'.format(vulnerability.name, vulnerability.http_url), "vulnerability_information": tpl, "severity_rating": severity_value, "impact": "More information about the impact and vulnerability can be found here: \n" + vulnerability.reference if vulnerability.reference else "NA", } } } r = requests.post( 'https://api.hackerone.com/v1/hackers/reports', auth=(hackerone.username, hackerone.api_key), json=data, headers=headers ) response = r.json() status_code = r.status_code if status_code == 201: vulnerability.hackerone_report_id = response['data']["id"] vulnerability.open_status = False vulnerability.save() return status_code else: logger.error('No team handle found.') status_code = 111 return status_code #-------------# # Utils tasks # #-------------# @app.task(name='parse_nmap_results', bind=False, queue='parse_nmap_results_queue') def parse_nmap_results(xml_file, output_file=None): """Parse results from nmap output file. Args: xml_file (str): nmap XML report file path. Returns: list: List of vulnerabilities found from nmap results. """ with open(xml_file, encoding='utf8') as f: content = f.read() try: nmap_results = xmltodict.parse(content) # parse XML to dict except Exception as e: logger.exception(e) logger.error(f'Cannot parse {xml_file} to valid JSON. Skipping.') return [] # Write JSON to output file if output_file: with open(output_file, 'w') as f: json.dump(nmap_results, f, indent=4) logger.warning(json.dumps(nmap_results, indent=4)) hosts = ( nmap_results .get('nmaprun', {}) .get('host', {}) ) all_vulns = [] if isinstance(hosts, dict): hosts = [hosts] for host in hosts: # Grab hostname / IP from output hostnames_dict = host.get('hostnames', {}) if hostnames_dict: # Ensure that hostnames['hostname'] is a list for consistency hostnames_list = hostnames_dict['hostname'] if isinstance(hostnames_dict['hostname'], list) else [hostnames_dict['hostname']] # Extract all the @name values from the list of dictionaries hostnames = [entry.get('@name') for entry in hostnames_list] else: hostnames = [host.get('address')['@addr']] # Iterate over each hostname for each port for hostname in hostnames: # Grab ports from output ports = host.get('ports', {}).get('port', []) if isinstance(ports, dict): ports = [ports] for port in ports: url_vulns = [] port_number = port['@portid'] url = sanitize_url(f'{hostname}:{port_number}') logger.info(f'Parsing nmap results for {hostname}:{port_number} ...') if not port_number or not port_number.isdigit(): continue port_protocol = port['@protocol'] scripts = port.get('script', []) if isinstance(scripts, dict): scripts = [scripts] for script in scripts: script_id = script['@id'] script_output = script['@output'] script_output_table = script.get('table', []) logger.debug(f'Ran nmap script "{script_id}" on {port_number}/{port_protocol}:\n{script_output}\n') if script_id == 'vulscan': vulns = parse_nmap_vulscan_output(script_output) url_vulns.extend(vulns) elif script_id == 'vulners': vulns = parse_nmap_vulners_output(script_output) url_vulns.extend(vulns) # elif script_id == 'http-server-header': # TODO: nmap can help find technologies as well using the http-server-header script # regex = r'(\w+)/([\d.]+)\s?(?:\((\w+)\))?' # tech_name, tech_version, tech_os = re.match(regex, test_string).groups() # Technology.objects.get_or_create(...) # elif script_id == 'http_csrf': # vulns = parse_nmap_http_csrf_output(script_output) # url_vulns.extend(vulns) else: logger.warning(f'Script output parsing for script "{script_id}" is not supported yet.') # Add URL to vuln for vuln in url_vulns: # TODO: This should extend to any URL, not just HTTP vuln['http_url'] = url if 'http_path' in vuln: vuln['http_url'] += vuln['http_path'] all_vulns.append(vuln) return all_vulns def parse_nmap_http_csrf_output(script_output): pass def parse_nmap_vulscan_output(script_output): """Parse nmap vulscan script output. Args: script_output (str): Vulscan script output. Returns: list: List of Vulnerability dicts. """ data = {} vulns = [] provider_name = '' # Sort all vulns found by provider so that we can match each provider with # a function that pulls from its API to get more info about the # vulnerability. for line in script_output.splitlines(): if not line: continue if not line.startswith('['): # provider line if "No findings" in line: logger.info(f"No findings: {line}") continue elif ' - ' in line: provider_name, provider_url = tuple(line.split(' - ')) data[provider_name] = {'url': provider_url.rstrip(':'), 'entries': []} continue else: # Log a warning logger.warning(f"Unexpected line format: {line}") continue reg = r'\[(.*)\] (.*)' matches = re.match(reg, line) id, title = matches.groups() entry = {'id': id, 'title': title} data[provider_name]['entries'].append(entry) logger.warning('Vulscan parsed output:') logger.warning(pprint.pformat(data)) for provider_name in data: if provider_name == 'Exploit-DB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'IBM X-Force': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'MITRE CVE': logger.error(f'Provider {provider_name} is not supported YET.') for entry in data[provider_name]['entries']: cve_id = entry['id'] vuln = cve_to_vuln(cve_id) vulns.append(vuln) elif provider_name == 'OSVDB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'OpenVAS (Nessus)': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'SecurityFocus': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'VulDB': logger.error(f'Provider {provider_name} is not supported YET.') pass else: logger.error(f'Provider {provider_name} is not supported.') return vulns def parse_nmap_vulners_output(script_output, url=''): """Parse nmap vulners script output. TODO: Rework this as it's currently matching all CVEs no matter the confidence. Args: script_output (str): Script output. Returns: list: List of found vulnerabilities. """ vulns = [] # Check for CVE in script output CVE_REGEX = re.compile(r'.*(CVE-\d\d\d\d-\d+).*') matches = CVE_REGEX.findall(script_output) matches = list(dict.fromkeys(matches)) for cve_id in matches: # get CVE info vuln = cve_to_vuln(cve_id, vuln_type='nmap-vulners-nse') if vuln: vulns.append(vuln) return vulns def cve_to_vuln(cve_id, vuln_type=''): """Search for a CVE using CVESearch and return Vulnerability data. Args: cve_id (str): CVE ID in the form CVE-* Returns: dict: Vulnerability dict. """ cve_info = CVESearch('https://cve.circl.lu').id(cve_id) if not cve_info: logger.error(f'Could not fetch CVE info for cve {cve_id}. Skipping.') return None vuln_cve_id = cve_info['id'] vuln_name = vuln_cve_id vuln_description = cve_info.get('summary', 'none').replace(vuln_cve_id, '').strip() try: vuln_cvss = float(cve_info.get('cvss', -1)) except (ValueError, TypeError): vuln_cvss = -1 vuln_cwe_id = cve_info.get('cwe', '') exploit_ids = cve_info.get('refmap', {}).get('exploit-db', []) osvdb_ids = cve_info.get('refmap', {}).get('osvdb', []) references = cve_info.get('references', []) capec_objects = cve_info.get('capec', []) # Parse ovals for a better vuln name / type ovals = cve_info.get('oval', []) if ovals: vuln_name = ovals[0]['title'] vuln_type = ovals[0]['family'] # Set vulnerability severity based on CVSS score vuln_severity = 'info' if vuln_cvss < 4: vuln_severity = 'low' elif vuln_cvss < 7: vuln_severity = 'medium' elif vuln_cvss < 9: vuln_severity = 'high' else: vuln_severity = 'critical' # Build console warning message msg = f'{vuln_name} | {vuln_severity.upper()} | {vuln_cve_id} | {vuln_cwe_id} | {vuln_cvss}' for id in osvdb_ids: msg += f'\n\tOSVDB: {id}' for exploit_id in exploit_ids: msg += f'\n\tEXPLOITDB: {exploit_id}' logger.warning(msg) vuln = { 'name': vuln_name, 'type': vuln_type, 'severity': NUCLEI_SEVERITY_MAP[vuln_severity], 'description': vuln_description, 'cvss_score': vuln_cvss, 'references': references, 'cve_ids': [vuln_cve_id], 'cwe_ids': [vuln_cwe_id] } return vuln def parse_s3scanner_result(line): ''' Parses and returns s3Scanner Data ''' bucket = line['bucket'] return { 'name': bucket['name'], 'region': bucket['region'], 'provider': bucket['provider'], 'owner_display_name': bucket['owner_display_name'], 'owner_id': bucket['owner_id'], 'perm_auth_users_read': bucket['perm_auth_users_read'], 'perm_auth_users_write': bucket['perm_auth_users_write'], 'perm_auth_users_read_acl': bucket['perm_auth_users_read_acl'], 'perm_auth_users_write_acl': bucket['perm_auth_users_write_acl'], 'perm_auth_users_full_control': bucket['perm_auth_users_full_control'], 'perm_all_users_read': bucket['perm_all_users_read'], 'perm_all_users_write': bucket['perm_all_users_write'], 'perm_all_users_read_acl': bucket['perm_all_users_read_acl'], 'perm_all_users_write_acl': bucket['perm_all_users_write_acl'], 'perm_all_users_full_control': bucket['perm_all_users_full_control'], 'num_objects': bucket['num_objects'], 'size': bucket['bucket_size'] } def parse_nuclei_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ return { 'name': line['info'].get('name', ''), 'type': line['type'], 'severity': NUCLEI_SEVERITY_MAP[line['info'].get('severity', 'unknown')], 'template': line['template'], 'template_url': line['template-url'], 'template_id': line['template-id'], 'description': line['info'].get('description', ''), 'matcher_name': line.get('matcher-name', ''), 'curl_command': line.get('curl-command'), 'request': line.get('request'), 'response': line.get('response'), 'extracted_results': line.get('extracted-results', []), 'cvss_metrics': line['info'].get('classification', {}).get('cvss-metrics', ''), 'cvss_score': line['info'].get('classification', {}).get('cvss-score'), 'cve_ids': line['info'].get('classification', {}).get('cve_id', []) or [], 'cwe_ids': line['info'].get('classification', {}).get('cwe_id', []) or [], 'references': line['info'].get('reference', []) or [], 'tags': line['info'].get('tags', []), 'source': NUCLEI, } def parse_dalfox_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ description = '' description += f" Evidence: {line.get('evidence')} <br>" if line.get('evidence') else '' description += f" Message: {line.get('message')} <br>" if line.get('message') else '' description += f" Payload: {line.get('message_str')} <br>" if line.get('message_str') else '' description += f" Vulnerable Parameter: {line.get('param')} <br>" if line.get('param') else '' return { 'name': 'XSS (Cross Site Scripting)', 'type': 'XSS', 'severity': DALFOX_SEVERITY_MAP[line.get('severity', 'unknown')], 'description': description, 'source': DALFOX, 'cwe_ids': [line.get('cwe')] } def parse_crlfuzz_result(url): """Parse CRLF results Args: url (str): CRLF Vulnerable URL Returns: dict: Vulnerability data. """ return { 'name': 'CRLF (HTTP Response Splitting)', 'type': 'CRLF', 'severity': 2, 'description': 'A CRLF (HTTP Response Splitting) vulnerability has been discovered.', 'source': CRLFUZZ, } def record_exists(model, data, exclude_keys=[]): """ Check if a record already exists in the database based on the given data. Args: model (django.db.models.Model): The Django model to check against. data (dict): Data dictionary containing fields and values. exclude_keys (list): List of keys to exclude from the lookup. Returns: bool: True if the record exists, False otherwise. """ # Extract the keys that will be used for the lookup lookup_fields = {key: data[key] for key in data if key not in exclude_keys} # Return True if a record exists based on the lookup fields, False otherwise return model.objects.filter(**lookup_fields).exists() @app.task(name='geo_localize', bind=False, queue='geo_localize_queue') def geo_localize(host, ip_id=None): """Uses geoiplookup to find location associated with host. Args: host (str): Hostname. ip_id (int): IpAddress object id. Returns: startScan.models.CountryISO: CountryISO object from DB or None. """ if validators.ipv6(host): logger.info(f'Ipv6 "{host}" is not supported by geoiplookup. Skipping.') return None cmd = f'geoiplookup {host}' _, out = run_command(cmd) if 'IP Address not found' not in out and "can't resolve hostname" not in out: country_iso = out.split(':')[1].strip().split(',')[0] country_name = out.split(':')[1].strip().split(',')[1].strip() geo_object, _ = CountryISO.objects.get_or_create( iso=country_iso, name=country_name ) geo_json = { 'iso': country_iso, 'name': country_name } if ip_id: ip = IpAddress.objects.get(pk=ip_id) ip.geo_iso = geo_object ip.save() return geo_json logger.info(f'Geo IP lookup failed for host "{host}"') return None @app.task(name='query_whois', bind=False, queue='query_whois_queue') def query_whois(ip_domain, force_reload_whois=False): """Query WHOIS information for an IP or a domain name. Args: ip_domain (str): IP address or domain name. save_domain (bool): Whether to save domain or not, default False Returns: dict: WHOIS information. """ if not force_reload_whois and Domain.objects.filter(name=ip_domain).exists() and Domain.objects.get(name=ip_domain).domain_info: domain = Domain.objects.get(name=ip_domain) if not domain.insert_date: domain.insert_date = timezone.now() domain.save() domain_info_db = domain.domain_info domain_info = DottedDict( dnssec=domain_info_db.dnssec, created=domain_info_db.created, updated=domain_info_db.updated, expires=domain_info_db.expires, geolocation_iso=domain_info_db.geolocation_iso, status=[status['name'] for status in DomainWhoisStatusSerializer(domain_info_db.status, many=True).data], whois_server=domain_info_db.whois_server, ns_records=[ns['name'] for ns in NameServersSerializer(domain_info_db.name_servers, many=True).data], registrar_name=domain_info_db.registrar.name, registrar_phone=domain_info_db.registrar.phone, registrar_email=domain_info_db.registrar.email, registrar_url=domain_info_db.registrar.url, registrant_name=domain_info_db.registrant.name, registrant_id=domain_info_db.registrant.id_str, registrant_organization=domain_info_db.registrant.organization, registrant_city=domain_info_db.registrant.city, registrant_state=domain_info_db.registrant.state, registrant_zip_code=domain_info_db.registrant.zip_code, registrant_country=domain_info_db.registrant.country, registrant_phone=domain_info_db.registrant.phone, registrant_fax=domain_info_db.registrant.fax, registrant_email=domain_info_db.registrant.email, registrant_address=domain_info_db.registrant.address, admin_name=domain_info_db.admin.name, admin_id=domain_info_db.admin.id_str, admin_organization=domain_info_db.admin.organization, admin_city=domain_info_db.admin.city, admin_state=domain_info_db.admin.state, admin_zip_code=domain_info_db.admin.zip_code, admin_country=domain_info_db.admin.country, admin_phone=domain_info_db.admin.phone, admin_fax=domain_info_db.admin.fax, admin_email=domain_info_db.admin.email, admin_address=domain_info_db.admin.address, tech_name=domain_info_db.tech.name, tech_id=domain_info_db.tech.id_str, tech_organization=domain_info_db.tech.organization, tech_city=domain_info_db.tech.city, tech_state=domain_info_db.tech.state, tech_zip_code=domain_info_db.tech.zip_code, tech_country=domain_info_db.tech.country, tech_phone=domain_info_db.tech.phone, tech_fax=domain_info_db.tech.fax, tech_email=domain_info_db.tech.email, tech_address=domain_info_db.tech.address, related_tlds=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_tlds, many=True).data], related_domains=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_domains, many=True).data], historical_ips=[ip for ip in HistoricalIPSerializer(domain_info_db.historical_ips, many=True).data], ) if domain_info_db.dns_records: a_records = [] txt_records = [] mx_records = [] dns_records = [{'name': dns['name'], 'type': dns['type']} for dns in DomainDNSRecordSerializer(domain_info_db.dns_records, many=True).data] for dns in dns_records: if dns['type'] == 'a': a_records.append(dns['name']) elif dns['type'] == 'txt': txt_records.append(dns['name']) elif dns['type'] == 'mx': mx_records.append(dns['name']) domain_info.a_records = a_records domain_info.txt_records = txt_records domain_info.mx_records = mx_records else: logger.info(f'Domain info for "{ip_domain}" not found in DB, querying whois') domain_info = DottedDict() # find domain historical ip try: historical_ips = get_domain_historical_ip_address(ip_domain) domain_info.historical_ips = historical_ips except Exception as e: logger.error(f'HistoricalIP for {ip_domain} not found!\nError: {str(e)}') historical_ips = [] # find associated domains using ip_domain try: related_domains = reverse_whois(ip_domain.split('.')[0]) except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] # find related tlds using TLSx try: related_tlds = [] output_path = '/tmp/ip_domain_tlsx.txt' tlsx_command = f'tlsx -san -cn -silent -ro -host {ip_domain} -o {output_path}' run_command( tlsx_command, shell=True, ) tlsx_output = [] with open(output_path) as f: tlsx_output = f.readlines() tldextract_target = tldextract.extract(ip_domain) for doms in tlsx_output: doms = doms.strip() tldextract_res = tldextract.extract(doms) if ip_domain != doms and tldextract_res.domain == tldextract_target.domain and tldextract_res.subdomain == '': related_tlds.append(doms) related_tlds = list(set(related_tlds)) domain_info.related_tlds = related_tlds except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] related_domains_list = [] if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_related = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_related) related_domains_list.append(_domain['name']) for _domain in related_tlds: domain_related = RelatedDomain.objects.get_or_create( name=_domain, )[0] db_domain_info.related_tlds.add(domain_related) for _ip in historical_ips: historical_ip = HistoricalIP.objects.get_or_create( ip=_ip['ip'], owner=_ip['owner'], location=_ip['location'], last_seen=_ip['last_seen'], )[0] db_domain_info.historical_ips.add(historical_ip) domain.domain_info = db_domain_info domain.save() command = f'netlas host {ip_domain} -f json' # check if netlas key is provided netlas_key = get_netlas_key() command += f' -a {netlas_key}' if netlas_key else '' result = subprocess.check_output(command.split()).decode('utf-8') if 'Failed to parse response data' in result: # do fallback return { 'status': False, 'ip_domain': ip_domain, 'result': "Netlas limit exceeded.", 'message': 'Netlas limit exceeded.' } try: result = json.loads(result) logger.info(result) whois = result.get('whois') if result.get('whois') else {} domain_info.created = whois.get('created_date') domain_info.expires = whois.get('expiration_date') domain_info.updated = whois.get('updated_date') domain_info.whois_server = whois.get('whois_server') if 'registrant' in whois: registrant = whois.get('registrant') domain_info.registrant_name = registrant.get('name') domain_info.registrant_country = registrant.get('country') domain_info.registrant_id = registrant.get('id') domain_info.registrant_state = registrant.get('province') domain_info.registrant_city = registrant.get('city') domain_info.registrant_phone = registrant.get('phone') domain_info.registrant_address = registrant.get('street') domain_info.registrant_organization = registrant.get('organization') domain_info.registrant_fax = registrant.get('fax') domain_info.registrant_zip_code = registrant.get('postal_code') email_search = EMAIL_REGEX.search(str(registrant.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.registrant_email = field_content if 'administrative' in whois: administrative = whois.get('administrative') domain_info.admin_name = administrative.get('name') domain_info.admin_country = administrative.get('country') domain_info.admin_id = administrative.get('id') domain_info.admin_state = administrative.get('province') domain_info.admin_city = administrative.get('city') domain_info.admin_phone = administrative.get('phone') domain_info.admin_address = administrative.get('street') domain_info.admin_organization = administrative.get('organization') domain_info.admin_fax = administrative.get('fax') domain_info.admin_zip_code = administrative.get('postal_code') mail_search = EMAIL_REGEX.search(str(administrative.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.admin_email = field_content if 'technical' in whois: technical = whois.get('technical') domain_info.tech_name = technical.get('name') domain_info.tech_country = technical.get('country') domain_info.tech_state = technical.get('province') domain_info.tech_id = technical.get('id') domain_info.tech_city = technical.get('city') domain_info.tech_phone = technical.get('phone') domain_info.tech_address = technical.get('street') domain_info.tech_organization = technical.get('organization') domain_info.tech_fax = technical.get('fax') domain_info.tech_zip_code = technical.get('postal_code') mail_search = EMAIL_REGEX.search(str(technical.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.tech_email = field_content if 'dns' in result: dns = result.get('dns') domain_info.mx_records = dns.get('mx') domain_info.txt_records = dns.get('txt') domain_info.a_records = dns.get('a') domain_info.ns_records = whois.get('name_servers') domain_info.dnssec = True if whois.get('dnssec') else False domain_info.status = whois.get('status') if 'registrar' in whois: registrar = whois.get('registrar') domain_info.registrar_name = registrar.get('name') domain_info.registrar_email = registrar.get('email') domain_info.registrar_phone = registrar.get('phone') domain_info.registrar_url = registrar.get('url') # find associated domains if registrant email is found related_domains = reverse_whois(domain_info.get('registrant_email')) if domain_info.get('registrant_email') else [] for _domain in related_domains: related_domains_list.append(_domain['name']) # remove duplicate domains from related domains list related_domains_list = list(set(related_domains_list)) domain_info.related_domains = related_domains_list # save to db if domain exists if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_rel = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_rel) db_domain_info.dnssec = domain_info.get('dnssec') #dates db_domain_info.created = domain_info.get('created') db_domain_info.updated = domain_info.get('updated') db_domain_info.expires = domain_info.get('expires') #registrar db_domain_info.registrar = Registrar.objects.get_or_create( name=domain_info.get('registrar_name'), email=domain_info.get('registrar_email'), phone=domain_info.get('registrar_phone'), url=domain_info.get('registrar_url'), )[0] db_domain_info.registrant = DomainRegistration.objects.get_or_create( name=domain_info.get('registrant_name'), organization=domain_info.get('registrant_organization'), address=domain_info.get('registrant_address'), city=domain_info.get('registrant_city'), state=domain_info.get('registrant_state'), zip_code=domain_info.get('registrant_zip_code'), country=domain_info.get('registrant_country'), email=domain_info.get('registrant_email'), phone=domain_info.get('registrant_phone'), fax=domain_info.get('registrant_fax'), id_str=domain_info.get('registrant_id'), )[0] db_domain_info.admin = DomainRegistration.objects.get_or_create( name=domain_info.get('admin_name'), organization=domain_info.get('admin_organization'), address=domain_info.get('admin_address'), city=domain_info.get('admin_city'), state=domain_info.get('admin_state'), zip_code=domain_info.get('admin_zip_code'), country=domain_info.get('admin_country'), email=domain_info.get('admin_email'), phone=domain_info.get('admin_phone'), fax=domain_info.get('admin_fax'), id_str=domain_info.get('admin_id'), )[0] db_domain_info.tech = DomainRegistration.objects.get_or_create( name=domain_info.get('tech_name'), organization=domain_info.get('tech_organization'), address=domain_info.get('tech_address'), city=domain_info.get('tech_city'), state=domain_info.get('tech_state'), zip_code=domain_info.get('tech_zip_code'), country=domain_info.get('tech_country'), email=domain_info.get('tech_email'), phone=domain_info.get('tech_phone'), fax=domain_info.get('tech_fax'), id_str=domain_info.get('tech_id'), )[0] for status in domain_info.get('status') or []: _status = WhoisStatus.objects.get_or_create( name=status )[0] _status.save() db_domain_info.status.add(_status) for ns in domain_info.get('ns_records') or []: _ns = NameServer.objects.get_or_create( name=ns )[0] _ns.save() db_domain_info.name_servers.add(_ns) for a in domain_info.get('a_records') or []: _a = DNSRecord.objects.get_or_create( name=a, type='a' )[0] _a.save() db_domain_info.dns_records.add(_a) for mx in domain_info.get('mx_records') or []: _mx = DNSRecord.objects.get_or_create( name=mx, type='mx' )[0] _mx.save() db_domain_info.dns_records.add(_mx) for txt in domain_info.get('txt_records') or []: _txt = DNSRecord.objects.get_or_create( name=txt, type='txt' )[0] _txt.save() db_domain_info.dns_records.add(_txt) db_domain_info.geolocation_iso = domain_info.get('registrant_country') db_domain_info.whois_server = domain_info.get('whois_server') db_domain_info.save() domain.domain_info = db_domain_info domain.save() except Exception as e: return { 'status': False, 'ip_domain': ip_domain, 'result': "unable to fetch records from WHOIS database.", 'message': str(e) } return { 'status': True, 'ip_domain': ip_domain, 'dnssec': domain_info.get('dnssec'), 'created': domain_info.get('created'), 'updated': domain_info.get('updated'), 'expires': domain_info.get('expires'), 'geolocation_iso': domain_info.get('registrant_country'), 'domain_statuses': domain_info.get('status'), 'whois_server': domain_info.get('whois_server'), 'dns': { 'a': domain_info.get('a_records'), 'mx': domain_info.get('mx_records'), 'txt': domain_info.get('txt_records'), }, 'registrar': { 'name': domain_info.get('registrar_name'), 'phone': domain_info.get('registrar_phone'), 'email': domain_info.get('registrar_email'), 'url': domain_info.get('registrar_url'), }, 'registrant': { 'name': domain_info.get('registrant_name'), 'id': domain_info.get('registrant_id'), 'organization': domain_info.get('registrant_organization'), 'address': domain_info.get('registrant_address'), 'city': domain_info.get('registrant_city'), 'state': domain_info.get('registrant_state'), 'zipcode': domain_info.get('registrant_zip_code'), 'country': domain_info.get('registrant_country'), 'phone': domain_info.get('registrant_phone'), 'fax': domain_info.get('registrant_fax'), 'email': domain_info.get('registrant_email'), }, 'admin': { 'name': domain_info.get('admin_name'), 'id': domain_info.get('admin_id'), 'organization': domain_info.get('admin_organization'), 'address':domain_info.get('admin_address'), 'city': domain_info.get('admin_city'), 'state': domain_info.get('admin_state'), 'zipcode': domain_info.get('admin_zip_code'), 'country': domain_info.get('admin_country'), 'phone': domain_info.get('admin_phone'), 'fax': domain_info.get('admin_fax'), 'email': domain_info.get('admin_email'), }, 'technical_contact': { 'name': domain_info.get('tech_name'), 'id': domain_info.get('tech_id'), 'organization': domain_info.get('tech_organization'), 'address': domain_info.get('tech_address'), 'city': domain_info.get('tech_city'), 'state': domain_info.get('tech_state'), 'zipcode': domain_info.get('tech_zip_code'), 'country': domain_info.get('tech_country'), 'phone': domain_info.get('tech_phone'), 'fax': domain_info.get('tech_fax'), 'email': domain_info.get('tech_email'), }, 'nameservers': domain_info.get('ns_records'), # 'similar_domains': domain_info.get('similar_domains'), 'related_domains': domain_info.get('related_domains'), 'related_tlds': domain_info.get('related_tlds'), 'historical_ips': domain_info.get('historical_ips'), } @app.task(name='remove_duplicate_endpoints', bind=False, queue='remove_duplicate_endpoints_queue') def remove_duplicate_endpoints( scan_history_id, domain_id, subdomain_id=None, filter_ids=[], filter_status=[200, 301, 404], duplicate_removal_fields=ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS ): """Remove duplicate endpoints. Check for implicit redirections by comparing endpoints: - [x] `content_length` similarities indicating redirections - [x] `page_title` (check for same page title) - [ ] Sign-in / login page (check for endpoints with the same words) Args: scan_history_id: ScanHistory id. domain_id (int): Domain id. subdomain_id (int, optional): Subdomain id. filter_ids (list): List of endpoint ids to filter on. filter_status (list): List of HTTP status codes to filter on. duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates """ logger.info(f'Removing duplicate endpoints based on {duplicate_removal_fields}') endpoints = ( EndPoint.objects .filter(scan_history__id=scan_history_id) .filter(target_domain__id=domain_id) ) if filter_status: endpoints = endpoints.filter(http_status__in=filter_status) if subdomain_id: endpoints = endpoints.filter(subdomain__id=subdomain_id) if filter_ids: endpoints = endpoints.filter(id__in=filter_ids) for field_name in duplicate_removal_fields: cl_query = ( endpoints .values_list(field_name) .annotate(mc=Count(field_name)) .order_by('-mc') ) for (field_value, count) in cl_query: if count > DELETE_DUPLICATES_THRESHOLD: eps_to_delete = ( endpoints .filter(**{field_name: field_value}) .order_by('discovered_date') .all()[1:] ) msg = f'Deleting {len(eps_to_delete)} endpoints [reason: same {field_name} {field_value}]' for ep in eps_to_delete: url = urlparse(ep.http_url) if url.path in ['', '/', '/login']: # try do not delete the original page that other pages redirect to continue msg += f'\n\t {ep.http_url} [{ep.http_status}] [{field_name}={field_value}]' ep.delete() logger.warning(msg) @app.task(name='run_command', bind=False, queue='run_command_queue') def run_command(cmd, cwd=None, shell=False, history_file=None, scan_id=None, activity_id=None): """Run a given command using subprocess module. Args: cmd (str): Command to run. cwd (str): Current working directory. echo (bool): Log command. shell (bool): Run within separate shell if True. history_file (str): Write command + output to history file. Returns: tuple: Tuple with return_code, output. """ logger.info(cmd) logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Run the command using subprocess popen = subprocess.Popen( cmd if shell else cmd.split(), shell=shell, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, cwd=cwd, universal_newlines=True) output = '' for stdout_line in iter(popen.stdout.readline, ""): item = stdout_line.strip() output += '\n' + item logger.debug(item) popen.stdout.close() popen.wait() return_code = popen.returncode command_obj.output = output command_obj.return_code = return_code command_obj.save() if history_file: mode = 'a' if not os.path.exists(history_file): mode = 'w' with open(history_file, mode) as f: f.write(f'\n{cmd}\n{return_code}\n{output}\n------------------\n') return return_code, output #-------------# # Other utils # #-------------# def stream_command(cmd, cwd=None, shell=False, history_file=None, encoding='utf-8', scan_id=None, activity_id=None, trunc_char=None): # Log cmd logger.info(cmd) # logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Sanitize the cmd command = cmd if shell else cmd.split() # Run the command using subprocess process = subprocess.Popen( command, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, universal_newlines=True, shell=shell) # Log the output in real-time to the database output = "" # Process the output for line in iter(lambda: process.stdout.readline(), b''): if not line: break line = line.strip() ansi_escape = re.compile(r'\x1B(?:[@-Z\\-_]|\[[0-?]*[ -/]*[@-~])') line = ansi_escape.sub('', line) line = line.replace('\\x0d\\x0a', '\n') if trunc_char and line.endswith(trunc_char): line = line[:-1] item = line # Try to parse the line as JSON try: item = json.loads(line) except json.JSONDecodeError: pass # Yield the line #logger.debug(item) yield item # Add the log line to the output output += line + "\n" # Update the command record in the database command_obj.output = output command_obj.save() # Retrieve the return code and output process.wait() return_code = process.returncode # Update the return code and final output in the database command_obj.return_code = return_code command_obj.save() # Append the command, return code and output to the history file if history_file is not None: with open(history_file, "a") as f: f.write(f"{cmd}\n{return_code}\n{output}\n") def process_httpx_response(line): """TODO: implement this""" def extract_httpx_url(line): """Extract final URL from httpx results. Always follow redirects to find the last URL. Args: line (dict): URL data output by httpx. Returns: tuple: (final_url, redirect_bool) tuple. """ status_code = line.get('status_code', 0) final_url = line.get('final_url') location = line.get('location') chain_status_codes = line.get('chain_status_codes', []) # Final URL is already looking nice, if it exists return it if final_url: return final_url, False http_url = line['url'] # fallback to url field # Handle redirects manually REDIRECT_STATUS_CODES = [301, 302] is_redirect = ( status_code in REDIRECT_STATUS_CODES or any(x in REDIRECT_STATUS_CODES for x in chain_status_codes) ) if is_redirect and location: if location.startswith(('http', 'https')): http_url = location else: http_url = f'{http_url}/{location.lstrip("/")}' # Sanitize URL http_url = sanitize_url(http_url) return http_url, is_redirect #-------------# # OSInt utils # #-------------# def get_and_save_dork_results(lookup_target, results_dir, type, lookup_keywords=None, lookup_extensions=None, delay=3, page_count=2, scan_history=None): """ Uses gofuzz to dork and store information Args: lookup_target (str): target to look into such as stackoverflow or even the target itself results_dir (str): Results directory type (str): Dork Type Title lookup_keywords (str): comma separated keywords or paths to look for lookup_extensions (str): comma separated extensions to look for delay (int): delay between each requests page_count (int): pages in google to extract information scan_history (startScan.ScanHistory): Scan History Object """ results = [] gofuzz_command = f'{GOFUZZ_EXEC_PATH} -t {lookup_target} -d {delay} -p {page_count}' if lookup_extensions: gofuzz_command += f' -e {lookup_extensions}' elif lookup_keywords: gofuzz_command += f' -w {lookup_keywords}' output_file = f'{results_dir}/gofuzz.txt' gofuzz_command += f' -o {output_file}' history_file = f'{results_dir}/commands.txt' try: run_command( gofuzz_command, shell=False, history_file=history_file, scan_id=scan_history.id, ) if not os.path.isfile(output_file): return with open(output_file) as f: for line in f.readlines(): url = line.strip() if url: results.append(url) dork, created = Dork.objects.get_or_create( type=type, url=url ) if scan_history: scan_history.dorks.add(dork) # remove output file os.remove(output_file) except Exception as e: logger.exception(e) return results def get_and_save_emails(scan_history, activity_id, results_dir): """Get and save emails from Google, Bing and Baidu. Args: scan_history (startScan.ScanHistory): Scan history object. activity_id: ScanActivity Object results_dir (str): Results directory. Returns: list: List of emails found. """ emails = [] # Proxy settings # get_random_proxy() # Gather emails from Google, Bing and Baidu output_file = f'{results_dir}/emails_tmp.txt' history_file = f'{results_dir}/commands.txt' command = f'python3 /usr/src/github/Infoga/infoga.py --domain {scan_history.domain.name} --source all --report {output_file}' try: run_command( command, shell=False, history_file=history_file, scan_id=scan_history.id, activity_id=activity_id) if not os.path.isfile(output_file): logger.info('No Email results') return [] with open(output_file) as f: for line in f.readlines(): if 'Email' in line: split_email = line.split(' ')[2] emails.append(split_email) output_path = f'{results_dir}/emails.txt' with open(output_path, 'w') as output_file: for email_address in emails: save_email(email_address, scan_history) output_file.write(f'{email_address}\n') except Exception as e: logger.exception(e) return emails def save_metadata_info(meta_dict): """Extract metadata from Google Search. Args: meta_dict (dict): Info dict. Returns: list: List of startScan.MetaFinderDocument objects. """ logger.warning(f'Getting metadata for {meta_dict.osint_target}') scan_history = ScanHistory.objects.get(id=meta_dict.scan_id) # Proxy settings get_random_proxy() # Get metadata result = extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if not result: logger.error(f'No metadata result from Google Search for {meta_dict.osint_target}.') return [] # Add metadata info to DB results = [] for metadata_name, data in result.get_metadata().items(): subdomain = Subdomain.objects.get( scan_history=meta_dict.scan_id, name=meta_dict.osint_target) metadata = DottedDict({k: v for k, v in data.items()}) meta_finder_document = MetaFinderDocument( subdomain=subdomain, target_domain=meta_dict.domain, scan_history=scan_history, url=metadata.url, doc_name=metadata_name, http_status=metadata.status_code, producer=metadata.metadata.get('Producer'), creator=metadata.metadata.get('Creator'), creation_date=metadata.metadata.get('CreationDate'), modified_date=metadata.metadata.get('ModDate'), author=metadata.metadata.get('Author'), title=metadata.metadata.get('Title'), os=metadata.metadata.get('OSInfo')) meta_finder_document.save() results.append(data) return results #-----------------# # Utils functions # #-----------------# def create_scan_activity(scan_history_id, message, status): scan_activity = ScanActivity() scan_activity.scan_of = ScanHistory.objects.get(pk=scan_history_id) scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id #--------------------# # Database functions # #--------------------# def save_vulnerability(**vuln_data): references = vuln_data.pop('references', []) cve_ids = vuln_data.pop('cve_ids', []) cwe_ids = vuln_data.pop('cwe_ids', []) tags = vuln_data.pop('tags', []) subscan = vuln_data.pop('subscan', None) # remove nulls vuln_data = replace_nulls(vuln_data) # Create vulnerability vuln, created = Vulnerability.objects.get_or_create(**vuln_data) if created: vuln.discovered_date = timezone.now() vuln.open_status = True vuln.save() # Save vuln tags for tag_name in tags or []: tag, created = VulnerabilityTags.objects.get_or_create(name=tag_name) if tag: vuln.tags.add(tag) vuln.save() # Save CVEs for cve_id in cve_ids or []: cve, created = CveId.objects.get_or_create(name=cve_id) if cve: vuln.cve_ids.add(cve) vuln.save() # Save CWEs for cve_id in cwe_ids or []: cwe, created = CweId.objects.get_or_create(name=cve_id) if cwe: vuln.cwe_ids.add(cwe) vuln.save() # Save vuln reference for url in references or []: ref, created = VulnerabilityReference.objects.get_or_create(url=url) if created: vuln.references.add(ref) vuln.save() # Save subscan id in vuln object if subscan: vuln.vuln_subscan_ids.add(subscan) vuln.save() return vuln, created def save_endpoint( http_url, ctx={}, crawl=False, is_default=False, **endpoint_data): """Get or create EndPoint object. If crawl is True, also crawl the endpoint HTTP URL with httpx. Args: http_url (str): Input HTTP URL. is_default (bool): If the url is a default url for SubDomains. scan_history (startScan.models.ScanHistory): ScanHistory object. domain (startScan.models.Domain): Domain object. subdomain (starScan.models.Subdomain): Subdomain object. results_dir (str, optional): Results directory. crawl (bool, optional): Run httpx on endpoint if True. Default: False. force (bool, optional): Force crawl even if ENABLE_HTTP_CRAWL mode is on. subscan (startScan.models.SubScan, optional): SubScan object. Returns: tuple: (startScan.models.EndPoint, created) where `created` is a boolean indicating if the object is new or already existed. """ # remove nulls endpoint_data = replace_nulls(endpoint_data) scheme = urlparse(http_url).scheme endpoint = None created = False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in http_url: logger.error(f"{http_url} is not a URL of domain {domain.name}. Skipping.") return None, False if crawl: ctx['track'] = False results = http_crawl( urls=[http_url], method='HEAD', ctx=ctx) if results: endpoint_data = results[0] endpoint_id = endpoint_data['endpoint_id'] created = endpoint_data['endpoint_created'] endpoint = EndPoint.objects.get(pk=endpoint_id) elif not scheme: return None, False else: # add dumb endpoint without probing it scan = ScanHistory.objects.filter(pk=ctx.get('scan_history_id')).first() domain = Domain.objects.filter(pk=ctx.get('domain_id')).first() if not validators.url(http_url): return None, False http_url = sanitize_url(http_url) # Try to get the first matching record (prevent duplicate error) endpoints = EndPoint.objects.filter( scan_history=scan, target_domain=domain, http_url=http_url, **endpoint_data ) if endpoints.exists(): endpoint = endpoints.first() created = False else: # No existing record, create a new one endpoint = EndPoint.objects.create( scan_history=scan, target_domain=domain, http_url=http_url, **endpoint_data ) created = True if created: endpoint.is_default = is_default endpoint.discovered_date = timezone.now() endpoint.save() subscan_id = ctx.get('subscan_id') if subscan_id: endpoint.endpoint_subscan_ids.add(subscan_id) endpoint.save() return endpoint, created def save_subdomain(subdomain_name, ctx={}): """Get or create Subdomain object. Args: subdomain_name (str): Subdomain name. scan_history (startScan.models.ScanHistory): ScanHistory object. Returns: tuple: (startScan.models.Subdomain, created) where `created` is a boolean indicating if the object has been created in DB. """ scan_id = ctx.get('scan_history_id') subscan_id = ctx.get('subscan_id') out_of_scope_subdomains = ctx.get('out_of_scope_subdomains', []) valid_domain = ( validators.domain(subdomain_name) or validators.ipv4(subdomain_name) or validators.ipv6(subdomain_name) ) if not valid_domain: logger.error(f'{subdomain_name} is not an invalid domain. Skipping.') return None, False if subdomain_name in out_of_scope_subdomains: logger.error(f'{subdomain_name} is out-of-scope. Skipping.') return None, False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in subdomain_name: logger.error(f"{subdomain_name} is not a subdomain of domain {domain.name}. Skipping.") return None, False scan = ScanHistory.objects.filter(pk=scan_id).first() domain = scan.domain if scan else None subdomain, created = Subdomain.objects.get_or_create( scan_history=scan, target_domain=domain, name=subdomain_name) if created: # logger.warning(f'Found new subdomain {subdomain_name}') subdomain.discovered_date = timezone.now() if subscan_id: subdomain.subdomain_subscan_ids.add(subscan_id) subdomain.save() return subdomain, created def save_email(email_address, scan_history=None): if not validators.email(email_address): logger.info(f'Email {email_address} is invalid. Skipping.') return None, False email, created = Email.objects.get_or_create(address=email_address) # if created: # logger.warning(f'Found new email address {email_address}') # Add email to ScanHistory if scan_history: scan_history.emails.add(email) scan_history.save() return email, created def save_employee(name, designation, scan_history=None): employee, created = Employee.objects.get_or_create( name=name, designation=designation) # if created: # logger.warning(f'Found new employee {name}') # Add employee to ScanHistory if scan_history: scan_history.employees.add(employee) scan_history.save() return employee, created def save_ip_address(ip_address, subdomain=None, subscan=None, **kwargs): if not (validators.ipv4(ip_address) or validators.ipv6(ip_address)): logger.info(f'IP {ip_address} is not a valid IP. Skipping.') return None, False ip, created = IpAddress.objects.get_or_create(address=ip_address) # if created: # logger.warning(f'Found new IP {ip_address}') # Set extra attributes for key, value in kwargs.items(): setattr(ip, key, value) ip.save() # Add IP to subdomain if subdomain: subdomain.ip_addresses.add(ip) subdomain.save() # Add subscan to IP if subscan: ip.ip_subscan_ids.add(subscan) # Geo-localize IP asynchronously if created: geo_localize.delay(ip_address, ip.id) return ip, created def save_imported_subdomains(subdomains, ctx={}): """Take a list of subdomains imported and write them to from_imported.txt. Args: subdomains (list): List of subdomain names. scan_history (startScan.models.ScanHistory): ScanHistory instance. domain (startScan.models.Domain): Domain instance. results_dir (str): Results directory. """ domain_id = ctx['domain_id'] domain = Domain.objects.get(pk=domain_id) results_dir = ctx.get('results_dir', RENGINE_RESULTS) # Validate each subdomain and de-duplicate entries subdomains = list(set([ subdomain for subdomain in subdomains if validators.domain(subdomain) and domain.name == get_domain_from_subdomain(subdomain) ])) if not subdomains: return logger.warning(f'Found {len(subdomains)} imported subdomains.') with open(f'{results_dir}/from_imported.txt', 'w+') as output_file: for name in subdomains: subdomain_name = name.strip() subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain.is_imported_subdomain = True subdomain.save() output_file.write(f'{subdomain}\n') @app.task(name='query_reverse_whois', bind=False, queue='query_reverse_whois_queue') def query_reverse_whois(lookup_keyword): """Queries Reverse WHOIS information for an organization or email address. Args: lookup_keyword (str): Registrar Name or email Returns: dict: Reverse WHOIS information. """ return get_associated_domains(lookup_keyword) @app.task(name='query_ip_history', bind=False, queue='query_ip_history_queue') def query_ip_history(domain): """Queries the IP history for a domain Args: domain (str): domain_name Returns: list: list of historical ip addresses """ return get_domain_historical_ip_address(domain) @app.task(name='gpt_vulnerability_description', bind=False, queue='gpt_queue') def gpt_vulnerability_description(vulnerability_id): """Generate and store Vulnerability Description using GPT. Args: vulnerability_id (Vulnerability Model ID): Vulnerability ID to fetch Description. """ logger.info('Getting GPT Vulnerability Description') try: lookup_vulnerability = Vulnerability.objects.get(id=vulnerability_id) lookup_url = urlparse(lookup_vulnerability.http_url) path = lookup_url.path except Exception as e: return { 'status': False, 'error': str(e) } # check in db GPTVulnerabilityReport model if vulnerability description and path matches stored = GPTVulnerabilityReport.objects.filter(url_path=path).filter(title=lookup_vulnerability.name).first() if stored: response = { 'status': True, 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: vulnerability_description = get_gpt_vuln_input_description( lookup_vulnerability.name, path ) # one can add more description here later gpt_generator = GPTVulnerabilityReportGenerator() response = gpt_generator.get_vulnerability_description(vulnerability_description) add_gpt_description_db( lookup_vulnerability.name, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) # for all vulnerabilities with the same vulnerability name this description has to be stored. # also the consition is that the url must contain a part of this. for vuln in Vulnerability.objects.filter(name=lookup_vulnerability.name, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() return response

psyray

7c01a46cea370e74385682ba7c28eaf4e58f5d69

2e089dc62f1bd64aa481750da10fa750e3aa232d

Is this comment still needed?

AnonymousWP

yogeshojha/rengine

Fix crash on saving endpoint (FFUF related only)

Fix #1006 I've added : - a **try except** block to catch error on duplicate record returned by **get_or_create** in **saving_endpoint** method - a **check** on endpoint existence in **dir_file_fuzz** method Errors are logged to the console with the URL.  Tested and working Now we need to find why there are duplicates endpoints in the db But it's another issue

2023-11-22 02:57:45+00:00

2023-11-27 12:37:27+00:00

web/reNgine/tasks.py

import csv import json import os import pprint import subprocess import time import validators import whatportis import xmltodict import yaml import tldextract import concurrent.futures from datetime import datetime from urllib.parse import urlparse from api.serializers import SubdomainSerializer from celery import chain, chord, group from celery.result import allow_join_result from celery.utils.log import get_task_logger from django.db.models import Count from dotted_dict import DottedDict from django.utils import timezone from pycvesearch import CVESearch from metafinder.extractor import extract_metadata_from_google_search from reNgine.celery import app from reNgine.gpt import GPTVulnerabilityReportGenerator from reNgine.celery_custom_task import RengineTask from reNgine.common_func import * from reNgine.definitions import * from reNgine.settings import * from reNgine.gpt import * from reNgine.utilities import * from scanEngine.models import (EngineType, InstalledExternalTool, Notification, Proxy) from startScan.models import * from startScan.models import EndPoint, Subdomain, Vulnerability from targetApp.models import Domain """ Celery tasks. """ logger = get_task_logger(__name__) #----------------------# # Scan / Subscan tasks # #----------------------# @app.task(name='initiate_scan', bind=False, queue='initiate_scan_queue') def initiate_scan( scan_history_id, domain_id, engine_id=None, scan_type=LIVE_SCAN, results_dir=RENGINE_RESULTS, imported_subdomains=[], out_of_scope_subdomains=[], url_filter=''): """Initiate a new scan. Args: scan_history_id (int): ScanHistory id. domain_id (int): Domain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. imported_subdomains (list): Imported subdomains. out_of_scope_subdomains (list): Out-of-scope subdomains. url_filter (str): URL path. Default: '' """ # Get scan history scan = ScanHistory.objects.get(pk=scan_history_id) # Get scan engine engine_id = engine_id or scan.scan_type.id # scan history engine_id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, []) # Get domain and set last_scan_date domain = Domain.objects.get(pk=domain_id) domain.last_scan_date = timezone.now() domain.save() # Get path filter url_filter = url_filter.rstrip('/') # Get or create ScanHistory() object if scan_type == LIVE_SCAN: # immediate scan = ScanHistory.objects.get(pk=scan_history_id) scan.scan_status = RUNNING_TASK elif scan_type == SCHEDULED_SCAN: # scheduled scan = ScanHistory() scan.scan_status = INITIATED_TASK scan.scan_type = engine scan.celery_ids = [initiate_scan.request.id] scan.domain = domain scan.start_scan_date = timezone.now() scan.tasks = engine.tasks scan.results_dir = f'{results_dir}/{domain.name}_{scan.id}' add_gf_patterns = gf_patterns and 'fetch_url' in engine.tasks if add_gf_patterns: scan.used_gf_patterns = ','.join(gf_patterns) scan.save() # Create scan results dir os.makedirs(scan.results_dir) # Build task context ctx = { 'scan_history_id': scan_history_id, 'engine_id': engine_id, 'domain_id': domain.id, 'results_dir': scan.results_dir, 'url_filter': url_filter, 'yaml_configuration': config, 'out_of_scope_subdomains': out_of_scope_subdomains } ctx_str = json.dumps(ctx, indent=2) # Send start notif logger.warning(f'Starting scan {scan_history_id} with context:\n{ctx_str}') send_scan_notif.delay( scan_history_id, subscan_id=None, engine_id=engine_id, status=CELERY_TASK_STATUS_MAP[scan.scan_status]) # Save imported subdomains in DB save_imported_subdomains(imported_subdomains, ctx=ctx) # Create initial subdomain in DB: make a copy of domain as a subdomain so # that other tasks using subdomains can use it. subdomain_name = domain.name subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) # If enable_http_crawl is set, create an initial root HTTP endpoint so that # HTTP crawling can start somewhere http_url = f'{domain.name}{url_filter}' if url_filter else domain.name endpoint, _ = save_endpoint( http_url, ctx=ctx, crawl=enable_http_crawl, is_default=True, subdomain=subdomain ) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build Celery tasks, crafted according to the dependency graph below: # subdomain_discovery --> port_scan --> fetch_url --> dir_file_fuzz # osint vulnerability_scan # osint dalfox xss scan # screenshot # waf_detection workflow = chain( group( subdomain_discovery.si(ctx=ctx, description='Subdomain discovery'), osint.si(ctx=ctx, description='OS Intelligence') ), port_scan.si(ctx=ctx, description='Port scan'), fetch_url.si(ctx=ctx, description='Fetch URL'), group( dir_file_fuzz.si(ctx=ctx, description='Directories & files fuzz'), vulnerability_scan.si(ctx=ctx, description='Vulnerability scan'), screenshot.si(ctx=ctx, description='Screenshot'), waf_detection.si(ctx=ctx, description='WAF detection') ) ) # Build callback callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery chord logger.info(f'Running Celery workflow with {len(workflow.tasks) + 1} tasks') task = chain(workflow, callback).on_error(callback).delay() scan.celery_ids.append(task.id) scan.save() return { 'success': True, 'task_id': task.id } @app.task(name='initiate_subscan', bind=False, queue='subscan_queue') def initiate_subscan( scan_history_id, subdomain_id, engine_id=None, scan_type=None, results_dir=RENGINE_RESULTS, url_filter=''): """Initiate a new subscan. Args: scan_history_id (int): ScanHistory id. subdomain_id (int): Subdomain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. url_filter (str): URL path. Default: '' """ # Get Subdomain, Domain and ScanHistory subdomain = Subdomain.objects.get(pk=subdomain_id) scan = ScanHistory.objects.get(pk=subdomain.scan_history.id) domain = Domain.objects.get(pk=subdomain.target_domain.id) # Get EngineType engine_id = engine_id or scan.scan_type.id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Create scan activity of SubScan Model subscan = SubScan( start_scan_date=timezone.now(), celery_ids=[initiate_subscan.request.id], scan_history=scan, subdomain=subdomain, type=scan_type, status=RUNNING_TASK, engine=engine) subscan.save() # Get YAML configuration config = yaml.safe_load(engine.yaml_configuration) # Create results directory results_dir = f'{scan.results_dir}/subscans/{subscan.id}' os.makedirs(results_dir, exist_ok=True) # Run task method = globals().get(scan_type) if not method: logger.warning(f'Task {scan_type} is not supported by reNgine. Skipping') return scan.tasks.append(scan_type) scan.save() # Send start notif send_scan_notif.delay( scan.id, subscan_id=subscan.id, engine_id=engine_id, status='RUNNING') # Build context ctx = { 'scan_history_id': scan.id, 'subscan_id': subscan.id, 'engine_id': engine_id, 'domain_id': domain.id, 'subdomain_id': subdomain.id, 'yaml_configuration': config, 'results_dir': results_dir, 'url_filter': url_filter } # Create initial endpoints in DB: find domain HTTP endpoint so that HTTP # crawling can start somewhere base_url = f'{subdomain.name}{url_filter}' if url_filter else subdomain.name endpoint, _ = save_endpoint( base_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build header + callback workflow = method.si(ctx=ctx) callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery tasks task = chain(workflow, callback).on_error(callback).delay() subscan.celery_ids.append(task.id) subscan.save() return { 'success': True, 'task_id': task.id } @app.task(name='report', bind=False, queue='report_queue') def report(ctx={}, description=None): """Report task running after all other tasks. Mark ScanHistory or SubScan object as completed and update with final status, log run details and send notification. Args: description (str, optional): Task description shown in UI. """ # Get objects subscan_id = ctx.get('subscan_id') scan_id = ctx.get('scan_history_id') engine_id = ctx.get('engine_id') scan = ScanHistory.objects.filter(pk=scan_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() # Get failed tasks tasks = ScanActivity.objects.filter(scan_of=scan).all() if subscan: tasks = tasks.filter(celery_id__in=subscan.celery_ids) failed_tasks = tasks.filter(status=FAILED_TASK) # Get task status failed_count = failed_tasks.count() status = SUCCESS_TASK if failed_count == 0 else FAILED_TASK status_h = 'SUCCESS' if failed_count == 0 else 'FAILED' # Update scan / subscan status if subscan: subscan.stop_scan_date = timezone.now() subscan.status = status subscan.save() else: scan.scan_status = status scan.stop_scan_date = timezone.now() scan.save() # Send scan status notif send_scan_notif.delay( scan_history_id=scan_id, subscan_id=subscan_id, engine_id=engine_id, status=status_h) #------------------------- # # Tracked reNgine tasks # #--------------------------# @app.task(name='subdomain_discovery', queue='main_scan_queue', base=RengineTask, bind=True) def subdomain_discovery( self, host=None, ctx=None, description=None): """Uses a set of tools (see SUBDOMAIN_SCAN_DEFAULT_TOOLS) to scan all subdomains associated with a domain. Args: host (str): Hostname to scan. Returns: subdomains (list): List of subdomain names. """ if not host: host = self.subdomain.name if self.subdomain else self.domain.name if self.url_filter: logger.warning(f'Ignoring subdomains scan as an URL path filter was passed ({self.url_filter}).') return # Config config = self.yaml_configuration.get(SUBDOMAIN_DISCOVERY) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL) or self.yaml_configuration.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) tools = config.get(USES_TOOLS, SUBDOMAIN_SCAN_DEFAULT_TOOLS) default_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=True).filter(is_subdomain_gathering=True)] custom_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=False).filter(is_subdomain_gathering=True)] send_subdomain_changes, send_interesting = False, False notif = Notification.objects.first() if notif: send_subdomain_changes = notif.send_subdomain_changes_notif send_interesting = notif.send_interesting_notif # Gather tools to run for subdomain scan if ALL in tools: tools = SUBDOMAIN_SCAN_DEFAULT_TOOLS + custom_subdomain_tools tools = [t.lower() for t in tools] # Make exception for amass since tool name is amass, but command is amass-active/passive default_subdomain_tools.append('amass-passive') default_subdomain_tools.append('amass-active') # Run tools for tool in tools: cmd = None logger.info(f'Scanning subdomains for {host} with {tool}') proxy = get_random_proxy() if tool in default_subdomain_tools: if tool == 'amass-passive': cmd = f'amass enum -passive -d {host} -o {self.results_dir}/subdomains_amass.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' elif tool == 'amass-active': use_amass_config = config.get(USE_AMASS_CONFIG, False) amass_wordlist_name = config.get(AMASS_WORDLIST, 'deepmagic.com-prefixes-top50000') wordlist_path = f'/usr/src/wordlist/{amass_wordlist_name}.txt' cmd = f'amass enum -active -d {host} -o {self.results_dir}/subdomains_amass_active.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' cmd += f' -brute -w {wordlist_path}' elif tool == 'sublist3r': cmd = f'python3 /usr/src/github/Sublist3r/sublist3r.py -d {host} -t {threads} -o {self.results_dir}/subdomains_sublister.txt' elif tool == 'subfinder': cmd = f'subfinder -d {host} -o {self.results_dir}/subdomains_subfinder.txt' use_subfinder_config = config.get(USE_SUBFINDER_CONFIG, False) cmd += ' -config /root/.config/subfinder/config.yaml' if use_subfinder_config else '' cmd += f' -proxy {proxy}' if proxy else '' cmd += f' -timeout {timeout}' if timeout else '' cmd += f' -t {threads}' if threads else '' cmd += f' -silent' elif tool == 'oneforall': cmd = f'python3 /usr/src/github/OneForAll/oneforall.py --target {host} run' cmd_extract = f'cut -d\',\' -f6 /usr/src/github/OneForAll/results/{host}.csv > {self.results_dir}/subdomains_oneforall.txt' cmd_rm = f'rm -rf /usr/src/github/OneForAll/results/{host}.csv' cmd += f' && {cmd_extract} && {cmd_rm}' elif tool == 'ctfr': results_file = self.results_dir + '/subdomains_ctfr.txt' cmd = f'python3 /usr/src/github/ctfr/ctfr.py -d {host} -o {results_file}' cmd_extract = f"cat {results_file} | sed 's/\*.//g' | tail -n +12 | uniq | sort > {results_file}" cmd += f' && {cmd_extract}' elif tool == 'tlsx': results_file = self.results_dir + '/subdomains_tlsx.txt' cmd = f'tlsx -san -cn -silent -ro -host {host}' cmd += f" | sed -n '/^\([a-zA-Z0-9]\([-a-zA-Z0-9]*[a-zA-Z0-9]\)\?\.\)\+{host}$/p' | uniq | sort" cmd += f' > {results_file}' elif tool == 'netlas': results_file = self.results_dir + '/subdomains_netlas.txt' cmd = f'netlas search -d domain -i domain domain:"*.{host}" -f json' netlas_key = get_netlas_key() cmd += f' -a {netlas_key}' if netlas_key else '' cmd_extract = f"grep -oE '([a-zA-Z0-9]([-a-zA-Z0-9]*[a-zA-Z0-9])?\.)+{host}'" cmd += f' | {cmd_extract} > {results_file}' elif tool in custom_subdomain_tools: tool_query = InstalledExternalTool.objects.filter(name__icontains=tool.lower()) if not tool_query.exists(): logger.error(f'Missing {{TARGET}} and {{OUTPUT}} placeholders in {tool} configuration. Skipping.') continue custom_tool = tool_query.first() cmd = custom_tool.subdomain_gathering_command if '{TARGET}' in cmd and '{OUTPUT}' in cmd: cmd = cmd.replace('{TARGET}', host) cmd = cmd.replace('{OUTPUT}', f'{self.results_dir}/subdomains_{tool}.txt') cmd = cmd.replace('{PATH}', custom_tool.github_clone_path) if '{PATH}' in cmd else cmd else: logger.warning( f'Subdomain discovery tool "{tool}" is not supported by reNgine. Skipping.') continue # Run tool try: run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) except Exception as e: logger.error( f'Subdomain discovery tool "{tool}" raised an exception') logger.exception(e) # Gather all the tools' results in one single file. Write subdomains into # separate files, and sort all subdomains. run_command( f'cat {self.results_dir}/subdomains_*.txt > {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {self.output_path} -o {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) with open(self.output_path) as f: lines = f.readlines() # Parse the output_file file and store Subdomain and EndPoint objects found # in db. subdomain_count = 0 subdomains = [] urls = [] for line in lines: subdomain_name = line.strip() valid_url = bool(validators.url(subdomain_name)) valid_domain = ( bool(validators.domain(subdomain_name)) or bool(validators.ipv4(subdomain_name)) or bool(validators.ipv6(subdomain_name)) or valid_url ) if not valid_domain: logger.error(f'Subdomain {subdomain_name} is not a valid domain, IP or URL. Skipping.') continue if valid_url: subdomain_name = urlparse(subdomain_name).netloc if subdomain_name in self.out_of_scope_subdomains: logger.error(f'Subdomain {subdomain_name} is out of scope. Skipping.') continue # Add subdomain subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain_count += 1 subdomains.append(subdomain) urls.append(subdomain.name) # Bulk crawl subdomains if enable_http_crawl: ctx['track'] = True http_crawl(urls, ctx=ctx, is_ran_from_subdomain_scan=True) # Find root subdomain endpoints for subdomain in subdomains: pass # Send notifications subdomains_str = '\n'.join([f'• `{subdomain.name}`' for subdomain in subdomains]) self.notify(fields={ 'Subdomain count': len(subdomains), 'Subdomains': subdomains_str, }) if send_subdomain_changes and self.scan_id and self.domain_id: added = get_new_added_subdomain(self.scan_id, self.domain_id) removed = get_removed_subdomain(self.scan_id, self.domain_id) if added: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in added]) self.notify(fields={'Added subdomains': subdomains_str}) if removed: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in removed]) self.notify(fields={'Removed subdomains': subdomains_str}) if send_interesting and self.scan_id and self.domain_id: interesting_subdomains = get_interesting_subdomains(self.scan_id, self.domain_id) if interesting_subdomains: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in interesting_subdomains]) self.notify(fields={'Interesting subdomains': subdomains_str}) return SubdomainSerializer(subdomains, many=True).data @app.task(name='osint', queue='main_scan_queue', base=RengineTask, bind=True) def osint(self, host=None, ctx={}, description=None): """Run Open-Source Intelligence tools on selected domain. Args: host (str): Hostname to scan. Returns: dict: Results from osint discovery and dorking. """ config = self.yaml_configuration.get(OSINT) or OSINT_DEFAULT_CONFIG results = {} grouped_tasks = [] if 'discover' in config: ctx['track'] = False # results = osint_discovery(host=host, ctx=ctx) _task = osint_discovery.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, activity_id=self.activity_id, results_dir=self.results_dir, ctx=ctx ) grouped_tasks.append(_task) if OSINT_DORK in config or OSINT_CUSTOM_DORK in config: _task = dorking.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, results_dir=self.results_dir ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('OSINT Tasks finished...') # with open(self.output_path, 'w') as f: # json.dump(results, f, indent=4) # # return results @app.task(name='osint_discovery', queue='osint_discovery_queue', bind=False) def osint_discovery(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run OSINT discovery. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: dict: osint metadat and theHarvester and h8mail results. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) osint_lookup = config.get(OSINT_DISCOVER, []) osint_intensity = config.get(INTENSITY, 'normal') documents_limit = config.get(OSINT_DOCUMENTS_LIMIT, 50) results = {} meta_info = [] emails = [] creds = [] # Get and save meta info if 'metainfo' in osint_lookup: if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': host, 'domain': host, 'scan_id': scan_history_id, 'documents_limit': documents_limit }) meta_info.append(save_metadata_info(meta_dict)) # TODO: disabled for now # elif osint_intensity == 'deep': # subdomains = Subdomain.objects # if self.scan: # subdomains = subdomains.filter(scan_history=self.scan) # for subdomain in subdomains: # meta_dict = DottedDict({ # 'osint_target': subdomain.name, # 'domain': self.domain, # 'scan_id': self.scan_id, # 'documents_limit': documents_limit # }) # meta_info.append(save_metadata_info(meta_dict)) grouped_tasks = [] if 'emails' in osint_lookup: emails = get_and_save_emails(scan_history, activity_id, results_dir) emails_str = '\n'.join([f'• `{email}`' for email in emails]) # self.notify(fields={'Emails': emails_str}) # ctx['track'] = False _task = h8mail.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) if 'employees' in osint_lookup: ctx['track'] = False _task = theHarvester.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) # results['emails'] = results.get('emails', []) + emails # results['creds'] = creds # results['meta_info'] = meta_info return results @app.task(name='dorking', bind=False, queue='dorking_queue') def dorking(config, host, scan_history_id, results_dir): """Run Google dorks. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: list: Dorking results for each dork ran. """ # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh scan_history = ScanHistory.objects.get(pk=scan_history_id) dorks = config.get(OSINT_DORK, []) custom_dorks = config.get(OSINT_CUSTOM_DORK, []) results = [] # custom dorking has higher priority try: for custom_dork in custom_dorks: lookup_target = custom_dork.get('lookup_site') # replace with original host if _target_ lookup_target = host if lookup_target == '_target_' else lookup_target if 'lookup_extensions' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_extensions=custom_dork.get('lookup_extensions'), scan_history=scan_history ) elif 'lookup_keywords' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_keywords=custom_dork.get('lookup_keywords'), scan_history=scan_history ) except Exception as e: logger.exception(e) # default dorking try: for dork in dorks: logger.info(f'Getting dork information for {dork}') if dork == 'stackoverflow': results = get_and_save_dork_results( lookup_target='stackoverflow.com', results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'login_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/login/,login.html', page_count=5, scan_history=scan_history ) elif dork == 'admin_panels': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/admin/,admin.html', page_count=5, scan_history=scan_history ) elif dork == 'dashboard_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/dashboard/,dashboard.html', page_count=5, scan_history=scan_history ) elif dork == 'social_media' : social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'reddit.com' ] for site in social_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'project_management' : project_websites = [ 'trello.com', 'atlassian.net' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'code_sharing' : project_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'config_files' : config_file_exts = [ 'env', 'xml', 'conf', 'toml', 'yml', 'yaml', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(config_file_exts), page_count=4, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keyword = 'Jenkins' results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=lookup_keyword, page_count=1, scan_history=scan_history ) elif dork == 'wordpress_files' : lookup_keywords = [ '/wp-content/', '/wp-includes/' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'php_error' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'exposed_documents' : docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(docs_file_ext), page_count=7, scan_history=scan_history ) elif dork == 'db_files' : file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) elif dork == 'git_exposed' : file_ext = [ 'git', ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) except Exception as e: logger.exception(e) return results @app.task(name='theHarvester', queue='theHarvester_queue', bind=False) def theHarvester(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run theHarvester to get save emails, hosts, employees found in domain. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: dict: Dict of emails, employees, hosts and ips found during crawling. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) output_path_json = f'{results_dir}/theHarvester.json' theHarvester_dir = '/usr/src/github/theHarvester' history_file = f'{results_dir}/commands.txt' cmd = f'python3 {theHarvester_dir}/theHarvester.py -d {host} -b all -f {output_path_json}' # Update proxies.yaml proxy_query = Proxy.objects.all() if proxy_query.exists(): proxy = proxy_query.first() if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(f'{theHarvester_dir}/proxies.yaml', 'w') as file: yaml.dump(yaml_data, file) # Run cmd run_command( cmd, shell=False, cwd=theHarvester_dir, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) # Get file location if not os.path.isfile(output_path_json): logger.error(f'Could not open {output_path_json}') return {} # Load theHarvester results with open(output_path_json, 'r') as f: data = json.load(f) # Re-indent theHarvester JSON with open(output_path_json, 'w') as f: json.dump(data, f, indent=4) emails = data.get('emails', []) for email_address in emails: email, _ = save_email(email_address, scan_history=scan_history) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) linkedin_people = data.get('linkedin_people', []) for people in linkedin_people: employee, _ = save_employee( people, designation='linkedin', scan_history=scan_history) # if employee: # self.notify(fields={'LinkedIn people': f'• {employee.name}'}) twitter_people = data.get('twitter_people', []) for people in twitter_people: employee, _ = save_employee( people, designation='twitter', scan_history=scan_history) # if employee: # self.notify(fields={'Twitter people': f'• {employee.name}'}) hosts = data.get('hosts', []) urls = [] for host in hosts: split = tuple(host.split(':')) http_url = split[0] subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) endpoint, _ = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain) # if endpoint: # urls.append(endpoint.http_url) # self.notify(fields={'Hosts': f'• {endpoint.http_url}'}) # if enable_http_crawl: # ctx['track'] = False # http_crawl(urls, ctx=ctx) # TODO: Lots of ips unrelated with our domain are found, disabling # this for now. # ips = data.get('ips', []) # for ip_address in ips: # ip, created = save_ip_address( # ip_address, # subscan=subscan) # if ip: # send_task_notif.delay( # 'osint', # scan_history_id=scan_history_id, # subscan_id=subscan_id, # severity='success', # update_fields={'IPs': f'{ip.address}'}) return data @app.task(name='h8mail', queue='h8mail_queue', bind=False) def h8mail(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run h8mail. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: list[dict]: List of credentials info. """ logger.warning('Getting leaked credentials') scan_history = ScanHistory.objects.get(pk=scan_history_id) input_path = f'{results_dir}/emails.txt' output_file = f'{results_dir}/h8mail.json' cmd = f'h8mail -t {input_path} --json {output_file}' history_file = f'{results_dir}/commands.txt' run_command( cmd, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) with open(output_file) as f: data = json.load(f) creds = data.get('targets', []) # TODO: go through h8mail output and save emails to DB for cred in creds: logger.warning(cred) email_address = cred['target'] pwn_num = cred['pwn_num'] pwn_data = cred.get('data', []) email, created = save_email(email_address, scan_history=scan) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) return creds @app.task(name='screenshot', queue='main_scan_queue', base=RengineTask, bind=True) def screenshot(self, ctx={}, description=None): """Uses EyeWitness to gather screenshot of a domain and/or url. Args: description (str, optional): Task description shown in UI. """ # Config screenshots_path = f'{self.results_dir}/screenshots' output_path = f'{self.results_dir}/screenshots/{self.filename}' alive_endpoints_file = f'{self.results_dir}/endpoints_alive.txt' config = self.yaml_configuration.get(SCREENSHOT) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT + 5) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) # If intensity is normal, grab only the root endpoints of each subdomain strict = True if intensity == 'normal' else False # Get URLs to take screenshot of get_http_urls( is_alive=enable_http_crawl, strict=strict, write_filepath=alive_endpoints_file, get_only_default_urls=True, ctx=ctx ) # Send start notif notification = Notification.objects.first() send_output_file = notification.send_scan_output_file if notification else False # Run cmd cmd = f'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py -f {alive_endpoints_file} -d {screenshots_path} --no-prompt' cmd += f' --timeout {timeout}' if timeout > 0 else '' cmd += f' --threads {threads}' if threads > 0 else '' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(output_path): logger.error(f'Could not load EyeWitness results at {output_path} for {self.domain.name}.') return # Loop through results and save objects in DB screenshot_paths = [] with open(output_path, 'r') as file: reader = csv.reader(file) for row in reader: "Protocol,Port,Domain,Request Status,Screenshot Path, Source Path" protocol, port, subdomain_name, status, screenshot_path, source_path = tuple(row) logger.info(f'{protocol}:{port}:{subdomain_name}:{status}') subdomain_query = Subdomain.objects.filter(name=subdomain_name) if self.scan: subdomain_query = subdomain_query.filter(scan_history=self.scan) if status == 'Successful' and subdomain_query.exists(): subdomain = subdomain_query.first() screenshot_paths.append(screenshot_path) subdomain.screenshot_path = screenshot_path.replace('/usr/src/scan_results/', '') subdomain.save() logger.warning(f'Added screenshot for {subdomain.name} to DB') # Remove all db, html extra files in screenshot results run_command( 'rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format(screenshots_path), shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'rm -rf {screenshots_path}/source', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Send finish notifs screenshots_str = '• ' + '\n• '.join([f'`{path}`' for path in screenshot_paths]) self.notify(fields={'Screenshots': screenshots_str}) if send_output_file: for path in screenshot_paths: title = get_output_file_name( self.scan_id, self.subscan_id, self.filename) send_file_to_discord.delay(path, title) @app.task(name='port_scan', queue='main_scan_queue', base=RengineTask, bind=True) def port_scan(self, hosts=[], ctx={}, description=None): """Run port scan. Args: hosts (list, optional): Hosts to run port scan on. description (str, optional): Task description shown in UI. Returns: list: List of open ports (dict). """ input_file = f'{self.results_dir}/input_subdomains_port_scan.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(PORT_SCAN) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) exclude_ports = config.get(NAABU_EXCLUDE_PORTS, []) exclude_subdomains = config.get(NAABU_EXCLUDE_SUBDOMAINS, False) ports = config.get(PORTS, NAABU_DEFAULT_PORTS) ports = [str(port) for port in ports] rate_limit = config.get(NAABU_RATE) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) passive = config.get(NAABU_PASSIVE, False) use_naabu_config = config.get(USE_NAABU_CONFIG, False) exclude_ports_str = ','.join(return_iterable(exclude_ports)) # nmap args nmap_enabled = config.get(ENABLE_NMAP, False) nmap_cmd = config.get(NMAP_COMMAND, '') nmap_script = config.get(NMAP_SCRIPT, '') nmap_script = ','.join(return_iterable(nmap_script)) nmap_script_args = config.get(NMAP_SCRIPT_ARGS) if hosts: with open(input_file, 'w') as f: f.write('\n'.join(hosts)) else: hosts = get_subdomains( write_filepath=input_file, exclude_subdomains=exclude_subdomains, ctx=ctx) # Build cmd cmd = 'naabu -json -exclude-cdn' cmd += f' -list {input_file}' if len(hosts) > 0 else f' -host {hosts[0]}' if 'full' in ports or 'all' in ports: ports_str = ' -p "-"' elif 'top-100' in ports: ports_str = ' -top-ports 100' elif 'top-1000' in ports: ports_str = ' -top-ports 1000' else: ports_str = ','.join(ports) ports_str = f' -p {ports_str}' cmd += ports_str cmd += ' -config /root/.config/naabu/config.yaml' if use_naabu_config else '' cmd += f' -proxy "{proxy}"' if proxy else '' cmd += f' -c {threads}' if threads else '' cmd += f' -rate {rate_limit}' if rate_limit > 0 else '' cmd += f' -timeout {timeout*1000}' if timeout > 0 else '' cmd += f' -passive' if passive else '' cmd += f' -exclude-ports {exclude_ports_str}' if exclude_ports else '' cmd += f' -silent' # Execute cmd and gather results results = [] urls = [] ports_data = {} for line in stream_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) port_number = line['port'] ip_address = line['ip'] host = line.get('host') or ip_address if port_number == 0: continue # Grab subdomain subdomain = Subdomain.objects.filter( name=host, target_domain=self.domain, scan_history=self.scan ).first() # Add IP DB ip, _ = save_ip_address(ip_address, subdomain, subscan=self.subscan) if self.subscan: ip.ip_subscan_ids.add(self.subscan) ip.save() # Add endpoint to DB # port 80 and 443 not needed as http crawl already does that. if port_number not in [80, 443]: http_url = f'{host}:{port_number}' endpoint, _ = save_endpoint( http_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint: http_url = endpoint.http_url urls.append(http_url) # Add Port in DB port_details = whatportis.get_ports(str(port_number)) service_name = port_details[0].name if len(port_details) > 0 else 'unknown' description = port_details[0].description if len(port_details) > 0 else '' # get or create port port, created = Port.objects.get_or_create( number=port_number, service_name=service_name, description=description ) if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port.save() ip.ports.add(port) ip.save() if host in ports_data: ports_data[host].append(port_number) else: ports_data[host] = [port_number] # Send notification logger.warning(f'Found opened port {port_number} on {ip_address} ({host})') if len(ports_data) == 0: logger.info('Finished running naabu port scan - No open ports found.') if nmap_enabled: logger.info('Nmap scans skipped') return ports_data # Send notification fields_str = '' for host, ports in ports_data.items(): ports_str = ', '.join([f'`{port}`' for port in ports]) fields_str += f'• `{host}`: {ports_str}\n' self.notify(fields={'Ports discovered': fields_str}) # Save output to file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) logger.info('Finished running naabu port scan.') # Process nmap results: 1 process per host sigs = [] if nmap_enabled: logger.warning(f'Starting nmap scans ...') logger.warning(ports_data) for host, port_list in ports_data.items(): ports_str = '_'.join([str(p) for p in port_list]) ctx_nmap = ctx.copy() ctx_nmap['description'] = get_task_title(f'nmap_{host}', self.scan_id, self.subscan_id) ctx_nmap['track'] = False sig = nmap.si( cmd=nmap_cmd, ports=port_list, host=host, script=nmap_script, script_args=nmap_script_args, max_rate=rate_limit, ctx=ctx_nmap) sigs.append(sig) task = group(sigs).apply_async() with allow_join_result(): results = task.get() return ports_data @app.task(name='nmap', queue='main_scan_queue', base=RengineTask, bind=True) def nmap( self, cmd=None, ports=[], host=None, input_file=None, script=None, script_args=None, max_rate=None, ctx={}, description=None): """Run nmap on a host. Args: cmd (str, optional): Existing nmap command to complete. ports (list, optional): List of ports to scan. host (str, optional): Host to scan. input_file (str, optional): Input hosts file. script (str, optional): NSE script to run. script_args (str, optional): NSE script args. max_rate (int): Max rate. description (str, optional): Task description shown in UI. """ notif = Notification.objects.first() ports_str = ','.join(str(port) for port in ports) self.filename = self.filename.replace('.txt', '.xml') filename_vulns = self.filename.replace('.xml', '_vulns.json') output_file = self.output_path output_file_xml = f'{self.results_dir}/{host}_{self.filename}' vulns_file = f'{self.results_dir}/{host}_{filename_vulns}' logger.warning(f'Running nmap on {host}:{ports}') # Build cmd nmap_cmd = get_nmap_cmd( cmd=cmd, ports=ports_str, script=script, script_args=script_args, max_rate=max_rate, host=host, input_file=input_file, output_file=output_file_xml) # Run cmd run_command( nmap_cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Get nmap XML results and convert to JSON vulns = parse_nmap_results(output_file_xml, output_file) with open(vulns_file, 'w') as f: json.dump(vulns, f, indent=4) # Save vulnerabilities found by nmap vulns_str = '' for vuln_data in vulns: # URL is not necessarily an HTTP URL when running nmap (can be any # other vulnerable protocols). Look for existing endpoint and use its # URL as vulnerability.http_url if it exists. url = vuln_data['http_url'] endpoint = EndPoint.objects.filter(http_url__contains=url).first() if endpoint: vuln_data['http_url'] = endpoint.http_url vuln, created = save_vulnerability( target_domain=self.domain, subdomain=self.subdomain, scan_history=self.scan, subscan=self.subscan, endpoint=endpoint, **vuln_data) vulns_str += f'• {str(vuln)}\n' if created: logger.warning(str(vuln)) # Send only 1 notif for all vulns to reduce number of notifs if notif and notif.send_vuln_notif and vulns_str: logger.warning(vulns_str) self.notify(fields={'CVEs': vulns_str}) return vulns @app.task(name='waf_detection', queue='main_scan_queue', base=RengineTask, bind=True) def waf_detection(self, ctx={}, description=None): """ Uses wafw00f to check for the presence of a WAF. Args: description (str, optional): Task description shown in UI. Returns: list: List of startScan.models.Waf objects. """ input_path = f'{self.results_dir}/input_endpoints_waf_detection.txt' config = self.yaml_configuration.get(WAF_DETECTION) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Get alive endpoints from DB get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) cmd = f'wafw00f -i {input_path} -o {self.output_path}' run_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(self.output_path): logger.error(f'Could not find {self.output_path}') return with open(self.output_path) as file: wafs = file.readlines() for line in wafs: line = " ".join(line.split()) splitted = line.split(' ', 1) waf_info = splitted[1].strip() waf_name = waf_info[:waf_info.find('(')].strip() waf_manufacturer = waf_info[waf_info.find('(')+1:waf_info.find(')')].strip().replace('.', '') http_url = sanitize_url(splitted[0].strip()) if not waf_name or waf_name == 'None': continue # Add waf to db waf, _ = Waf.objects.get_or_create( name=waf_name, manufacturer=waf_manufacturer ) # Add waf info to Subdomain in DB subdomain = get_subdomain_from_url(http_url) logger.info(f'Wafw00f Subdomain : {subdomain}') subdomain_query, _ = Subdomain.objects.get_or_create(scan_history=self.scan, name=subdomain) subdomain_query.waf.add(waf) subdomain_query.save() return wafs @app.task(name='dir_file_fuzz', queue='main_scan_queue', base=RengineTask, bind=True) def dir_file_fuzz(self, ctx={}, description=None): """Perform directory scan, and currently uses `ffuf` as a default tool. Args: description (str, optional): Task description shown in UI. Returns: list: List of URLs discovered. """ # Config cmd = 'ffuf' config = self.yaml_configuration.get(DIR_FILE_FUZZ) or {} custom_header = self.yaml_configuration.get(CUSTOM_HEADER) auto_calibration = config.get(AUTO_CALIBRATION, True) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) extensions = config.get(EXTENSIONS, DEFAULT_DIR_FILE_FUZZ_EXTENSIONS) # prepend . on extensions extensions = [ext if ext.startswith('.') else '.' + ext for ext in extensions] extensions_str = ','.join(map(str, extensions)) follow_redirect = config.get(FOLLOW_REDIRECT, FFUF_DEFAULT_FOLLOW_REDIRECT) max_time = config.get(MAX_TIME, 0) match_http_status = config.get(MATCH_HTTP_STATUS, FFUF_DEFAULT_MATCH_HTTP_STATUS) mc = ','.join([str(c) for c in match_http_status]) recursive_level = config.get(RECURSIVE_LEVEL, FFUF_DEFAULT_RECURSIVE_LEVEL) stop_on_error = config.get(STOP_ON_ERROR, False) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) wordlist_name = config.get(WORDLIST, 'dicc') delay = rate_limit / (threads * 100) # calculate request pause delay from rate_limit and number of threads input_path = f'{self.results_dir}/input_dir_file_fuzz.txt' # Get wordlist wordlist_name = 'dicc' if wordlist_name == 'default' else wordlist_name wordlist_path = f'/usr/src/wordlist/{wordlist_name}.txt' # Build command cmd += f' -w {wordlist_path}' cmd += f' -e {extensions_str}' if extensions else '' cmd += f' -maxtime {max_time}' if max_time > 0 else '' cmd += f' -p {delay}' if delay > 0 else '' cmd += f' -recursion -recursion-depth {recursive_level} ' if recursive_level > 0 else '' cmd += f' -t {threads}' if threads and threads > 0 else '' cmd += f' -timeout {timeout}' if timeout and timeout > 0 else '' cmd += ' -se' if stop_on_error else '' cmd += ' -fr' if follow_redirect else '' cmd += ' -ac' if auto_calibration else '' cmd += f' -mc {mc}' if mc else '' cmd += f' -H "{custom_header}"' if custom_header else '' # Grab URLs to fuzz urls = get_http_urls( is_alive=True, ignore_files=False, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) logger.warning(urls) # Loop through URLs and run command results = [] for url in urls: ''' Above while fetching urls, we are not ignoring files, because some default urls may redirect to https://example.com/login.php so, ignore_files is set to False but, during fuzzing, we will only need part of the path, in above example it is still a good idea to ffuf base url https://example.com so files from base url ''' url_parse = urlparse(url) url = url_parse.scheme + '://' + url_parse.netloc url += '/FUZZ' # TODO: fuzz not only URL but also POST / PUT / headers proxy = get_random_proxy() # Build final cmd fcmd = cmd fcmd += f' -x {proxy}' if proxy else '' fcmd += f' -u {url} -json' # Initialize DirectoryScan object dirscan = DirectoryScan() dirscan.scanned_date = timezone.now() dirscan.command_line = fcmd dirscan.save() # Loop through results and populate EndPoint and DirectoryFile in DB results = [] for line in stream_command( fcmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) name = line['input'].get('FUZZ') length = line['length'] status = line['status'] words = line['words'] url = line['url'] lines = line['lines'] content_type = line['content-type'] duration = line['duration'] if not name: logger.error(f'FUZZ not found for "{url}"') continue endpoint, created = save_endpoint(url, crawl=False, ctx=ctx) # endpoint.is_default = False endpoint.http_status = status endpoint.content_length = length endpoint.response_time = duration / 1000000000 endpoint.save() if created: urls.append(endpoint.http_url) endpoint.status = status endpoint.content_type = content_type endpoint.content_length = length dfile, created = DirectoryFile.objects.get_or_create( name=name, length=length, words=words, lines=lines, content_type=content_type, url=url) dfile.http_status = status dfile.save() # if created: # logger.warning(f'Found new directory or file {url}') dirscan.directory_files.add(dfile) dirscan.save() if self.subscan: dirscan.dir_subscan_ids.add(self.subscan) subdomain_name = get_subdomain_from_url(endpoint.http_url) subdomain = Subdomain.objects.get(name=subdomain_name, scan_history=self.scan) subdomain.directories.add(dirscan) subdomain.save() # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl(urls, ctx=ctx) return results @app.task(name='fetch_url', queue='main_scan_queue', base=RengineTask, bind=True) def fetch_url(self, urls=[], ctx={}, description=None): """Fetch URLs using different tools like gauplus, gau, gospider, waybackurls ... Args: urls (list): List of URLs to start from. description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/input_endpoints_fetch_url.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(FETCH_URL) or {} should_remove_duplicate_endpoints = config.get(REMOVE_DUPLICATE_ENDPOINTS, True) duplicate_removal_fields = config.get(DUPLICATE_REMOVAL_FIELDS, ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, DEFAULT_GF_PATTERNS) ignore_file_extension = config.get(IGNORE_FILE_EXTENSION, DEFAULT_IGNORE_FILE_EXTENSIONS) tools = config.get(USES_TOOLS, ENDPOINT_SCAN_DEFAULT_TOOLS) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) domain_request_headers = self.domain.request_headers if self.domain else None custom_header = domain_request_headers or self.yaml_configuration.get(CUSTOM_HEADER) exclude_subdomains = config.get(EXCLUDED_SUBDOMAINS, False) # Get URLs to scan and save to input file if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, exclude_subdomains=exclude_subdomains, get_only_default_urls=True, ctx=ctx ) # Domain regex host = self.domain.name if self.domain else urlparse(urls[0]).netloc host_regex = f"\'https?://([a-z0-9]+[.])*{host}.*\'" # Tools cmds cmd_map = { 'gau': f'gau', 'gauplus': f'gauplus -random-agent', 'hakrawler': 'hakrawler -subs -u', 'waybackurls': 'waybackurls', 'gospider': f'gospider -S {input_path} --js -d 2 --sitemap --robots -w -r', 'katana': f'katana -list {input_path} -silent -jc -kf all -d 3 -fs rdn', } if proxy: cmd_map['gau'] += f' --proxy "{proxy}"' cmd_map['gauplus'] += f' -p "{proxy}"' cmd_map['gospider'] += f' -p {proxy}' cmd_map['hakrawler'] += f' -proxy {proxy}' cmd_map['katana'] += f' -proxy {proxy}' if threads > 0: cmd_map['gau'] += f' --threads {threads}' cmd_map['gauplus'] += f' -t {threads}' cmd_map['gospider'] += f' -t {threads}' cmd_map['katana'] += f' -c {threads}' if custom_header: header_string = ';;'.join([ f'{key}: {value}' for key, value in custom_header.items() ]) cmd_map['hakrawler'] += f' -h {header_string}' cmd_map['katana'] += f' -H {header_string}' header_flags = [':'.join(h) for h in header_string.split(';;')] for flag in header_flags: cmd_map['gospider'] += f' -H {flag}' cat_input = f'cat {input_path}' grep_output = f'grep -Eo {host_regex}' cmd_map = { tool: f'{cat_input} | {cmd} | {grep_output} > {self.results_dir}/urls_{tool}.txt' for tool, cmd in cmd_map.items() } tasks = group( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for tool, cmd in cmd_map.items() if tool in tools ) # Cleanup task sort_output = [ f'cat {self.results_dir}/urls_* > {self.output_path}', f'cat {input_path} >> {self.output_path}', f'sort -u {self.output_path} -o {self.output_path}', ] if ignore_file_extension: ignore_exts = '|'.join(ignore_file_extension) grep_ext_filtered_output = [ f'cat {self.output_path} | grep -Eiv "\\.({ignore_exts}).*" > {self.results_dir}/urls_filtered.txt', f'mv {self.results_dir}/urls_filtered.txt {self.output_path}' ] sort_output.extend(grep_ext_filtered_output) cleanup = chain( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for cmd in sort_output ) # Run all commands task = chord(tasks)(cleanup) with allow_join_result(): task.get() # Store all the endpoints and run httpx with open(self.output_path) as f: discovered_urls = f.readlines() self.notify(fields={'Discovered URLs': len(discovered_urls)}) # Some tools can have an URL in the format <URL>] - <PATH> or <URL> - <PATH>, add them # to the final URL list all_urls = [] for url in discovered_urls: url = url.strip() urlpath = None base_url = None if '] ' in url: # found JS scraped endpoint e.g from gospider split = tuple(url.split('] ')) if not len(split) == 2: logger.warning(f'URL format not recognized for "{url}". Skipping.') continue base_url, urlpath = split urlpath = urlpath.lstrip('- ') elif ' - ' in url: # found JS scraped endpoint e.g from gospider base_url, urlpath = tuple(url.split(' - ')) if base_url and urlpath: subdomain = urlparse(base_url) url = f'{subdomain.scheme}://{subdomain.netloc}{self.url_filter}' if not validators.url(url): logger.warning(f'Invalid URL "{url}". Skipping.') if url not in all_urls: all_urls.append(url) # Filter out URLs if a path filter was passed if self.url_filter: all_urls = [url for url in all_urls if self.url_filter in url] # Write result to output path with open(self.output_path, 'w') as f: f.write('\n'.join(all_urls)) logger.warning(f'Found {len(all_urls)} usable URLs') # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl( all_urls, ctx=ctx, should_remove_duplicate_endpoints=should_remove_duplicate_endpoints, duplicate_removal_fields=duplicate_removal_fields ) #-------------------# # GF PATTERNS MATCH # #-------------------# # Combine old gf patterns with new ones if gf_patterns: self.scan.used_gf_patterns = ','.join(gf_patterns) self.scan.save() # Run gf patterns on saved endpoints # TODO: refactor to Celery task for gf_pattern in gf_patterns: # TODO: js var is causing issues, removing for now if gf_pattern == 'jsvar': logger.info('Ignoring jsvar as it is causing issues.') continue # Run gf on current pattern logger.warning(f'Running gf on pattern "{gf_pattern}"') gf_output_file = f'{self.results_dir}/gf_patterns_{gf_pattern}.txt' cmd = f'cat {self.output_path} | gf {gf_pattern} | grep -Eo {host_regex} >> {gf_output_file}' run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Check output file if not os.path.exists(gf_output_file): logger.error(f'Could not find GF output file {gf_output_file}. Skipping GF pattern "{gf_pattern}"') continue # Read output file line by line and with open(gf_output_file, 'r') as f: lines = f.readlines() # Add endpoints / subdomains to DB for url in lines: http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue endpoint, created = save_endpoint( http_url, crawl=False, subdomain=subdomain, ctx=ctx) if not endpoint: continue earlier_pattern = None if not created: earlier_pattern = endpoint.matched_gf_patterns pattern = f'{earlier_pattern},{gf_pattern}' if earlier_pattern else gf_pattern endpoint.matched_gf_patterns = pattern endpoint.save() return all_urls def parse_curl_output(response): # TODO: Enrich from other cURL fields. CURL_REGEX_HTTP_STATUS = f'HTTP\/(?:(?:\d\.?)+)\s(\d+)\s(?:\w+)' http_status = 0 if response: failed = False regex = re.compile(CURL_REGEX_HTTP_STATUS, re.MULTILINE) try: http_status = int(regex.findall(response)[0]) except (KeyError, TypeError, IndexError): pass return { 'http_status': http_status, } @app.task(name='vulnerability_scan', queue='main_scan_queue', bind=True, base=RengineTask) def vulnerability_scan(self, urls=[], ctx={}, description=None): """ This function will serve as an entrypoint to vulnerability scan. All other vulnerability scan will be run from here including nuclei, crlfuzz, etc """ logger.info('Running Vulnerability Scan Queue') config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_run_nuclei = config.get(RUN_NUCLEI, True) should_run_crlfuzz = config.get(RUN_CRLFUZZ, False) should_run_dalfox = config.get(RUN_DALFOX, False) should_run_s3scanner = config.get(RUN_S3SCANNER, True) grouped_tasks = [] if should_run_nuclei: _task = nuclei_scan.si( urls=urls, ctx=ctx, description=f'Nuclei Scan' ) grouped_tasks.append(_task) if should_run_crlfuzz: _task = crlfuzz_scan.si( urls=urls, ctx=ctx, description=f'CRLFuzz Scan' ) grouped_tasks.append(_task) if should_run_dalfox: _task = dalfox_xss_scan.si( urls=urls, ctx=ctx, description=f'Dalfox XSS Scan' ) grouped_tasks.append(_task) if should_run_s3scanner: _task = s3scanner.si( ctx=ctx, description=f'Misconfigured S3 Buckets Scanner' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan completed...') # return results return None @app.task(name='nuclei_individual_severity_module', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_individual_severity_module(self, cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx={}, description=None): ''' This celery task will run vulnerability scan in parallel. All severities supplied should run in parallel as grouped tasks. ''' results = [] logger.info(f'Running vulnerability scan with severity: {severity}') cmd += f' -severity {severity}' # Send start notification notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) # Gather nuclei results vuln_data = parse_nuclei_result(line) # Get corresponding subdomain http_url = sanitize_url(line.get('matched-at')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) # Look for duplicate vulnerabilities by excluding records that might change but are irrelevant. object_comparison_exclude = ['response', 'curl_command', 'tags', 'references', 'cve_ids', 'cwe_ids'] # Add subdomain and target domain to the duplicate check vuln_data_copy = vuln_data.copy() vuln_data_copy['subdomain'] = subdomain vuln_data_copy['target_domain'] = self.domain # Check if record exists, if exists do not save it if record_exists(Vulnerability, data=vuln_data_copy, exclude_keys=object_comparison_exclude): logger.warning(f'Nuclei vulnerability of severity {severity} : {vuln_data_copy["name"]} for {subdomain_name} already exists') continue # Get or create EndPoint object response = line.get('response') httpx_crawl = False if response else enable_http_crawl # avoid yet another httpx crawl endpoint, _ = save_endpoint( http_url, crawl=httpx_crawl, subdomain=subdomain, ctx=ctx) if endpoint: http_url = endpoint.http_url if not httpx_crawl: output = parse_curl_output(response) endpoint.http_status = output['http_status'] endpoint.save() # Get or create Vulnerability object vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, subdomain=subdomain, **vuln_data) if not vuln: continue # Print vuln severity = line['info'].get('severity', 'unknown') logger.warning(str(vuln)) # Send notification for all vulnerabilities except info url = vuln.http_url or vuln.subdomain send_vuln = ( notif and notif.send_vuln_notif and vuln and severity in ['low', 'medium', 'high', 'critical']) if send_vuln: fields = { 'Severity': f'**{severity.upper()}**', 'URL': http_url, 'Subdomain': subdomain_name, 'Name': vuln.name, 'Type': vuln.type, 'Description': vuln.description, 'Template': vuln.template_url, 'Tags': vuln.get_tags_str(), 'CVEs': vuln.get_cve_str(), 'CWEs': vuln.get_cwe_str(), 'References': vuln.get_refs_str() } severity_map = { 'low': 'info', 'medium': 'warning', 'high': 'error', 'critical': 'error' } self.notify( f'vulnerability_scan_#{vuln.id}', severity_map[severity], fields, add_meta_info=False) # Send report to hackerone hackerone_query = Hackerone.objects.all() send_report = ( hackerone_query.exists() and severity not in ('info', 'low') and vuln.target_domain.h1_team_handle ) if send_report: hackerone = hackerone_query.first() if hackerone.send_critical and severity == 'critical': send_hackerone_report.delay(vuln.id) elif hackerone.send_high and severity == 'high': send_hackerone_report.delay(vuln.id) elif hackerone.send_medium and severity == 'medium': send_hackerone_report.delay(vuln.id) # Write results to JSON file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) # Send finish notif if send_status: vulns = Vulnerability.objects.filter(scan_history__id=self.scan_id) info_count = vulns.filter(severity=0).count() low_count = vulns.filter(severity=1).count() medium_count = vulns.filter(severity=2).count() high_count = vulns.filter(severity=3).count() critical_count = vulns.filter(severity=4).count() unknown_count = vulns.filter(severity=-1).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count + unknown_count fields = { 'Total': vulnerability_count, 'Critical': critical_count, 'High': high_count, 'Medium': medium_count, 'Low': low_count, 'Info': info_count, 'Unknown': unknown_count } self.notify(fields=fields) # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=NUCLEI ).exclude( severity=0 ) # find all unique vulnerabilities based on path and title # all unique vulnerability will go thru gpt function and get report # once report is got, it will be matched with other vulnerabilities and saved unique_vulns = set() for vuln in vulns: unique_vulns.add((vuln.name, vuln.get_path())) unique_vulns = list(unique_vulns) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in unique_vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return None def get_vulnerability_gpt_report(vuln): title = vuln[0] path = vuln[1] logger.info(f'Getting GPT Report for {title}, PATH: {path}') # check if in db already exists stored = GPTVulnerabilityReport.objects.filter( url_path=path ).filter( title=title ).first() if stored: response = { 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: report = GPTVulnerabilityReportGenerator() vulnerability_description = get_gpt_vuln_input_description( title, path ) response = report.get_vulnerability_description(vulnerability_description) add_gpt_description_db( title, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) for vuln in Vulnerability.objects.filter(name=title, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() def add_gpt_description_db(title, path, description, impact, remediation, references): gpt_report = GPTVulnerabilityReport() gpt_report.url_path = path gpt_report.title = title gpt_report.description = description gpt_report.impact = impact gpt_report.remediation = remediation gpt_report.save() for url in references: ref, created = VulnerabilityReference.objects.get_or_create(url=url) gpt_report.references.add(ref) gpt_report.save() @app.task(name='nuclei_scan', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_scan(self, urls=[], ctx={}, description=None): """HTTP vulnerability scan using Nuclei Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. Notes: Unfurl the urls to keep only domain and path, will be sent to vuln scan and ignore certain file extensions. Thanks: https://github.com/six2dez/reconftw """ # Config config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} input_path = f'{self.results_dir}/input_endpoints_vulnerability_scan.txt' enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) concurrency = config.get(NUCLEI_CONCURRENCY) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) retries = config.get(RETRIES) or self.yaml_configuration.get(RETRIES, DEFAULT_RETRIES) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) custom_header = config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) should_fetch_gpt_report = config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) proxy = get_random_proxy() nuclei_specific_config = config.get('nuclei', {}) use_nuclei_conf = nuclei_specific_config.get(USE_CONFIG, False) severities = nuclei_specific_config.get(NUCLEI_SEVERITY, NUCLEI_DEFAULT_SEVERITIES) tags = nuclei_specific_config.get(NUCLEI_TAGS, []) tags = ','.join(tags) nuclei_templates = nuclei_specific_config.get(NUCLEI_TEMPLATE) custom_nuclei_templates = nuclei_specific_config.get(NUCLEI_CUSTOM_TEMPLATE) # severities_str = ','.join(severities) # Get alive endpoints if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=enable_http_crawl, ignore_files=True, write_filepath=input_path, ctx=ctx ) if intensity == 'normal': # reduce number of endpoints to scan unfurl_filter = f'{self.results_dir}/urls_unfurled.txt' run_command( f"cat {input_path} | unfurl -u format %s://%d%p |uro > {unfurl_filter}", shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {unfurl_filter} -o {unfurl_filter}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) input_path = unfurl_filter # Build templates # logger.info('Updating Nuclei templates ...') run_command( 'nuclei -update-templates', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) templates = [] if not (nuclei_templates or custom_nuclei_templates): templates.append(NUCLEI_DEFAULT_TEMPLATES_PATH) if nuclei_templates: if ALL in nuclei_templates: template = NUCLEI_DEFAULT_TEMPLATES_PATH templates.append(template) else: templates.extend(nuclei_templates) if custom_nuclei_templates: custom_nuclei_template_paths = [f'{str(elem)}.yaml' for elem in custom_nuclei_templates] template = templates.extend(custom_nuclei_template_paths) # Build CMD cmd = 'nuclei -j' cmd += ' -config /root/.config/nuclei/config.yaml' if use_nuclei_conf else '' cmd += f' -irr' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -l {input_path}' cmd += f' -c {str(concurrency)}' if concurrency > 0 else '' cmd += f' -proxy {proxy} ' if proxy else '' cmd += f' -retries {retries}' if retries > 0 else '' cmd += f' -rl {rate_limit}' if rate_limit > 0 else '' # cmd += f' -severity {severities_str}' cmd += f' -timeout {str(timeout)}' if timeout and timeout > 0 else '' cmd += f' -tags {tags}' if tags else '' cmd += f' -silent' for tpl in templates: cmd += f' -t {tpl}' grouped_tasks = [] custom_ctx = ctx for severity in severities: custom_ctx['track'] = True _task = nuclei_individual_severity_module.si( cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx=custom_ctx, description=f'Nuclei Scan with severity {severity}' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan with all severities completed...') return None @app.task(name='dalfox_xss_scan', queue='main_scan_queue', base=RengineTask, bind=True) def dalfox_xss_scan(self, urls=[], ctx={}, description=None): """XSS Scan using dalfox Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) dalfox_config = vuln_config.get(DALFOX) or {} custom_header = dalfox_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() is_waf_evasion = dalfox_config.get(WAF_EVASION, False) blind_xss_server = dalfox_config.get(BLIND_XSS_SERVER) user_agent = dalfox_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) timeout = dalfox_config.get(TIMEOUT) delay = dalfox_config.get(DELAY) threads = dalfox_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_dalfox_xss.txt' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=False, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'dalfox --silence --no-color --no-spinner' cmd += f' --only-poc r ' cmd += f' --ignore-return 302,404,403' cmd += f' --skip-bav' cmd += f' file {input_path}' cmd += f' --proxy {proxy}' if proxy else '' cmd += f' --waf-evasion' if is_waf_evasion else '' cmd += f' -b {blind_xss_server}' if blind_xss_server else '' cmd += f' --delay {delay}' if delay else '' cmd += f' --timeout {timeout}' if timeout else '' cmd += f' --user-agent {user_agent}' if user_agent else '' cmd += f' --header {custom_header}' if custom_header else '' cmd += f' --worker {threads}' if threads else '' cmd += f' --format json' results = [] for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id, trunc_char=',' ): if not isinstance(line, dict): continue results.append(line) vuln_data = parse_dalfox_result(line) http_url = sanitize_url(line.get('data')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Dalfox Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=DALFOX ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='crlfuzz_scan', queue='main_scan_queue', base=RengineTask, bind=True) def crlfuzz_scan(self, urls=[], ctx={}, description=None): """CRLF Fuzzing with CRLFuzz Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) custom_header = vuln_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() user_agent = vuln_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) threads = vuln_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_crlf.txt' output_path = f'{self.results_dir}/{self.filename}' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=True, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'crlfuzz -s' cmd += f' -l {input_path}' cmd += f' -x {proxy}' if proxy else '' cmd += f' --H {custom_header}' if custom_header else '' cmd += f' -o {output_path}' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id ) if not os.path.isfile(output_path): logger.info('No Results from CRLFuzz') return crlfs = [] results = [] with open(output_path, 'r') as file: crlfs = file.readlines() for crlf in crlfs: url = crlf.strip() vuln_data = parse_crlfuzz_result(url) http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting CRLFuzz Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=CRLFUZZ ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='s3scanner', queue='main_scan_queue', base=RengineTask, bind=True) def s3scanner(self, ctx={}, description=None): """Bucket Scanner Args: ctx (dict): Context description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/#{self.scan_id}_subdomain_discovery.txt' vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} s3_config = vuln_config.get(S3SCANNER) or {} threads = s3_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) providers = s3_config.get(PROVIDERS, S3SCANNER_DEFAULT_PROVIDERS) scan_history = ScanHistory.objects.filter(pk=self.scan_id).first() for provider in providers: cmd = f's3scanner -bucket-file {input_path} -enumerate -provider {provider} -threads {threads} -json' for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue if line.get('bucket', {}).get('exists', 0) == 1: result = parse_s3scanner_result(line) s3bucket, created = S3Bucket.objects.get_or_create(**result) scan_history.buckets.add(s3bucket) logger.info(f"s3 bucket added {result['provider']}-{result['name']}-{result['region']}") @app.task(name='http_crawl', queue='main_scan_queue', base=RengineTask, bind=True) def http_crawl( self, urls=[], method=None, recrawl=False, ctx={}, track=True, description=None, is_ran_from_subdomain_scan=False, should_remove_duplicate_endpoints=True, duplicate_removal_fields=[]): """Use httpx to query HTTP URLs for important info like page titles, http status, etc... Args: urls (list, optional): A set of URLs to check. Overrides default behavior which queries all endpoints related to this scan. method (str): HTTP method to use (GET, HEAD, POST, PUT, DELETE). recrawl (bool, optional): If False, filter out URLs that have already been crawled. should_remove_duplicate_endpoints (bool): Whether to remove duplicate endpoints duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates Returns: list: httpx results. """ logger.info('Initiating HTTP Crawl') if is_ran_from_subdomain_scan: logger.info('Running From Subdomain Scan...') cmd = '/go/bin/httpx' cfg = self.yaml_configuration.get(HTTP_CRAWL) or {} custom_header = cfg.get(CUSTOM_HEADER, '') threads = cfg.get(THREADS, DEFAULT_THREADS) follow_redirect = cfg.get(FOLLOW_REDIRECT, True) self.output_path = None input_path = f'{self.results_dir}/httpx_input.txt' history_file = f'{self.results_dir}/commands.txt' if urls: # direct passing URLs to check if self.url_filter: urls = [u for u in urls if self.url_filter in u] with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_uncrawled=not recrawl, write_filepath=input_path, ctx=ctx ) # logger.debug(urls) # If no URLs found, skip it if not urls: return # Re-adjust thread number if few URLs to avoid spinning up a monster to # kill a fly. if len(urls) < threads: threads = len(urls) # Get random proxy proxy = get_random_proxy() # Run command cmd += f' -cl -ct -rt -location -td -websocket -cname -asn -cdn -probe -random-agent' cmd += f' -t {threads}' if threads > 0 else '' cmd += f' --http-proxy {proxy}' if proxy else '' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -json' cmd += f' -u {urls[0]}' if len(urls) == 1 else f' -l {input_path}' cmd += f' -x {method}' if method else '' cmd += f' -silent' if follow_redirect: cmd += ' -fr' results = [] endpoint_ids = [] for line in stream_command( cmd, history_file=history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not line or not isinstance(line, dict): continue logger.debug(line) # No response from endpoint if line.get('failed', False): continue # Parse httpx output host = line.get('host', '') content_length = line.get('content_length', 0) http_status = line.get('status_code') http_url, is_redirect = extract_httpx_url(line) page_title = line.get('title') webserver = line.get('webserver') cdn = line.get('cdn', False) rt = line.get('time') techs = line.get('tech', []) cname = line.get('cname', '') content_type = line.get('content_type', '') response_time = -1 if rt: response_time = float(''.join(ch for ch in rt if not ch.isalpha())) if rt[-2:] == 'ms': response_time = response_time / 1000 # Create Subdomain object in DB subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue # Save default HTTP URL to endpoint object in DB endpoint, created = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain, is_default=is_ran_from_subdomain_scan ) if not endpoint: continue endpoint.http_status = http_status endpoint.page_title = page_title endpoint.content_length = content_length endpoint.webserver = webserver endpoint.response_time = response_time endpoint.content_type = content_type endpoint.save() endpoint_str = f'{http_url} [{http_status}] `{content_length}B` `{webserver}` `{rt}`' logger.warning(endpoint_str) if endpoint and endpoint.is_alive and endpoint.http_status != 403: self.notify( fields={'Alive endpoint': f'• {endpoint_str}'}, add_meta_info=False) # Add endpoint to results line['_cmd'] = cmd line['final_url'] = http_url line['endpoint_id'] = endpoint.id line['endpoint_created'] = created line['is_redirect'] = is_redirect results.append(line) # Add technology objects to DB for technology in techs: tech, _ = Technology.objects.get_or_create(name=technology) endpoint.techs.add(tech) if is_ran_from_subdomain_scan: subdomain.technologies.add(tech) subdomain.save() endpoint.save() techs_str = ', '.join([f'`{tech}`' for tech in techs]) self.notify( fields={'Technologies': techs_str}, add_meta_info=False) # Add IP objects for 'a' records to DB a_records = line.get('a', []) for ip_address in a_records: ip, created = save_ip_address( ip_address, subdomain, subscan=self.subscan, cdn=cdn) ips_str = '• ' + '\n• '.join([f'`{ip}`' for ip in a_records]) self.notify( fields={'IPs': ips_str}, add_meta_info=False) # Add IP object for host in DB if host: ip, created = save_ip_address( host, subdomain, subscan=self.subscan, cdn=cdn) self.notify( fields={'IPs': f'• `{ip.address}`'}, add_meta_info=False) # Save subdomain and endpoint if is_ran_from_subdomain_scan: # save subdomain stuffs subdomain.http_url = http_url subdomain.http_status = http_status subdomain.page_title = page_title subdomain.content_length = content_length subdomain.webserver = webserver subdomain.response_time = response_time subdomain.content_type = content_type subdomain.cname = ','.join(cname) subdomain.is_cdn = cdn if cdn: subdomain.cdn_name = line.get('cdn_name') subdomain.save() endpoint.save() endpoint_ids.append(endpoint.id) if should_remove_duplicate_endpoints: # Remove 'fake' alive endpoints that are just redirects to the same page remove_duplicate_endpoints( self.scan_id, self.domain_id, self.subdomain_id, filter_ids=endpoint_ids ) # Remove input file run_command( f'rm {input_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) return results #---------------------# # Notifications tasks # #---------------------# @app.task(name='send_notif', bind=False, queue='send_notif_queue') def send_notif( message, scan_history_id=None, subscan_id=None, **options): if not 'title' in options: message = enrich_notification(message, scan_history_id, subscan_id) send_discord_message(message, **options) send_slack_message(message) send_telegram_message(message) @app.task(name='send_scan_notif', bind=False, queue='send_scan_notif_queue') def send_scan_notif( scan_history_id, subscan_id=None, engine_id=None, status='RUNNING'): """Send scan status notification. Works for scan or a subscan if subscan_id is passed. Args: scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Get domain, engine, scan_history objects engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() tasks = ScanActivity.objects.filter(scan_of=scan) if scan else 0 # Build notif options url = get_scan_url(scan_history_id, subscan_id) title = get_scan_title(scan_history_id, subscan_id) fields = get_scan_fields(engine, scan, subscan, status, tasks) severity = None msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) if status: severity = STATUS_TO_SEVERITIES.get(status) opts = { 'title': title, 'url': url, 'fields': fields, 'severity': severity } logger.warning(f'Sending notification "{title}" [{severity}]') # Send notification send_notif( msg, scan_history_id, subscan_id, **opts) @app.task(name='send_task_notif', bind=False, queue='send_task_notif_queue') def send_task_notif( task_name, status=None, result=None, output_path=None, traceback=None, scan_history_id=None, engine_id=None, subscan_id=None, severity=None, add_meta_info=True, update_fields={}): """Send task status notification. Args: task_name (str): Task name. status (str, optional): Task status. result (str, optional): Task result. output_path (str, optional): Task output path. traceback (str, optional): Task traceback. scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. severity (str, optional): Severity (will be mapped to notif colors) add_meta_info (bool, optional): Wheter to add scan / subscan info to notif. update_fields (dict, optional): Fields key / value to update. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Build fields url = None fields = {} if add_meta_info: engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() url = get_scan_url(scan_history_id) if status: fields['Status'] = f'**{status}**' if engine: fields['Engine'] = engine.engine_name if scan: fields['Scan ID'] = f'[#{scan.id}]({url})' if subscan: url = get_scan_url(scan_history_id, subscan_id) fields['Subscan ID'] = f'[#{subscan.id}]({url})' title = get_task_title(task_name, scan_history_id, subscan_id) if status: severity = STATUS_TO_SEVERITIES.get(status) msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) # Add fields to update for k, v in update_fields.items(): fields[k] = v # Add traceback to notif if traceback and notif.send_scan_tracebacks: fields['Traceback'] = f'```\n{traceback}\n```' # Add files to notif files = [] attach_file = ( notif.send_scan_output_file and output_path and result and not traceback ) if attach_file: output_title = output_path.split('/')[-1] files = [(output_path, output_title)] # Send notif opts = { 'title': title, 'url': url, 'files': files, 'severity': severity, 'fields': fields, 'fields_append': update_fields.keys() } send_notif( msg, scan_history_id=scan_history_id, subscan_id=subscan_id, **opts) @app.task(name='send_file_to_discord', bind=False, queue='send_file_to_discord_queue') def send_file_to_discord(file_path, title=None): notif = Notification.objects.first() do_send = notif and notif.send_to_discord and notif.discord_hook_url if not do_send: return False webhook = DiscordWebhook( url=notif.discord_hook_url, rate_limit_retry=True, username=title or "reNgine Discord Plugin" ) with open(file_path, "rb") as f: head, tail = os.path.split(file_path) webhook.add_file(file=f.read(), filename=tail) webhook.execute() @app.task(name='send_hackerone_report', bind=False, queue='send_hackerone_report_queue') def send_hackerone_report(vulnerability_id): """Send HackerOne vulnerability report. Args: vulnerability_id (int): Vulnerability id. Returns: int: HTTP response status code. """ vulnerability = Vulnerability.objects.get(id=vulnerability_id) severities = {v: k for k,v in NUCLEI_SEVERITY_MAP.items()} headers = { 'Content-Type': 'application/json', 'Accept': 'application/json' } # can only send vulnerability report if team_handle exists if len(vulnerability.target_domain.h1_team_handle) !=0: hackerone_query = Hackerone.objects.all() if hackerone_query.exists(): hackerone = Hackerone.objects.first() severity_value = severities[vulnerability.severity] tpl = hackerone.report_template # Replace syntax of report template with actual content tpl = tpl.replace('{vulnerability_name}', vulnerability.name) tpl = tpl.replace('{vulnerable_url}', vulnerability.http_url) tpl = tpl.replace('{vulnerability_severity}', severity_value) tpl = tpl.replace('{vulnerability_description}', vulnerability.description if vulnerability.description else '') tpl = tpl.replace('{vulnerability_extracted_results}', vulnerability.extracted_results if vulnerability.extracted_results else '') tpl = tpl.replace('{vulnerability_reference}', vulnerability.reference if vulnerability.reference else '') data = { "data": { "type": "report", "attributes": { "team_handle": vulnerability.target_domain.h1_team_handle, "title": '{} found in {}'.format(vulnerability.name, vulnerability.http_url), "vulnerability_information": tpl, "severity_rating": severity_value, "impact": "More information about the impact and vulnerability can be found here: \n" + vulnerability.reference if vulnerability.reference else "NA", } } } r = requests.post( 'https://api.hackerone.com/v1/hackers/reports', auth=(hackerone.username, hackerone.api_key), json=data, headers=headers ) response = r.json() status_code = r.status_code if status_code == 201: vulnerability.hackerone_report_id = response['data']["id"] vulnerability.open_status = False vulnerability.save() return status_code else: logger.error('No team handle found.') status_code = 111 return status_code #-------------# # Utils tasks # #-------------# @app.task(name='parse_nmap_results', bind=False, queue='parse_nmap_results_queue') def parse_nmap_results(xml_file, output_file=None): """Parse results from nmap output file. Args: xml_file (str): nmap XML report file path. Returns: list: List of vulnerabilities found from nmap results. """ with open(xml_file, encoding='utf8') as f: content = f.read() try: nmap_results = xmltodict.parse(content) # parse XML to dict except Exception as e: logger.exception(e) logger.error(f'Cannot parse {xml_file} to valid JSON. Skipping.') return [] # Write JSON to output file if output_file: with open(output_file, 'w') as f: json.dump(nmap_results, f, indent=4) logger.warning(json.dumps(nmap_results, indent=4)) hosts = ( nmap_results .get('nmaprun', {}) .get('host', {}) ) all_vulns = [] if isinstance(hosts, dict): hosts = [hosts] for host in hosts: # Grab hostname / IP from output hostnames_dict = host.get('hostnames', {}) if hostnames_dict: # Ensure that hostnames['hostname'] is a list for consistency hostnames_list = hostnames_dict['hostname'] if isinstance(hostnames_dict['hostname'], list) else [hostnames_dict['hostname']] # Extract all the @name values from the list of dictionaries hostnames = [entry.get('@name') for entry in hostnames_list] else: hostnames = [host.get('address')['@addr']] # Iterate over each hostname for each port for hostname in hostnames: # Grab ports from output ports = host.get('ports', {}).get('port', []) if isinstance(ports, dict): ports = [ports] for port in ports: url_vulns = [] port_number = port['@portid'] url = sanitize_url(f'{hostname}:{port_number}') logger.info(f'Parsing nmap results for {hostname}:{port_number} ...') if not port_number or not port_number.isdigit(): continue port_protocol = port['@protocol'] scripts = port.get('script', []) if isinstance(scripts, dict): scripts = [scripts] for script in scripts: script_id = script['@id'] script_output = script['@output'] script_output_table = script.get('table', []) logger.debug(f'Ran nmap script "{script_id}" on {port_number}/{port_protocol}:\n{script_output}\n') if script_id == 'vulscan': vulns = parse_nmap_vulscan_output(script_output) url_vulns.extend(vulns) elif script_id == 'vulners': vulns = parse_nmap_vulners_output(script_output) url_vulns.extend(vulns) # elif script_id == 'http-server-header': # TODO: nmap can help find technologies as well using the http-server-header script # regex = r'(\w+)/([\d.]+)\s?(?:\((\w+)\))?' # tech_name, tech_version, tech_os = re.match(regex, test_string).groups() # Technology.objects.get_or_create(...) # elif script_id == 'http_csrf': # vulns = parse_nmap_http_csrf_output(script_output) # url_vulns.extend(vulns) else: logger.warning(f'Script output parsing for script "{script_id}" is not supported yet.') # Add URL to vuln for vuln in url_vulns: # TODO: This should extend to any URL, not just HTTP vuln['http_url'] = url if 'http_path' in vuln: vuln['http_url'] += vuln['http_path'] all_vulns.append(vuln) return all_vulns def parse_nmap_http_csrf_output(script_output): pass def parse_nmap_vulscan_output(script_output): """Parse nmap vulscan script output. Args: script_output (str): Vulscan script output. Returns: list: List of Vulnerability dicts. """ data = {} vulns = [] provider_name = '' # Sort all vulns found by provider so that we can match each provider with # a function that pulls from its API to get more info about the # vulnerability. for line in script_output.splitlines(): if not line: continue if not line.startswith('['): # provider line if "No findings" in line: logger.info(f"No findings: {line}") continue elif ' - ' in line: provider_name, provider_url = tuple(line.split(' - ')) data[provider_name] = {'url': provider_url.rstrip(':'), 'entries': []} continue else: # Log a warning logger.warning(f"Unexpected line format: {line}") continue reg = r'\[(.*)\] (.*)' matches = re.match(reg, line) id, title = matches.groups() entry = {'id': id, 'title': title} data[provider_name]['entries'].append(entry) logger.warning('Vulscan parsed output:') logger.warning(pprint.pformat(data)) for provider_name in data: if provider_name == 'Exploit-DB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'IBM X-Force': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'MITRE CVE': logger.error(f'Provider {provider_name} is not supported YET.') for entry in data[provider_name]['entries']: cve_id = entry['id'] vuln = cve_to_vuln(cve_id) vulns.append(vuln) elif provider_name == 'OSVDB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'OpenVAS (Nessus)': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'SecurityFocus': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'VulDB': logger.error(f'Provider {provider_name} is not supported YET.') pass else: logger.error(f'Provider {provider_name} is not supported.') return vulns def parse_nmap_vulners_output(script_output, url=''): """Parse nmap vulners script output. TODO: Rework this as it's currently matching all CVEs no matter the confidence. Args: script_output (str): Script output. Returns: list: List of found vulnerabilities. """ vulns = [] # Check for CVE in script output CVE_REGEX = re.compile(r'.*(CVE-\d\d\d\d-\d+).*') matches = CVE_REGEX.findall(script_output) matches = list(dict.fromkeys(matches)) for cve_id in matches: # get CVE info vuln = cve_to_vuln(cve_id, vuln_type='nmap-vulners-nse') if vuln: vulns.append(vuln) return vulns def cve_to_vuln(cve_id, vuln_type=''): """Search for a CVE using CVESearch and return Vulnerability data. Args: cve_id (str): CVE ID in the form CVE-* Returns: dict: Vulnerability dict. """ cve_info = CVESearch('https://cve.circl.lu').id(cve_id) if not cve_info: logger.error(f'Could not fetch CVE info for cve {cve_id}. Skipping.') return None vuln_cve_id = cve_info['id'] vuln_name = vuln_cve_id vuln_description = cve_info.get('summary', 'none').replace(vuln_cve_id, '').strip() try: vuln_cvss = float(cve_info.get('cvss', -1)) except (ValueError, TypeError): vuln_cvss = -1 vuln_cwe_id = cve_info.get('cwe', '') exploit_ids = cve_info.get('refmap', {}).get('exploit-db', []) osvdb_ids = cve_info.get('refmap', {}).get('osvdb', []) references = cve_info.get('references', []) capec_objects = cve_info.get('capec', []) # Parse ovals for a better vuln name / type ovals = cve_info.get('oval', []) if ovals: vuln_name = ovals[0]['title'] vuln_type = ovals[0]['family'] # Set vulnerability severity based on CVSS score vuln_severity = 'info' if vuln_cvss < 4: vuln_severity = 'low' elif vuln_cvss < 7: vuln_severity = 'medium' elif vuln_cvss < 9: vuln_severity = 'high' else: vuln_severity = 'critical' # Build console warning message msg = f'{vuln_name} | {vuln_severity.upper()} | {vuln_cve_id} | {vuln_cwe_id} | {vuln_cvss}' for id in osvdb_ids: msg += f'\n\tOSVDB: {id}' for exploit_id in exploit_ids: msg += f'\n\tEXPLOITDB: {exploit_id}' logger.warning(msg) vuln = { 'name': vuln_name, 'type': vuln_type, 'severity': NUCLEI_SEVERITY_MAP[vuln_severity], 'description': vuln_description, 'cvss_score': vuln_cvss, 'references': references, 'cve_ids': [vuln_cve_id], 'cwe_ids': [vuln_cwe_id] } return vuln def parse_s3scanner_result(line): ''' Parses and returns s3Scanner Data ''' bucket = line['bucket'] return { 'name': bucket['name'], 'region': bucket['region'], 'provider': bucket['provider'], 'owner_display_name': bucket['owner_display_name'], 'owner_id': bucket['owner_id'], 'perm_auth_users_read': bucket['perm_auth_users_read'], 'perm_auth_users_write': bucket['perm_auth_users_write'], 'perm_auth_users_read_acl': bucket['perm_auth_users_read_acl'], 'perm_auth_users_write_acl': bucket['perm_auth_users_write_acl'], 'perm_auth_users_full_control': bucket['perm_auth_users_full_control'], 'perm_all_users_read': bucket['perm_all_users_read'], 'perm_all_users_write': bucket['perm_all_users_write'], 'perm_all_users_read_acl': bucket['perm_all_users_read_acl'], 'perm_all_users_write_acl': bucket['perm_all_users_write_acl'], 'perm_all_users_full_control': bucket['perm_all_users_full_control'], 'num_objects': bucket['num_objects'], 'size': bucket['bucket_size'] } def parse_nuclei_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ return { 'name': line['info'].get('name', ''), 'type': line['type'], 'severity': NUCLEI_SEVERITY_MAP[line['info'].get('severity', 'unknown')], 'template': line['template'], 'template_url': line['template-url'], 'template_id': line['template-id'], 'description': line['info'].get('description', ''), 'matcher_name': line.get('matcher-name', ''), 'curl_command': line.get('curl-command'), 'request': line.get('request'), 'response': line.get('response'), 'extracted_results': line.get('extracted-results', []), 'cvss_metrics': line['info'].get('classification', {}).get('cvss-metrics', ''), 'cvss_score': line['info'].get('classification', {}).get('cvss-score'), 'cve_ids': line['info'].get('classification', {}).get('cve_id', []) or [], 'cwe_ids': line['info'].get('classification', {}).get('cwe_id', []) or [], 'references': line['info'].get('reference', []) or [], 'tags': line['info'].get('tags', []), 'source': NUCLEI, } def parse_dalfox_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ description = '' description += f" Evidence: {line.get('evidence')} <br>" if line.get('evidence') else '' description += f" Message: {line.get('message')} <br>" if line.get('message') else '' description += f" Payload: {line.get('message_str')} <br>" if line.get('message_str') else '' description += f" Vulnerable Parameter: {line.get('param')} <br>" if line.get('param') else '' return { 'name': 'XSS (Cross Site Scripting)', 'type': 'XSS', 'severity': DALFOX_SEVERITY_MAP[line.get('severity', 'unknown')], 'description': description, 'source': DALFOX, 'cwe_ids': [line.get('cwe')] } def parse_crlfuzz_result(url): """Parse CRLF results Args: url (str): CRLF Vulnerable URL Returns: dict: Vulnerability data. """ return { 'name': 'CRLF (HTTP Response Splitting)', 'type': 'CRLF', 'severity': 2, 'description': 'A CRLF (HTTP Response Splitting) vulnerability has been discovered.', 'source': CRLFUZZ, } def record_exists(model, data, exclude_keys=[]): """ Check if a record already exists in the database based on the given data. Args: model (django.db.models.Model): The Django model to check against. data (dict): Data dictionary containing fields and values. exclude_keys (list): List of keys to exclude from the lookup. Returns: bool: True if the record exists, False otherwise. """ # Extract the keys that will be used for the lookup lookup_fields = {key: data[key] for key in data if key not in exclude_keys} # Return True if a record exists based on the lookup fields, False otherwise return model.objects.filter(**lookup_fields).exists() @app.task(name='geo_localize', bind=False, queue='geo_localize_queue') def geo_localize(host, ip_id=None): """Uses geoiplookup to find location associated with host. Args: host (str): Hostname. ip_id (int): IpAddress object id. Returns: startScan.models.CountryISO: CountryISO object from DB or None. """ if validators.ipv6(host): logger.info(f'Ipv6 "{host}" is not supported by geoiplookup. Skipping.') return None cmd = f'geoiplookup {host}' _, out = run_command(cmd) if 'IP Address not found' not in out and "can't resolve hostname" not in out: country_iso = out.split(':')[1].strip().split(',')[0] country_name = out.split(':')[1].strip().split(',')[1].strip() geo_object, _ = CountryISO.objects.get_or_create( iso=country_iso, name=country_name ) geo_json = { 'iso': country_iso, 'name': country_name } if ip_id: ip = IpAddress.objects.get(pk=ip_id) ip.geo_iso = geo_object ip.save() return geo_json logger.info(f'Geo IP lookup failed for host "{host}"') return None @app.task(name='query_whois', bind=False, queue='query_whois_queue') def query_whois(ip_domain, force_reload_whois=False): """Query WHOIS information for an IP or a domain name. Args: ip_domain (str): IP address or domain name. save_domain (bool): Whether to save domain or not, default False Returns: dict: WHOIS information. """ if not force_reload_whois and Domain.objects.filter(name=ip_domain).exists() and Domain.objects.get(name=ip_domain).domain_info: domain = Domain.objects.get(name=ip_domain) if not domain.insert_date: domain.insert_date = timezone.now() domain.save() domain_info_db = domain.domain_info domain_info = DottedDict( dnssec=domain_info_db.dnssec, created=domain_info_db.created, updated=domain_info_db.updated, expires=domain_info_db.expires, geolocation_iso=domain_info_db.geolocation_iso, status=[status['name'] for status in DomainWhoisStatusSerializer(domain_info_db.status, many=True).data], whois_server=domain_info_db.whois_server, ns_records=[ns['name'] for ns in NameServersSerializer(domain_info_db.name_servers, many=True).data], registrar_name=domain_info_db.registrar.name, registrar_phone=domain_info_db.registrar.phone, registrar_email=domain_info_db.registrar.email, registrar_url=domain_info_db.registrar.url, registrant_name=domain_info_db.registrant.name, registrant_id=domain_info_db.registrant.id_str, registrant_organization=domain_info_db.registrant.organization, registrant_city=domain_info_db.registrant.city, registrant_state=domain_info_db.registrant.state, registrant_zip_code=domain_info_db.registrant.zip_code, registrant_country=domain_info_db.registrant.country, registrant_phone=domain_info_db.registrant.phone, registrant_fax=domain_info_db.registrant.fax, registrant_email=domain_info_db.registrant.email, registrant_address=domain_info_db.registrant.address, admin_name=domain_info_db.admin.name, admin_id=domain_info_db.admin.id_str, admin_organization=domain_info_db.admin.organization, admin_city=domain_info_db.admin.city, admin_state=domain_info_db.admin.state, admin_zip_code=domain_info_db.admin.zip_code, admin_country=domain_info_db.admin.country, admin_phone=domain_info_db.admin.phone, admin_fax=domain_info_db.admin.fax, admin_email=domain_info_db.admin.email, admin_address=domain_info_db.admin.address, tech_name=domain_info_db.tech.name, tech_id=domain_info_db.tech.id_str, tech_organization=domain_info_db.tech.organization, tech_city=domain_info_db.tech.city, tech_state=domain_info_db.tech.state, tech_zip_code=domain_info_db.tech.zip_code, tech_country=domain_info_db.tech.country, tech_phone=domain_info_db.tech.phone, tech_fax=domain_info_db.tech.fax, tech_email=domain_info_db.tech.email, tech_address=domain_info_db.tech.address, related_tlds=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_tlds, many=True).data], related_domains=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_domains, many=True).data], historical_ips=[ip for ip in HistoricalIPSerializer(domain_info_db.historical_ips, many=True).data], ) if domain_info_db.dns_records: a_records = [] txt_records = [] mx_records = [] dns_records = [{'name': dns['name'], 'type': dns['type']} for dns in DomainDNSRecordSerializer(domain_info_db.dns_records, many=True).data] for dns in dns_records: if dns['type'] == 'a': a_records.append(dns['name']) elif dns['type'] == 'txt': txt_records.append(dns['name']) elif dns['type'] == 'mx': mx_records.append(dns['name']) domain_info.a_records = a_records domain_info.txt_records = txt_records domain_info.mx_records = mx_records else: logger.info(f'Domain info for "{ip_domain}" not found in DB, querying whois') domain_info = DottedDict() # find domain historical ip try: historical_ips = get_domain_historical_ip_address(ip_domain) domain_info.historical_ips = historical_ips except Exception as e: logger.error(f'HistoricalIP for {ip_domain} not found!\nError: {str(e)}') historical_ips = [] # find associated domains using ip_domain try: related_domains = reverse_whois(ip_domain.split('.')[0]) except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] # find related tlds using TLSx try: related_tlds = [] output_path = '/tmp/ip_domain_tlsx.txt' tlsx_command = f'tlsx -san -cn -silent -ro -host {ip_domain} -o {output_path}' run_command( tlsx_command, shell=True, ) tlsx_output = [] with open(output_path) as f: tlsx_output = f.readlines() tldextract_target = tldextract.extract(ip_domain) for doms in tlsx_output: doms = doms.strip() tldextract_res = tldextract.extract(doms) if ip_domain != doms and tldextract_res.domain == tldextract_target.domain and tldextract_res.subdomain == '': related_tlds.append(doms) related_tlds = list(set(related_tlds)) domain_info.related_tlds = related_tlds except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] related_domains_list = [] if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_related = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_related) related_domains_list.append(_domain['name']) for _domain in related_tlds: domain_related = RelatedDomain.objects.get_or_create( name=_domain, )[0] db_domain_info.related_tlds.add(domain_related) for _ip in historical_ips: historical_ip = HistoricalIP.objects.get_or_create( ip=_ip['ip'], owner=_ip['owner'], location=_ip['location'], last_seen=_ip['last_seen'], )[0] db_domain_info.historical_ips.add(historical_ip) domain.domain_info = db_domain_info domain.save() command = f'netlas host {ip_domain} -f json' # check if netlas key is provided netlas_key = get_netlas_key() command += f' -a {netlas_key}' if netlas_key else '' result = subprocess.check_output(command.split()).decode('utf-8') if 'Failed to parse response data' in result: # do fallback return { 'status': False, 'ip_domain': ip_domain, 'result': "Netlas limit exceeded.", 'message': 'Netlas limit exceeded.' } try: result = json.loads(result) logger.info(result) whois = result.get('whois') if result.get('whois') else {} domain_info.created = whois.get('created_date') domain_info.expires = whois.get('expiration_date') domain_info.updated = whois.get('updated_date') domain_info.whois_server = whois.get('whois_server') if 'registrant' in whois: registrant = whois.get('registrant') domain_info.registrant_name = registrant.get('name') domain_info.registrant_country = registrant.get('country') domain_info.registrant_id = registrant.get('id') domain_info.registrant_state = registrant.get('province') domain_info.registrant_city = registrant.get('city') domain_info.registrant_phone = registrant.get('phone') domain_info.registrant_address = registrant.get('street') domain_info.registrant_organization = registrant.get('organization') domain_info.registrant_fax = registrant.get('fax') domain_info.registrant_zip_code = registrant.get('postal_code') email_search = EMAIL_REGEX.search(str(registrant.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.registrant_email = field_content if 'administrative' in whois: administrative = whois.get('administrative') domain_info.admin_name = administrative.get('name') domain_info.admin_country = administrative.get('country') domain_info.admin_id = administrative.get('id') domain_info.admin_state = administrative.get('province') domain_info.admin_city = administrative.get('city') domain_info.admin_phone = administrative.get('phone') domain_info.admin_address = administrative.get('street') domain_info.admin_organization = administrative.get('organization') domain_info.admin_fax = administrative.get('fax') domain_info.admin_zip_code = administrative.get('postal_code') mail_search = EMAIL_REGEX.search(str(administrative.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.admin_email = field_content if 'technical' in whois: technical = whois.get('technical') domain_info.tech_name = technical.get('name') domain_info.tech_country = technical.get('country') domain_info.tech_state = technical.get('province') domain_info.tech_id = technical.get('id') domain_info.tech_city = technical.get('city') domain_info.tech_phone = technical.get('phone') domain_info.tech_address = technical.get('street') domain_info.tech_organization = technical.get('organization') domain_info.tech_fax = technical.get('fax') domain_info.tech_zip_code = technical.get('postal_code') mail_search = EMAIL_REGEX.search(str(technical.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.tech_email = field_content if 'dns' in result: dns = result.get('dns') domain_info.mx_records = dns.get('mx') domain_info.txt_records = dns.get('txt') domain_info.a_records = dns.get('a') domain_info.ns_records = whois.get('name_servers') domain_info.dnssec = True if whois.get('dnssec') else False domain_info.status = whois.get('status') if 'registrar' in whois: registrar = whois.get('registrar') domain_info.registrar_name = registrar.get('name') domain_info.registrar_email = registrar.get('email') domain_info.registrar_phone = registrar.get('phone') domain_info.registrar_url = registrar.get('url') # find associated domains if registrant email is found related_domains = reverse_whois(domain_info.get('registrant_email')) if domain_info.get('registrant_email') else [] for _domain in related_domains: related_domains_list.append(_domain['name']) # remove duplicate domains from related domains list related_domains_list = list(set(related_domains_list)) domain_info.related_domains = related_domains_list # save to db if domain exists if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_rel = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_rel) db_domain_info.dnssec = domain_info.get('dnssec') #dates db_domain_info.created = domain_info.get('created') db_domain_info.updated = domain_info.get('updated') db_domain_info.expires = domain_info.get('expires') #registrar db_domain_info.registrar = Registrar.objects.get_or_create( name=domain_info.get('registrar_name'), email=domain_info.get('registrar_email'), phone=domain_info.get('registrar_phone'), url=domain_info.get('registrar_url'), )[0] db_domain_info.registrant = DomainRegistration.objects.get_or_create( name=domain_info.get('registrant_name'), organization=domain_info.get('registrant_organization'), address=domain_info.get('registrant_address'), city=domain_info.get('registrant_city'), state=domain_info.get('registrant_state'), zip_code=domain_info.get('registrant_zip_code'), country=domain_info.get('registrant_country'), email=domain_info.get('registrant_email'), phone=domain_info.get('registrant_phone'), fax=domain_info.get('registrant_fax'), id_str=domain_info.get('registrant_id'), )[0] db_domain_info.admin = DomainRegistration.objects.get_or_create( name=domain_info.get('admin_name'), organization=domain_info.get('admin_organization'), address=domain_info.get('admin_address'), city=domain_info.get('admin_city'), state=domain_info.get('admin_state'), zip_code=domain_info.get('admin_zip_code'), country=domain_info.get('admin_country'), email=domain_info.get('admin_email'), phone=domain_info.get('admin_phone'), fax=domain_info.get('admin_fax'), id_str=domain_info.get('admin_id'), )[0] db_domain_info.tech = DomainRegistration.objects.get_or_create( name=domain_info.get('tech_name'), organization=domain_info.get('tech_organization'), address=domain_info.get('tech_address'), city=domain_info.get('tech_city'), state=domain_info.get('tech_state'), zip_code=domain_info.get('tech_zip_code'), country=domain_info.get('tech_country'), email=domain_info.get('tech_email'), phone=domain_info.get('tech_phone'), fax=domain_info.get('tech_fax'), id_str=domain_info.get('tech_id'), )[0] for status in domain_info.get('status') or []: _status = WhoisStatus.objects.get_or_create( name=status )[0] _status.save() db_domain_info.status.add(_status) for ns in domain_info.get('ns_records') or []: _ns = NameServer.objects.get_or_create( name=ns )[0] _ns.save() db_domain_info.name_servers.add(_ns) for a in domain_info.get('a_records') or []: _a = DNSRecord.objects.get_or_create( name=a, type='a' )[0] _a.save() db_domain_info.dns_records.add(_a) for mx in domain_info.get('mx_records') or []: _mx = DNSRecord.objects.get_or_create( name=mx, type='mx' )[0] _mx.save() db_domain_info.dns_records.add(_mx) for txt in domain_info.get('txt_records') or []: _txt = DNSRecord.objects.get_or_create( name=txt, type='txt' )[0] _txt.save() db_domain_info.dns_records.add(_txt) db_domain_info.geolocation_iso = domain_info.get('registrant_country') db_domain_info.whois_server = domain_info.get('whois_server') db_domain_info.save() domain.domain_info = db_domain_info domain.save() except Exception as e: return { 'status': False, 'ip_domain': ip_domain, 'result': "unable to fetch records from WHOIS database.", 'message': str(e) } return { 'status': True, 'ip_domain': ip_domain, 'dnssec': domain_info.get('dnssec'), 'created': domain_info.get('created'), 'updated': domain_info.get('updated'), 'expires': domain_info.get('expires'), 'geolocation_iso': domain_info.get('registrant_country'), 'domain_statuses': domain_info.get('status'), 'whois_server': domain_info.get('whois_server'), 'dns': { 'a': domain_info.get('a_records'), 'mx': domain_info.get('mx_records'), 'txt': domain_info.get('txt_records'), }, 'registrar': { 'name': domain_info.get('registrar_name'), 'phone': domain_info.get('registrar_phone'), 'email': domain_info.get('registrar_email'), 'url': domain_info.get('registrar_url'), }, 'registrant': { 'name': domain_info.get('registrant_name'), 'id': domain_info.get('registrant_id'), 'organization': domain_info.get('registrant_organization'), 'address': domain_info.get('registrant_address'), 'city': domain_info.get('registrant_city'), 'state': domain_info.get('registrant_state'), 'zipcode': domain_info.get('registrant_zip_code'), 'country': domain_info.get('registrant_country'), 'phone': domain_info.get('registrant_phone'), 'fax': domain_info.get('registrant_fax'), 'email': domain_info.get('registrant_email'), }, 'admin': { 'name': domain_info.get('admin_name'), 'id': domain_info.get('admin_id'), 'organization': domain_info.get('admin_organization'), 'address':domain_info.get('admin_address'), 'city': domain_info.get('admin_city'), 'state': domain_info.get('admin_state'), 'zipcode': domain_info.get('admin_zip_code'), 'country': domain_info.get('admin_country'), 'phone': domain_info.get('admin_phone'), 'fax': domain_info.get('admin_fax'), 'email': domain_info.get('admin_email'), }, 'technical_contact': { 'name': domain_info.get('tech_name'), 'id': domain_info.get('tech_id'), 'organization': domain_info.get('tech_organization'), 'address': domain_info.get('tech_address'), 'city': domain_info.get('tech_city'), 'state': domain_info.get('tech_state'), 'zipcode': domain_info.get('tech_zip_code'), 'country': domain_info.get('tech_country'), 'phone': domain_info.get('tech_phone'), 'fax': domain_info.get('tech_fax'), 'email': domain_info.get('tech_email'), }, 'nameservers': domain_info.get('ns_records'), # 'similar_domains': domain_info.get('similar_domains'), 'related_domains': domain_info.get('related_domains'), 'related_tlds': domain_info.get('related_tlds'), 'historical_ips': domain_info.get('historical_ips'), } @app.task(name='remove_duplicate_endpoints', bind=False, queue='remove_duplicate_endpoints_queue') def remove_duplicate_endpoints( scan_history_id, domain_id, subdomain_id=None, filter_ids=[], filter_status=[200, 301, 404], duplicate_removal_fields=ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS ): """Remove duplicate endpoints. Check for implicit redirections by comparing endpoints: - [x] `content_length` similarities indicating redirections - [x] `page_title` (check for same page title) - [ ] Sign-in / login page (check for endpoints with the same words) Args: scan_history_id: ScanHistory id. domain_id (int): Domain id. subdomain_id (int, optional): Subdomain id. filter_ids (list): List of endpoint ids to filter on. filter_status (list): List of HTTP status codes to filter on. duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates """ logger.info(f'Removing duplicate endpoints based on {duplicate_removal_fields}') endpoints = ( EndPoint.objects .filter(scan_history__id=scan_history_id) .filter(target_domain__id=domain_id) ) if filter_status: endpoints = endpoints.filter(http_status__in=filter_status) if subdomain_id: endpoints = endpoints.filter(subdomain__id=subdomain_id) if filter_ids: endpoints = endpoints.filter(id__in=filter_ids) for field_name in duplicate_removal_fields: cl_query = ( endpoints .values_list(field_name) .annotate(mc=Count(field_name)) .order_by('-mc') ) for (field_value, count) in cl_query: if count > DELETE_DUPLICATES_THRESHOLD: eps_to_delete = ( endpoints .filter(**{field_name: field_value}) .order_by('discovered_date') .all()[1:] ) msg = f'Deleting {len(eps_to_delete)} endpoints [reason: same {field_name} {field_value}]' for ep in eps_to_delete: url = urlparse(ep.http_url) if url.path in ['', '/', '/login']: # try do not delete the original page that other pages redirect to continue msg += f'\n\t {ep.http_url} [{ep.http_status}] [{field_name}={field_value}]' ep.delete() logger.warning(msg) @app.task(name='run_command', bind=False, queue='run_command_queue') def run_command(cmd, cwd=None, shell=False, history_file=None, scan_id=None, activity_id=None): """Run a given command using subprocess module. Args: cmd (str): Command to run. cwd (str): Current working directory. echo (bool): Log command. shell (bool): Run within separate shell if True. history_file (str): Write command + output to history file. Returns: tuple: Tuple with return_code, output. """ logger.info(cmd) logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Run the command using subprocess popen = subprocess.Popen( cmd if shell else cmd.split(), shell=shell, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, cwd=cwd, universal_newlines=True) output = '' for stdout_line in iter(popen.stdout.readline, ""): item = stdout_line.strip() output += '\n' + item logger.debug(item) popen.stdout.close() popen.wait() return_code = popen.returncode command_obj.output = output command_obj.return_code = return_code command_obj.save() if history_file: mode = 'a' if not os.path.exists(history_file): mode = 'w' with open(history_file, mode) as f: f.write(f'\n{cmd}\n{return_code}\n{output}\n------------------\n') return return_code, output #-------------# # Other utils # #-------------# def stream_command(cmd, cwd=None, shell=False, history_file=None, encoding='utf-8', scan_id=None, activity_id=None, trunc_char=None): # Log cmd logger.info(cmd) # logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Sanitize the cmd command = cmd if shell else cmd.split() # Run the command using subprocess process = subprocess.Popen( command, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, universal_newlines=True, shell=shell) # Log the output in real-time to the database output = "" # Process the output for line in iter(lambda: process.stdout.readline(), b''): if not line: break line = line.strip() ansi_escape = re.compile(r'\x1B(?:[@-Z\\-_]|\[[0-?]*[ -/]*[@-~])') line = ansi_escape.sub('', line) line = line.replace('\\x0d\\x0a', '\n') if trunc_char and line.endswith(trunc_char): line = line[:-1] item = line # Try to parse the line as JSON try: item = json.loads(line) except json.JSONDecodeError: pass # Yield the line #logger.debug(item) yield item # Add the log line to the output output += line + "\n" # Update the command record in the database command_obj.output = output command_obj.save() # Retrieve the return code and output process.wait() return_code = process.returncode # Update the return code and final output in the database command_obj.return_code = return_code command_obj.save() # Append the command, return code and output to the history file if history_file is not None: with open(history_file, "a") as f: f.write(f"{cmd}\n{return_code}\n{output}\n") def process_httpx_response(line): """TODO: implement this""" def extract_httpx_url(line): """Extract final URL from httpx results. Always follow redirects to find the last URL. Args: line (dict): URL data output by httpx. Returns: tuple: (final_url, redirect_bool) tuple. """ status_code = line.get('status_code', 0) final_url = line.get('final_url') location = line.get('location') chain_status_codes = line.get('chain_status_codes', []) # Final URL is already looking nice, if it exists return it if final_url: return final_url, False http_url = line['url'] # fallback to url field # Handle redirects manually REDIRECT_STATUS_CODES = [301, 302] is_redirect = ( status_code in REDIRECT_STATUS_CODES or any(x in REDIRECT_STATUS_CODES for x in chain_status_codes) ) if is_redirect and location: if location.startswith(('http', 'https')): http_url = location else: http_url = f'{http_url}/{location.lstrip("/")}' # Sanitize URL http_url = sanitize_url(http_url) return http_url, is_redirect #-------------# # OSInt utils # #-------------# def get_and_save_dork_results(lookup_target, results_dir, type, lookup_keywords=None, lookup_extensions=None, delay=3, page_count=2, scan_history=None): """ Uses gofuzz to dork and store information Args: lookup_target (str): target to look into such as stackoverflow or even the target itself results_dir (str): Results directory type (str): Dork Type Title lookup_keywords (str): comma separated keywords or paths to look for lookup_extensions (str): comma separated extensions to look for delay (int): delay between each requests page_count (int): pages in google to extract information scan_history (startScan.ScanHistory): Scan History Object """ results = [] gofuzz_command = f'{GOFUZZ_EXEC_PATH} -t {lookup_target} -d {delay} -p {page_count}' if lookup_extensions: gofuzz_command += f' -e {lookup_extensions}' elif lookup_keywords: gofuzz_command += f' -w {lookup_keywords}' output_file = f'{results_dir}/gofuzz.txt' gofuzz_command += f' -o {output_file}' history_file = f'{results_dir}/commands.txt' try: run_command( gofuzz_command, shell=False, history_file=history_file, scan_id=scan_history.id, ) if not os.path.isfile(output_file): return with open(output_file) as f: for line in f.readlines(): url = line.strip() if url: results.append(url) dork, created = Dork.objects.get_or_create( type=type, url=url ) if scan_history: scan_history.dorks.add(dork) # remove output file os.remove(output_file) except Exception as e: logger.exception(e) return results def get_and_save_emails(scan_history, activity_id, results_dir): """Get and save emails from Google, Bing and Baidu. Args: scan_history (startScan.ScanHistory): Scan history object. activity_id: ScanActivity Object results_dir (str): Results directory. Returns: list: List of emails found. """ emails = [] # Proxy settings # get_random_proxy() # Gather emails from Google, Bing and Baidu output_file = f'{results_dir}/emails_tmp.txt' history_file = f'{results_dir}/commands.txt' command = f'python3 /usr/src/github/Infoga/infoga.py --domain {scan_history.domain.name} --source all --report {output_file}' try: run_command( command, shell=False, history_file=history_file, scan_id=scan_history.id, activity_id=activity_id) if not os.path.isfile(output_file): logger.info('No Email results') return [] with open(output_file) as f: for line in f.readlines(): if 'Email' in line: split_email = line.split(' ')[2] emails.append(split_email) output_path = f'{results_dir}/emails.txt' with open(output_path, 'w') as output_file: for email_address in emails: save_email(email_address, scan_history) output_file.write(f'{email_address}\n') except Exception as e: logger.exception(e) return emails def save_metadata_info(meta_dict): """Extract metadata from Google Search. Args: meta_dict (dict): Info dict. Returns: list: List of startScan.MetaFinderDocument objects. """ logger.warning(f'Getting metadata for {meta_dict.osint_target}') scan_history = ScanHistory.objects.get(id=meta_dict.scan_id) # Proxy settings get_random_proxy() # Get metadata result = extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if not result: logger.error(f'No metadata result from Google Search for {meta_dict.osint_target}.') return [] # Add metadata info to DB results = [] for metadata_name, data in result.get_metadata().items(): subdomain = Subdomain.objects.get( scan_history=meta_dict.scan_id, name=meta_dict.osint_target) metadata = DottedDict({k: v for k, v in data.items()}) meta_finder_document = MetaFinderDocument( subdomain=subdomain, target_domain=meta_dict.domain, scan_history=scan_history, url=metadata.url, doc_name=metadata_name, http_status=metadata.status_code, producer=metadata.metadata.get('Producer'), creator=metadata.metadata.get('Creator'), creation_date=metadata.metadata.get('CreationDate'), modified_date=metadata.metadata.get('ModDate'), author=metadata.metadata.get('Author'), title=metadata.metadata.get('Title'), os=metadata.metadata.get('OSInfo')) meta_finder_document.save() results.append(data) return results #-----------------# # Utils functions # #-----------------# def create_scan_activity(scan_history_id, message, status): scan_activity = ScanActivity() scan_activity.scan_of = ScanHistory.objects.get(pk=scan_history_id) scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id #--------------------# # Database functions # #--------------------# def save_vulnerability(**vuln_data): references = vuln_data.pop('references', []) cve_ids = vuln_data.pop('cve_ids', []) cwe_ids = vuln_data.pop('cwe_ids', []) tags = vuln_data.pop('tags', []) subscan = vuln_data.pop('subscan', None) # remove nulls vuln_data = replace_nulls(vuln_data) # Create vulnerability vuln, created = Vulnerability.objects.get_or_create(**vuln_data) if created: vuln.discovered_date = timezone.now() vuln.open_status = True vuln.save() # Save vuln tags for tag_name in tags or []: tag, created = VulnerabilityTags.objects.get_or_create(name=tag_name) if tag: vuln.tags.add(tag) vuln.save() # Save CVEs for cve_id in cve_ids or []: cve, created = CveId.objects.get_or_create(name=cve_id) if cve: vuln.cve_ids.add(cve) vuln.save() # Save CWEs for cve_id in cwe_ids or []: cwe, created = CweId.objects.get_or_create(name=cve_id) if cwe: vuln.cwe_ids.add(cwe) vuln.save() # Save vuln reference for url in references or []: ref, created = VulnerabilityReference.objects.get_or_create(url=url) if created: vuln.references.add(ref) vuln.save() # Save subscan id in vuln object if subscan: vuln.vuln_subscan_ids.add(subscan) vuln.save() return vuln, created def save_endpoint( http_url, ctx={}, crawl=False, is_default=False, **endpoint_data): """Get or create EndPoint object. If crawl is True, also crawl the endpoint HTTP URL with httpx. Args: http_url (str): Input HTTP URL. is_default (bool): If the url is a default url for SubDomains. scan_history (startScan.models.ScanHistory): ScanHistory object. domain (startScan.models.Domain): Domain object. subdomain (starScan.models.Subdomain): Subdomain object. results_dir (str, optional): Results directory. crawl (bool, optional): Run httpx on endpoint if True. Default: False. force (bool, optional): Force crawl even if ENABLE_HTTP_CRAWL mode is on. subscan (startScan.models.SubScan, optional): SubScan object. Returns: tuple: (startScan.models.EndPoint, created) where `created` is a boolean indicating if the object is new or already existed. """ # remove nulls endpoint_data = replace_nulls(endpoint_data) scheme = urlparse(http_url).scheme endpoint = None created = False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in http_url: logger.error(f"{http_url} is not a URL of domain {domain.name}. Skipping.") return None, False if crawl: ctx['track'] = False results = http_crawl( urls=[http_url], method='HEAD', ctx=ctx) if results: endpoint_data = results[0] endpoint_id = endpoint_data['endpoint_id'] created = endpoint_data['endpoint_created'] endpoint = EndPoint.objects.get(pk=endpoint_id) elif not scheme: return None, False else: # add dumb endpoint without probing it scan = ScanHistory.objects.filter(pk=ctx.get('scan_history_id')).first() domain = Domain.objects.filter(pk=ctx.get('domain_id')).first() if not validators.url(http_url): return None, False http_url = sanitize_url(http_url) endpoint, created = EndPoint.objects.get_or_create( scan_history=scan, target_domain=domain, http_url=http_url, **endpoint_data) if created: endpoint.is_default = is_default endpoint.discovered_date = timezone.now() endpoint.save() subscan_id = ctx.get('subscan_id') if subscan_id: endpoint.endpoint_subscan_ids.add(subscan_id) endpoint.save() return endpoint, created def save_subdomain(subdomain_name, ctx={}): """Get or create Subdomain object. Args: subdomain_name (str): Subdomain name. scan_history (startScan.models.ScanHistory): ScanHistory object. Returns: tuple: (startScan.models.Subdomain, created) where `created` is a boolean indicating if the object has been created in DB. """ scan_id = ctx.get('scan_history_id') subscan_id = ctx.get('subscan_id') out_of_scope_subdomains = ctx.get('out_of_scope_subdomains', []) valid_domain = ( validators.domain(subdomain_name) or validators.ipv4(subdomain_name) or validators.ipv6(subdomain_name) ) if not valid_domain: logger.error(f'{subdomain_name} is not an invalid domain. Skipping.') return None, False if subdomain_name in out_of_scope_subdomains: logger.error(f'{subdomain_name} is out-of-scope. Skipping.') return None, False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in subdomain_name: logger.error(f"{subdomain_name} is not a subdomain of domain {domain.name}. Skipping.") return None, False scan = ScanHistory.objects.filter(pk=scan_id).first() domain = scan.domain if scan else None subdomain, created = Subdomain.objects.get_or_create( scan_history=scan, target_domain=domain, name=subdomain_name) if created: # logger.warning(f'Found new subdomain {subdomain_name}') subdomain.discovered_date = timezone.now() if subscan_id: subdomain.subdomain_subscan_ids.add(subscan_id) subdomain.save() return subdomain, created def save_email(email_address, scan_history=None): if not validators.email(email_address): logger.info(f'Email {email_address} is invalid. Skipping.') return None, False email, created = Email.objects.get_or_create(address=email_address) # if created: # logger.warning(f'Found new email address {email_address}') # Add email to ScanHistory if scan_history: scan_history.emails.add(email) scan_history.save() return email, created def save_employee(name, designation, scan_history=None): employee, created = Employee.objects.get_or_create( name=name, designation=designation) # if created: # logger.warning(f'Found new employee {name}') # Add employee to ScanHistory if scan_history: scan_history.employees.add(employee) scan_history.save() return employee, created def save_ip_address(ip_address, subdomain=None, subscan=None, **kwargs): if not (validators.ipv4(ip_address) or validators.ipv6(ip_address)): logger.info(f'IP {ip_address} is not a valid IP. Skipping.') return None, False ip, created = IpAddress.objects.get_or_create(address=ip_address) # if created: # logger.warning(f'Found new IP {ip_address}') # Set extra attributes for key, value in kwargs.items(): setattr(ip, key, value) ip.save() # Add IP to subdomain if subdomain: subdomain.ip_addresses.add(ip) subdomain.save() # Add subscan to IP if subscan: ip.ip_subscan_ids.add(subscan) # Geo-localize IP asynchronously if created: geo_localize.delay(ip_address, ip.id) return ip, created def save_imported_subdomains(subdomains, ctx={}): """Take a list of subdomains imported and write them to from_imported.txt. Args: subdomains (list): List of subdomain names. scan_history (startScan.models.ScanHistory): ScanHistory instance. domain (startScan.models.Domain): Domain instance. results_dir (str): Results directory. """ domain_id = ctx['domain_id'] domain = Domain.objects.get(pk=domain_id) results_dir = ctx.get('results_dir', RENGINE_RESULTS) # Validate each subdomain and de-duplicate entries subdomains = list(set([ subdomain for subdomain in subdomains if validators.domain(subdomain) and domain.name == get_domain_from_subdomain(subdomain) ])) if not subdomains: return logger.warning(f'Found {len(subdomains)} imported subdomains.') with open(f'{results_dir}/from_imported.txt', 'w+') as output_file: for name in subdomains: subdomain_name = name.strip() subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain.is_imported_subdomain = True subdomain.save() output_file.write(f'{subdomain}\n') @app.task(name='query_reverse_whois', bind=False, queue='query_reverse_whois_queue') def query_reverse_whois(lookup_keyword): """Queries Reverse WHOIS information for an organization or email address. Args: lookup_keyword (str): Registrar Name or email Returns: dict: Reverse WHOIS information. """ return get_associated_domains(lookup_keyword) @app.task(name='query_ip_history', bind=False, queue='query_ip_history_queue') def query_ip_history(domain): """Queries the IP history for a domain Args: domain (str): domain_name Returns: list: list of historical ip addresses """ return get_domain_historical_ip_address(domain) @app.task(name='gpt_vulnerability_description', bind=False, queue='gpt_queue') def gpt_vulnerability_description(vulnerability_id): """Generate and store Vulnerability Description using GPT. Args: vulnerability_id (Vulnerability Model ID): Vulnerability ID to fetch Description. """ logger.info('Getting GPT Vulnerability Description') try: lookup_vulnerability = Vulnerability.objects.get(id=vulnerability_id) lookup_url = urlparse(lookup_vulnerability.http_url) path = lookup_url.path except Exception as e: return { 'status': False, 'error': str(e) } # check in db GPTVulnerabilityReport model if vulnerability description and path matches stored = GPTVulnerabilityReport.objects.filter(url_path=path).filter(title=lookup_vulnerability.name).first() if stored: response = { 'status': True, 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: vulnerability_description = get_gpt_vuln_input_description( lookup_vulnerability.name, path ) # one can add more description here later gpt_generator = GPTVulnerabilityReportGenerator() response = gpt_generator.get_vulnerability_description(vulnerability_description) add_gpt_description_db( lookup_vulnerability.name, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) # for all vulnerabilities with the same vulnerability name this description has to be stored. # also the consition is that the url must contain a part of this. for vuln in Vulnerability.objects.filter(name=lookup_vulnerability.name, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() return response

import csv import json import os import pprint import subprocess import time import validators import whatportis import xmltodict import yaml import tldextract import concurrent.futures from datetime import datetime from urllib.parse import urlparse from api.serializers import SubdomainSerializer from celery import chain, chord, group from celery.result import allow_join_result from celery.utils.log import get_task_logger from django.db.models import Count from dotted_dict import DottedDict from django.utils import timezone from pycvesearch import CVESearch from metafinder.extractor import extract_metadata_from_google_search from reNgine.celery import app from reNgine.gpt import GPTVulnerabilityReportGenerator from reNgine.celery_custom_task import RengineTask from reNgine.common_func import * from reNgine.definitions import * from reNgine.settings import * from reNgine.gpt import * from reNgine.utilities import * from scanEngine.models import (EngineType, InstalledExternalTool, Notification, Proxy) from startScan.models import * from startScan.models import EndPoint, Subdomain, Vulnerability from targetApp.models import Domain """ Celery tasks. """ logger = get_task_logger(__name__) #----------------------# # Scan / Subscan tasks # #----------------------# @app.task(name='initiate_scan', bind=False, queue='initiate_scan_queue') def initiate_scan( scan_history_id, domain_id, engine_id=None, scan_type=LIVE_SCAN, results_dir=RENGINE_RESULTS, imported_subdomains=[], out_of_scope_subdomains=[], url_filter=''): """Initiate a new scan. Args: scan_history_id (int): ScanHistory id. domain_id (int): Domain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. imported_subdomains (list): Imported subdomains. out_of_scope_subdomains (list): Out-of-scope subdomains. url_filter (str): URL path. Default: '' """ # Get scan history scan = ScanHistory.objects.get(pk=scan_history_id) # Get scan engine engine_id = engine_id or scan.scan_type.id # scan history engine_id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, []) # Get domain and set last_scan_date domain = Domain.objects.get(pk=domain_id) domain.last_scan_date = timezone.now() domain.save() # Get path filter url_filter = url_filter.rstrip('/') # Get or create ScanHistory() object if scan_type == LIVE_SCAN: # immediate scan = ScanHistory.objects.get(pk=scan_history_id) scan.scan_status = RUNNING_TASK elif scan_type == SCHEDULED_SCAN: # scheduled scan = ScanHistory() scan.scan_status = INITIATED_TASK scan.scan_type = engine scan.celery_ids = [initiate_scan.request.id] scan.domain = domain scan.start_scan_date = timezone.now() scan.tasks = engine.tasks scan.results_dir = f'{results_dir}/{domain.name}_{scan.id}' add_gf_patterns = gf_patterns and 'fetch_url' in engine.tasks if add_gf_patterns: scan.used_gf_patterns = ','.join(gf_patterns) scan.save() # Create scan results dir os.makedirs(scan.results_dir) # Build task context ctx = { 'scan_history_id': scan_history_id, 'engine_id': engine_id, 'domain_id': domain.id, 'results_dir': scan.results_dir, 'url_filter': url_filter, 'yaml_configuration': config, 'out_of_scope_subdomains': out_of_scope_subdomains } ctx_str = json.dumps(ctx, indent=2) # Send start notif logger.warning(f'Starting scan {scan_history_id} with context:\n{ctx_str}') send_scan_notif.delay( scan_history_id, subscan_id=None, engine_id=engine_id, status=CELERY_TASK_STATUS_MAP[scan.scan_status]) # Save imported subdomains in DB save_imported_subdomains(imported_subdomains, ctx=ctx) # Create initial subdomain in DB: make a copy of domain as a subdomain so # that other tasks using subdomains can use it. subdomain_name = domain.name subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) # If enable_http_crawl is set, create an initial root HTTP endpoint so that # HTTP crawling can start somewhere http_url = f'{domain.name}{url_filter}' if url_filter else domain.name endpoint, _ = save_endpoint( http_url, ctx=ctx, crawl=enable_http_crawl, is_default=True, subdomain=subdomain ) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build Celery tasks, crafted according to the dependency graph below: # subdomain_discovery --> port_scan --> fetch_url --> dir_file_fuzz # osint vulnerability_scan # osint dalfox xss scan # screenshot # waf_detection workflow = chain( group( subdomain_discovery.si(ctx=ctx, description='Subdomain discovery'), osint.si(ctx=ctx, description='OS Intelligence') ), port_scan.si(ctx=ctx, description='Port scan'), fetch_url.si(ctx=ctx, description='Fetch URL'), group( dir_file_fuzz.si(ctx=ctx, description='Directories & files fuzz'), vulnerability_scan.si(ctx=ctx, description='Vulnerability scan'), screenshot.si(ctx=ctx, description='Screenshot'), waf_detection.si(ctx=ctx, description='WAF detection') ) ) # Build callback callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery chord logger.info(f'Running Celery workflow with {len(workflow.tasks) + 1} tasks') task = chain(workflow, callback).on_error(callback).delay() scan.celery_ids.append(task.id) scan.save() return { 'success': True, 'task_id': task.id } @app.task(name='initiate_subscan', bind=False, queue='subscan_queue') def initiate_subscan( scan_history_id, subdomain_id, engine_id=None, scan_type=None, results_dir=RENGINE_RESULTS, url_filter=''): """Initiate a new subscan. Args: scan_history_id (int): ScanHistory id. subdomain_id (int): Subdomain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. url_filter (str): URL path. Default: '' """ # Get Subdomain, Domain and ScanHistory subdomain = Subdomain.objects.get(pk=subdomain_id) scan = ScanHistory.objects.get(pk=subdomain.scan_history.id) domain = Domain.objects.get(pk=subdomain.target_domain.id) # Get EngineType engine_id = engine_id or scan.scan_type.id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Create scan activity of SubScan Model subscan = SubScan( start_scan_date=timezone.now(), celery_ids=[initiate_subscan.request.id], scan_history=scan, subdomain=subdomain, type=scan_type, status=RUNNING_TASK, engine=engine) subscan.save() # Get YAML configuration config = yaml.safe_load(engine.yaml_configuration) # Create results directory results_dir = f'{scan.results_dir}/subscans/{subscan.id}' os.makedirs(results_dir, exist_ok=True) # Run task method = globals().get(scan_type) if not method: logger.warning(f'Task {scan_type} is not supported by reNgine. Skipping') return scan.tasks.append(scan_type) scan.save() # Send start notif send_scan_notif.delay( scan.id, subscan_id=subscan.id, engine_id=engine_id, status='RUNNING') # Build context ctx = { 'scan_history_id': scan.id, 'subscan_id': subscan.id, 'engine_id': engine_id, 'domain_id': domain.id, 'subdomain_id': subdomain.id, 'yaml_configuration': config, 'results_dir': results_dir, 'url_filter': url_filter } # Create initial endpoints in DB: find domain HTTP endpoint so that HTTP # crawling can start somewhere base_url = f'{subdomain.name}{url_filter}' if url_filter else subdomain.name endpoint, _ = save_endpoint( base_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build header + callback workflow = method.si(ctx=ctx) callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery tasks task = chain(workflow, callback).on_error(callback).delay() subscan.celery_ids.append(task.id) subscan.save() return { 'success': True, 'task_id': task.id } @app.task(name='report', bind=False, queue='report_queue') def report(ctx={}, description=None): """Report task running after all other tasks. Mark ScanHistory or SubScan object as completed and update with final status, log run details and send notification. Args: description (str, optional): Task description shown in UI. """ # Get objects subscan_id = ctx.get('subscan_id') scan_id = ctx.get('scan_history_id') engine_id = ctx.get('engine_id') scan = ScanHistory.objects.filter(pk=scan_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() # Get failed tasks tasks = ScanActivity.objects.filter(scan_of=scan).all() if subscan: tasks = tasks.filter(celery_id__in=subscan.celery_ids) failed_tasks = tasks.filter(status=FAILED_TASK) # Get task status failed_count = failed_tasks.count() status = SUCCESS_TASK if failed_count == 0 else FAILED_TASK status_h = 'SUCCESS' if failed_count == 0 else 'FAILED' # Update scan / subscan status if subscan: subscan.stop_scan_date = timezone.now() subscan.status = status subscan.save() else: scan.scan_status = status scan.stop_scan_date = timezone.now() scan.save() # Send scan status notif send_scan_notif.delay( scan_history_id=scan_id, subscan_id=subscan_id, engine_id=engine_id, status=status_h) #------------------------- # # Tracked reNgine tasks # #--------------------------# @app.task(name='subdomain_discovery', queue='main_scan_queue', base=RengineTask, bind=True) def subdomain_discovery( self, host=None, ctx=None, description=None): """Uses a set of tools (see SUBDOMAIN_SCAN_DEFAULT_TOOLS) to scan all subdomains associated with a domain. Args: host (str): Hostname to scan. Returns: subdomains (list): List of subdomain names. """ if not host: host = self.subdomain.name if self.subdomain else self.domain.name if self.url_filter: logger.warning(f'Ignoring subdomains scan as an URL path filter was passed ({self.url_filter}).') return # Config config = self.yaml_configuration.get(SUBDOMAIN_DISCOVERY) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL) or self.yaml_configuration.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) tools = config.get(USES_TOOLS, SUBDOMAIN_SCAN_DEFAULT_TOOLS) default_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=True).filter(is_subdomain_gathering=True)] custom_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=False).filter(is_subdomain_gathering=True)] send_subdomain_changes, send_interesting = False, False notif = Notification.objects.first() if notif: send_subdomain_changes = notif.send_subdomain_changes_notif send_interesting = notif.send_interesting_notif # Gather tools to run for subdomain scan if ALL in tools: tools = SUBDOMAIN_SCAN_DEFAULT_TOOLS + custom_subdomain_tools tools = [t.lower() for t in tools] # Make exception for amass since tool name is amass, but command is amass-active/passive default_subdomain_tools.append('amass-passive') default_subdomain_tools.append('amass-active') # Run tools for tool in tools: cmd = None logger.info(f'Scanning subdomains for {host} with {tool}') proxy = get_random_proxy() if tool in default_subdomain_tools: if tool == 'amass-passive': cmd = f'amass enum -passive -d {host} -o {self.results_dir}/subdomains_amass.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' elif tool == 'amass-active': use_amass_config = config.get(USE_AMASS_CONFIG, False) amass_wordlist_name = config.get(AMASS_WORDLIST, 'deepmagic.com-prefixes-top50000') wordlist_path = f'/usr/src/wordlist/{amass_wordlist_name}.txt' cmd = f'amass enum -active -d {host} -o {self.results_dir}/subdomains_amass_active.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' cmd += f' -brute -w {wordlist_path}' elif tool == 'sublist3r': cmd = f'python3 /usr/src/github/Sublist3r/sublist3r.py -d {host} -t {threads} -o {self.results_dir}/subdomains_sublister.txt' elif tool == 'subfinder': cmd = f'subfinder -d {host} -o {self.results_dir}/subdomains_subfinder.txt' use_subfinder_config = config.get(USE_SUBFINDER_CONFIG, False) cmd += ' -config /root/.config/subfinder/config.yaml' if use_subfinder_config else '' cmd += f' -proxy {proxy}' if proxy else '' cmd += f' -timeout {timeout}' if timeout else '' cmd += f' -t {threads}' if threads else '' cmd += f' -silent' elif tool == 'oneforall': cmd = f'python3 /usr/src/github/OneForAll/oneforall.py --target {host} run' cmd_extract = f'cut -d\',\' -f6 /usr/src/github/OneForAll/results/{host}.csv > {self.results_dir}/subdomains_oneforall.txt' cmd_rm = f'rm -rf /usr/src/github/OneForAll/results/{host}.csv' cmd += f' && {cmd_extract} && {cmd_rm}' elif tool == 'ctfr': results_file = self.results_dir + '/subdomains_ctfr.txt' cmd = f'python3 /usr/src/github/ctfr/ctfr.py -d {host} -o {results_file}' cmd_extract = f"cat {results_file} | sed 's/\*.//g' | tail -n +12 | uniq | sort > {results_file}" cmd += f' && {cmd_extract}' elif tool == 'tlsx': results_file = self.results_dir + '/subdomains_tlsx.txt' cmd = f'tlsx -san -cn -silent -ro -host {host}' cmd += f" | sed -n '/^\([a-zA-Z0-9]\([-a-zA-Z0-9]*[a-zA-Z0-9]\)\?\.\)\+{host}$/p' | uniq | sort" cmd += f' > {results_file}' elif tool == 'netlas': results_file = self.results_dir + '/subdomains_netlas.txt' cmd = f'netlas search -d domain -i domain domain:"*.{host}" -f json' netlas_key = get_netlas_key() cmd += f' -a {netlas_key}' if netlas_key else '' cmd_extract = f"grep -oE '([a-zA-Z0-9]([-a-zA-Z0-9]*[a-zA-Z0-9])?\.)+{host}'" cmd += f' | {cmd_extract} > {results_file}' elif tool in custom_subdomain_tools: tool_query = InstalledExternalTool.objects.filter(name__icontains=tool.lower()) if not tool_query.exists(): logger.error(f'Missing {{TARGET}} and {{OUTPUT}} placeholders in {tool} configuration. Skipping.') continue custom_tool = tool_query.first() cmd = custom_tool.subdomain_gathering_command if '{TARGET}' in cmd and '{OUTPUT}' in cmd: cmd = cmd.replace('{TARGET}', host) cmd = cmd.replace('{OUTPUT}', f'{self.results_dir}/subdomains_{tool}.txt') cmd = cmd.replace('{PATH}', custom_tool.github_clone_path) if '{PATH}' in cmd else cmd else: logger.warning( f'Subdomain discovery tool "{tool}" is not supported by reNgine. Skipping.') continue # Run tool try: run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) except Exception as e: logger.error( f'Subdomain discovery tool "{tool}" raised an exception') logger.exception(e) # Gather all the tools' results in one single file. Write subdomains into # separate files, and sort all subdomains. run_command( f'cat {self.results_dir}/subdomains_*.txt > {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {self.output_path} -o {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) with open(self.output_path) as f: lines = f.readlines() # Parse the output_file file and store Subdomain and EndPoint objects found # in db. subdomain_count = 0 subdomains = [] urls = [] for line in lines: subdomain_name = line.strip() valid_url = bool(validators.url(subdomain_name)) valid_domain = ( bool(validators.domain(subdomain_name)) or bool(validators.ipv4(subdomain_name)) or bool(validators.ipv6(subdomain_name)) or valid_url ) if not valid_domain: logger.error(f'Subdomain {subdomain_name} is not a valid domain, IP or URL. Skipping.') continue if valid_url: subdomain_name = urlparse(subdomain_name).netloc if subdomain_name in self.out_of_scope_subdomains: logger.error(f'Subdomain {subdomain_name} is out of scope. Skipping.') continue # Add subdomain subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain_count += 1 subdomains.append(subdomain) urls.append(subdomain.name) # Bulk crawl subdomains if enable_http_crawl: ctx['track'] = True http_crawl(urls, ctx=ctx, is_ran_from_subdomain_scan=True) # Find root subdomain endpoints for subdomain in subdomains: pass # Send notifications subdomains_str = '\n'.join([f'• `{subdomain.name}`' for subdomain in subdomains]) self.notify(fields={ 'Subdomain count': len(subdomains), 'Subdomains': subdomains_str, }) if send_subdomain_changes and self.scan_id and self.domain_id: added = get_new_added_subdomain(self.scan_id, self.domain_id) removed = get_removed_subdomain(self.scan_id, self.domain_id) if added: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in added]) self.notify(fields={'Added subdomains': subdomains_str}) if removed: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in removed]) self.notify(fields={'Removed subdomains': subdomains_str}) if send_interesting and self.scan_id and self.domain_id: interesting_subdomains = get_interesting_subdomains(self.scan_id, self.domain_id) if interesting_subdomains: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in interesting_subdomains]) self.notify(fields={'Interesting subdomains': subdomains_str}) return SubdomainSerializer(subdomains, many=True).data @app.task(name='osint', queue='main_scan_queue', base=RengineTask, bind=True) def osint(self, host=None, ctx={}, description=None): """Run Open-Source Intelligence tools on selected domain. Args: host (str): Hostname to scan. Returns: dict: Results from osint discovery and dorking. """ config = self.yaml_configuration.get(OSINT) or OSINT_DEFAULT_CONFIG results = {} grouped_tasks = [] if 'discover' in config: ctx['track'] = False # results = osint_discovery(host=host, ctx=ctx) _task = osint_discovery.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, activity_id=self.activity_id, results_dir=self.results_dir, ctx=ctx ) grouped_tasks.append(_task) if OSINT_DORK in config or OSINT_CUSTOM_DORK in config: _task = dorking.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, results_dir=self.results_dir ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('OSINT Tasks finished...') # with open(self.output_path, 'w') as f: # json.dump(results, f, indent=4) # # return results @app.task(name='osint_discovery', queue='osint_discovery_queue', bind=False) def osint_discovery(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run OSINT discovery. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: dict: osint metadat and theHarvester and h8mail results. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) osint_lookup = config.get(OSINT_DISCOVER, []) osint_intensity = config.get(INTENSITY, 'normal') documents_limit = config.get(OSINT_DOCUMENTS_LIMIT, 50) results = {} meta_info = [] emails = [] creds = [] # Get and save meta info if 'metainfo' in osint_lookup: if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': host, 'domain': host, 'scan_id': scan_history_id, 'documents_limit': documents_limit }) meta_info.append(save_metadata_info(meta_dict)) # TODO: disabled for now # elif osint_intensity == 'deep': # subdomains = Subdomain.objects # if self.scan: # subdomains = subdomains.filter(scan_history=self.scan) # for subdomain in subdomains: # meta_dict = DottedDict({ # 'osint_target': subdomain.name, # 'domain': self.domain, # 'scan_id': self.scan_id, # 'documents_limit': documents_limit # }) # meta_info.append(save_metadata_info(meta_dict)) grouped_tasks = [] if 'emails' in osint_lookup: emails = get_and_save_emails(scan_history, activity_id, results_dir) emails_str = '\n'.join([f'• `{email}`' for email in emails]) # self.notify(fields={'Emails': emails_str}) # ctx['track'] = False _task = h8mail.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) if 'employees' in osint_lookup: ctx['track'] = False _task = theHarvester.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) # results['emails'] = results.get('emails', []) + emails # results['creds'] = creds # results['meta_info'] = meta_info return results @app.task(name='dorking', bind=False, queue='dorking_queue') def dorking(config, host, scan_history_id, results_dir): """Run Google dorks. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: list: Dorking results for each dork ran. """ # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh scan_history = ScanHistory.objects.get(pk=scan_history_id) dorks = config.get(OSINT_DORK, []) custom_dorks = config.get(OSINT_CUSTOM_DORK, []) results = [] # custom dorking has higher priority try: for custom_dork in custom_dorks: lookup_target = custom_dork.get('lookup_site') # replace with original host if _target_ lookup_target = host if lookup_target == '_target_' else lookup_target if 'lookup_extensions' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_extensions=custom_dork.get('lookup_extensions'), scan_history=scan_history ) elif 'lookup_keywords' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_keywords=custom_dork.get('lookup_keywords'), scan_history=scan_history ) except Exception as e: logger.exception(e) # default dorking try: for dork in dorks: logger.info(f'Getting dork information for {dork}') if dork == 'stackoverflow': results = get_and_save_dork_results( lookup_target='stackoverflow.com', results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'login_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/login/,login.html', page_count=5, scan_history=scan_history ) elif dork == 'admin_panels': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/admin/,admin.html', page_count=5, scan_history=scan_history ) elif dork == 'dashboard_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/dashboard/,dashboard.html', page_count=5, scan_history=scan_history ) elif dork == 'social_media' : social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'reddit.com' ] for site in social_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'project_management' : project_websites = [ 'trello.com', 'atlassian.net' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'code_sharing' : project_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'config_files' : config_file_exts = [ 'env', 'xml', 'conf', 'toml', 'yml', 'yaml', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(config_file_exts), page_count=4, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keyword = 'Jenkins' results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=lookup_keyword, page_count=1, scan_history=scan_history ) elif dork == 'wordpress_files' : lookup_keywords = [ '/wp-content/', '/wp-includes/' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'php_error' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'exposed_documents' : docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(docs_file_ext), page_count=7, scan_history=scan_history ) elif dork == 'db_files' : file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) elif dork == 'git_exposed' : file_ext = [ 'git', ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) except Exception as e: logger.exception(e) return results @app.task(name='theHarvester', queue='theHarvester_queue', bind=False) def theHarvester(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run theHarvester to get save emails, hosts, employees found in domain. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: dict: Dict of emails, employees, hosts and ips found during crawling. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) output_path_json = f'{results_dir}/theHarvester.json' theHarvester_dir = '/usr/src/github/theHarvester' history_file = f'{results_dir}/commands.txt' cmd = f'python3 {theHarvester_dir}/theHarvester.py -d {host} -b all -f {output_path_json}' # Update proxies.yaml proxy_query = Proxy.objects.all() if proxy_query.exists(): proxy = proxy_query.first() if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(f'{theHarvester_dir}/proxies.yaml', 'w') as file: yaml.dump(yaml_data, file) # Run cmd run_command( cmd, shell=False, cwd=theHarvester_dir, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) # Get file location if not os.path.isfile(output_path_json): logger.error(f'Could not open {output_path_json}') return {} # Load theHarvester results with open(output_path_json, 'r') as f: data = json.load(f) # Re-indent theHarvester JSON with open(output_path_json, 'w') as f: json.dump(data, f, indent=4) emails = data.get('emails', []) for email_address in emails: email, _ = save_email(email_address, scan_history=scan_history) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) linkedin_people = data.get('linkedin_people', []) for people in linkedin_people: employee, _ = save_employee( people, designation='linkedin', scan_history=scan_history) # if employee: # self.notify(fields={'LinkedIn people': f'• {employee.name}'}) twitter_people = data.get('twitter_people', []) for people in twitter_people: employee, _ = save_employee( people, designation='twitter', scan_history=scan_history) # if employee: # self.notify(fields={'Twitter people': f'• {employee.name}'}) hosts = data.get('hosts', []) urls = [] for host in hosts: split = tuple(host.split(':')) http_url = split[0] subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) endpoint, _ = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain) # if endpoint: # urls.append(endpoint.http_url) # self.notify(fields={'Hosts': f'• {endpoint.http_url}'}) # if enable_http_crawl: # ctx['track'] = False # http_crawl(urls, ctx=ctx) # TODO: Lots of ips unrelated with our domain are found, disabling # this for now. # ips = data.get('ips', []) # for ip_address in ips: # ip, created = save_ip_address( # ip_address, # subscan=subscan) # if ip: # send_task_notif.delay( # 'osint', # scan_history_id=scan_history_id, # subscan_id=subscan_id, # severity='success', # update_fields={'IPs': f'{ip.address}'}) return data @app.task(name='h8mail', queue='h8mail_queue', bind=False) def h8mail(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run h8mail. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: list[dict]: List of credentials info. """ logger.warning('Getting leaked credentials') scan_history = ScanHistory.objects.get(pk=scan_history_id) input_path = f'{results_dir}/emails.txt' output_file = f'{results_dir}/h8mail.json' cmd = f'h8mail -t {input_path} --json {output_file}' history_file = f'{results_dir}/commands.txt' run_command( cmd, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) with open(output_file) as f: data = json.load(f) creds = data.get('targets', []) # TODO: go through h8mail output and save emails to DB for cred in creds: logger.warning(cred) email_address = cred['target'] pwn_num = cred['pwn_num'] pwn_data = cred.get('data', []) email, created = save_email(email_address, scan_history=scan) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) return creds @app.task(name='screenshot', queue='main_scan_queue', base=RengineTask, bind=True) def screenshot(self, ctx={}, description=None): """Uses EyeWitness to gather screenshot of a domain and/or url. Args: description (str, optional): Task description shown in UI. """ # Config screenshots_path = f'{self.results_dir}/screenshots' output_path = f'{self.results_dir}/screenshots/{self.filename}' alive_endpoints_file = f'{self.results_dir}/endpoints_alive.txt' config = self.yaml_configuration.get(SCREENSHOT) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT + 5) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) # If intensity is normal, grab only the root endpoints of each subdomain strict = True if intensity == 'normal' else False # Get URLs to take screenshot of get_http_urls( is_alive=enable_http_crawl, strict=strict, write_filepath=alive_endpoints_file, get_only_default_urls=True, ctx=ctx ) # Send start notif notification = Notification.objects.first() send_output_file = notification.send_scan_output_file if notification else False # Run cmd cmd = f'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py -f {alive_endpoints_file} -d {screenshots_path} --no-prompt' cmd += f' --timeout {timeout}' if timeout > 0 else '' cmd += f' --threads {threads}' if threads > 0 else '' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(output_path): logger.error(f'Could not load EyeWitness results at {output_path} for {self.domain.name}.') return # Loop through results and save objects in DB screenshot_paths = [] with open(output_path, 'r') as file: reader = csv.reader(file) for row in reader: "Protocol,Port,Domain,Request Status,Screenshot Path, Source Path" protocol, port, subdomain_name, status, screenshot_path, source_path = tuple(row) logger.info(f'{protocol}:{port}:{subdomain_name}:{status}') subdomain_query = Subdomain.objects.filter(name=subdomain_name) if self.scan: subdomain_query = subdomain_query.filter(scan_history=self.scan) if status == 'Successful' and subdomain_query.exists(): subdomain = subdomain_query.first() screenshot_paths.append(screenshot_path) subdomain.screenshot_path = screenshot_path.replace('/usr/src/scan_results/', '') subdomain.save() logger.warning(f'Added screenshot for {subdomain.name} to DB') # Remove all db, html extra files in screenshot results run_command( 'rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format(screenshots_path), shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'rm -rf {screenshots_path}/source', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Send finish notifs screenshots_str = '• ' + '\n• '.join([f'`{path}`' for path in screenshot_paths]) self.notify(fields={'Screenshots': screenshots_str}) if send_output_file: for path in screenshot_paths: title = get_output_file_name( self.scan_id, self.subscan_id, self.filename) send_file_to_discord.delay(path, title) @app.task(name='port_scan', queue='main_scan_queue', base=RengineTask, bind=True) def port_scan(self, hosts=[], ctx={}, description=None): """Run port scan. Args: hosts (list, optional): Hosts to run port scan on. description (str, optional): Task description shown in UI. Returns: list: List of open ports (dict). """ input_file = f'{self.results_dir}/input_subdomains_port_scan.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(PORT_SCAN) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) exclude_ports = config.get(NAABU_EXCLUDE_PORTS, []) exclude_subdomains = config.get(NAABU_EXCLUDE_SUBDOMAINS, False) ports = config.get(PORTS, NAABU_DEFAULT_PORTS) ports = [str(port) for port in ports] rate_limit = config.get(NAABU_RATE) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) passive = config.get(NAABU_PASSIVE, False) use_naabu_config = config.get(USE_NAABU_CONFIG, False) exclude_ports_str = ','.join(return_iterable(exclude_ports)) # nmap args nmap_enabled = config.get(ENABLE_NMAP, False) nmap_cmd = config.get(NMAP_COMMAND, '') nmap_script = config.get(NMAP_SCRIPT, '') nmap_script = ','.join(return_iterable(nmap_script)) nmap_script_args = config.get(NMAP_SCRIPT_ARGS) if hosts: with open(input_file, 'w') as f: f.write('\n'.join(hosts)) else: hosts = get_subdomains( write_filepath=input_file, exclude_subdomains=exclude_subdomains, ctx=ctx) # Build cmd cmd = 'naabu -json -exclude-cdn' cmd += f' -list {input_file}' if len(hosts) > 0 else f' -host {hosts[0]}' if 'full' in ports or 'all' in ports: ports_str = ' -p "-"' elif 'top-100' in ports: ports_str = ' -top-ports 100' elif 'top-1000' in ports: ports_str = ' -top-ports 1000' else: ports_str = ','.join(ports) ports_str = f' -p {ports_str}' cmd += ports_str cmd += ' -config /root/.config/naabu/config.yaml' if use_naabu_config else '' cmd += f' -proxy "{proxy}"' if proxy else '' cmd += f' -c {threads}' if threads else '' cmd += f' -rate {rate_limit}' if rate_limit > 0 else '' cmd += f' -timeout {timeout*1000}' if timeout > 0 else '' cmd += f' -passive' if passive else '' cmd += f' -exclude-ports {exclude_ports_str}' if exclude_ports else '' cmd += f' -silent' # Execute cmd and gather results results = [] urls = [] ports_data = {} for line in stream_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) port_number = line['port'] ip_address = line['ip'] host = line.get('host') or ip_address if port_number == 0: continue # Grab subdomain subdomain = Subdomain.objects.filter( name=host, target_domain=self.domain, scan_history=self.scan ).first() # Add IP DB ip, _ = save_ip_address(ip_address, subdomain, subscan=self.subscan) if self.subscan: ip.ip_subscan_ids.add(self.subscan) ip.save() # Add endpoint to DB # port 80 and 443 not needed as http crawl already does that. if port_number not in [80, 443]: http_url = f'{host}:{port_number}' endpoint, _ = save_endpoint( http_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint: http_url = endpoint.http_url urls.append(http_url) # Add Port in DB port_details = whatportis.get_ports(str(port_number)) service_name = port_details[0].name if len(port_details) > 0 else 'unknown' description = port_details[0].description if len(port_details) > 0 else '' # get or create port port, created = Port.objects.get_or_create( number=port_number, service_name=service_name, description=description ) if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port.save() ip.ports.add(port) ip.save() if host in ports_data: ports_data[host].append(port_number) else: ports_data[host] = [port_number] # Send notification logger.warning(f'Found opened port {port_number} on {ip_address} ({host})') if len(ports_data) == 0: logger.info('Finished running naabu port scan - No open ports found.') if nmap_enabled: logger.info('Nmap scans skipped') return ports_data # Send notification fields_str = '' for host, ports in ports_data.items(): ports_str = ', '.join([f'`{port}`' for port in ports]) fields_str += f'• `{host}`: {ports_str}\n' self.notify(fields={'Ports discovered': fields_str}) # Save output to file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) logger.info('Finished running naabu port scan.') # Process nmap results: 1 process per host sigs = [] if nmap_enabled: logger.warning(f'Starting nmap scans ...') logger.warning(ports_data) for host, port_list in ports_data.items(): ports_str = '_'.join([str(p) for p in port_list]) ctx_nmap = ctx.copy() ctx_nmap['description'] = get_task_title(f'nmap_{host}', self.scan_id, self.subscan_id) ctx_nmap['track'] = False sig = nmap.si( cmd=nmap_cmd, ports=port_list, host=host, script=nmap_script, script_args=nmap_script_args, max_rate=rate_limit, ctx=ctx_nmap) sigs.append(sig) task = group(sigs).apply_async() with allow_join_result(): results = task.get() return ports_data @app.task(name='nmap', queue='main_scan_queue', base=RengineTask, bind=True) def nmap( self, cmd=None, ports=[], host=None, input_file=None, script=None, script_args=None, max_rate=None, ctx={}, description=None): """Run nmap on a host. Args: cmd (str, optional): Existing nmap command to complete. ports (list, optional): List of ports to scan. host (str, optional): Host to scan. input_file (str, optional): Input hosts file. script (str, optional): NSE script to run. script_args (str, optional): NSE script args. max_rate (int): Max rate. description (str, optional): Task description shown in UI. """ notif = Notification.objects.first() ports_str = ','.join(str(port) for port in ports) self.filename = self.filename.replace('.txt', '.xml') filename_vulns = self.filename.replace('.xml', '_vulns.json') output_file = self.output_path output_file_xml = f'{self.results_dir}/{host}_{self.filename}' vulns_file = f'{self.results_dir}/{host}_{filename_vulns}' logger.warning(f'Running nmap on {host}:{ports}') # Build cmd nmap_cmd = get_nmap_cmd( cmd=cmd, ports=ports_str, script=script, script_args=script_args, max_rate=max_rate, host=host, input_file=input_file, output_file=output_file_xml) # Run cmd run_command( nmap_cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Get nmap XML results and convert to JSON vulns = parse_nmap_results(output_file_xml, output_file) with open(vulns_file, 'w') as f: json.dump(vulns, f, indent=4) # Save vulnerabilities found by nmap vulns_str = '' for vuln_data in vulns: # URL is not necessarily an HTTP URL when running nmap (can be any # other vulnerable protocols). Look for existing endpoint and use its # URL as vulnerability.http_url if it exists. url = vuln_data['http_url'] endpoint = EndPoint.objects.filter(http_url__contains=url).first() if endpoint: vuln_data['http_url'] = endpoint.http_url vuln, created = save_vulnerability( target_domain=self.domain, subdomain=self.subdomain, scan_history=self.scan, subscan=self.subscan, endpoint=endpoint, **vuln_data) vulns_str += f'• {str(vuln)}\n' if created: logger.warning(str(vuln)) # Send only 1 notif for all vulns to reduce number of notifs if notif and notif.send_vuln_notif and vulns_str: logger.warning(vulns_str) self.notify(fields={'CVEs': vulns_str}) return vulns @app.task(name='waf_detection', queue='main_scan_queue', base=RengineTask, bind=True) def waf_detection(self, ctx={}, description=None): """ Uses wafw00f to check for the presence of a WAF. Args: description (str, optional): Task description shown in UI. Returns: list: List of startScan.models.Waf objects. """ input_path = f'{self.results_dir}/input_endpoints_waf_detection.txt' config = self.yaml_configuration.get(WAF_DETECTION) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Get alive endpoints from DB get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) cmd = f'wafw00f -i {input_path} -o {self.output_path}' run_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(self.output_path): logger.error(f'Could not find {self.output_path}') return with open(self.output_path) as file: wafs = file.readlines() for line in wafs: line = " ".join(line.split()) splitted = line.split(' ', 1) waf_info = splitted[1].strip() waf_name = waf_info[:waf_info.find('(')].strip() waf_manufacturer = waf_info[waf_info.find('(')+1:waf_info.find(')')].strip().replace('.', '') http_url = sanitize_url(splitted[0].strip()) if not waf_name or waf_name == 'None': continue # Add waf to db waf, _ = Waf.objects.get_or_create( name=waf_name, manufacturer=waf_manufacturer ) # Add waf info to Subdomain in DB subdomain = get_subdomain_from_url(http_url) logger.info(f'Wafw00f Subdomain : {subdomain}') subdomain_query, _ = Subdomain.objects.get_or_create(scan_history=self.scan, name=subdomain) subdomain_query.waf.add(waf) subdomain_query.save() return wafs @app.task(name='dir_file_fuzz', queue='main_scan_queue', base=RengineTask, bind=True) def dir_file_fuzz(self, ctx={}, description=None): """Perform directory scan, and currently uses `ffuf` as a default tool. Args: description (str, optional): Task description shown in UI. Returns: list: List of URLs discovered. """ # Config cmd = 'ffuf' config = self.yaml_configuration.get(DIR_FILE_FUZZ) or {} custom_header = self.yaml_configuration.get(CUSTOM_HEADER) auto_calibration = config.get(AUTO_CALIBRATION, True) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) extensions = config.get(EXTENSIONS, DEFAULT_DIR_FILE_FUZZ_EXTENSIONS) # prepend . on extensions extensions = [ext if ext.startswith('.') else '.' + ext for ext in extensions] extensions_str = ','.join(map(str, extensions)) follow_redirect = config.get(FOLLOW_REDIRECT, FFUF_DEFAULT_FOLLOW_REDIRECT) max_time = config.get(MAX_TIME, 0) match_http_status = config.get(MATCH_HTTP_STATUS, FFUF_DEFAULT_MATCH_HTTP_STATUS) mc = ','.join([str(c) for c in match_http_status]) recursive_level = config.get(RECURSIVE_LEVEL, FFUF_DEFAULT_RECURSIVE_LEVEL) stop_on_error = config.get(STOP_ON_ERROR, False) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) wordlist_name = config.get(WORDLIST, 'dicc') delay = rate_limit / (threads * 100) # calculate request pause delay from rate_limit and number of threads input_path = f'{self.results_dir}/input_dir_file_fuzz.txt' # Get wordlist wordlist_name = 'dicc' if wordlist_name == 'default' else wordlist_name wordlist_path = f'/usr/src/wordlist/{wordlist_name}.txt' # Build command cmd += f' -w {wordlist_path}' cmd += f' -e {extensions_str}' if extensions else '' cmd += f' -maxtime {max_time}' if max_time > 0 else '' cmd += f' -p {delay}' if delay > 0 else '' cmd += f' -recursion -recursion-depth {recursive_level} ' if recursive_level > 0 else '' cmd += f' -t {threads}' if threads and threads > 0 else '' cmd += f' -timeout {timeout}' if timeout and timeout > 0 else '' cmd += ' -se' if stop_on_error else '' cmd += ' -fr' if follow_redirect else '' cmd += ' -ac' if auto_calibration else '' cmd += f' -mc {mc}' if mc else '' cmd += f' -H "{custom_header}"' if custom_header else '' # Grab URLs to fuzz urls = get_http_urls( is_alive=True, ignore_files=False, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) logger.warning(urls) # Loop through URLs and run command results = [] for url in urls: ''' Above while fetching urls, we are not ignoring files, because some default urls may redirect to https://example.com/login.php so, ignore_files is set to False but, during fuzzing, we will only need part of the path, in above example it is still a good idea to ffuf base url https://example.com so files from base url ''' url_parse = urlparse(url) url = url_parse.scheme + '://' + url_parse.netloc url += '/FUZZ' # TODO: fuzz not only URL but also POST / PUT / headers proxy = get_random_proxy() # Build final cmd fcmd = cmd fcmd += f' -x {proxy}' if proxy else '' fcmd += f' -u {url} -json' # Initialize DirectoryScan object dirscan = DirectoryScan() dirscan.scanned_date = timezone.now() dirscan.command_line = fcmd dirscan.save() # Loop through results and populate EndPoint and DirectoryFile in DB results = [] for line in stream_command( fcmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) name = line['input'].get('FUZZ') length = line['length'] status = line['status'] words = line['words'] url = line['url'] lines = line['lines'] content_type = line['content-type'] duration = line['duration'] if not name: logger.error(f'FUZZ not found for "{url}"') continue endpoint, created = save_endpoint(url, crawl=False, ctx=ctx) endpoint.http_status = status endpoint.content_length = length endpoint.response_time = duration / 1000000000 endpoint.save() if created: urls.append(endpoint.http_url) endpoint.status = status endpoint.content_type = content_type endpoint.content_length = length dfile, created = DirectoryFile.objects.get_or_create( name=name, length=length, words=words, lines=lines, content_type=content_type, url=url) dfile.http_status = status dfile.save() # if created: # logger.warning(f'Found new directory or file {url}') dirscan.directory_files.add(dfile) dirscan.save() if self.subscan: dirscan.dir_subscan_ids.add(self.subscan) subdomain_name = get_subdomain_from_url(endpoint.http_url) subdomain = Subdomain.objects.get(name=subdomain_name, scan_history=self.scan) subdomain.directories.add(dirscan) subdomain.save() # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl(urls, ctx=ctx) return results @app.task(name='fetch_url', queue='main_scan_queue', base=RengineTask, bind=True) def fetch_url(self, urls=[], ctx={}, description=None): """Fetch URLs using different tools like gauplus, gau, gospider, waybackurls ... Args: urls (list): List of URLs to start from. description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/input_endpoints_fetch_url.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(FETCH_URL) or {} should_remove_duplicate_endpoints = config.get(REMOVE_DUPLICATE_ENDPOINTS, True) duplicate_removal_fields = config.get(DUPLICATE_REMOVAL_FIELDS, ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, DEFAULT_GF_PATTERNS) ignore_file_extension = config.get(IGNORE_FILE_EXTENSION, DEFAULT_IGNORE_FILE_EXTENSIONS) tools = config.get(USES_TOOLS, ENDPOINT_SCAN_DEFAULT_TOOLS) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) domain_request_headers = self.domain.request_headers if self.domain else None custom_header = domain_request_headers or self.yaml_configuration.get(CUSTOM_HEADER) exclude_subdomains = config.get(EXCLUDED_SUBDOMAINS, False) # Get URLs to scan and save to input file if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, exclude_subdomains=exclude_subdomains, get_only_default_urls=True, ctx=ctx ) # Domain regex host = self.domain.name if self.domain else urlparse(urls[0]).netloc host_regex = f"\'https?://([a-z0-9]+[.])*{host}.*\'" # Tools cmds cmd_map = { 'gau': f'gau', 'gauplus': f'gauplus -random-agent', 'hakrawler': 'hakrawler -subs -u', 'waybackurls': 'waybackurls', 'gospider': f'gospider -S {input_path} --js -d 2 --sitemap --robots -w -r', 'katana': f'katana -list {input_path} -silent -jc -kf all -d 3 -fs rdn', } if proxy: cmd_map['gau'] += f' --proxy "{proxy}"' cmd_map['gauplus'] += f' -p "{proxy}"' cmd_map['gospider'] += f' -p {proxy}' cmd_map['hakrawler'] += f' -proxy {proxy}' cmd_map['katana'] += f' -proxy {proxy}' if threads > 0: cmd_map['gau'] += f' --threads {threads}' cmd_map['gauplus'] += f' -t {threads}' cmd_map['gospider'] += f' -t {threads}' cmd_map['katana'] += f' -c {threads}' if custom_header: header_string = ';;'.join([ f'{key}: {value}' for key, value in custom_header.items() ]) cmd_map['hakrawler'] += f' -h {header_string}' cmd_map['katana'] += f' -H {header_string}' header_flags = [':'.join(h) for h in header_string.split(';;')] for flag in header_flags: cmd_map['gospider'] += f' -H {flag}' cat_input = f'cat {input_path}' grep_output = f'grep -Eo {host_regex}' cmd_map = { tool: f'{cat_input} | {cmd} | {grep_output} > {self.results_dir}/urls_{tool}.txt' for tool, cmd in cmd_map.items() } tasks = group( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for tool, cmd in cmd_map.items() if tool in tools ) # Cleanup task sort_output = [ f'cat {self.results_dir}/urls_* > {self.output_path}', f'cat {input_path} >> {self.output_path}', f'sort -u {self.output_path} -o {self.output_path}', ] if ignore_file_extension: ignore_exts = '|'.join(ignore_file_extension) grep_ext_filtered_output = [ f'cat {self.output_path} | grep -Eiv "\\.({ignore_exts}).*" > {self.results_dir}/urls_filtered.txt', f'mv {self.results_dir}/urls_filtered.txt {self.output_path}' ] sort_output.extend(grep_ext_filtered_output) cleanup = chain( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for cmd in sort_output ) # Run all commands task = chord(tasks)(cleanup) with allow_join_result(): task.get() # Store all the endpoints and run httpx with open(self.output_path) as f: discovered_urls = f.readlines() self.notify(fields={'Discovered URLs': len(discovered_urls)}) # Some tools can have an URL in the format <URL>] - <PATH> or <URL> - <PATH>, add them # to the final URL list all_urls = [] for url in discovered_urls: url = url.strip() urlpath = None base_url = None if '] ' in url: # found JS scraped endpoint e.g from gospider split = tuple(url.split('] ')) if not len(split) == 2: logger.warning(f'URL format not recognized for "{url}". Skipping.') continue base_url, urlpath = split urlpath = urlpath.lstrip('- ') elif ' - ' in url: # found JS scraped endpoint e.g from gospider base_url, urlpath = tuple(url.split(' - ')) if base_url and urlpath: subdomain = urlparse(base_url) url = f'{subdomain.scheme}://{subdomain.netloc}{self.url_filter}' if not validators.url(url): logger.warning(f'Invalid URL "{url}". Skipping.') if url not in all_urls: all_urls.append(url) # Filter out URLs if a path filter was passed if self.url_filter: all_urls = [url for url in all_urls if self.url_filter in url] # Write result to output path with open(self.output_path, 'w') as f: f.write('\n'.join(all_urls)) logger.warning(f'Found {len(all_urls)} usable URLs') # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl( all_urls, ctx=ctx, should_remove_duplicate_endpoints=should_remove_duplicate_endpoints, duplicate_removal_fields=duplicate_removal_fields ) #-------------------# # GF PATTERNS MATCH # #-------------------# # Combine old gf patterns with new ones if gf_patterns: self.scan.used_gf_patterns = ','.join(gf_patterns) self.scan.save() # Run gf patterns on saved endpoints # TODO: refactor to Celery task for gf_pattern in gf_patterns: # TODO: js var is causing issues, removing for now if gf_pattern == 'jsvar': logger.info('Ignoring jsvar as it is causing issues.') continue # Run gf on current pattern logger.warning(f'Running gf on pattern "{gf_pattern}"') gf_output_file = f'{self.results_dir}/gf_patterns_{gf_pattern}.txt' cmd = f'cat {self.output_path} | gf {gf_pattern} | grep -Eo {host_regex} >> {gf_output_file}' run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Check output file if not os.path.exists(gf_output_file): logger.error(f'Could not find GF output file {gf_output_file}. Skipping GF pattern "{gf_pattern}"') continue # Read output file line by line and with open(gf_output_file, 'r') as f: lines = f.readlines() # Add endpoints / subdomains to DB for url in lines: http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue endpoint, created = save_endpoint( http_url, crawl=False, subdomain=subdomain, ctx=ctx) if not endpoint: continue earlier_pattern = None if not created: earlier_pattern = endpoint.matched_gf_patterns pattern = f'{earlier_pattern},{gf_pattern}' if earlier_pattern else gf_pattern endpoint.matched_gf_patterns = pattern endpoint.save() return all_urls def parse_curl_output(response): # TODO: Enrich from other cURL fields. CURL_REGEX_HTTP_STATUS = f'HTTP\/(?:(?:\d\.?)+)\s(\d+)\s(?:\w+)' http_status = 0 if response: failed = False regex = re.compile(CURL_REGEX_HTTP_STATUS, re.MULTILINE) try: http_status = int(regex.findall(response)[0]) except (KeyError, TypeError, IndexError): pass return { 'http_status': http_status, } @app.task(name='vulnerability_scan', queue='main_scan_queue', bind=True, base=RengineTask) def vulnerability_scan(self, urls=[], ctx={}, description=None): """ This function will serve as an entrypoint to vulnerability scan. All other vulnerability scan will be run from here including nuclei, crlfuzz, etc """ logger.info('Running Vulnerability Scan Queue') config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_run_nuclei = config.get(RUN_NUCLEI, True) should_run_crlfuzz = config.get(RUN_CRLFUZZ, False) should_run_dalfox = config.get(RUN_DALFOX, False) should_run_s3scanner = config.get(RUN_S3SCANNER, True) grouped_tasks = [] if should_run_nuclei: _task = nuclei_scan.si( urls=urls, ctx=ctx, description=f'Nuclei Scan' ) grouped_tasks.append(_task) if should_run_crlfuzz: _task = crlfuzz_scan.si( urls=urls, ctx=ctx, description=f'CRLFuzz Scan' ) grouped_tasks.append(_task) if should_run_dalfox: _task = dalfox_xss_scan.si( urls=urls, ctx=ctx, description=f'Dalfox XSS Scan' ) grouped_tasks.append(_task) if should_run_s3scanner: _task = s3scanner.si( ctx=ctx, description=f'Misconfigured S3 Buckets Scanner' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan completed...') # return results return None @app.task(name='nuclei_individual_severity_module', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_individual_severity_module(self, cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx={}, description=None): ''' This celery task will run vulnerability scan in parallel. All severities supplied should run in parallel as grouped tasks. ''' results = [] logger.info(f'Running vulnerability scan with severity: {severity}') cmd += f' -severity {severity}' # Send start notification notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) # Gather nuclei results vuln_data = parse_nuclei_result(line) # Get corresponding subdomain http_url = sanitize_url(line.get('matched-at')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) # Look for duplicate vulnerabilities by excluding records that might change but are irrelevant. object_comparison_exclude = ['response', 'curl_command', 'tags', 'references', 'cve_ids', 'cwe_ids'] # Add subdomain and target domain to the duplicate check vuln_data_copy = vuln_data.copy() vuln_data_copy['subdomain'] = subdomain vuln_data_copy['target_domain'] = self.domain # Check if record exists, if exists do not save it if record_exists(Vulnerability, data=vuln_data_copy, exclude_keys=object_comparison_exclude): logger.warning(f'Nuclei vulnerability of severity {severity} : {vuln_data_copy["name"]} for {subdomain_name} already exists') continue # Get or create EndPoint object response = line.get('response') httpx_crawl = False if response else enable_http_crawl # avoid yet another httpx crawl endpoint, _ = save_endpoint( http_url, crawl=httpx_crawl, subdomain=subdomain, ctx=ctx) if endpoint: http_url = endpoint.http_url if not httpx_crawl: output = parse_curl_output(response) endpoint.http_status = output['http_status'] endpoint.save() # Get or create Vulnerability object vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, subdomain=subdomain, **vuln_data) if not vuln: continue # Print vuln severity = line['info'].get('severity', 'unknown') logger.warning(str(vuln)) # Send notification for all vulnerabilities except info url = vuln.http_url or vuln.subdomain send_vuln = ( notif and notif.send_vuln_notif and vuln and severity in ['low', 'medium', 'high', 'critical']) if send_vuln: fields = { 'Severity': f'**{severity.upper()}**', 'URL': http_url, 'Subdomain': subdomain_name, 'Name': vuln.name, 'Type': vuln.type, 'Description': vuln.description, 'Template': vuln.template_url, 'Tags': vuln.get_tags_str(), 'CVEs': vuln.get_cve_str(), 'CWEs': vuln.get_cwe_str(), 'References': vuln.get_refs_str() } severity_map = { 'low': 'info', 'medium': 'warning', 'high': 'error', 'critical': 'error' } self.notify( f'vulnerability_scan_#{vuln.id}', severity_map[severity], fields, add_meta_info=False) # Send report to hackerone hackerone_query = Hackerone.objects.all() send_report = ( hackerone_query.exists() and severity not in ('info', 'low') and vuln.target_domain.h1_team_handle ) if send_report: hackerone = hackerone_query.first() if hackerone.send_critical and severity == 'critical': send_hackerone_report.delay(vuln.id) elif hackerone.send_high and severity == 'high': send_hackerone_report.delay(vuln.id) elif hackerone.send_medium and severity == 'medium': send_hackerone_report.delay(vuln.id) # Write results to JSON file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) # Send finish notif if send_status: vulns = Vulnerability.objects.filter(scan_history__id=self.scan_id) info_count = vulns.filter(severity=0).count() low_count = vulns.filter(severity=1).count() medium_count = vulns.filter(severity=2).count() high_count = vulns.filter(severity=3).count() critical_count = vulns.filter(severity=4).count() unknown_count = vulns.filter(severity=-1).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count + unknown_count fields = { 'Total': vulnerability_count, 'Critical': critical_count, 'High': high_count, 'Medium': medium_count, 'Low': low_count, 'Info': info_count, 'Unknown': unknown_count } self.notify(fields=fields) # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=NUCLEI ).exclude( severity=0 ) # find all unique vulnerabilities based on path and title # all unique vulnerability will go thru gpt function and get report # once report is got, it will be matched with other vulnerabilities and saved unique_vulns = set() for vuln in vulns: unique_vulns.add((vuln.name, vuln.get_path())) unique_vulns = list(unique_vulns) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in unique_vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return None def get_vulnerability_gpt_report(vuln): title = vuln[0] path = vuln[1] logger.info(f'Getting GPT Report for {title}, PATH: {path}') # check if in db already exists stored = GPTVulnerabilityReport.objects.filter( url_path=path ).filter( title=title ).first() if stored: response = { 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: report = GPTVulnerabilityReportGenerator() vulnerability_description = get_gpt_vuln_input_description( title, path ) response = report.get_vulnerability_description(vulnerability_description) add_gpt_description_db( title, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) for vuln in Vulnerability.objects.filter(name=title, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() def add_gpt_description_db(title, path, description, impact, remediation, references): gpt_report = GPTVulnerabilityReport() gpt_report.url_path = path gpt_report.title = title gpt_report.description = description gpt_report.impact = impact gpt_report.remediation = remediation gpt_report.save() for url in references: ref, created = VulnerabilityReference.objects.get_or_create(url=url) gpt_report.references.add(ref) gpt_report.save() @app.task(name='nuclei_scan', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_scan(self, urls=[], ctx={}, description=None): """HTTP vulnerability scan using Nuclei Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. Notes: Unfurl the urls to keep only domain and path, will be sent to vuln scan and ignore certain file extensions. Thanks: https://github.com/six2dez/reconftw """ # Config config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} input_path = f'{self.results_dir}/input_endpoints_vulnerability_scan.txt' enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) concurrency = config.get(NUCLEI_CONCURRENCY) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) retries = config.get(RETRIES) or self.yaml_configuration.get(RETRIES, DEFAULT_RETRIES) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) custom_header = config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) should_fetch_gpt_report = config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) proxy = get_random_proxy() nuclei_specific_config = config.get('nuclei', {}) use_nuclei_conf = nuclei_specific_config.get(USE_CONFIG, False) severities = nuclei_specific_config.get(NUCLEI_SEVERITY, NUCLEI_DEFAULT_SEVERITIES) tags = nuclei_specific_config.get(NUCLEI_TAGS, []) tags = ','.join(tags) nuclei_templates = nuclei_specific_config.get(NUCLEI_TEMPLATE) custom_nuclei_templates = nuclei_specific_config.get(NUCLEI_CUSTOM_TEMPLATE) # severities_str = ','.join(severities) # Get alive endpoints if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=enable_http_crawl, ignore_files=True, write_filepath=input_path, ctx=ctx ) if intensity == 'normal': # reduce number of endpoints to scan unfurl_filter = f'{self.results_dir}/urls_unfurled.txt' run_command( f"cat {input_path} | unfurl -u format %s://%d%p |uro > {unfurl_filter}", shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {unfurl_filter} -o {unfurl_filter}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) input_path = unfurl_filter # Build templates # logger.info('Updating Nuclei templates ...') run_command( 'nuclei -update-templates', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) templates = [] if not (nuclei_templates or custom_nuclei_templates): templates.append(NUCLEI_DEFAULT_TEMPLATES_PATH) if nuclei_templates: if ALL in nuclei_templates: template = NUCLEI_DEFAULT_TEMPLATES_PATH templates.append(template) else: templates.extend(nuclei_templates) if custom_nuclei_templates: custom_nuclei_template_paths = [f'{str(elem)}.yaml' for elem in custom_nuclei_templates] template = templates.extend(custom_nuclei_template_paths) # Build CMD cmd = 'nuclei -j' cmd += ' -config /root/.config/nuclei/config.yaml' if use_nuclei_conf else '' cmd += f' -irr' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -l {input_path}' cmd += f' -c {str(concurrency)}' if concurrency > 0 else '' cmd += f' -proxy {proxy} ' if proxy else '' cmd += f' -retries {retries}' if retries > 0 else '' cmd += f' -rl {rate_limit}' if rate_limit > 0 else '' # cmd += f' -severity {severities_str}' cmd += f' -timeout {str(timeout)}' if timeout and timeout > 0 else '' cmd += f' -tags {tags}' if tags else '' cmd += f' -silent' for tpl in templates: cmd += f' -t {tpl}' grouped_tasks = [] custom_ctx = ctx for severity in severities: custom_ctx['track'] = True _task = nuclei_individual_severity_module.si( cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx=custom_ctx, description=f'Nuclei Scan with severity {severity}' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan with all severities completed...') return None @app.task(name='dalfox_xss_scan', queue='main_scan_queue', base=RengineTask, bind=True) def dalfox_xss_scan(self, urls=[], ctx={}, description=None): """XSS Scan using dalfox Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) dalfox_config = vuln_config.get(DALFOX) or {} custom_header = dalfox_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() is_waf_evasion = dalfox_config.get(WAF_EVASION, False) blind_xss_server = dalfox_config.get(BLIND_XSS_SERVER) user_agent = dalfox_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) timeout = dalfox_config.get(TIMEOUT) delay = dalfox_config.get(DELAY) threads = dalfox_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_dalfox_xss.txt' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=False, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'dalfox --silence --no-color --no-spinner' cmd += f' --only-poc r ' cmd += f' --ignore-return 302,404,403' cmd += f' --skip-bav' cmd += f' file {input_path}' cmd += f' --proxy {proxy}' if proxy else '' cmd += f' --waf-evasion' if is_waf_evasion else '' cmd += f' -b {blind_xss_server}' if blind_xss_server else '' cmd += f' --delay {delay}' if delay else '' cmd += f' --timeout {timeout}' if timeout else '' cmd += f' --user-agent {user_agent}' if user_agent else '' cmd += f' --header {custom_header}' if custom_header else '' cmd += f' --worker {threads}' if threads else '' cmd += f' --format json' results = [] for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id, trunc_char=',' ): if not isinstance(line, dict): continue results.append(line) vuln_data = parse_dalfox_result(line) http_url = sanitize_url(line.get('data')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Dalfox Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=DALFOX ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='crlfuzz_scan', queue='main_scan_queue', base=RengineTask, bind=True) def crlfuzz_scan(self, urls=[], ctx={}, description=None): """CRLF Fuzzing with CRLFuzz Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) custom_header = vuln_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() user_agent = vuln_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) threads = vuln_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_crlf.txt' output_path = f'{self.results_dir}/{self.filename}' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=True, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'crlfuzz -s' cmd += f' -l {input_path}' cmd += f' -x {proxy}' if proxy else '' cmd += f' --H {custom_header}' if custom_header else '' cmd += f' -o {output_path}' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id ) if not os.path.isfile(output_path): logger.info('No Results from CRLFuzz') return crlfs = [] results = [] with open(output_path, 'r') as file: crlfs = file.readlines() for crlf in crlfs: url = crlf.strip() vuln_data = parse_crlfuzz_result(url) http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting CRLFuzz Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=CRLFUZZ ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='s3scanner', queue='main_scan_queue', base=RengineTask, bind=True) def s3scanner(self, ctx={}, description=None): """Bucket Scanner Args: ctx (dict): Context description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/#{self.scan_id}_subdomain_discovery.txt' vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} s3_config = vuln_config.get(S3SCANNER) or {} threads = s3_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) providers = s3_config.get(PROVIDERS, S3SCANNER_DEFAULT_PROVIDERS) scan_history = ScanHistory.objects.filter(pk=self.scan_id).first() for provider in providers: cmd = f's3scanner -bucket-file {input_path} -enumerate -provider {provider} -threads {threads} -json' for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue if line.get('bucket', {}).get('exists', 0) == 1: result = parse_s3scanner_result(line) s3bucket, created = S3Bucket.objects.get_or_create(**result) scan_history.buckets.add(s3bucket) logger.info(f"s3 bucket added {result['provider']}-{result['name']}-{result['region']}") @app.task(name='http_crawl', queue='main_scan_queue', base=RengineTask, bind=True) def http_crawl( self, urls=[], method=None, recrawl=False, ctx={}, track=True, description=None, is_ran_from_subdomain_scan=False, should_remove_duplicate_endpoints=True, duplicate_removal_fields=[]): """Use httpx to query HTTP URLs for important info like page titles, http status, etc... Args: urls (list, optional): A set of URLs to check. Overrides default behavior which queries all endpoints related to this scan. method (str): HTTP method to use (GET, HEAD, POST, PUT, DELETE). recrawl (bool, optional): If False, filter out URLs that have already been crawled. should_remove_duplicate_endpoints (bool): Whether to remove duplicate endpoints duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates Returns: list: httpx results. """ logger.info('Initiating HTTP Crawl') if is_ran_from_subdomain_scan: logger.info('Running From Subdomain Scan...') cmd = '/go/bin/httpx' cfg = self.yaml_configuration.get(HTTP_CRAWL) or {} custom_header = cfg.get(CUSTOM_HEADER, '') threads = cfg.get(THREADS, DEFAULT_THREADS) follow_redirect = cfg.get(FOLLOW_REDIRECT, True) self.output_path = None input_path = f'{self.results_dir}/httpx_input.txt' history_file = f'{self.results_dir}/commands.txt' if urls: # direct passing URLs to check if self.url_filter: urls = [u for u in urls if self.url_filter in u] with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_uncrawled=not recrawl, write_filepath=input_path, ctx=ctx ) # logger.debug(urls) # If no URLs found, skip it if not urls: return # Re-adjust thread number if few URLs to avoid spinning up a monster to # kill a fly. if len(urls) < threads: threads = len(urls) # Get random proxy proxy = get_random_proxy() # Run command cmd += f' -cl -ct -rt -location -td -websocket -cname -asn -cdn -probe -random-agent' cmd += f' -t {threads}' if threads > 0 else '' cmd += f' --http-proxy {proxy}' if proxy else '' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -json' cmd += f' -u {urls[0]}' if len(urls) == 1 else f' -l {input_path}' cmd += f' -x {method}' if method else '' cmd += f' -silent' if follow_redirect: cmd += ' -fr' results = [] endpoint_ids = [] for line in stream_command( cmd, history_file=history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not line or not isinstance(line, dict): continue logger.debug(line) # No response from endpoint if line.get('failed', False): continue # Parse httpx output host = line.get('host', '') content_length = line.get('content_length', 0) http_status = line.get('status_code') http_url, is_redirect = extract_httpx_url(line) page_title = line.get('title') webserver = line.get('webserver') cdn = line.get('cdn', False) rt = line.get('time') techs = line.get('tech', []) cname = line.get('cname', '') content_type = line.get('content_type', '') response_time = -1 if rt: response_time = float(''.join(ch for ch in rt if not ch.isalpha())) if rt[-2:] == 'ms': response_time = response_time / 1000 # Create Subdomain object in DB subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue # Save default HTTP URL to endpoint object in DB endpoint, created = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain, is_default=is_ran_from_subdomain_scan ) if not endpoint: continue endpoint.http_status = http_status endpoint.page_title = page_title endpoint.content_length = content_length endpoint.webserver = webserver endpoint.response_time = response_time endpoint.content_type = content_type endpoint.save() endpoint_str = f'{http_url} [{http_status}] `{content_length}B` `{webserver}` `{rt}`' logger.warning(endpoint_str) if endpoint and endpoint.is_alive and endpoint.http_status != 403: self.notify( fields={'Alive endpoint': f'• {endpoint_str}'}, add_meta_info=False) # Add endpoint to results line['_cmd'] = cmd line['final_url'] = http_url line['endpoint_id'] = endpoint.id line['endpoint_created'] = created line['is_redirect'] = is_redirect results.append(line) # Add technology objects to DB for technology in techs: tech, _ = Technology.objects.get_or_create(name=technology) endpoint.techs.add(tech) if is_ran_from_subdomain_scan: subdomain.technologies.add(tech) subdomain.save() endpoint.save() techs_str = ', '.join([f'`{tech}`' for tech in techs]) self.notify( fields={'Technologies': techs_str}, add_meta_info=False) # Add IP objects for 'a' records to DB a_records = line.get('a', []) for ip_address in a_records: ip, created = save_ip_address( ip_address, subdomain, subscan=self.subscan, cdn=cdn) ips_str = '• ' + '\n• '.join([f'`{ip}`' for ip in a_records]) self.notify( fields={'IPs': ips_str}, add_meta_info=False) # Add IP object for host in DB if host: ip, created = save_ip_address( host, subdomain, subscan=self.subscan, cdn=cdn) self.notify( fields={'IPs': f'• `{ip.address}`'}, add_meta_info=False) # Save subdomain and endpoint if is_ran_from_subdomain_scan: # save subdomain stuffs subdomain.http_url = http_url subdomain.http_status = http_status subdomain.page_title = page_title subdomain.content_length = content_length subdomain.webserver = webserver subdomain.response_time = response_time subdomain.content_type = content_type subdomain.cname = ','.join(cname) subdomain.is_cdn = cdn if cdn: subdomain.cdn_name = line.get('cdn_name') subdomain.save() endpoint.save() endpoint_ids.append(endpoint.id) if should_remove_duplicate_endpoints: # Remove 'fake' alive endpoints that are just redirects to the same page remove_duplicate_endpoints( self.scan_id, self.domain_id, self.subdomain_id, filter_ids=endpoint_ids ) # Remove input file run_command( f'rm {input_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) return results #---------------------# # Notifications tasks # #---------------------# @app.task(name='send_notif', bind=False, queue='send_notif_queue') def send_notif( message, scan_history_id=None, subscan_id=None, **options): if not 'title' in options: message = enrich_notification(message, scan_history_id, subscan_id) send_discord_message(message, **options) send_slack_message(message) send_telegram_message(message) @app.task(name='send_scan_notif', bind=False, queue='send_scan_notif_queue') def send_scan_notif( scan_history_id, subscan_id=None, engine_id=None, status='RUNNING'): """Send scan status notification. Works for scan or a subscan if subscan_id is passed. Args: scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Get domain, engine, scan_history objects engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() tasks = ScanActivity.objects.filter(scan_of=scan) if scan else 0 # Build notif options url = get_scan_url(scan_history_id, subscan_id) title = get_scan_title(scan_history_id, subscan_id) fields = get_scan_fields(engine, scan, subscan, status, tasks) severity = None msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) if status: severity = STATUS_TO_SEVERITIES.get(status) opts = { 'title': title, 'url': url, 'fields': fields, 'severity': severity } logger.warning(f'Sending notification "{title}" [{severity}]') # Send notification send_notif( msg, scan_history_id, subscan_id, **opts) @app.task(name='send_task_notif', bind=False, queue='send_task_notif_queue') def send_task_notif( task_name, status=None, result=None, output_path=None, traceback=None, scan_history_id=None, engine_id=None, subscan_id=None, severity=None, add_meta_info=True, update_fields={}): """Send task status notification. Args: task_name (str): Task name. status (str, optional): Task status. result (str, optional): Task result. output_path (str, optional): Task output path. traceback (str, optional): Task traceback. scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. severity (str, optional): Severity (will be mapped to notif colors) add_meta_info (bool, optional): Wheter to add scan / subscan info to notif. update_fields (dict, optional): Fields key / value to update. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Build fields url = None fields = {} if add_meta_info: engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() url = get_scan_url(scan_history_id) if status: fields['Status'] = f'**{status}**' if engine: fields['Engine'] = engine.engine_name if scan: fields['Scan ID'] = f'[#{scan.id}]({url})' if subscan: url = get_scan_url(scan_history_id, subscan_id) fields['Subscan ID'] = f'[#{subscan.id}]({url})' title = get_task_title(task_name, scan_history_id, subscan_id) if status: severity = STATUS_TO_SEVERITIES.get(status) msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) # Add fields to update for k, v in update_fields.items(): fields[k] = v # Add traceback to notif if traceback and notif.send_scan_tracebacks: fields['Traceback'] = f'```\n{traceback}\n```' # Add files to notif files = [] attach_file = ( notif.send_scan_output_file and output_path and result and not traceback ) if attach_file: output_title = output_path.split('/')[-1] files = [(output_path, output_title)] # Send notif opts = { 'title': title, 'url': url, 'files': files, 'severity': severity, 'fields': fields, 'fields_append': update_fields.keys() } send_notif( msg, scan_history_id=scan_history_id, subscan_id=subscan_id, **opts) @app.task(name='send_file_to_discord', bind=False, queue='send_file_to_discord_queue') def send_file_to_discord(file_path, title=None): notif = Notification.objects.first() do_send = notif and notif.send_to_discord and notif.discord_hook_url if not do_send: return False webhook = DiscordWebhook( url=notif.discord_hook_url, rate_limit_retry=True, username=title or "reNgine Discord Plugin" ) with open(file_path, "rb") as f: head, tail = os.path.split(file_path) webhook.add_file(file=f.read(), filename=tail) webhook.execute() @app.task(name='send_hackerone_report', bind=False, queue='send_hackerone_report_queue') def send_hackerone_report(vulnerability_id): """Send HackerOne vulnerability report. Args: vulnerability_id (int): Vulnerability id. Returns: int: HTTP response status code. """ vulnerability = Vulnerability.objects.get(id=vulnerability_id) severities = {v: k for k,v in NUCLEI_SEVERITY_MAP.items()} headers = { 'Content-Type': 'application/json', 'Accept': 'application/json' } # can only send vulnerability report if team_handle exists if len(vulnerability.target_domain.h1_team_handle) !=0: hackerone_query = Hackerone.objects.all() if hackerone_query.exists(): hackerone = Hackerone.objects.first() severity_value = severities[vulnerability.severity] tpl = hackerone.report_template # Replace syntax of report template with actual content tpl = tpl.replace('{vulnerability_name}', vulnerability.name) tpl = tpl.replace('{vulnerable_url}', vulnerability.http_url) tpl = tpl.replace('{vulnerability_severity}', severity_value) tpl = tpl.replace('{vulnerability_description}', vulnerability.description if vulnerability.description else '') tpl = tpl.replace('{vulnerability_extracted_results}', vulnerability.extracted_results if vulnerability.extracted_results else '') tpl = tpl.replace('{vulnerability_reference}', vulnerability.reference if vulnerability.reference else '') data = { "data": { "type": "report", "attributes": { "team_handle": vulnerability.target_domain.h1_team_handle, "title": '{} found in {}'.format(vulnerability.name, vulnerability.http_url), "vulnerability_information": tpl, "severity_rating": severity_value, "impact": "More information about the impact and vulnerability can be found here: \n" + vulnerability.reference if vulnerability.reference else "NA", } } } r = requests.post( 'https://api.hackerone.com/v1/hackers/reports', auth=(hackerone.username, hackerone.api_key), json=data, headers=headers ) response = r.json() status_code = r.status_code if status_code == 201: vulnerability.hackerone_report_id = response['data']["id"] vulnerability.open_status = False vulnerability.save() return status_code else: logger.error('No team handle found.') status_code = 111 return status_code #-------------# # Utils tasks # #-------------# @app.task(name='parse_nmap_results', bind=False, queue='parse_nmap_results_queue') def parse_nmap_results(xml_file, output_file=None): """Parse results from nmap output file. Args: xml_file (str): nmap XML report file path. Returns: list: List of vulnerabilities found from nmap results. """ with open(xml_file, encoding='utf8') as f: content = f.read() try: nmap_results = xmltodict.parse(content) # parse XML to dict except Exception as e: logger.exception(e) logger.error(f'Cannot parse {xml_file} to valid JSON. Skipping.') return [] # Write JSON to output file if output_file: with open(output_file, 'w') as f: json.dump(nmap_results, f, indent=4) logger.warning(json.dumps(nmap_results, indent=4)) hosts = ( nmap_results .get('nmaprun', {}) .get('host', {}) ) all_vulns = [] if isinstance(hosts, dict): hosts = [hosts] for host in hosts: # Grab hostname / IP from output hostnames_dict = host.get('hostnames', {}) if hostnames_dict: # Ensure that hostnames['hostname'] is a list for consistency hostnames_list = hostnames_dict['hostname'] if isinstance(hostnames_dict['hostname'], list) else [hostnames_dict['hostname']] # Extract all the @name values from the list of dictionaries hostnames = [entry.get('@name') for entry in hostnames_list] else: hostnames = [host.get('address')['@addr']] # Iterate over each hostname for each port for hostname in hostnames: # Grab ports from output ports = host.get('ports', {}).get('port', []) if isinstance(ports, dict): ports = [ports] for port in ports: url_vulns = [] port_number = port['@portid'] url = sanitize_url(f'{hostname}:{port_number}') logger.info(f'Parsing nmap results for {hostname}:{port_number} ...') if not port_number or not port_number.isdigit(): continue port_protocol = port['@protocol'] scripts = port.get('script', []) if isinstance(scripts, dict): scripts = [scripts] for script in scripts: script_id = script['@id'] script_output = script['@output'] script_output_table = script.get('table', []) logger.debug(f'Ran nmap script "{script_id}" on {port_number}/{port_protocol}:\n{script_output}\n') if script_id == 'vulscan': vulns = parse_nmap_vulscan_output(script_output) url_vulns.extend(vulns) elif script_id == 'vulners': vulns = parse_nmap_vulners_output(script_output) url_vulns.extend(vulns) # elif script_id == 'http-server-header': # TODO: nmap can help find technologies as well using the http-server-header script # regex = r'(\w+)/([\d.]+)\s?(?:\((\w+)\))?' # tech_name, tech_version, tech_os = re.match(regex, test_string).groups() # Technology.objects.get_or_create(...) # elif script_id == 'http_csrf': # vulns = parse_nmap_http_csrf_output(script_output) # url_vulns.extend(vulns) else: logger.warning(f'Script output parsing for script "{script_id}" is not supported yet.') # Add URL to vuln for vuln in url_vulns: # TODO: This should extend to any URL, not just HTTP vuln['http_url'] = url if 'http_path' in vuln: vuln['http_url'] += vuln['http_path'] all_vulns.append(vuln) return all_vulns def parse_nmap_http_csrf_output(script_output): pass def parse_nmap_vulscan_output(script_output): """Parse nmap vulscan script output. Args: script_output (str): Vulscan script output. Returns: list: List of Vulnerability dicts. """ data = {} vulns = [] provider_name = '' # Sort all vulns found by provider so that we can match each provider with # a function that pulls from its API to get more info about the # vulnerability. for line in script_output.splitlines(): if not line: continue if not line.startswith('['): # provider line if "No findings" in line: logger.info(f"No findings: {line}") continue elif ' - ' in line: provider_name, provider_url = tuple(line.split(' - ')) data[provider_name] = {'url': provider_url.rstrip(':'), 'entries': []} continue else: # Log a warning logger.warning(f"Unexpected line format: {line}") continue reg = r'\[(.*)\] (.*)' matches = re.match(reg, line) id, title = matches.groups() entry = {'id': id, 'title': title} data[provider_name]['entries'].append(entry) logger.warning('Vulscan parsed output:') logger.warning(pprint.pformat(data)) for provider_name in data: if provider_name == 'Exploit-DB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'IBM X-Force': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'MITRE CVE': logger.error(f'Provider {provider_name} is not supported YET.') for entry in data[provider_name]['entries']: cve_id = entry['id'] vuln = cve_to_vuln(cve_id) vulns.append(vuln) elif provider_name == 'OSVDB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'OpenVAS (Nessus)': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'SecurityFocus': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'VulDB': logger.error(f'Provider {provider_name} is not supported YET.') pass else: logger.error(f'Provider {provider_name} is not supported.') return vulns def parse_nmap_vulners_output(script_output, url=''): """Parse nmap vulners script output. TODO: Rework this as it's currently matching all CVEs no matter the confidence. Args: script_output (str): Script output. Returns: list: List of found vulnerabilities. """ vulns = [] # Check for CVE in script output CVE_REGEX = re.compile(r'.*(CVE-\d\d\d\d-\d+).*') matches = CVE_REGEX.findall(script_output) matches = list(dict.fromkeys(matches)) for cve_id in matches: # get CVE info vuln = cve_to_vuln(cve_id, vuln_type='nmap-vulners-nse') if vuln: vulns.append(vuln) return vulns def cve_to_vuln(cve_id, vuln_type=''): """Search for a CVE using CVESearch and return Vulnerability data. Args: cve_id (str): CVE ID in the form CVE-* Returns: dict: Vulnerability dict. """ cve_info = CVESearch('https://cve.circl.lu').id(cve_id) if not cve_info: logger.error(f'Could not fetch CVE info for cve {cve_id}. Skipping.') return None vuln_cve_id = cve_info['id'] vuln_name = vuln_cve_id vuln_description = cve_info.get('summary', 'none').replace(vuln_cve_id, '').strip() try: vuln_cvss = float(cve_info.get('cvss', -1)) except (ValueError, TypeError): vuln_cvss = -1 vuln_cwe_id = cve_info.get('cwe', '') exploit_ids = cve_info.get('refmap', {}).get('exploit-db', []) osvdb_ids = cve_info.get('refmap', {}).get('osvdb', []) references = cve_info.get('references', []) capec_objects = cve_info.get('capec', []) # Parse ovals for a better vuln name / type ovals = cve_info.get('oval', []) if ovals: vuln_name = ovals[0]['title'] vuln_type = ovals[0]['family'] # Set vulnerability severity based on CVSS score vuln_severity = 'info' if vuln_cvss < 4: vuln_severity = 'low' elif vuln_cvss < 7: vuln_severity = 'medium' elif vuln_cvss < 9: vuln_severity = 'high' else: vuln_severity = 'critical' # Build console warning message msg = f'{vuln_name} | {vuln_severity.upper()} | {vuln_cve_id} | {vuln_cwe_id} | {vuln_cvss}' for id in osvdb_ids: msg += f'\n\tOSVDB: {id}' for exploit_id in exploit_ids: msg += f'\n\tEXPLOITDB: {exploit_id}' logger.warning(msg) vuln = { 'name': vuln_name, 'type': vuln_type, 'severity': NUCLEI_SEVERITY_MAP[vuln_severity], 'description': vuln_description, 'cvss_score': vuln_cvss, 'references': references, 'cve_ids': [vuln_cve_id], 'cwe_ids': [vuln_cwe_id] } return vuln def parse_s3scanner_result(line): ''' Parses and returns s3Scanner Data ''' bucket = line['bucket'] return { 'name': bucket['name'], 'region': bucket['region'], 'provider': bucket['provider'], 'owner_display_name': bucket['owner_display_name'], 'owner_id': bucket['owner_id'], 'perm_auth_users_read': bucket['perm_auth_users_read'], 'perm_auth_users_write': bucket['perm_auth_users_write'], 'perm_auth_users_read_acl': bucket['perm_auth_users_read_acl'], 'perm_auth_users_write_acl': bucket['perm_auth_users_write_acl'], 'perm_auth_users_full_control': bucket['perm_auth_users_full_control'], 'perm_all_users_read': bucket['perm_all_users_read'], 'perm_all_users_write': bucket['perm_all_users_write'], 'perm_all_users_read_acl': bucket['perm_all_users_read_acl'], 'perm_all_users_write_acl': bucket['perm_all_users_write_acl'], 'perm_all_users_full_control': bucket['perm_all_users_full_control'], 'num_objects': bucket['num_objects'], 'size': bucket['bucket_size'] } def parse_nuclei_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ return { 'name': line['info'].get('name', ''), 'type': line['type'], 'severity': NUCLEI_SEVERITY_MAP[line['info'].get('severity', 'unknown')], 'template': line['template'], 'template_url': line['template-url'], 'template_id': line['template-id'], 'description': line['info'].get('description', ''), 'matcher_name': line.get('matcher-name', ''), 'curl_command': line.get('curl-command'), 'request': line.get('request'), 'response': line.get('response'), 'extracted_results': line.get('extracted-results', []), 'cvss_metrics': line['info'].get('classification', {}).get('cvss-metrics', ''), 'cvss_score': line['info'].get('classification', {}).get('cvss-score'), 'cve_ids': line['info'].get('classification', {}).get('cve_id', []) or [], 'cwe_ids': line['info'].get('classification', {}).get('cwe_id', []) or [], 'references': line['info'].get('reference', []) or [], 'tags': line['info'].get('tags', []), 'source': NUCLEI, } def parse_dalfox_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ description = '' description += f" Evidence: {line.get('evidence')} <br>" if line.get('evidence') else '' description += f" Message: {line.get('message')} <br>" if line.get('message') else '' description += f" Payload: {line.get('message_str')} <br>" if line.get('message_str') else '' description += f" Vulnerable Parameter: {line.get('param')} <br>" if line.get('param') else '' return { 'name': 'XSS (Cross Site Scripting)', 'type': 'XSS', 'severity': DALFOX_SEVERITY_MAP[line.get('severity', 'unknown')], 'description': description, 'source': DALFOX, 'cwe_ids': [line.get('cwe')] } def parse_crlfuzz_result(url): """Parse CRLF results Args: url (str): CRLF Vulnerable URL Returns: dict: Vulnerability data. """ return { 'name': 'CRLF (HTTP Response Splitting)', 'type': 'CRLF', 'severity': 2, 'description': 'A CRLF (HTTP Response Splitting) vulnerability has been discovered.', 'source': CRLFUZZ, } def record_exists(model, data, exclude_keys=[]): """ Check if a record already exists in the database based on the given data. Args: model (django.db.models.Model): The Django model to check against. data (dict): Data dictionary containing fields and values. exclude_keys (list): List of keys to exclude from the lookup. Returns: bool: True if the record exists, False otherwise. """ # Extract the keys that will be used for the lookup lookup_fields = {key: data[key] for key in data if key not in exclude_keys} # Return True if a record exists based on the lookup fields, False otherwise return model.objects.filter(**lookup_fields).exists() @app.task(name='geo_localize', bind=False, queue='geo_localize_queue') def geo_localize(host, ip_id=None): """Uses geoiplookup to find location associated with host. Args: host (str): Hostname. ip_id (int): IpAddress object id. Returns: startScan.models.CountryISO: CountryISO object from DB or None. """ if validators.ipv6(host): logger.info(f'Ipv6 "{host}" is not supported by geoiplookup. Skipping.') return None cmd = f'geoiplookup {host}' _, out = run_command(cmd) if 'IP Address not found' not in out and "can't resolve hostname" not in out: country_iso = out.split(':')[1].strip().split(',')[0] country_name = out.split(':')[1].strip().split(',')[1].strip() geo_object, _ = CountryISO.objects.get_or_create( iso=country_iso, name=country_name ) geo_json = { 'iso': country_iso, 'name': country_name } if ip_id: ip = IpAddress.objects.get(pk=ip_id) ip.geo_iso = geo_object ip.save() return geo_json logger.info(f'Geo IP lookup failed for host "{host}"') return None @app.task(name='query_whois', bind=False, queue='query_whois_queue') def query_whois(ip_domain, force_reload_whois=False): """Query WHOIS information for an IP or a domain name. Args: ip_domain (str): IP address or domain name. save_domain (bool): Whether to save domain or not, default False Returns: dict: WHOIS information. """ if not force_reload_whois and Domain.objects.filter(name=ip_domain).exists() and Domain.objects.get(name=ip_domain).domain_info: domain = Domain.objects.get(name=ip_domain) if not domain.insert_date: domain.insert_date = timezone.now() domain.save() domain_info_db = domain.domain_info domain_info = DottedDict( dnssec=domain_info_db.dnssec, created=domain_info_db.created, updated=domain_info_db.updated, expires=domain_info_db.expires, geolocation_iso=domain_info_db.geolocation_iso, status=[status['name'] for status in DomainWhoisStatusSerializer(domain_info_db.status, many=True).data], whois_server=domain_info_db.whois_server, ns_records=[ns['name'] for ns in NameServersSerializer(domain_info_db.name_servers, many=True).data], registrar_name=domain_info_db.registrar.name, registrar_phone=domain_info_db.registrar.phone, registrar_email=domain_info_db.registrar.email, registrar_url=domain_info_db.registrar.url, registrant_name=domain_info_db.registrant.name, registrant_id=domain_info_db.registrant.id_str, registrant_organization=domain_info_db.registrant.organization, registrant_city=domain_info_db.registrant.city, registrant_state=domain_info_db.registrant.state, registrant_zip_code=domain_info_db.registrant.zip_code, registrant_country=domain_info_db.registrant.country, registrant_phone=domain_info_db.registrant.phone, registrant_fax=domain_info_db.registrant.fax, registrant_email=domain_info_db.registrant.email, registrant_address=domain_info_db.registrant.address, admin_name=domain_info_db.admin.name, admin_id=domain_info_db.admin.id_str, admin_organization=domain_info_db.admin.organization, admin_city=domain_info_db.admin.city, admin_state=domain_info_db.admin.state, admin_zip_code=domain_info_db.admin.zip_code, admin_country=domain_info_db.admin.country, admin_phone=domain_info_db.admin.phone, admin_fax=domain_info_db.admin.fax, admin_email=domain_info_db.admin.email, admin_address=domain_info_db.admin.address, tech_name=domain_info_db.tech.name, tech_id=domain_info_db.tech.id_str, tech_organization=domain_info_db.tech.organization, tech_city=domain_info_db.tech.city, tech_state=domain_info_db.tech.state, tech_zip_code=domain_info_db.tech.zip_code, tech_country=domain_info_db.tech.country, tech_phone=domain_info_db.tech.phone, tech_fax=domain_info_db.tech.fax, tech_email=domain_info_db.tech.email, tech_address=domain_info_db.tech.address, related_tlds=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_tlds, many=True).data], related_domains=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_domains, many=True).data], historical_ips=[ip for ip in HistoricalIPSerializer(domain_info_db.historical_ips, many=True).data], ) if domain_info_db.dns_records: a_records = [] txt_records = [] mx_records = [] dns_records = [{'name': dns['name'], 'type': dns['type']} for dns in DomainDNSRecordSerializer(domain_info_db.dns_records, many=True).data] for dns in dns_records: if dns['type'] == 'a': a_records.append(dns['name']) elif dns['type'] == 'txt': txt_records.append(dns['name']) elif dns['type'] == 'mx': mx_records.append(dns['name']) domain_info.a_records = a_records domain_info.txt_records = txt_records domain_info.mx_records = mx_records else: logger.info(f'Domain info for "{ip_domain}" not found in DB, querying whois') domain_info = DottedDict() # find domain historical ip try: historical_ips = get_domain_historical_ip_address(ip_domain) domain_info.historical_ips = historical_ips except Exception as e: logger.error(f'HistoricalIP for {ip_domain} not found!\nError: {str(e)}') historical_ips = [] # find associated domains using ip_domain try: related_domains = reverse_whois(ip_domain.split('.')[0]) except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] # find related tlds using TLSx try: related_tlds = [] output_path = '/tmp/ip_domain_tlsx.txt' tlsx_command = f'tlsx -san -cn -silent -ro -host {ip_domain} -o {output_path}' run_command( tlsx_command, shell=True, ) tlsx_output = [] with open(output_path) as f: tlsx_output = f.readlines() tldextract_target = tldextract.extract(ip_domain) for doms in tlsx_output: doms = doms.strip() tldextract_res = tldextract.extract(doms) if ip_domain != doms and tldextract_res.domain == tldextract_target.domain and tldextract_res.subdomain == '': related_tlds.append(doms) related_tlds = list(set(related_tlds)) domain_info.related_tlds = related_tlds except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] related_domains_list = [] if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_related = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_related) related_domains_list.append(_domain['name']) for _domain in related_tlds: domain_related = RelatedDomain.objects.get_or_create( name=_domain, )[0] db_domain_info.related_tlds.add(domain_related) for _ip in historical_ips: historical_ip = HistoricalIP.objects.get_or_create( ip=_ip['ip'], owner=_ip['owner'], location=_ip['location'], last_seen=_ip['last_seen'], )[0] db_domain_info.historical_ips.add(historical_ip) domain.domain_info = db_domain_info domain.save() command = f'netlas host {ip_domain} -f json' # check if netlas key is provided netlas_key = get_netlas_key() command += f' -a {netlas_key}' if netlas_key else '' result = subprocess.check_output(command.split()).decode('utf-8') if 'Failed to parse response data' in result: # do fallback return { 'status': False, 'ip_domain': ip_domain, 'result': "Netlas limit exceeded.", 'message': 'Netlas limit exceeded.' } try: result = json.loads(result) logger.info(result) whois = result.get('whois') if result.get('whois') else {} domain_info.created = whois.get('created_date') domain_info.expires = whois.get('expiration_date') domain_info.updated = whois.get('updated_date') domain_info.whois_server = whois.get('whois_server') if 'registrant' in whois: registrant = whois.get('registrant') domain_info.registrant_name = registrant.get('name') domain_info.registrant_country = registrant.get('country') domain_info.registrant_id = registrant.get('id') domain_info.registrant_state = registrant.get('province') domain_info.registrant_city = registrant.get('city') domain_info.registrant_phone = registrant.get('phone') domain_info.registrant_address = registrant.get('street') domain_info.registrant_organization = registrant.get('organization') domain_info.registrant_fax = registrant.get('fax') domain_info.registrant_zip_code = registrant.get('postal_code') email_search = EMAIL_REGEX.search(str(registrant.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.registrant_email = field_content if 'administrative' in whois: administrative = whois.get('administrative') domain_info.admin_name = administrative.get('name') domain_info.admin_country = administrative.get('country') domain_info.admin_id = administrative.get('id') domain_info.admin_state = administrative.get('province') domain_info.admin_city = administrative.get('city') domain_info.admin_phone = administrative.get('phone') domain_info.admin_address = administrative.get('street') domain_info.admin_organization = administrative.get('organization') domain_info.admin_fax = administrative.get('fax') domain_info.admin_zip_code = administrative.get('postal_code') mail_search = EMAIL_REGEX.search(str(administrative.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.admin_email = field_content if 'technical' in whois: technical = whois.get('technical') domain_info.tech_name = technical.get('name') domain_info.tech_country = technical.get('country') domain_info.tech_state = technical.get('province') domain_info.tech_id = technical.get('id') domain_info.tech_city = technical.get('city') domain_info.tech_phone = technical.get('phone') domain_info.tech_address = technical.get('street') domain_info.tech_organization = technical.get('organization') domain_info.tech_fax = technical.get('fax') domain_info.tech_zip_code = technical.get('postal_code') mail_search = EMAIL_REGEX.search(str(technical.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.tech_email = field_content if 'dns' in result: dns = result.get('dns') domain_info.mx_records = dns.get('mx') domain_info.txt_records = dns.get('txt') domain_info.a_records = dns.get('a') domain_info.ns_records = whois.get('name_servers') domain_info.dnssec = True if whois.get('dnssec') else False domain_info.status = whois.get('status') if 'registrar' in whois: registrar = whois.get('registrar') domain_info.registrar_name = registrar.get('name') domain_info.registrar_email = registrar.get('email') domain_info.registrar_phone = registrar.get('phone') domain_info.registrar_url = registrar.get('url') # find associated domains if registrant email is found related_domains = reverse_whois(domain_info.get('registrant_email')) if domain_info.get('registrant_email') else [] for _domain in related_domains: related_domains_list.append(_domain['name']) # remove duplicate domains from related domains list related_domains_list = list(set(related_domains_list)) domain_info.related_domains = related_domains_list # save to db if domain exists if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_rel = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_rel) db_domain_info.dnssec = domain_info.get('dnssec') #dates db_domain_info.created = domain_info.get('created') db_domain_info.updated = domain_info.get('updated') db_domain_info.expires = domain_info.get('expires') #registrar db_domain_info.registrar = Registrar.objects.get_or_create( name=domain_info.get('registrar_name'), email=domain_info.get('registrar_email'), phone=domain_info.get('registrar_phone'), url=domain_info.get('registrar_url'), )[0] db_domain_info.registrant = DomainRegistration.objects.get_or_create( name=domain_info.get('registrant_name'), organization=domain_info.get('registrant_organization'), address=domain_info.get('registrant_address'), city=domain_info.get('registrant_city'), state=domain_info.get('registrant_state'), zip_code=domain_info.get('registrant_zip_code'), country=domain_info.get('registrant_country'), email=domain_info.get('registrant_email'), phone=domain_info.get('registrant_phone'), fax=domain_info.get('registrant_fax'), id_str=domain_info.get('registrant_id'), )[0] db_domain_info.admin = DomainRegistration.objects.get_or_create( name=domain_info.get('admin_name'), organization=domain_info.get('admin_organization'), address=domain_info.get('admin_address'), city=domain_info.get('admin_city'), state=domain_info.get('admin_state'), zip_code=domain_info.get('admin_zip_code'), country=domain_info.get('admin_country'), email=domain_info.get('admin_email'), phone=domain_info.get('admin_phone'), fax=domain_info.get('admin_fax'), id_str=domain_info.get('admin_id'), )[0] db_domain_info.tech = DomainRegistration.objects.get_or_create( name=domain_info.get('tech_name'), organization=domain_info.get('tech_organization'), address=domain_info.get('tech_address'), city=domain_info.get('tech_city'), state=domain_info.get('tech_state'), zip_code=domain_info.get('tech_zip_code'), country=domain_info.get('tech_country'), email=domain_info.get('tech_email'), phone=domain_info.get('tech_phone'), fax=domain_info.get('tech_fax'), id_str=domain_info.get('tech_id'), )[0] for status in domain_info.get('status') or []: _status = WhoisStatus.objects.get_or_create( name=status )[0] _status.save() db_domain_info.status.add(_status) for ns in domain_info.get('ns_records') or []: _ns = NameServer.objects.get_or_create( name=ns )[0] _ns.save() db_domain_info.name_servers.add(_ns) for a in domain_info.get('a_records') or []: _a = DNSRecord.objects.get_or_create( name=a, type='a' )[0] _a.save() db_domain_info.dns_records.add(_a) for mx in domain_info.get('mx_records') or []: _mx = DNSRecord.objects.get_or_create( name=mx, type='mx' )[0] _mx.save() db_domain_info.dns_records.add(_mx) for txt in domain_info.get('txt_records') or []: _txt = DNSRecord.objects.get_or_create( name=txt, type='txt' )[0] _txt.save() db_domain_info.dns_records.add(_txt) db_domain_info.geolocation_iso = domain_info.get('registrant_country') db_domain_info.whois_server = domain_info.get('whois_server') db_domain_info.save() domain.domain_info = db_domain_info domain.save() except Exception as e: return { 'status': False, 'ip_domain': ip_domain, 'result': "unable to fetch records from WHOIS database.", 'message': str(e) } return { 'status': True, 'ip_domain': ip_domain, 'dnssec': domain_info.get('dnssec'), 'created': domain_info.get('created'), 'updated': domain_info.get('updated'), 'expires': domain_info.get('expires'), 'geolocation_iso': domain_info.get('registrant_country'), 'domain_statuses': domain_info.get('status'), 'whois_server': domain_info.get('whois_server'), 'dns': { 'a': domain_info.get('a_records'), 'mx': domain_info.get('mx_records'), 'txt': domain_info.get('txt_records'), }, 'registrar': { 'name': domain_info.get('registrar_name'), 'phone': domain_info.get('registrar_phone'), 'email': domain_info.get('registrar_email'), 'url': domain_info.get('registrar_url'), }, 'registrant': { 'name': domain_info.get('registrant_name'), 'id': domain_info.get('registrant_id'), 'organization': domain_info.get('registrant_organization'), 'address': domain_info.get('registrant_address'), 'city': domain_info.get('registrant_city'), 'state': domain_info.get('registrant_state'), 'zipcode': domain_info.get('registrant_zip_code'), 'country': domain_info.get('registrant_country'), 'phone': domain_info.get('registrant_phone'), 'fax': domain_info.get('registrant_fax'), 'email': domain_info.get('registrant_email'), }, 'admin': { 'name': domain_info.get('admin_name'), 'id': domain_info.get('admin_id'), 'organization': domain_info.get('admin_organization'), 'address':domain_info.get('admin_address'), 'city': domain_info.get('admin_city'), 'state': domain_info.get('admin_state'), 'zipcode': domain_info.get('admin_zip_code'), 'country': domain_info.get('admin_country'), 'phone': domain_info.get('admin_phone'), 'fax': domain_info.get('admin_fax'), 'email': domain_info.get('admin_email'), }, 'technical_contact': { 'name': domain_info.get('tech_name'), 'id': domain_info.get('tech_id'), 'organization': domain_info.get('tech_organization'), 'address': domain_info.get('tech_address'), 'city': domain_info.get('tech_city'), 'state': domain_info.get('tech_state'), 'zipcode': domain_info.get('tech_zip_code'), 'country': domain_info.get('tech_country'), 'phone': domain_info.get('tech_phone'), 'fax': domain_info.get('tech_fax'), 'email': domain_info.get('tech_email'), }, 'nameservers': domain_info.get('ns_records'), # 'similar_domains': domain_info.get('similar_domains'), 'related_domains': domain_info.get('related_domains'), 'related_tlds': domain_info.get('related_tlds'), 'historical_ips': domain_info.get('historical_ips'), } @app.task(name='remove_duplicate_endpoints', bind=False, queue='remove_duplicate_endpoints_queue') def remove_duplicate_endpoints( scan_history_id, domain_id, subdomain_id=None, filter_ids=[], filter_status=[200, 301, 404], duplicate_removal_fields=ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS ): """Remove duplicate endpoints. Check for implicit redirections by comparing endpoints: - [x] `content_length` similarities indicating redirections - [x] `page_title` (check for same page title) - [ ] Sign-in / login page (check for endpoints with the same words) Args: scan_history_id: ScanHistory id. domain_id (int): Domain id. subdomain_id (int, optional): Subdomain id. filter_ids (list): List of endpoint ids to filter on. filter_status (list): List of HTTP status codes to filter on. duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates """ logger.info(f'Removing duplicate endpoints based on {duplicate_removal_fields}') endpoints = ( EndPoint.objects .filter(scan_history__id=scan_history_id) .filter(target_domain__id=domain_id) ) if filter_status: endpoints = endpoints.filter(http_status__in=filter_status) if subdomain_id: endpoints = endpoints.filter(subdomain__id=subdomain_id) if filter_ids: endpoints = endpoints.filter(id__in=filter_ids) for field_name in duplicate_removal_fields: cl_query = ( endpoints .values_list(field_name) .annotate(mc=Count(field_name)) .order_by('-mc') ) for (field_value, count) in cl_query: if count > DELETE_DUPLICATES_THRESHOLD: eps_to_delete = ( endpoints .filter(**{field_name: field_value}) .order_by('discovered_date') .all()[1:] ) msg = f'Deleting {len(eps_to_delete)} endpoints [reason: same {field_name} {field_value}]' for ep in eps_to_delete: url = urlparse(ep.http_url) if url.path in ['', '/', '/login']: # try do not delete the original page that other pages redirect to continue msg += f'\n\t {ep.http_url} [{ep.http_status}] [{field_name}={field_value}]' ep.delete() logger.warning(msg) @app.task(name='run_command', bind=False, queue='run_command_queue') def run_command(cmd, cwd=None, shell=False, history_file=None, scan_id=None, activity_id=None): """Run a given command using subprocess module. Args: cmd (str): Command to run. cwd (str): Current working directory. echo (bool): Log command. shell (bool): Run within separate shell if True. history_file (str): Write command + output to history file. Returns: tuple: Tuple with return_code, output. """ logger.info(cmd) logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Run the command using subprocess popen = subprocess.Popen( cmd if shell else cmd.split(), shell=shell, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, cwd=cwd, universal_newlines=True) output = '' for stdout_line in iter(popen.stdout.readline, ""): item = stdout_line.strip() output += '\n' + item logger.debug(item) popen.stdout.close() popen.wait() return_code = popen.returncode command_obj.output = output command_obj.return_code = return_code command_obj.save() if history_file: mode = 'a' if not os.path.exists(history_file): mode = 'w' with open(history_file, mode) as f: f.write(f'\n{cmd}\n{return_code}\n{output}\n------------------\n') return return_code, output #-------------# # Other utils # #-------------# def stream_command(cmd, cwd=None, shell=False, history_file=None, encoding='utf-8', scan_id=None, activity_id=None, trunc_char=None): # Log cmd logger.info(cmd) # logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Sanitize the cmd command = cmd if shell else cmd.split() # Run the command using subprocess process = subprocess.Popen( command, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, universal_newlines=True, shell=shell) # Log the output in real-time to the database output = "" # Process the output for line in iter(lambda: process.stdout.readline(), b''): if not line: break line = line.strip() ansi_escape = re.compile(r'\x1B(?:[@-Z\\-_]|\[[0-?]*[ -/]*[@-~])') line = ansi_escape.sub('', line) line = line.replace('\\x0d\\x0a', '\n') if trunc_char and line.endswith(trunc_char): line = line[:-1] item = line # Try to parse the line as JSON try: item = json.loads(line) except json.JSONDecodeError: pass # Yield the line #logger.debug(item) yield item # Add the log line to the output output += line + "\n" # Update the command record in the database command_obj.output = output command_obj.save() # Retrieve the return code and output process.wait() return_code = process.returncode # Update the return code and final output in the database command_obj.return_code = return_code command_obj.save() # Append the command, return code and output to the history file if history_file is not None: with open(history_file, "a") as f: f.write(f"{cmd}\n{return_code}\n{output}\n") def process_httpx_response(line): """TODO: implement this""" def extract_httpx_url(line): """Extract final URL from httpx results. Always follow redirects to find the last URL. Args: line (dict): URL data output by httpx. Returns: tuple: (final_url, redirect_bool) tuple. """ status_code = line.get('status_code', 0) final_url = line.get('final_url') location = line.get('location') chain_status_codes = line.get('chain_status_codes', []) # Final URL is already looking nice, if it exists return it if final_url: return final_url, False http_url = line['url'] # fallback to url field # Handle redirects manually REDIRECT_STATUS_CODES = [301, 302] is_redirect = ( status_code in REDIRECT_STATUS_CODES or any(x in REDIRECT_STATUS_CODES for x in chain_status_codes) ) if is_redirect and location: if location.startswith(('http', 'https')): http_url = location else: http_url = f'{http_url}/{location.lstrip("/")}' # Sanitize URL http_url = sanitize_url(http_url) return http_url, is_redirect #-------------# # OSInt utils # #-------------# def get_and_save_dork_results(lookup_target, results_dir, type, lookup_keywords=None, lookup_extensions=None, delay=3, page_count=2, scan_history=None): """ Uses gofuzz to dork and store information Args: lookup_target (str): target to look into such as stackoverflow or even the target itself results_dir (str): Results directory type (str): Dork Type Title lookup_keywords (str): comma separated keywords or paths to look for lookup_extensions (str): comma separated extensions to look for delay (int): delay between each requests page_count (int): pages in google to extract information scan_history (startScan.ScanHistory): Scan History Object """ results = [] gofuzz_command = f'{GOFUZZ_EXEC_PATH} -t {lookup_target} -d {delay} -p {page_count}' if lookup_extensions: gofuzz_command += f' -e {lookup_extensions}' elif lookup_keywords: gofuzz_command += f' -w {lookup_keywords}' output_file = f'{results_dir}/gofuzz.txt' gofuzz_command += f' -o {output_file}' history_file = f'{results_dir}/commands.txt' try: run_command( gofuzz_command, shell=False, history_file=history_file, scan_id=scan_history.id, ) if not os.path.isfile(output_file): return with open(output_file) as f: for line in f.readlines(): url = line.strip() if url: results.append(url) dork, created = Dork.objects.get_or_create( type=type, url=url ) if scan_history: scan_history.dorks.add(dork) # remove output file os.remove(output_file) except Exception as e: logger.exception(e) return results def get_and_save_emails(scan_history, activity_id, results_dir): """Get and save emails from Google, Bing and Baidu. Args: scan_history (startScan.ScanHistory): Scan history object. activity_id: ScanActivity Object results_dir (str): Results directory. Returns: list: List of emails found. """ emails = [] # Proxy settings # get_random_proxy() # Gather emails from Google, Bing and Baidu output_file = f'{results_dir}/emails_tmp.txt' history_file = f'{results_dir}/commands.txt' command = f'python3 /usr/src/github/Infoga/infoga.py --domain {scan_history.domain.name} --source all --report {output_file}' try: run_command( command, shell=False, history_file=history_file, scan_id=scan_history.id, activity_id=activity_id) if not os.path.isfile(output_file): logger.info('No Email results') return [] with open(output_file) as f: for line in f.readlines(): if 'Email' in line: split_email = line.split(' ')[2] emails.append(split_email) output_path = f'{results_dir}/emails.txt' with open(output_path, 'w') as output_file: for email_address in emails: save_email(email_address, scan_history) output_file.write(f'{email_address}\n') except Exception as e: logger.exception(e) return emails def save_metadata_info(meta_dict): """Extract metadata from Google Search. Args: meta_dict (dict): Info dict. Returns: list: List of startScan.MetaFinderDocument objects. """ logger.warning(f'Getting metadata for {meta_dict.osint_target}') scan_history = ScanHistory.objects.get(id=meta_dict.scan_id) # Proxy settings get_random_proxy() # Get metadata result = extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if not result: logger.error(f'No metadata result from Google Search for {meta_dict.osint_target}.') return [] # Add metadata info to DB results = [] for metadata_name, data in result.get_metadata().items(): subdomain = Subdomain.objects.get( scan_history=meta_dict.scan_id, name=meta_dict.osint_target) metadata = DottedDict({k: v for k, v in data.items()}) meta_finder_document = MetaFinderDocument( subdomain=subdomain, target_domain=meta_dict.domain, scan_history=scan_history, url=metadata.url, doc_name=metadata_name, http_status=metadata.status_code, producer=metadata.metadata.get('Producer'), creator=metadata.metadata.get('Creator'), creation_date=metadata.metadata.get('CreationDate'), modified_date=metadata.metadata.get('ModDate'), author=metadata.metadata.get('Author'), title=metadata.metadata.get('Title'), os=metadata.metadata.get('OSInfo')) meta_finder_document.save() results.append(data) return results #-----------------# # Utils functions # #-----------------# def create_scan_activity(scan_history_id, message, status): scan_activity = ScanActivity() scan_activity.scan_of = ScanHistory.objects.get(pk=scan_history_id) scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id #--------------------# # Database functions # #--------------------# def save_vulnerability(**vuln_data): references = vuln_data.pop('references', []) cve_ids = vuln_data.pop('cve_ids', []) cwe_ids = vuln_data.pop('cwe_ids', []) tags = vuln_data.pop('tags', []) subscan = vuln_data.pop('subscan', None) # remove nulls vuln_data = replace_nulls(vuln_data) # Create vulnerability vuln, created = Vulnerability.objects.get_or_create(**vuln_data) if created: vuln.discovered_date = timezone.now() vuln.open_status = True vuln.save() # Save vuln tags for tag_name in tags or []: tag, created = VulnerabilityTags.objects.get_or_create(name=tag_name) if tag: vuln.tags.add(tag) vuln.save() # Save CVEs for cve_id in cve_ids or []: cve, created = CveId.objects.get_or_create(name=cve_id) if cve: vuln.cve_ids.add(cve) vuln.save() # Save CWEs for cve_id in cwe_ids or []: cwe, created = CweId.objects.get_or_create(name=cve_id) if cwe: vuln.cwe_ids.add(cwe) vuln.save() # Save vuln reference for url in references or []: ref, created = VulnerabilityReference.objects.get_or_create(url=url) if created: vuln.references.add(ref) vuln.save() # Save subscan id in vuln object if subscan: vuln.vuln_subscan_ids.add(subscan) vuln.save() return vuln, created def save_endpoint( http_url, ctx={}, crawl=False, is_default=False, **endpoint_data): """Get or create EndPoint object. If crawl is True, also crawl the endpoint HTTP URL with httpx. Args: http_url (str): Input HTTP URL. is_default (bool): If the url is a default url for SubDomains. scan_history (startScan.models.ScanHistory): ScanHistory object. domain (startScan.models.Domain): Domain object. subdomain (starScan.models.Subdomain): Subdomain object. results_dir (str, optional): Results directory. crawl (bool, optional): Run httpx on endpoint if True. Default: False. force (bool, optional): Force crawl even if ENABLE_HTTP_CRAWL mode is on. subscan (startScan.models.SubScan, optional): SubScan object. Returns: tuple: (startScan.models.EndPoint, created) where `created` is a boolean indicating if the object is new or already existed. """ # remove nulls endpoint_data = replace_nulls(endpoint_data) scheme = urlparse(http_url).scheme endpoint = None created = False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in http_url: logger.error(f"{http_url} is not a URL of domain {domain.name}. Skipping.") return None, False if crawl: ctx['track'] = False results = http_crawl( urls=[http_url], method='HEAD', ctx=ctx) if results: endpoint_data = results[0] endpoint_id = endpoint_data['endpoint_id'] created = endpoint_data['endpoint_created'] endpoint = EndPoint.objects.get(pk=endpoint_id) elif not scheme: return None, False else: # add dumb endpoint without probing it scan = ScanHistory.objects.filter(pk=ctx.get('scan_history_id')).first() domain = Domain.objects.filter(pk=ctx.get('domain_id')).first() if not validators.url(http_url): return None, False http_url = sanitize_url(http_url) # Try to get the first matching record (prevent duplicate error) endpoints = EndPoint.objects.filter( scan_history=scan, target_domain=domain, http_url=http_url, **endpoint_data ) if endpoints.exists(): endpoint = endpoints.first() created = False else: # No existing record, create a new one endpoint = EndPoint.objects.create( scan_history=scan, target_domain=domain, http_url=http_url, **endpoint_data ) created = True if created: endpoint.is_default = is_default endpoint.discovered_date = timezone.now() endpoint.save() subscan_id = ctx.get('subscan_id') if subscan_id: endpoint.endpoint_subscan_ids.add(subscan_id) endpoint.save() return endpoint, created def save_subdomain(subdomain_name, ctx={}): """Get or create Subdomain object. Args: subdomain_name (str): Subdomain name. scan_history (startScan.models.ScanHistory): ScanHistory object. Returns: tuple: (startScan.models.Subdomain, created) where `created` is a boolean indicating if the object has been created in DB. """ scan_id = ctx.get('scan_history_id') subscan_id = ctx.get('subscan_id') out_of_scope_subdomains = ctx.get('out_of_scope_subdomains', []) valid_domain = ( validators.domain(subdomain_name) or validators.ipv4(subdomain_name) or validators.ipv6(subdomain_name) ) if not valid_domain: logger.error(f'{subdomain_name} is not an invalid domain. Skipping.') return None, False if subdomain_name in out_of_scope_subdomains: logger.error(f'{subdomain_name} is out-of-scope. Skipping.') return None, False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in subdomain_name: logger.error(f"{subdomain_name} is not a subdomain of domain {domain.name}. Skipping.") return None, False scan = ScanHistory.objects.filter(pk=scan_id).first() domain = scan.domain if scan else None subdomain, created = Subdomain.objects.get_or_create( scan_history=scan, target_domain=domain, name=subdomain_name) if created: # logger.warning(f'Found new subdomain {subdomain_name}') subdomain.discovered_date = timezone.now() if subscan_id: subdomain.subdomain_subscan_ids.add(subscan_id) subdomain.save() return subdomain, created def save_email(email_address, scan_history=None): if not validators.email(email_address): logger.info(f'Email {email_address} is invalid. Skipping.') return None, False email, created = Email.objects.get_or_create(address=email_address) # if created: # logger.warning(f'Found new email address {email_address}') # Add email to ScanHistory if scan_history: scan_history.emails.add(email) scan_history.save() return email, created def save_employee(name, designation, scan_history=None): employee, created = Employee.objects.get_or_create( name=name, designation=designation) # if created: # logger.warning(f'Found new employee {name}') # Add employee to ScanHistory if scan_history: scan_history.employees.add(employee) scan_history.save() return employee, created def save_ip_address(ip_address, subdomain=None, subscan=None, **kwargs): if not (validators.ipv4(ip_address) or validators.ipv6(ip_address)): logger.info(f'IP {ip_address} is not a valid IP. Skipping.') return None, False ip, created = IpAddress.objects.get_or_create(address=ip_address) # if created: # logger.warning(f'Found new IP {ip_address}') # Set extra attributes for key, value in kwargs.items(): setattr(ip, key, value) ip.save() # Add IP to subdomain if subdomain: subdomain.ip_addresses.add(ip) subdomain.save() # Add subscan to IP if subscan: ip.ip_subscan_ids.add(subscan) # Geo-localize IP asynchronously if created: geo_localize.delay(ip_address, ip.id) return ip, created def save_imported_subdomains(subdomains, ctx={}): """Take a list of subdomains imported and write them to from_imported.txt. Args: subdomains (list): List of subdomain names. scan_history (startScan.models.ScanHistory): ScanHistory instance. domain (startScan.models.Domain): Domain instance. results_dir (str): Results directory. """ domain_id = ctx['domain_id'] domain = Domain.objects.get(pk=domain_id) results_dir = ctx.get('results_dir', RENGINE_RESULTS) # Validate each subdomain and de-duplicate entries subdomains = list(set([ subdomain for subdomain in subdomains if validators.domain(subdomain) and domain.name == get_domain_from_subdomain(subdomain) ])) if not subdomains: return logger.warning(f'Found {len(subdomains)} imported subdomains.') with open(f'{results_dir}/from_imported.txt', 'w+') as output_file: for name in subdomains: subdomain_name = name.strip() subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain.is_imported_subdomain = True subdomain.save() output_file.write(f'{subdomain}\n') @app.task(name='query_reverse_whois', bind=False, queue='query_reverse_whois_queue') def query_reverse_whois(lookup_keyword): """Queries Reverse WHOIS information for an organization or email address. Args: lookup_keyword (str): Registrar Name or email Returns: dict: Reverse WHOIS information. """ return get_associated_domains(lookup_keyword) @app.task(name='query_ip_history', bind=False, queue='query_ip_history_queue') def query_ip_history(domain): """Queries the IP history for a domain Args: domain (str): domain_name Returns: list: list of historical ip addresses """ return get_domain_historical_ip_address(domain) @app.task(name='gpt_vulnerability_description', bind=False, queue='gpt_queue') def gpt_vulnerability_description(vulnerability_id): """Generate and store Vulnerability Description using GPT. Args: vulnerability_id (Vulnerability Model ID): Vulnerability ID to fetch Description. """ logger.info('Getting GPT Vulnerability Description') try: lookup_vulnerability = Vulnerability.objects.get(id=vulnerability_id) lookup_url = urlparse(lookup_vulnerability.http_url) path = lookup_url.path except Exception as e: return { 'status': False, 'error': str(e) } # check in db GPTVulnerabilityReport model if vulnerability description and path matches stored = GPTVulnerabilityReport.objects.filter(url_path=path).filter(title=lookup_vulnerability.name).first() if stored: response = { 'status': True, 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: vulnerability_description = get_gpt_vuln_input_description( lookup_vulnerability.name, path ) # one can add more description here later gpt_generator = GPTVulnerabilityReportGenerator() response = gpt_generator.get_vulnerability_description(vulnerability_description) add_gpt_description_db( lookup_vulnerability.name, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) # for all vulnerabilities with the same vulnerability name this description has to be stored. # also the consition is that the url must contain a part of this. for vuln in Vulnerability.objects.filter(name=lookup_vulnerability.name, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() return response

psyray

7c01a46cea370e74385682ba7c28eaf4e58f5d69

2e089dc62f1bd64aa481750da10fa750e3aa232d

Don't know, I could delete it

psyray

yogeshojha/rengine

Fix crash on saving endpoint (FFUF related only)

Fix #1006 I've added : - a **try except** block to catch error on duplicate record returned by **get_or_create** in **saving_endpoint** method - a **check** on endpoint existence in **dir_file_fuzz** method Errors are logged to the console with the URL.  Tested and working Now we need to find why there are duplicates endpoints in the db But it's another issue

2023-11-22 02:57:45+00:00

2023-11-27 12:37:27+00:00

web/reNgine/tasks.py

import csv import json import os import pprint import subprocess import time import validators import whatportis import xmltodict import yaml import tldextract import concurrent.futures from datetime import datetime from urllib.parse import urlparse from api.serializers import SubdomainSerializer from celery import chain, chord, group from celery.result import allow_join_result from celery.utils.log import get_task_logger from django.db.models import Count from dotted_dict import DottedDict from django.utils import timezone from pycvesearch import CVESearch from metafinder.extractor import extract_metadata_from_google_search from reNgine.celery import app from reNgine.gpt import GPTVulnerabilityReportGenerator from reNgine.celery_custom_task import RengineTask from reNgine.common_func import * from reNgine.definitions import * from reNgine.settings import * from reNgine.gpt import * from reNgine.utilities import * from scanEngine.models import (EngineType, InstalledExternalTool, Notification, Proxy) from startScan.models import * from startScan.models import EndPoint, Subdomain, Vulnerability from targetApp.models import Domain """ Celery tasks. """ logger = get_task_logger(__name__) #----------------------# # Scan / Subscan tasks # #----------------------# @app.task(name='initiate_scan', bind=False, queue='initiate_scan_queue') def initiate_scan( scan_history_id, domain_id, engine_id=None, scan_type=LIVE_SCAN, results_dir=RENGINE_RESULTS, imported_subdomains=[], out_of_scope_subdomains=[], url_filter=''): """Initiate a new scan. Args: scan_history_id (int): ScanHistory id. domain_id (int): Domain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. imported_subdomains (list): Imported subdomains. out_of_scope_subdomains (list): Out-of-scope subdomains. url_filter (str): URL path. Default: '' """ # Get scan history scan = ScanHistory.objects.get(pk=scan_history_id) # Get scan engine engine_id = engine_id or scan.scan_type.id # scan history engine_id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, []) # Get domain and set last_scan_date domain = Domain.objects.get(pk=domain_id) domain.last_scan_date = timezone.now() domain.save() # Get path filter url_filter = url_filter.rstrip('/') # Get or create ScanHistory() object if scan_type == LIVE_SCAN: # immediate scan = ScanHistory.objects.get(pk=scan_history_id) scan.scan_status = RUNNING_TASK elif scan_type == SCHEDULED_SCAN: # scheduled scan = ScanHistory() scan.scan_status = INITIATED_TASK scan.scan_type = engine scan.celery_ids = [initiate_scan.request.id] scan.domain = domain scan.start_scan_date = timezone.now() scan.tasks = engine.tasks scan.results_dir = f'{results_dir}/{domain.name}_{scan.id}' add_gf_patterns = gf_patterns and 'fetch_url' in engine.tasks if add_gf_patterns: scan.used_gf_patterns = ','.join(gf_patterns) scan.save() # Create scan results dir os.makedirs(scan.results_dir) # Build task context ctx = { 'scan_history_id': scan_history_id, 'engine_id': engine_id, 'domain_id': domain.id, 'results_dir': scan.results_dir, 'url_filter': url_filter, 'yaml_configuration': config, 'out_of_scope_subdomains': out_of_scope_subdomains } ctx_str = json.dumps(ctx, indent=2) # Send start notif logger.warning(f'Starting scan {scan_history_id} with context:\n{ctx_str}') send_scan_notif.delay( scan_history_id, subscan_id=None, engine_id=engine_id, status=CELERY_TASK_STATUS_MAP[scan.scan_status]) # Save imported subdomains in DB save_imported_subdomains(imported_subdomains, ctx=ctx) # Create initial subdomain in DB: make a copy of domain as a subdomain so # that other tasks using subdomains can use it. subdomain_name = domain.name subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) # If enable_http_crawl is set, create an initial root HTTP endpoint so that # HTTP crawling can start somewhere http_url = f'{domain.name}{url_filter}' if url_filter else domain.name endpoint, _ = save_endpoint( http_url, ctx=ctx, crawl=enable_http_crawl, is_default=True, subdomain=subdomain ) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build Celery tasks, crafted according to the dependency graph below: # subdomain_discovery --> port_scan --> fetch_url --> dir_file_fuzz # osint vulnerability_scan # osint dalfox xss scan # screenshot # waf_detection workflow = chain( group( subdomain_discovery.si(ctx=ctx, description='Subdomain discovery'), osint.si(ctx=ctx, description='OS Intelligence') ), port_scan.si(ctx=ctx, description='Port scan'), fetch_url.si(ctx=ctx, description='Fetch URL'), group( dir_file_fuzz.si(ctx=ctx, description='Directories & files fuzz'), vulnerability_scan.si(ctx=ctx, description='Vulnerability scan'), screenshot.si(ctx=ctx, description='Screenshot'), waf_detection.si(ctx=ctx, description='WAF detection') ) ) # Build callback callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery chord logger.info(f'Running Celery workflow with {len(workflow.tasks) + 1} tasks') task = chain(workflow, callback).on_error(callback).delay() scan.celery_ids.append(task.id) scan.save() return { 'success': True, 'task_id': task.id } @app.task(name='initiate_subscan', bind=False, queue='subscan_queue') def initiate_subscan( scan_history_id, subdomain_id, engine_id=None, scan_type=None, results_dir=RENGINE_RESULTS, url_filter=''): """Initiate a new subscan. Args: scan_history_id (int): ScanHistory id. subdomain_id (int): Subdomain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. url_filter (str): URL path. Default: '' """ # Get Subdomain, Domain and ScanHistory subdomain = Subdomain.objects.get(pk=subdomain_id) scan = ScanHistory.objects.get(pk=subdomain.scan_history.id) domain = Domain.objects.get(pk=subdomain.target_domain.id) # Get EngineType engine_id = engine_id or scan.scan_type.id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Create scan activity of SubScan Model subscan = SubScan( start_scan_date=timezone.now(), celery_ids=[initiate_subscan.request.id], scan_history=scan, subdomain=subdomain, type=scan_type, status=RUNNING_TASK, engine=engine) subscan.save() # Get YAML configuration config = yaml.safe_load(engine.yaml_configuration) # Create results directory results_dir = f'{scan.results_dir}/subscans/{subscan.id}' os.makedirs(results_dir, exist_ok=True) # Run task method = globals().get(scan_type) if not method: logger.warning(f'Task {scan_type} is not supported by reNgine. Skipping') return scan.tasks.append(scan_type) scan.save() # Send start notif send_scan_notif.delay( scan.id, subscan_id=subscan.id, engine_id=engine_id, status='RUNNING') # Build context ctx = { 'scan_history_id': scan.id, 'subscan_id': subscan.id, 'engine_id': engine_id, 'domain_id': domain.id, 'subdomain_id': subdomain.id, 'yaml_configuration': config, 'results_dir': results_dir, 'url_filter': url_filter } # Create initial endpoints in DB: find domain HTTP endpoint so that HTTP # crawling can start somewhere base_url = f'{subdomain.name}{url_filter}' if url_filter else subdomain.name endpoint, _ = save_endpoint( base_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build header + callback workflow = method.si(ctx=ctx) callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery tasks task = chain(workflow, callback).on_error(callback).delay() subscan.celery_ids.append(task.id) subscan.save() return { 'success': True, 'task_id': task.id } @app.task(name='report', bind=False, queue='report_queue') def report(ctx={}, description=None): """Report task running after all other tasks. Mark ScanHistory or SubScan object as completed and update with final status, log run details and send notification. Args: description (str, optional): Task description shown in UI. """ # Get objects subscan_id = ctx.get('subscan_id') scan_id = ctx.get('scan_history_id') engine_id = ctx.get('engine_id') scan = ScanHistory.objects.filter(pk=scan_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() # Get failed tasks tasks = ScanActivity.objects.filter(scan_of=scan).all() if subscan: tasks = tasks.filter(celery_id__in=subscan.celery_ids) failed_tasks = tasks.filter(status=FAILED_TASK) # Get task status failed_count = failed_tasks.count() status = SUCCESS_TASK if failed_count == 0 else FAILED_TASK status_h = 'SUCCESS' if failed_count == 0 else 'FAILED' # Update scan / subscan status if subscan: subscan.stop_scan_date = timezone.now() subscan.status = status subscan.save() else: scan.scan_status = status scan.stop_scan_date = timezone.now() scan.save() # Send scan status notif send_scan_notif.delay( scan_history_id=scan_id, subscan_id=subscan_id, engine_id=engine_id, status=status_h) #------------------------- # # Tracked reNgine tasks # #--------------------------# @app.task(name='subdomain_discovery', queue='main_scan_queue', base=RengineTask, bind=True) def subdomain_discovery( self, host=None, ctx=None, description=None): """Uses a set of tools (see SUBDOMAIN_SCAN_DEFAULT_TOOLS) to scan all subdomains associated with a domain. Args: host (str): Hostname to scan. Returns: subdomains (list): List of subdomain names. """ if not host: host = self.subdomain.name if self.subdomain else self.domain.name if self.url_filter: logger.warning(f'Ignoring subdomains scan as an URL path filter was passed ({self.url_filter}).') return # Config config = self.yaml_configuration.get(SUBDOMAIN_DISCOVERY) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL) or self.yaml_configuration.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) tools = config.get(USES_TOOLS, SUBDOMAIN_SCAN_DEFAULT_TOOLS) default_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=True).filter(is_subdomain_gathering=True)] custom_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=False).filter(is_subdomain_gathering=True)] send_subdomain_changes, send_interesting = False, False notif = Notification.objects.first() if notif: send_subdomain_changes = notif.send_subdomain_changes_notif send_interesting = notif.send_interesting_notif # Gather tools to run for subdomain scan if ALL in tools: tools = SUBDOMAIN_SCAN_DEFAULT_TOOLS + custom_subdomain_tools tools = [t.lower() for t in tools] # Make exception for amass since tool name is amass, but command is amass-active/passive default_subdomain_tools.append('amass-passive') default_subdomain_tools.append('amass-active') # Run tools for tool in tools: cmd = None logger.info(f'Scanning subdomains for {host} with {tool}') proxy = get_random_proxy() if tool in default_subdomain_tools: if tool == 'amass-passive': cmd = f'amass enum -passive -d {host} -o {self.results_dir}/subdomains_amass.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' elif tool == 'amass-active': use_amass_config = config.get(USE_AMASS_CONFIG, False) amass_wordlist_name = config.get(AMASS_WORDLIST, 'deepmagic.com-prefixes-top50000') wordlist_path = f'/usr/src/wordlist/{amass_wordlist_name}.txt' cmd = f'amass enum -active -d {host} -o {self.results_dir}/subdomains_amass_active.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' cmd += f' -brute -w {wordlist_path}' elif tool == 'sublist3r': cmd = f'python3 /usr/src/github/Sublist3r/sublist3r.py -d {host} -t {threads} -o {self.results_dir}/subdomains_sublister.txt' elif tool == 'subfinder': cmd = f'subfinder -d {host} -o {self.results_dir}/subdomains_subfinder.txt' use_subfinder_config = config.get(USE_SUBFINDER_CONFIG, False) cmd += ' -config /root/.config/subfinder/config.yaml' if use_subfinder_config else '' cmd += f' -proxy {proxy}' if proxy else '' cmd += f' -timeout {timeout}' if timeout else '' cmd += f' -t {threads}' if threads else '' cmd += f' -silent' elif tool == 'oneforall': cmd = f'python3 /usr/src/github/OneForAll/oneforall.py --target {host} run' cmd_extract = f'cut -d\',\' -f6 /usr/src/github/OneForAll/results/{host}.csv > {self.results_dir}/subdomains_oneforall.txt' cmd_rm = f'rm -rf /usr/src/github/OneForAll/results/{host}.csv' cmd += f' && {cmd_extract} && {cmd_rm}' elif tool == 'ctfr': results_file = self.results_dir + '/subdomains_ctfr.txt' cmd = f'python3 /usr/src/github/ctfr/ctfr.py -d {host} -o {results_file}' cmd_extract = f"cat {results_file} | sed 's/\*.//g' | tail -n +12 | uniq | sort > {results_file}" cmd += f' && {cmd_extract}' elif tool == 'tlsx': results_file = self.results_dir + '/subdomains_tlsx.txt' cmd = f'tlsx -san -cn -silent -ro -host {host}' cmd += f" | sed -n '/^\([a-zA-Z0-9]\([-a-zA-Z0-9]*[a-zA-Z0-9]\)\?\.\)\+{host}$/p' | uniq | sort" cmd += f' > {results_file}' elif tool == 'netlas': results_file = self.results_dir + '/subdomains_netlas.txt' cmd = f'netlas search -d domain -i domain domain:"*.{host}" -f json' netlas_key = get_netlas_key() cmd += f' -a {netlas_key}' if netlas_key else '' cmd_extract = f"grep -oE '([a-zA-Z0-9]([-a-zA-Z0-9]*[a-zA-Z0-9])?\.)+{host}'" cmd += f' | {cmd_extract} > {results_file}' elif tool in custom_subdomain_tools: tool_query = InstalledExternalTool.objects.filter(name__icontains=tool.lower()) if not tool_query.exists(): logger.error(f'Missing {{TARGET}} and {{OUTPUT}} placeholders in {tool} configuration. Skipping.') continue custom_tool = tool_query.first() cmd = custom_tool.subdomain_gathering_command if '{TARGET}' in cmd and '{OUTPUT}' in cmd: cmd = cmd.replace('{TARGET}', host) cmd = cmd.replace('{OUTPUT}', f'{self.results_dir}/subdomains_{tool}.txt') cmd = cmd.replace('{PATH}', custom_tool.github_clone_path) if '{PATH}' in cmd else cmd else: logger.warning( f'Subdomain discovery tool "{tool}" is not supported by reNgine. Skipping.') continue # Run tool try: run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) except Exception as e: logger.error( f'Subdomain discovery tool "{tool}" raised an exception') logger.exception(e) # Gather all the tools' results in one single file. Write subdomains into # separate files, and sort all subdomains. run_command( f'cat {self.results_dir}/subdomains_*.txt > {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {self.output_path} -o {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) with open(self.output_path) as f: lines = f.readlines() # Parse the output_file file and store Subdomain and EndPoint objects found # in db. subdomain_count = 0 subdomains = [] urls = [] for line in lines: subdomain_name = line.strip() valid_url = bool(validators.url(subdomain_name)) valid_domain = ( bool(validators.domain(subdomain_name)) or bool(validators.ipv4(subdomain_name)) or bool(validators.ipv6(subdomain_name)) or valid_url ) if not valid_domain: logger.error(f'Subdomain {subdomain_name} is not a valid domain, IP or URL. Skipping.') continue if valid_url: subdomain_name = urlparse(subdomain_name).netloc if subdomain_name in self.out_of_scope_subdomains: logger.error(f'Subdomain {subdomain_name} is out of scope. Skipping.') continue # Add subdomain subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain_count += 1 subdomains.append(subdomain) urls.append(subdomain.name) # Bulk crawl subdomains if enable_http_crawl: ctx['track'] = True http_crawl(urls, ctx=ctx, is_ran_from_subdomain_scan=True) # Find root subdomain endpoints for subdomain in subdomains: pass # Send notifications subdomains_str = '\n'.join([f'• `{subdomain.name}`' for subdomain in subdomains]) self.notify(fields={ 'Subdomain count': len(subdomains), 'Subdomains': subdomains_str, }) if send_subdomain_changes and self.scan_id and self.domain_id: added = get_new_added_subdomain(self.scan_id, self.domain_id) removed = get_removed_subdomain(self.scan_id, self.domain_id) if added: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in added]) self.notify(fields={'Added subdomains': subdomains_str}) if removed: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in removed]) self.notify(fields={'Removed subdomains': subdomains_str}) if send_interesting and self.scan_id and self.domain_id: interesting_subdomains = get_interesting_subdomains(self.scan_id, self.domain_id) if interesting_subdomains: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in interesting_subdomains]) self.notify(fields={'Interesting subdomains': subdomains_str}) return SubdomainSerializer(subdomains, many=True).data @app.task(name='osint', queue='main_scan_queue', base=RengineTask, bind=True) def osint(self, host=None, ctx={}, description=None): """Run Open-Source Intelligence tools on selected domain. Args: host (str): Hostname to scan. Returns: dict: Results from osint discovery and dorking. """ config = self.yaml_configuration.get(OSINT) or OSINT_DEFAULT_CONFIG results = {} grouped_tasks = [] if 'discover' in config: ctx['track'] = False # results = osint_discovery(host=host, ctx=ctx) _task = osint_discovery.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, activity_id=self.activity_id, results_dir=self.results_dir, ctx=ctx ) grouped_tasks.append(_task) if OSINT_DORK in config or OSINT_CUSTOM_DORK in config: _task = dorking.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, results_dir=self.results_dir ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('OSINT Tasks finished...') # with open(self.output_path, 'w') as f: # json.dump(results, f, indent=4) # # return results @app.task(name='osint_discovery', queue='osint_discovery_queue', bind=False) def osint_discovery(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run OSINT discovery. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: dict: osint metadat and theHarvester and h8mail results. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) osint_lookup = config.get(OSINT_DISCOVER, []) osint_intensity = config.get(INTENSITY, 'normal') documents_limit = config.get(OSINT_DOCUMENTS_LIMIT, 50) results = {} meta_info = [] emails = [] creds = [] # Get and save meta info if 'metainfo' in osint_lookup: if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': host, 'domain': host, 'scan_id': scan_history_id, 'documents_limit': documents_limit }) meta_info.append(save_metadata_info(meta_dict)) # TODO: disabled for now # elif osint_intensity == 'deep': # subdomains = Subdomain.objects # if self.scan: # subdomains = subdomains.filter(scan_history=self.scan) # for subdomain in subdomains: # meta_dict = DottedDict({ # 'osint_target': subdomain.name, # 'domain': self.domain, # 'scan_id': self.scan_id, # 'documents_limit': documents_limit # }) # meta_info.append(save_metadata_info(meta_dict)) grouped_tasks = [] if 'emails' in osint_lookup: emails = get_and_save_emails(scan_history, activity_id, results_dir) emails_str = '\n'.join([f'• `{email}`' for email in emails]) # self.notify(fields={'Emails': emails_str}) # ctx['track'] = False _task = h8mail.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) if 'employees' in osint_lookup: ctx['track'] = False _task = theHarvester.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) # results['emails'] = results.get('emails', []) + emails # results['creds'] = creds # results['meta_info'] = meta_info return results @app.task(name='dorking', bind=False, queue='dorking_queue') def dorking(config, host, scan_history_id, results_dir): """Run Google dorks. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: list: Dorking results for each dork ran. """ # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh scan_history = ScanHistory.objects.get(pk=scan_history_id) dorks = config.get(OSINT_DORK, []) custom_dorks = config.get(OSINT_CUSTOM_DORK, []) results = [] # custom dorking has higher priority try: for custom_dork in custom_dorks: lookup_target = custom_dork.get('lookup_site') # replace with original host if _target_ lookup_target = host if lookup_target == '_target_' else lookup_target if 'lookup_extensions' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_extensions=custom_dork.get('lookup_extensions'), scan_history=scan_history ) elif 'lookup_keywords' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_keywords=custom_dork.get('lookup_keywords'), scan_history=scan_history ) except Exception as e: logger.exception(e) # default dorking try: for dork in dorks: logger.info(f'Getting dork information for {dork}') if dork == 'stackoverflow': results = get_and_save_dork_results( lookup_target='stackoverflow.com', results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'login_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/login/,login.html', page_count=5, scan_history=scan_history ) elif dork == 'admin_panels': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/admin/,admin.html', page_count=5, scan_history=scan_history ) elif dork == 'dashboard_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/dashboard/,dashboard.html', page_count=5, scan_history=scan_history ) elif dork == 'social_media' : social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'reddit.com' ] for site in social_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'project_management' : project_websites = [ 'trello.com', 'atlassian.net' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'code_sharing' : project_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'config_files' : config_file_exts = [ 'env', 'xml', 'conf', 'toml', 'yml', 'yaml', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(config_file_exts), page_count=4, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keyword = 'Jenkins' results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=lookup_keyword, page_count=1, scan_history=scan_history ) elif dork == 'wordpress_files' : lookup_keywords = [ '/wp-content/', '/wp-includes/' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'php_error' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'exposed_documents' : docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(docs_file_ext), page_count=7, scan_history=scan_history ) elif dork == 'db_files' : file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) elif dork == 'git_exposed' : file_ext = [ 'git', ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) except Exception as e: logger.exception(e) return results @app.task(name='theHarvester', queue='theHarvester_queue', bind=False) def theHarvester(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run theHarvester to get save emails, hosts, employees found in domain. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: dict: Dict of emails, employees, hosts and ips found during crawling. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) output_path_json = f'{results_dir}/theHarvester.json' theHarvester_dir = '/usr/src/github/theHarvester' history_file = f'{results_dir}/commands.txt' cmd = f'python3 {theHarvester_dir}/theHarvester.py -d {host} -b all -f {output_path_json}' # Update proxies.yaml proxy_query = Proxy.objects.all() if proxy_query.exists(): proxy = proxy_query.first() if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(f'{theHarvester_dir}/proxies.yaml', 'w') as file: yaml.dump(yaml_data, file) # Run cmd run_command( cmd, shell=False, cwd=theHarvester_dir, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) # Get file location if not os.path.isfile(output_path_json): logger.error(f'Could not open {output_path_json}') return {} # Load theHarvester results with open(output_path_json, 'r') as f: data = json.load(f) # Re-indent theHarvester JSON with open(output_path_json, 'w') as f: json.dump(data, f, indent=4) emails = data.get('emails', []) for email_address in emails: email, _ = save_email(email_address, scan_history=scan_history) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) linkedin_people = data.get('linkedin_people', []) for people in linkedin_people: employee, _ = save_employee( people, designation='linkedin', scan_history=scan_history) # if employee: # self.notify(fields={'LinkedIn people': f'• {employee.name}'}) twitter_people = data.get('twitter_people', []) for people in twitter_people: employee, _ = save_employee( people, designation='twitter', scan_history=scan_history) # if employee: # self.notify(fields={'Twitter people': f'• {employee.name}'}) hosts = data.get('hosts', []) urls = [] for host in hosts: split = tuple(host.split(':')) http_url = split[0] subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) endpoint, _ = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain) # if endpoint: # urls.append(endpoint.http_url) # self.notify(fields={'Hosts': f'• {endpoint.http_url}'}) # if enable_http_crawl: # ctx['track'] = False # http_crawl(urls, ctx=ctx) # TODO: Lots of ips unrelated with our domain are found, disabling # this for now. # ips = data.get('ips', []) # for ip_address in ips: # ip, created = save_ip_address( # ip_address, # subscan=subscan) # if ip: # send_task_notif.delay( # 'osint', # scan_history_id=scan_history_id, # subscan_id=subscan_id, # severity='success', # update_fields={'IPs': f'{ip.address}'}) return data @app.task(name='h8mail', queue='h8mail_queue', bind=False) def h8mail(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run h8mail. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: list[dict]: List of credentials info. """ logger.warning('Getting leaked credentials') scan_history = ScanHistory.objects.get(pk=scan_history_id) input_path = f'{results_dir}/emails.txt' output_file = f'{results_dir}/h8mail.json' cmd = f'h8mail -t {input_path} --json {output_file}' history_file = f'{results_dir}/commands.txt' run_command( cmd, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) with open(output_file) as f: data = json.load(f) creds = data.get('targets', []) # TODO: go through h8mail output and save emails to DB for cred in creds: logger.warning(cred) email_address = cred['target'] pwn_num = cred['pwn_num'] pwn_data = cred.get('data', []) email, created = save_email(email_address, scan_history=scan) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) return creds @app.task(name='screenshot', queue='main_scan_queue', base=RengineTask, bind=True) def screenshot(self, ctx={}, description=None): """Uses EyeWitness to gather screenshot of a domain and/or url. Args: description (str, optional): Task description shown in UI. """ # Config screenshots_path = f'{self.results_dir}/screenshots' output_path = f'{self.results_dir}/screenshots/{self.filename}' alive_endpoints_file = f'{self.results_dir}/endpoints_alive.txt' config = self.yaml_configuration.get(SCREENSHOT) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT + 5) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) # If intensity is normal, grab only the root endpoints of each subdomain strict = True if intensity == 'normal' else False # Get URLs to take screenshot of get_http_urls( is_alive=enable_http_crawl, strict=strict, write_filepath=alive_endpoints_file, get_only_default_urls=True, ctx=ctx ) # Send start notif notification = Notification.objects.first() send_output_file = notification.send_scan_output_file if notification else False # Run cmd cmd = f'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py -f {alive_endpoints_file} -d {screenshots_path} --no-prompt' cmd += f' --timeout {timeout}' if timeout > 0 else '' cmd += f' --threads {threads}' if threads > 0 else '' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(output_path): logger.error(f'Could not load EyeWitness results at {output_path} for {self.domain.name}.') return # Loop through results and save objects in DB screenshot_paths = [] with open(output_path, 'r') as file: reader = csv.reader(file) for row in reader: "Protocol,Port,Domain,Request Status,Screenshot Path, Source Path" protocol, port, subdomain_name, status, screenshot_path, source_path = tuple(row) logger.info(f'{protocol}:{port}:{subdomain_name}:{status}') subdomain_query = Subdomain.objects.filter(name=subdomain_name) if self.scan: subdomain_query = subdomain_query.filter(scan_history=self.scan) if status == 'Successful' and subdomain_query.exists(): subdomain = subdomain_query.first() screenshot_paths.append(screenshot_path) subdomain.screenshot_path = screenshot_path.replace('/usr/src/scan_results/', '') subdomain.save() logger.warning(f'Added screenshot for {subdomain.name} to DB') # Remove all db, html extra files in screenshot results run_command( 'rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format(screenshots_path), shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'rm -rf {screenshots_path}/source', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Send finish notifs screenshots_str = '• ' + '\n• '.join([f'`{path}`' for path in screenshot_paths]) self.notify(fields={'Screenshots': screenshots_str}) if send_output_file: for path in screenshot_paths: title = get_output_file_name( self.scan_id, self.subscan_id, self.filename) send_file_to_discord.delay(path, title) @app.task(name='port_scan', queue='main_scan_queue', base=RengineTask, bind=True) def port_scan(self, hosts=[], ctx={}, description=None): """Run port scan. Args: hosts (list, optional): Hosts to run port scan on. description (str, optional): Task description shown in UI. Returns: list: List of open ports (dict). """ input_file = f'{self.results_dir}/input_subdomains_port_scan.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(PORT_SCAN) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) exclude_ports = config.get(NAABU_EXCLUDE_PORTS, []) exclude_subdomains = config.get(NAABU_EXCLUDE_SUBDOMAINS, False) ports = config.get(PORTS, NAABU_DEFAULT_PORTS) ports = [str(port) for port in ports] rate_limit = config.get(NAABU_RATE) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) passive = config.get(NAABU_PASSIVE, False) use_naabu_config = config.get(USE_NAABU_CONFIG, False) exclude_ports_str = ','.join(return_iterable(exclude_ports)) # nmap args nmap_enabled = config.get(ENABLE_NMAP, False) nmap_cmd = config.get(NMAP_COMMAND, '') nmap_script = config.get(NMAP_SCRIPT, '') nmap_script = ','.join(return_iterable(nmap_script)) nmap_script_args = config.get(NMAP_SCRIPT_ARGS) if hosts: with open(input_file, 'w') as f: f.write('\n'.join(hosts)) else: hosts = get_subdomains( write_filepath=input_file, exclude_subdomains=exclude_subdomains, ctx=ctx) # Build cmd cmd = 'naabu -json -exclude-cdn' cmd += f' -list {input_file}' if len(hosts) > 0 else f' -host {hosts[0]}' if 'full' in ports or 'all' in ports: ports_str = ' -p "-"' elif 'top-100' in ports: ports_str = ' -top-ports 100' elif 'top-1000' in ports: ports_str = ' -top-ports 1000' else: ports_str = ','.join(ports) ports_str = f' -p {ports_str}' cmd += ports_str cmd += ' -config /root/.config/naabu/config.yaml' if use_naabu_config else '' cmd += f' -proxy "{proxy}"' if proxy else '' cmd += f' -c {threads}' if threads else '' cmd += f' -rate {rate_limit}' if rate_limit > 0 else '' cmd += f' -timeout {timeout*1000}' if timeout > 0 else '' cmd += f' -passive' if passive else '' cmd += f' -exclude-ports {exclude_ports_str}' if exclude_ports else '' cmd += f' -silent' # Execute cmd and gather results results = [] urls = [] ports_data = {} for line in stream_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) port_number = line['port'] ip_address = line['ip'] host = line.get('host') or ip_address if port_number == 0: continue # Grab subdomain subdomain = Subdomain.objects.filter( name=host, target_domain=self.domain, scan_history=self.scan ).first() # Add IP DB ip, _ = save_ip_address(ip_address, subdomain, subscan=self.subscan) if self.subscan: ip.ip_subscan_ids.add(self.subscan) ip.save() # Add endpoint to DB # port 80 and 443 not needed as http crawl already does that. if port_number not in [80, 443]: http_url = f'{host}:{port_number}' endpoint, _ = save_endpoint( http_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint: http_url = endpoint.http_url urls.append(http_url) # Add Port in DB port_details = whatportis.get_ports(str(port_number)) service_name = port_details[0].name if len(port_details) > 0 else 'unknown' description = port_details[0].description if len(port_details) > 0 else '' # get or create port port, created = Port.objects.get_or_create( number=port_number, service_name=service_name, description=description ) if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port.save() ip.ports.add(port) ip.save() if host in ports_data: ports_data[host].append(port_number) else: ports_data[host] = [port_number] # Send notification logger.warning(f'Found opened port {port_number} on {ip_address} ({host})') if len(ports_data) == 0: logger.info('Finished running naabu port scan - No open ports found.') if nmap_enabled: logger.info('Nmap scans skipped') return ports_data # Send notification fields_str = '' for host, ports in ports_data.items(): ports_str = ', '.join([f'`{port}`' for port in ports]) fields_str += f'• `{host}`: {ports_str}\n' self.notify(fields={'Ports discovered': fields_str}) # Save output to file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) logger.info('Finished running naabu port scan.') # Process nmap results: 1 process per host sigs = [] if nmap_enabled: logger.warning(f'Starting nmap scans ...') logger.warning(ports_data) for host, port_list in ports_data.items(): ports_str = '_'.join([str(p) for p in port_list]) ctx_nmap = ctx.copy() ctx_nmap['description'] = get_task_title(f'nmap_{host}', self.scan_id, self.subscan_id) ctx_nmap['track'] = False sig = nmap.si( cmd=nmap_cmd, ports=port_list, host=host, script=nmap_script, script_args=nmap_script_args, max_rate=rate_limit, ctx=ctx_nmap) sigs.append(sig) task = group(sigs).apply_async() with allow_join_result(): results = task.get() return ports_data @app.task(name='nmap', queue='main_scan_queue', base=RengineTask, bind=True) def nmap( self, cmd=None, ports=[], host=None, input_file=None, script=None, script_args=None, max_rate=None, ctx={}, description=None): """Run nmap on a host. Args: cmd (str, optional): Existing nmap command to complete. ports (list, optional): List of ports to scan. host (str, optional): Host to scan. input_file (str, optional): Input hosts file. script (str, optional): NSE script to run. script_args (str, optional): NSE script args. max_rate (int): Max rate. description (str, optional): Task description shown in UI. """ notif = Notification.objects.first() ports_str = ','.join(str(port) for port in ports) self.filename = self.filename.replace('.txt', '.xml') filename_vulns = self.filename.replace('.xml', '_vulns.json') output_file = self.output_path output_file_xml = f'{self.results_dir}/{host}_{self.filename}' vulns_file = f'{self.results_dir}/{host}_{filename_vulns}' logger.warning(f'Running nmap on {host}:{ports}') # Build cmd nmap_cmd = get_nmap_cmd( cmd=cmd, ports=ports_str, script=script, script_args=script_args, max_rate=max_rate, host=host, input_file=input_file, output_file=output_file_xml) # Run cmd run_command( nmap_cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Get nmap XML results and convert to JSON vulns = parse_nmap_results(output_file_xml, output_file) with open(vulns_file, 'w') as f: json.dump(vulns, f, indent=4) # Save vulnerabilities found by nmap vulns_str = '' for vuln_data in vulns: # URL is not necessarily an HTTP URL when running nmap (can be any # other vulnerable protocols). Look for existing endpoint and use its # URL as vulnerability.http_url if it exists. url = vuln_data['http_url'] endpoint = EndPoint.objects.filter(http_url__contains=url).first() if endpoint: vuln_data['http_url'] = endpoint.http_url vuln, created = save_vulnerability( target_domain=self.domain, subdomain=self.subdomain, scan_history=self.scan, subscan=self.subscan, endpoint=endpoint, **vuln_data) vulns_str += f'• {str(vuln)}\n' if created: logger.warning(str(vuln)) # Send only 1 notif for all vulns to reduce number of notifs if notif and notif.send_vuln_notif and vulns_str: logger.warning(vulns_str) self.notify(fields={'CVEs': vulns_str}) return vulns @app.task(name='waf_detection', queue='main_scan_queue', base=RengineTask, bind=True) def waf_detection(self, ctx={}, description=None): """ Uses wafw00f to check for the presence of a WAF. Args: description (str, optional): Task description shown in UI. Returns: list: List of startScan.models.Waf objects. """ input_path = f'{self.results_dir}/input_endpoints_waf_detection.txt' config = self.yaml_configuration.get(WAF_DETECTION) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Get alive endpoints from DB get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) cmd = f'wafw00f -i {input_path} -o {self.output_path}' run_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(self.output_path): logger.error(f'Could not find {self.output_path}') return with open(self.output_path) as file: wafs = file.readlines() for line in wafs: line = " ".join(line.split()) splitted = line.split(' ', 1) waf_info = splitted[1].strip() waf_name = waf_info[:waf_info.find('(')].strip() waf_manufacturer = waf_info[waf_info.find('(')+1:waf_info.find(')')].strip().replace('.', '') http_url = sanitize_url(splitted[0].strip()) if not waf_name or waf_name == 'None': continue # Add waf to db waf, _ = Waf.objects.get_or_create( name=waf_name, manufacturer=waf_manufacturer ) # Add waf info to Subdomain in DB subdomain = get_subdomain_from_url(http_url) logger.info(f'Wafw00f Subdomain : {subdomain}') subdomain_query, _ = Subdomain.objects.get_or_create(scan_history=self.scan, name=subdomain) subdomain_query.waf.add(waf) subdomain_query.save() return wafs @app.task(name='dir_file_fuzz', queue='main_scan_queue', base=RengineTask, bind=True) def dir_file_fuzz(self, ctx={}, description=None): """Perform directory scan, and currently uses `ffuf` as a default tool. Args: description (str, optional): Task description shown in UI. Returns: list: List of URLs discovered. """ # Config cmd = 'ffuf' config = self.yaml_configuration.get(DIR_FILE_FUZZ) or {} custom_header = self.yaml_configuration.get(CUSTOM_HEADER) auto_calibration = config.get(AUTO_CALIBRATION, True) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) extensions = config.get(EXTENSIONS, DEFAULT_DIR_FILE_FUZZ_EXTENSIONS) # prepend . on extensions extensions = [ext if ext.startswith('.') else '.' + ext for ext in extensions] extensions_str = ','.join(map(str, extensions)) follow_redirect = config.get(FOLLOW_REDIRECT, FFUF_DEFAULT_FOLLOW_REDIRECT) max_time = config.get(MAX_TIME, 0) match_http_status = config.get(MATCH_HTTP_STATUS, FFUF_DEFAULT_MATCH_HTTP_STATUS) mc = ','.join([str(c) for c in match_http_status]) recursive_level = config.get(RECURSIVE_LEVEL, FFUF_DEFAULT_RECURSIVE_LEVEL) stop_on_error = config.get(STOP_ON_ERROR, False) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) wordlist_name = config.get(WORDLIST, 'dicc') delay = rate_limit / (threads * 100) # calculate request pause delay from rate_limit and number of threads input_path = f'{self.results_dir}/input_dir_file_fuzz.txt' # Get wordlist wordlist_name = 'dicc' if wordlist_name == 'default' else wordlist_name wordlist_path = f'/usr/src/wordlist/{wordlist_name}.txt' # Build command cmd += f' -w {wordlist_path}' cmd += f' -e {extensions_str}' if extensions else '' cmd += f' -maxtime {max_time}' if max_time > 0 else '' cmd += f' -p {delay}' if delay > 0 else '' cmd += f' -recursion -recursion-depth {recursive_level} ' if recursive_level > 0 else '' cmd += f' -t {threads}' if threads and threads > 0 else '' cmd += f' -timeout {timeout}' if timeout and timeout > 0 else '' cmd += ' -se' if stop_on_error else '' cmd += ' -fr' if follow_redirect else '' cmd += ' -ac' if auto_calibration else '' cmd += f' -mc {mc}' if mc else '' cmd += f' -H "{custom_header}"' if custom_header else '' # Grab URLs to fuzz urls = get_http_urls( is_alive=True, ignore_files=False, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) logger.warning(urls) # Loop through URLs and run command results = [] for url in urls: ''' Above while fetching urls, we are not ignoring files, because some default urls may redirect to https://example.com/login.php so, ignore_files is set to False but, during fuzzing, we will only need part of the path, in above example it is still a good idea to ffuf base url https://example.com so files from base url ''' url_parse = urlparse(url) url = url_parse.scheme + '://' + url_parse.netloc url += '/FUZZ' # TODO: fuzz not only URL but also POST / PUT / headers proxy = get_random_proxy() # Build final cmd fcmd = cmd fcmd += f' -x {proxy}' if proxy else '' fcmd += f' -u {url} -json' # Initialize DirectoryScan object dirscan = DirectoryScan() dirscan.scanned_date = timezone.now() dirscan.command_line = fcmd dirscan.save() # Loop through results and populate EndPoint and DirectoryFile in DB results = [] for line in stream_command( fcmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) name = line['input'].get('FUZZ') length = line['length'] status = line['status'] words = line['words'] url = line['url'] lines = line['lines'] content_type = line['content-type'] duration = line['duration'] if not name: logger.error(f'FUZZ not found for "{url}"') continue endpoint, created = save_endpoint(url, crawl=False, ctx=ctx) # endpoint.is_default = False endpoint.http_status = status endpoint.content_length = length endpoint.response_time = duration / 1000000000 endpoint.save() if created: urls.append(endpoint.http_url) endpoint.status = status endpoint.content_type = content_type endpoint.content_length = length dfile, created = DirectoryFile.objects.get_or_create( name=name, length=length, words=words, lines=lines, content_type=content_type, url=url) dfile.http_status = status dfile.save() # if created: # logger.warning(f'Found new directory or file {url}') dirscan.directory_files.add(dfile) dirscan.save() if self.subscan: dirscan.dir_subscan_ids.add(self.subscan) subdomain_name = get_subdomain_from_url(endpoint.http_url) subdomain = Subdomain.objects.get(name=subdomain_name, scan_history=self.scan) subdomain.directories.add(dirscan) subdomain.save() # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl(urls, ctx=ctx) return results @app.task(name='fetch_url', queue='main_scan_queue', base=RengineTask, bind=True) def fetch_url(self, urls=[], ctx={}, description=None): """Fetch URLs using different tools like gauplus, gau, gospider, waybackurls ... Args: urls (list): List of URLs to start from. description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/input_endpoints_fetch_url.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(FETCH_URL) or {} should_remove_duplicate_endpoints = config.get(REMOVE_DUPLICATE_ENDPOINTS, True) duplicate_removal_fields = config.get(DUPLICATE_REMOVAL_FIELDS, ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, DEFAULT_GF_PATTERNS) ignore_file_extension = config.get(IGNORE_FILE_EXTENSION, DEFAULT_IGNORE_FILE_EXTENSIONS) tools = config.get(USES_TOOLS, ENDPOINT_SCAN_DEFAULT_TOOLS) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) domain_request_headers = self.domain.request_headers if self.domain else None custom_header = domain_request_headers or self.yaml_configuration.get(CUSTOM_HEADER) exclude_subdomains = config.get(EXCLUDED_SUBDOMAINS, False) # Get URLs to scan and save to input file if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, exclude_subdomains=exclude_subdomains, get_only_default_urls=True, ctx=ctx ) # Domain regex host = self.domain.name if self.domain else urlparse(urls[0]).netloc host_regex = f"\'https?://([a-z0-9]+[.])*{host}.*\'" # Tools cmds cmd_map = { 'gau': f'gau', 'gauplus': f'gauplus -random-agent', 'hakrawler': 'hakrawler -subs -u', 'waybackurls': 'waybackurls', 'gospider': f'gospider -S {input_path} --js -d 2 --sitemap --robots -w -r', 'katana': f'katana -list {input_path} -silent -jc -kf all -d 3 -fs rdn', } if proxy: cmd_map['gau'] += f' --proxy "{proxy}"' cmd_map['gauplus'] += f' -p "{proxy}"' cmd_map['gospider'] += f' -p {proxy}' cmd_map['hakrawler'] += f' -proxy {proxy}' cmd_map['katana'] += f' -proxy {proxy}' if threads > 0: cmd_map['gau'] += f' --threads {threads}' cmd_map['gauplus'] += f' -t {threads}' cmd_map['gospider'] += f' -t {threads}' cmd_map['katana'] += f' -c {threads}' if custom_header: header_string = ';;'.join([ f'{key}: {value}' for key, value in custom_header.items() ]) cmd_map['hakrawler'] += f' -h {header_string}' cmd_map['katana'] += f' -H {header_string}' header_flags = [':'.join(h) for h in header_string.split(';;')] for flag in header_flags: cmd_map['gospider'] += f' -H {flag}' cat_input = f'cat {input_path}' grep_output = f'grep -Eo {host_regex}' cmd_map = { tool: f'{cat_input} | {cmd} | {grep_output} > {self.results_dir}/urls_{tool}.txt' for tool, cmd in cmd_map.items() } tasks = group( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for tool, cmd in cmd_map.items() if tool in tools ) # Cleanup task sort_output = [ f'cat {self.results_dir}/urls_* > {self.output_path}', f'cat {input_path} >> {self.output_path}', f'sort -u {self.output_path} -o {self.output_path}', ] if ignore_file_extension: ignore_exts = '|'.join(ignore_file_extension) grep_ext_filtered_output = [ f'cat {self.output_path} | grep -Eiv "\\.({ignore_exts}).*" > {self.results_dir}/urls_filtered.txt', f'mv {self.results_dir}/urls_filtered.txt {self.output_path}' ] sort_output.extend(grep_ext_filtered_output) cleanup = chain( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for cmd in sort_output ) # Run all commands task = chord(tasks)(cleanup) with allow_join_result(): task.get() # Store all the endpoints and run httpx with open(self.output_path) as f: discovered_urls = f.readlines() self.notify(fields={'Discovered URLs': len(discovered_urls)}) # Some tools can have an URL in the format <URL>] - <PATH> or <URL> - <PATH>, add them # to the final URL list all_urls = [] for url in discovered_urls: url = url.strip() urlpath = None base_url = None if '] ' in url: # found JS scraped endpoint e.g from gospider split = tuple(url.split('] ')) if not len(split) == 2: logger.warning(f'URL format not recognized for "{url}". Skipping.') continue base_url, urlpath = split urlpath = urlpath.lstrip('- ') elif ' - ' in url: # found JS scraped endpoint e.g from gospider base_url, urlpath = tuple(url.split(' - ')) if base_url and urlpath: subdomain = urlparse(base_url) url = f'{subdomain.scheme}://{subdomain.netloc}{self.url_filter}' if not validators.url(url): logger.warning(f'Invalid URL "{url}". Skipping.') if url not in all_urls: all_urls.append(url) # Filter out URLs if a path filter was passed if self.url_filter: all_urls = [url for url in all_urls if self.url_filter in url] # Write result to output path with open(self.output_path, 'w') as f: f.write('\n'.join(all_urls)) logger.warning(f'Found {len(all_urls)} usable URLs') # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl( all_urls, ctx=ctx, should_remove_duplicate_endpoints=should_remove_duplicate_endpoints, duplicate_removal_fields=duplicate_removal_fields ) #-------------------# # GF PATTERNS MATCH # #-------------------# # Combine old gf patterns with new ones if gf_patterns: self.scan.used_gf_patterns = ','.join(gf_patterns) self.scan.save() # Run gf patterns on saved endpoints # TODO: refactor to Celery task for gf_pattern in gf_patterns: # TODO: js var is causing issues, removing for now if gf_pattern == 'jsvar': logger.info('Ignoring jsvar as it is causing issues.') continue # Run gf on current pattern logger.warning(f'Running gf on pattern "{gf_pattern}"') gf_output_file = f'{self.results_dir}/gf_patterns_{gf_pattern}.txt' cmd = f'cat {self.output_path} | gf {gf_pattern} | grep -Eo {host_regex} >> {gf_output_file}' run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Check output file if not os.path.exists(gf_output_file): logger.error(f'Could not find GF output file {gf_output_file}. Skipping GF pattern "{gf_pattern}"') continue # Read output file line by line and with open(gf_output_file, 'r') as f: lines = f.readlines() # Add endpoints / subdomains to DB for url in lines: http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue endpoint, created = save_endpoint( http_url, crawl=False, subdomain=subdomain, ctx=ctx) if not endpoint: continue earlier_pattern = None if not created: earlier_pattern = endpoint.matched_gf_patterns pattern = f'{earlier_pattern},{gf_pattern}' if earlier_pattern else gf_pattern endpoint.matched_gf_patterns = pattern endpoint.save() return all_urls def parse_curl_output(response): # TODO: Enrich from other cURL fields. CURL_REGEX_HTTP_STATUS = f'HTTP\/(?:(?:\d\.?)+)\s(\d+)\s(?:\w+)' http_status = 0 if response: failed = False regex = re.compile(CURL_REGEX_HTTP_STATUS, re.MULTILINE) try: http_status = int(regex.findall(response)[0]) except (KeyError, TypeError, IndexError): pass return { 'http_status': http_status, } @app.task(name='vulnerability_scan', queue='main_scan_queue', bind=True, base=RengineTask) def vulnerability_scan(self, urls=[], ctx={}, description=None): """ This function will serve as an entrypoint to vulnerability scan. All other vulnerability scan will be run from here including nuclei, crlfuzz, etc """ logger.info('Running Vulnerability Scan Queue') config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_run_nuclei = config.get(RUN_NUCLEI, True) should_run_crlfuzz = config.get(RUN_CRLFUZZ, False) should_run_dalfox = config.get(RUN_DALFOX, False) should_run_s3scanner = config.get(RUN_S3SCANNER, True) grouped_tasks = [] if should_run_nuclei: _task = nuclei_scan.si( urls=urls, ctx=ctx, description=f'Nuclei Scan' ) grouped_tasks.append(_task) if should_run_crlfuzz: _task = crlfuzz_scan.si( urls=urls, ctx=ctx, description=f'CRLFuzz Scan' ) grouped_tasks.append(_task) if should_run_dalfox: _task = dalfox_xss_scan.si( urls=urls, ctx=ctx, description=f'Dalfox XSS Scan' ) grouped_tasks.append(_task) if should_run_s3scanner: _task = s3scanner.si( ctx=ctx, description=f'Misconfigured S3 Buckets Scanner' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan completed...') # return results return None @app.task(name='nuclei_individual_severity_module', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_individual_severity_module(self, cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx={}, description=None): ''' This celery task will run vulnerability scan in parallel. All severities supplied should run in parallel as grouped tasks. ''' results = [] logger.info(f'Running vulnerability scan with severity: {severity}') cmd += f' -severity {severity}' # Send start notification notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) # Gather nuclei results vuln_data = parse_nuclei_result(line) # Get corresponding subdomain http_url = sanitize_url(line.get('matched-at')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) # Look for duplicate vulnerabilities by excluding records that might change but are irrelevant. object_comparison_exclude = ['response', 'curl_command', 'tags', 'references', 'cve_ids', 'cwe_ids'] # Add subdomain and target domain to the duplicate check vuln_data_copy = vuln_data.copy() vuln_data_copy['subdomain'] = subdomain vuln_data_copy['target_domain'] = self.domain # Check if record exists, if exists do not save it if record_exists(Vulnerability, data=vuln_data_copy, exclude_keys=object_comparison_exclude): logger.warning(f'Nuclei vulnerability of severity {severity} : {vuln_data_copy["name"]} for {subdomain_name} already exists') continue # Get or create EndPoint object response = line.get('response') httpx_crawl = False if response else enable_http_crawl # avoid yet another httpx crawl endpoint, _ = save_endpoint( http_url, crawl=httpx_crawl, subdomain=subdomain, ctx=ctx) if endpoint: http_url = endpoint.http_url if not httpx_crawl: output = parse_curl_output(response) endpoint.http_status = output['http_status'] endpoint.save() # Get or create Vulnerability object vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, subdomain=subdomain, **vuln_data) if not vuln: continue # Print vuln severity = line['info'].get('severity', 'unknown') logger.warning(str(vuln)) # Send notification for all vulnerabilities except info url = vuln.http_url or vuln.subdomain send_vuln = ( notif and notif.send_vuln_notif and vuln and severity in ['low', 'medium', 'high', 'critical']) if send_vuln: fields = { 'Severity': f'**{severity.upper()}**', 'URL': http_url, 'Subdomain': subdomain_name, 'Name': vuln.name, 'Type': vuln.type, 'Description': vuln.description, 'Template': vuln.template_url, 'Tags': vuln.get_tags_str(), 'CVEs': vuln.get_cve_str(), 'CWEs': vuln.get_cwe_str(), 'References': vuln.get_refs_str() } severity_map = { 'low': 'info', 'medium': 'warning', 'high': 'error', 'critical': 'error' } self.notify( f'vulnerability_scan_#{vuln.id}', severity_map[severity], fields, add_meta_info=False) # Send report to hackerone hackerone_query = Hackerone.objects.all() send_report = ( hackerone_query.exists() and severity not in ('info', 'low') and vuln.target_domain.h1_team_handle ) if send_report: hackerone = hackerone_query.first() if hackerone.send_critical and severity == 'critical': send_hackerone_report.delay(vuln.id) elif hackerone.send_high and severity == 'high': send_hackerone_report.delay(vuln.id) elif hackerone.send_medium and severity == 'medium': send_hackerone_report.delay(vuln.id) # Write results to JSON file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) # Send finish notif if send_status: vulns = Vulnerability.objects.filter(scan_history__id=self.scan_id) info_count = vulns.filter(severity=0).count() low_count = vulns.filter(severity=1).count() medium_count = vulns.filter(severity=2).count() high_count = vulns.filter(severity=3).count() critical_count = vulns.filter(severity=4).count() unknown_count = vulns.filter(severity=-1).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count + unknown_count fields = { 'Total': vulnerability_count, 'Critical': critical_count, 'High': high_count, 'Medium': medium_count, 'Low': low_count, 'Info': info_count, 'Unknown': unknown_count } self.notify(fields=fields) # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=NUCLEI ).exclude( severity=0 ) # find all unique vulnerabilities based on path and title # all unique vulnerability will go thru gpt function and get report # once report is got, it will be matched with other vulnerabilities and saved unique_vulns = set() for vuln in vulns: unique_vulns.add((vuln.name, vuln.get_path())) unique_vulns = list(unique_vulns) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in unique_vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return None def get_vulnerability_gpt_report(vuln): title = vuln[0] path = vuln[1] logger.info(f'Getting GPT Report for {title}, PATH: {path}') # check if in db already exists stored = GPTVulnerabilityReport.objects.filter( url_path=path ).filter( title=title ).first() if stored: response = { 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: report = GPTVulnerabilityReportGenerator() vulnerability_description = get_gpt_vuln_input_description( title, path ) response = report.get_vulnerability_description(vulnerability_description) add_gpt_description_db( title, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) for vuln in Vulnerability.objects.filter(name=title, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() def add_gpt_description_db(title, path, description, impact, remediation, references): gpt_report = GPTVulnerabilityReport() gpt_report.url_path = path gpt_report.title = title gpt_report.description = description gpt_report.impact = impact gpt_report.remediation = remediation gpt_report.save() for url in references: ref, created = VulnerabilityReference.objects.get_or_create(url=url) gpt_report.references.add(ref) gpt_report.save() @app.task(name='nuclei_scan', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_scan(self, urls=[], ctx={}, description=None): """HTTP vulnerability scan using Nuclei Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. Notes: Unfurl the urls to keep only domain and path, will be sent to vuln scan and ignore certain file extensions. Thanks: https://github.com/six2dez/reconftw """ # Config config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} input_path = f'{self.results_dir}/input_endpoints_vulnerability_scan.txt' enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) concurrency = config.get(NUCLEI_CONCURRENCY) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) retries = config.get(RETRIES) or self.yaml_configuration.get(RETRIES, DEFAULT_RETRIES) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) custom_header = config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) should_fetch_gpt_report = config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) proxy = get_random_proxy() nuclei_specific_config = config.get('nuclei', {}) use_nuclei_conf = nuclei_specific_config.get(USE_CONFIG, False) severities = nuclei_specific_config.get(NUCLEI_SEVERITY, NUCLEI_DEFAULT_SEVERITIES) tags = nuclei_specific_config.get(NUCLEI_TAGS, []) tags = ','.join(tags) nuclei_templates = nuclei_specific_config.get(NUCLEI_TEMPLATE) custom_nuclei_templates = nuclei_specific_config.get(NUCLEI_CUSTOM_TEMPLATE) # severities_str = ','.join(severities) # Get alive endpoints if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=enable_http_crawl, ignore_files=True, write_filepath=input_path, ctx=ctx ) if intensity == 'normal': # reduce number of endpoints to scan unfurl_filter = f'{self.results_dir}/urls_unfurled.txt' run_command( f"cat {input_path} | unfurl -u format %s://%d%p |uro > {unfurl_filter}", shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {unfurl_filter} -o {unfurl_filter}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) input_path = unfurl_filter # Build templates # logger.info('Updating Nuclei templates ...') run_command( 'nuclei -update-templates', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) templates = [] if not (nuclei_templates or custom_nuclei_templates): templates.append(NUCLEI_DEFAULT_TEMPLATES_PATH) if nuclei_templates: if ALL in nuclei_templates: template = NUCLEI_DEFAULT_TEMPLATES_PATH templates.append(template) else: templates.extend(nuclei_templates) if custom_nuclei_templates: custom_nuclei_template_paths = [f'{str(elem)}.yaml' for elem in custom_nuclei_templates] template = templates.extend(custom_nuclei_template_paths) # Build CMD cmd = 'nuclei -j' cmd += ' -config /root/.config/nuclei/config.yaml' if use_nuclei_conf else '' cmd += f' -irr' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -l {input_path}' cmd += f' -c {str(concurrency)}' if concurrency > 0 else '' cmd += f' -proxy {proxy} ' if proxy else '' cmd += f' -retries {retries}' if retries > 0 else '' cmd += f' -rl {rate_limit}' if rate_limit > 0 else '' # cmd += f' -severity {severities_str}' cmd += f' -timeout {str(timeout)}' if timeout and timeout > 0 else '' cmd += f' -tags {tags}' if tags else '' cmd += f' -silent' for tpl in templates: cmd += f' -t {tpl}' grouped_tasks = [] custom_ctx = ctx for severity in severities: custom_ctx['track'] = True _task = nuclei_individual_severity_module.si( cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx=custom_ctx, description=f'Nuclei Scan with severity {severity}' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan with all severities completed...') return None @app.task(name='dalfox_xss_scan', queue='main_scan_queue', base=RengineTask, bind=True) def dalfox_xss_scan(self, urls=[], ctx={}, description=None): """XSS Scan using dalfox Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) dalfox_config = vuln_config.get(DALFOX) or {} custom_header = dalfox_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() is_waf_evasion = dalfox_config.get(WAF_EVASION, False) blind_xss_server = dalfox_config.get(BLIND_XSS_SERVER) user_agent = dalfox_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) timeout = dalfox_config.get(TIMEOUT) delay = dalfox_config.get(DELAY) threads = dalfox_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_dalfox_xss.txt' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=False, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'dalfox --silence --no-color --no-spinner' cmd += f' --only-poc r ' cmd += f' --ignore-return 302,404,403' cmd += f' --skip-bav' cmd += f' file {input_path}' cmd += f' --proxy {proxy}' if proxy else '' cmd += f' --waf-evasion' if is_waf_evasion else '' cmd += f' -b {blind_xss_server}' if blind_xss_server else '' cmd += f' --delay {delay}' if delay else '' cmd += f' --timeout {timeout}' if timeout else '' cmd += f' --user-agent {user_agent}' if user_agent else '' cmd += f' --header {custom_header}' if custom_header else '' cmd += f' --worker {threads}' if threads else '' cmd += f' --format json' results = [] for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id, trunc_char=',' ): if not isinstance(line, dict): continue results.append(line) vuln_data = parse_dalfox_result(line) http_url = sanitize_url(line.get('data')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Dalfox Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=DALFOX ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='crlfuzz_scan', queue='main_scan_queue', base=RengineTask, bind=True) def crlfuzz_scan(self, urls=[], ctx={}, description=None): """CRLF Fuzzing with CRLFuzz Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) custom_header = vuln_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() user_agent = vuln_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) threads = vuln_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_crlf.txt' output_path = f'{self.results_dir}/{self.filename}' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=True, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'crlfuzz -s' cmd += f' -l {input_path}' cmd += f' -x {proxy}' if proxy else '' cmd += f' --H {custom_header}' if custom_header else '' cmd += f' -o {output_path}' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id ) if not os.path.isfile(output_path): logger.info('No Results from CRLFuzz') return crlfs = [] results = [] with open(output_path, 'r') as file: crlfs = file.readlines() for crlf in crlfs: url = crlf.strip() vuln_data = parse_crlfuzz_result(url) http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting CRLFuzz Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=CRLFUZZ ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='s3scanner', queue='main_scan_queue', base=RengineTask, bind=True) def s3scanner(self, ctx={}, description=None): """Bucket Scanner Args: ctx (dict): Context description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/#{self.scan_id}_subdomain_discovery.txt' vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} s3_config = vuln_config.get(S3SCANNER) or {} threads = s3_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) providers = s3_config.get(PROVIDERS, S3SCANNER_DEFAULT_PROVIDERS) scan_history = ScanHistory.objects.filter(pk=self.scan_id).first() for provider in providers: cmd = f's3scanner -bucket-file {input_path} -enumerate -provider {provider} -threads {threads} -json' for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue if line.get('bucket', {}).get('exists', 0) == 1: result = parse_s3scanner_result(line) s3bucket, created = S3Bucket.objects.get_or_create(**result) scan_history.buckets.add(s3bucket) logger.info(f"s3 bucket added {result['provider']}-{result['name']}-{result['region']}") @app.task(name='http_crawl', queue='main_scan_queue', base=RengineTask, bind=True) def http_crawl( self, urls=[], method=None, recrawl=False, ctx={}, track=True, description=None, is_ran_from_subdomain_scan=False, should_remove_duplicate_endpoints=True, duplicate_removal_fields=[]): """Use httpx to query HTTP URLs for important info like page titles, http status, etc... Args: urls (list, optional): A set of URLs to check. Overrides default behavior which queries all endpoints related to this scan. method (str): HTTP method to use (GET, HEAD, POST, PUT, DELETE). recrawl (bool, optional): If False, filter out URLs that have already been crawled. should_remove_duplicate_endpoints (bool): Whether to remove duplicate endpoints duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates Returns: list: httpx results. """ logger.info('Initiating HTTP Crawl') if is_ran_from_subdomain_scan: logger.info('Running From Subdomain Scan...') cmd = '/go/bin/httpx' cfg = self.yaml_configuration.get(HTTP_CRAWL) or {} custom_header = cfg.get(CUSTOM_HEADER, '') threads = cfg.get(THREADS, DEFAULT_THREADS) follow_redirect = cfg.get(FOLLOW_REDIRECT, True) self.output_path = None input_path = f'{self.results_dir}/httpx_input.txt' history_file = f'{self.results_dir}/commands.txt' if urls: # direct passing URLs to check if self.url_filter: urls = [u for u in urls if self.url_filter in u] with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_uncrawled=not recrawl, write_filepath=input_path, ctx=ctx ) # logger.debug(urls) # If no URLs found, skip it if not urls: return # Re-adjust thread number if few URLs to avoid spinning up a monster to # kill a fly. if len(urls) < threads: threads = len(urls) # Get random proxy proxy = get_random_proxy() # Run command cmd += f' -cl -ct -rt -location -td -websocket -cname -asn -cdn -probe -random-agent' cmd += f' -t {threads}' if threads > 0 else '' cmd += f' --http-proxy {proxy}' if proxy else '' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -json' cmd += f' -u {urls[0]}' if len(urls) == 1 else f' -l {input_path}' cmd += f' -x {method}' if method else '' cmd += f' -silent' if follow_redirect: cmd += ' -fr' results = [] endpoint_ids = [] for line in stream_command( cmd, history_file=history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not line or not isinstance(line, dict): continue logger.debug(line) # No response from endpoint if line.get('failed', False): continue # Parse httpx output host = line.get('host', '') content_length = line.get('content_length', 0) http_status = line.get('status_code') http_url, is_redirect = extract_httpx_url(line) page_title = line.get('title') webserver = line.get('webserver') cdn = line.get('cdn', False) rt = line.get('time') techs = line.get('tech', []) cname = line.get('cname', '') content_type = line.get('content_type', '') response_time = -1 if rt: response_time = float(''.join(ch for ch in rt if not ch.isalpha())) if rt[-2:] == 'ms': response_time = response_time / 1000 # Create Subdomain object in DB subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue # Save default HTTP URL to endpoint object in DB endpoint, created = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain, is_default=is_ran_from_subdomain_scan ) if not endpoint: continue endpoint.http_status = http_status endpoint.page_title = page_title endpoint.content_length = content_length endpoint.webserver = webserver endpoint.response_time = response_time endpoint.content_type = content_type endpoint.save() endpoint_str = f'{http_url} [{http_status}] `{content_length}B` `{webserver}` `{rt}`' logger.warning(endpoint_str) if endpoint and endpoint.is_alive and endpoint.http_status != 403: self.notify( fields={'Alive endpoint': f'• {endpoint_str}'}, add_meta_info=False) # Add endpoint to results line['_cmd'] = cmd line['final_url'] = http_url line['endpoint_id'] = endpoint.id line['endpoint_created'] = created line['is_redirect'] = is_redirect results.append(line) # Add technology objects to DB for technology in techs: tech, _ = Technology.objects.get_or_create(name=technology) endpoint.techs.add(tech) if is_ran_from_subdomain_scan: subdomain.technologies.add(tech) subdomain.save() endpoint.save() techs_str = ', '.join([f'`{tech}`' for tech in techs]) self.notify( fields={'Technologies': techs_str}, add_meta_info=False) # Add IP objects for 'a' records to DB a_records = line.get('a', []) for ip_address in a_records: ip, created = save_ip_address( ip_address, subdomain, subscan=self.subscan, cdn=cdn) ips_str = '• ' + '\n• '.join([f'`{ip}`' for ip in a_records]) self.notify( fields={'IPs': ips_str}, add_meta_info=False) # Add IP object for host in DB if host: ip, created = save_ip_address( host, subdomain, subscan=self.subscan, cdn=cdn) self.notify( fields={'IPs': f'• `{ip.address}`'}, add_meta_info=False) # Save subdomain and endpoint if is_ran_from_subdomain_scan: # save subdomain stuffs subdomain.http_url = http_url subdomain.http_status = http_status subdomain.page_title = page_title subdomain.content_length = content_length subdomain.webserver = webserver subdomain.response_time = response_time subdomain.content_type = content_type subdomain.cname = ','.join(cname) subdomain.is_cdn = cdn if cdn: subdomain.cdn_name = line.get('cdn_name') subdomain.save() endpoint.save() endpoint_ids.append(endpoint.id) if should_remove_duplicate_endpoints: # Remove 'fake' alive endpoints that are just redirects to the same page remove_duplicate_endpoints( self.scan_id, self.domain_id, self.subdomain_id, filter_ids=endpoint_ids ) # Remove input file run_command( f'rm {input_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) return results #---------------------# # Notifications tasks # #---------------------# @app.task(name='send_notif', bind=False, queue='send_notif_queue') def send_notif( message, scan_history_id=None, subscan_id=None, **options): if not 'title' in options: message = enrich_notification(message, scan_history_id, subscan_id) send_discord_message(message, **options) send_slack_message(message) send_telegram_message(message) @app.task(name='send_scan_notif', bind=False, queue='send_scan_notif_queue') def send_scan_notif( scan_history_id, subscan_id=None, engine_id=None, status='RUNNING'): """Send scan status notification. Works for scan or a subscan if subscan_id is passed. Args: scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Get domain, engine, scan_history objects engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() tasks = ScanActivity.objects.filter(scan_of=scan) if scan else 0 # Build notif options url = get_scan_url(scan_history_id, subscan_id) title = get_scan_title(scan_history_id, subscan_id) fields = get_scan_fields(engine, scan, subscan, status, tasks) severity = None msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) if status: severity = STATUS_TO_SEVERITIES.get(status) opts = { 'title': title, 'url': url, 'fields': fields, 'severity': severity } logger.warning(f'Sending notification "{title}" [{severity}]') # Send notification send_notif( msg, scan_history_id, subscan_id, **opts) @app.task(name='send_task_notif', bind=False, queue='send_task_notif_queue') def send_task_notif( task_name, status=None, result=None, output_path=None, traceback=None, scan_history_id=None, engine_id=None, subscan_id=None, severity=None, add_meta_info=True, update_fields={}): """Send task status notification. Args: task_name (str): Task name. status (str, optional): Task status. result (str, optional): Task result. output_path (str, optional): Task output path. traceback (str, optional): Task traceback. scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. severity (str, optional): Severity (will be mapped to notif colors) add_meta_info (bool, optional): Wheter to add scan / subscan info to notif. update_fields (dict, optional): Fields key / value to update. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Build fields url = None fields = {} if add_meta_info: engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() url = get_scan_url(scan_history_id) if status: fields['Status'] = f'**{status}**' if engine: fields['Engine'] = engine.engine_name if scan: fields['Scan ID'] = f'[#{scan.id}]({url})' if subscan: url = get_scan_url(scan_history_id, subscan_id) fields['Subscan ID'] = f'[#{subscan.id}]({url})' title = get_task_title(task_name, scan_history_id, subscan_id) if status: severity = STATUS_TO_SEVERITIES.get(status) msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) # Add fields to update for k, v in update_fields.items(): fields[k] = v # Add traceback to notif if traceback and notif.send_scan_tracebacks: fields['Traceback'] = f'```\n{traceback}\n```' # Add files to notif files = [] attach_file = ( notif.send_scan_output_file and output_path and result and not traceback ) if attach_file: output_title = output_path.split('/')[-1] files = [(output_path, output_title)] # Send notif opts = { 'title': title, 'url': url, 'files': files, 'severity': severity, 'fields': fields, 'fields_append': update_fields.keys() } send_notif( msg, scan_history_id=scan_history_id, subscan_id=subscan_id, **opts) @app.task(name='send_file_to_discord', bind=False, queue='send_file_to_discord_queue') def send_file_to_discord(file_path, title=None): notif = Notification.objects.first() do_send = notif and notif.send_to_discord and notif.discord_hook_url if not do_send: return False webhook = DiscordWebhook( url=notif.discord_hook_url, rate_limit_retry=True, username=title or "reNgine Discord Plugin" ) with open(file_path, "rb") as f: head, tail = os.path.split(file_path) webhook.add_file(file=f.read(), filename=tail) webhook.execute() @app.task(name='send_hackerone_report', bind=False, queue='send_hackerone_report_queue') def send_hackerone_report(vulnerability_id): """Send HackerOne vulnerability report. Args: vulnerability_id (int): Vulnerability id. Returns: int: HTTP response status code. """ vulnerability = Vulnerability.objects.get(id=vulnerability_id) severities = {v: k for k,v in NUCLEI_SEVERITY_MAP.items()} headers = { 'Content-Type': 'application/json', 'Accept': 'application/json' } # can only send vulnerability report if team_handle exists if len(vulnerability.target_domain.h1_team_handle) !=0: hackerone_query = Hackerone.objects.all() if hackerone_query.exists(): hackerone = Hackerone.objects.first() severity_value = severities[vulnerability.severity] tpl = hackerone.report_template # Replace syntax of report template with actual content tpl = tpl.replace('{vulnerability_name}', vulnerability.name) tpl = tpl.replace('{vulnerable_url}', vulnerability.http_url) tpl = tpl.replace('{vulnerability_severity}', severity_value) tpl = tpl.replace('{vulnerability_description}', vulnerability.description if vulnerability.description else '') tpl = tpl.replace('{vulnerability_extracted_results}', vulnerability.extracted_results if vulnerability.extracted_results else '') tpl = tpl.replace('{vulnerability_reference}', vulnerability.reference if vulnerability.reference else '') data = { "data": { "type": "report", "attributes": { "team_handle": vulnerability.target_domain.h1_team_handle, "title": '{} found in {}'.format(vulnerability.name, vulnerability.http_url), "vulnerability_information": tpl, "severity_rating": severity_value, "impact": "More information about the impact and vulnerability can be found here: \n" + vulnerability.reference if vulnerability.reference else "NA", } } } r = requests.post( 'https://api.hackerone.com/v1/hackers/reports', auth=(hackerone.username, hackerone.api_key), json=data, headers=headers ) response = r.json() status_code = r.status_code if status_code == 201: vulnerability.hackerone_report_id = response['data']["id"] vulnerability.open_status = False vulnerability.save() return status_code else: logger.error('No team handle found.') status_code = 111 return status_code #-------------# # Utils tasks # #-------------# @app.task(name='parse_nmap_results', bind=False, queue='parse_nmap_results_queue') def parse_nmap_results(xml_file, output_file=None): """Parse results from nmap output file. Args: xml_file (str): nmap XML report file path. Returns: list: List of vulnerabilities found from nmap results. """ with open(xml_file, encoding='utf8') as f: content = f.read() try: nmap_results = xmltodict.parse(content) # parse XML to dict except Exception as e: logger.exception(e) logger.error(f'Cannot parse {xml_file} to valid JSON. Skipping.') return [] # Write JSON to output file if output_file: with open(output_file, 'w') as f: json.dump(nmap_results, f, indent=4) logger.warning(json.dumps(nmap_results, indent=4)) hosts = ( nmap_results .get('nmaprun', {}) .get('host', {}) ) all_vulns = [] if isinstance(hosts, dict): hosts = [hosts] for host in hosts: # Grab hostname / IP from output hostnames_dict = host.get('hostnames', {}) if hostnames_dict: # Ensure that hostnames['hostname'] is a list for consistency hostnames_list = hostnames_dict['hostname'] if isinstance(hostnames_dict['hostname'], list) else [hostnames_dict['hostname']] # Extract all the @name values from the list of dictionaries hostnames = [entry.get('@name') for entry in hostnames_list] else: hostnames = [host.get('address')['@addr']] # Iterate over each hostname for each port for hostname in hostnames: # Grab ports from output ports = host.get('ports', {}).get('port', []) if isinstance(ports, dict): ports = [ports] for port in ports: url_vulns = [] port_number = port['@portid'] url = sanitize_url(f'{hostname}:{port_number}') logger.info(f'Parsing nmap results for {hostname}:{port_number} ...') if not port_number or not port_number.isdigit(): continue port_protocol = port['@protocol'] scripts = port.get('script', []) if isinstance(scripts, dict): scripts = [scripts] for script in scripts: script_id = script['@id'] script_output = script['@output'] script_output_table = script.get('table', []) logger.debug(f'Ran nmap script "{script_id}" on {port_number}/{port_protocol}:\n{script_output}\n') if script_id == 'vulscan': vulns = parse_nmap_vulscan_output(script_output) url_vulns.extend(vulns) elif script_id == 'vulners': vulns = parse_nmap_vulners_output(script_output) url_vulns.extend(vulns) # elif script_id == 'http-server-header': # TODO: nmap can help find technologies as well using the http-server-header script # regex = r'(\w+)/([\d.]+)\s?(?:\((\w+)\))?' # tech_name, tech_version, tech_os = re.match(regex, test_string).groups() # Technology.objects.get_or_create(...) # elif script_id == 'http_csrf': # vulns = parse_nmap_http_csrf_output(script_output) # url_vulns.extend(vulns) else: logger.warning(f'Script output parsing for script "{script_id}" is not supported yet.') # Add URL to vuln for vuln in url_vulns: # TODO: This should extend to any URL, not just HTTP vuln['http_url'] = url if 'http_path' in vuln: vuln['http_url'] += vuln['http_path'] all_vulns.append(vuln) return all_vulns def parse_nmap_http_csrf_output(script_output): pass def parse_nmap_vulscan_output(script_output): """Parse nmap vulscan script output. Args: script_output (str): Vulscan script output. Returns: list: List of Vulnerability dicts. """ data = {} vulns = [] provider_name = '' # Sort all vulns found by provider so that we can match each provider with # a function that pulls from its API to get more info about the # vulnerability. for line in script_output.splitlines(): if not line: continue if not line.startswith('['): # provider line if "No findings" in line: logger.info(f"No findings: {line}") continue elif ' - ' in line: provider_name, provider_url = tuple(line.split(' - ')) data[provider_name] = {'url': provider_url.rstrip(':'), 'entries': []} continue else: # Log a warning logger.warning(f"Unexpected line format: {line}") continue reg = r'\[(.*)\] (.*)' matches = re.match(reg, line) id, title = matches.groups() entry = {'id': id, 'title': title} data[provider_name]['entries'].append(entry) logger.warning('Vulscan parsed output:') logger.warning(pprint.pformat(data)) for provider_name in data: if provider_name == 'Exploit-DB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'IBM X-Force': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'MITRE CVE': logger.error(f'Provider {provider_name} is not supported YET.') for entry in data[provider_name]['entries']: cve_id = entry['id'] vuln = cve_to_vuln(cve_id) vulns.append(vuln) elif provider_name == 'OSVDB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'OpenVAS (Nessus)': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'SecurityFocus': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'VulDB': logger.error(f'Provider {provider_name} is not supported YET.') pass else: logger.error(f'Provider {provider_name} is not supported.') return vulns def parse_nmap_vulners_output(script_output, url=''): """Parse nmap vulners script output. TODO: Rework this as it's currently matching all CVEs no matter the confidence. Args: script_output (str): Script output. Returns: list: List of found vulnerabilities. """ vulns = [] # Check for CVE in script output CVE_REGEX = re.compile(r'.*(CVE-\d\d\d\d-\d+).*') matches = CVE_REGEX.findall(script_output) matches = list(dict.fromkeys(matches)) for cve_id in matches: # get CVE info vuln = cve_to_vuln(cve_id, vuln_type='nmap-vulners-nse') if vuln: vulns.append(vuln) return vulns def cve_to_vuln(cve_id, vuln_type=''): """Search for a CVE using CVESearch and return Vulnerability data. Args: cve_id (str): CVE ID in the form CVE-* Returns: dict: Vulnerability dict. """ cve_info = CVESearch('https://cve.circl.lu').id(cve_id) if not cve_info: logger.error(f'Could not fetch CVE info for cve {cve_id}. Skipping.') return None vuln_cve_id = cve_info['id'] vuln_name = vuln_cve_id vuln_description = cve_info.get('summary', 'none').replace(vuln_cve_id, '').strip() try: vuln_cvss = float(cve_info.get('cvss', -1)) except (ValueError, TypeError): vuln_cvss = -1 vuln_cwe_id = cve_info.get('cwe', '') exploit_ids = cve_info.get('refmap', {}).get('exploit-db', []) osvdb_ids = cve_info.get('refmap', {}).get('osvdb', []) references = cve_info.get('references', []) capec_objects = cve_info.get('capec', []) # Parse ovals for a better vuln name / type ovals = cve_info.get('oval', []) if ovals: vuln_name = ovals[0]['title'] vuln_type = ovals[0]['family'] # Set vulnerability severity based on CVSS score vuln_severity = 'info' if vuln_cvss < 4: vuln_severity = 'low' elif vuln_cvss < 7: vuln_severity = 'medium' elif vuln_cvss < 9: vuln_severity = 'high' else: vuln_severity = 'critical' # Build console warning message msg = f'{vuln_name} | {vuln_severity.upper()} | {vuln_cve_id} | {vuln_cwe_id} | {vuln_cvss}' for id in osvdb_ids: msg += f'\n\tOSVDB: {id}' for exploit_id in exploit_ids: msg += f'\n\tEXPLOITDB: {exploit_id}' logger.warning(msg) vuln = { 'name': vuln_name, 'type': vuln_type, 'severity': NUCLEI_SEVERITY_MAP[vuln_severity], 'description': vuln_description, 'cvss_score': vuln_cvss, 'references': references, 'cve_ids': [vuln_cve_id], 'cwe_ids': [vuln_cwe_id] } return vuln def parse_s3scanner_result(line): ''' Parses and returns s3Scanner Data ''' bucket = line['bucket'] return { 'name': bucket['name'], 'region': bucket['region'], 'provider': bucket['provider'], 'owner_display_name': bucket['owner_display_name'], 'owner_id': bucket['owner_id'], 'perm_auth_users_read': bucket['perm_auth_users_read'], 'perm_auth_users_write': bucket['perm_auth_users_write'], 'perm_auth_users_read_acl': bucket['perm_auth_users_read_acl'], 'perm_auth_users_write_acl': bucket['perm_auth_users_write_acl'], 'perm_auth_users_full_control': bucket['perm_auth_users_full_control'], 'perm_all_users_read': bucket['perm_all_users_read'], 'perm_all_users_write': bucket['perm_all_users_write'], 'perm_all_users_read_acl': bucket['perm_all_users_read_acl'], 'perm_all_users_write_acl': bucket['perm_all_users_write_acl'], 'perm_all_users_full_control': bucket['perm_all_users_full_control'], 'num_objects': bucket['num_objects'], 'size': bucket['bucket_size'] } def parse_nuclei_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ return { 'name': line['info'].get('name', ''), 'type': line['type'], 'severity': NUCLEI_SEVERITY_MAP[line['info'].get('severity', 'unknown')], 'template': line['template'], 'template_url': line['template-url'], 'template_id': line['template-id'], 'description': line['info'].get('description', ''), 'matcher_name': line.get('matcher-name', ''), 'curl_command': line.get('curl-command'), 'request': line.get('request'), 'response': line.get('response'), 'extracted_results': line.get('extracted-results', []), 'cvss_metrics': line['info'].get('classification', {}).get('cvss-metrics', ''), 'cvss_score': line['info'].get('classification', {}).get('cvss-score'), 'cve_ids': line['info'].get('classification', {}).get('cve_id', []) or [], 'cwe_ids': line['info'].get('classification', {}).get('cwe_id', []) or [], 'references': line['info'].get('reference', []) or [], 'tags': line['info'].get('tags', []), 'source': NUCLEI, } def parse_dalfox_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ description = '' description += f" Evidence: {line.get('evidence')} <br>" if line.get('evidence') else '' description += f" Message: {line.get('message')} <br>" if line.get('message') else '' description += f" Payload: {line.get('message_str')} <br>" if line.get('message_str') else '' description += f" Vulnerable Parameter: {line.get('param')} <br>" if line.get('param') else '' return { 'name': 'XSS (Cross Site Scripting)', 'type': 'XSS', 'severity': DALFOX_SEVERITY_MAP[line.get('severity', 'unknown')], 'description': description, 'source': DALFOX, 'cwe_ids': [line.get('cwe')] } def parse_crlfuzz_result(url): """Parse CRLF results Args: url (str): CRLF Vulnerable URL Returns: dict: Vulnerability data. """ return { 'name': 'CRLF (HTTP Response Splitting)', 'type': 'CRLF', 'severity': 2, 'description': 'A CRLF (HTTP Response Splitting) vulnerability has been discovered.', 'source': CRLFUZZ, } def record_exists(model, data, exclude_keys=[]): """ Check if a record already exists in the database based on the given data. Args: model (django.db.models.Model): The Django model to check against. data (dict): Data dictionary containing fields and values. exclude_keys (list): List of keys to exclude from the lookup. Returns: bool: True if the record exists, False otherwise. """ # Extract the keys that will be used for the lookup lookup_fields = {key: data[key] for key in data if key not in exclude_keys} # Return True if a record exists based on the lookup fields, False otherwise return model.objects.filter(**lookup_fields).exists() @app.task(name='geo_localize', bind=False, queue='geo_localize_queue') def geo_localize(host, ip_id=None): """Uses geoiplookup to find location associated with host. Args: host (str): Hostname. ip_id (int): IpAddress object id. Returns: startScan.models.CountryISO: CountryISO object from DB or None. """ if validators.ipv6(host): logger.info(f'Ipv6 "{host}" is not supported by geoiplookup. Skipping.') return None cmd = f'geoiplookup {host}' _, out = run_command(cmd) if 'IP Address not found' not in out and "can't resolve hostname" not in out: country_iso = out.split(':')[1].strip().split(',')[0] country_name = out.split(':')[1].strip().split(',')[1].strip() geo_object, _ = CountryISO.objects.get_or_create( iso=country_iso, name=country_name ) geo_json = { 'iso': country_iso, 'name': country_name } if ip_id: ip = IpAddress.objects.get(pk=ip_id) ip.geo_iso = geo_object ip.save() return geo_json logger.info(f'Geo IP lookup failed for host "{host}"') return None @app.task(name='query_whois', bind=False, queue='query_whois_queue') def query_whois(ip_domain, force_reload_whois=False): """Query WHOIS information for an IP or a domain name. Args: ip_domain (str): IP address or domain name. save_domain (bool): Whether to save domain or not, default False Returns: dict: WHOIS information. """ if not force_reload_whois and Domain.objects.filter(name=ip_domain).exists() and Domain.objects.get(name=ip_domain).domain_info: domain = Domain.objects.get(name=ip_domain) if not domain.insert_date: domain.insert_date = timezone.now() domain.save() domain_info_db = domain.domain_info domain_info = DottedDict( dnssec=domain_info_db.dnssec, created=domain_info_db.created, updated=domain_info_db.updated, expires=domain_info_db.expires, geolocation_iso=domain_info_db.geolocation_iso, status=[status['name'] for status in DomainWhoisStatusSerializer(domain_info_db.status, many=True).data], whois_server=domain_info_db.whois_server, ns_records=[ns['name'] for ns in NameServersSerializer(domain_info_db.name_servers, many=True).data], registrar_name=domain_info_db.registrar.name, registrar_phone=domain_info_db.registrar.phone, registrar_email=domain_info_db.registrar.email, registrar_url=domain_info_db.registrar.url, registrant_name=domain_info_db.registrant.name, registrant_id=domain_info_db.registrant.id_str, registrant_organization=domain_info_db.registrant.organization, registrant_city=domain_info_db.registrant.city, registrant_state=domain_info_db.registrant.state, registrant_zip_code=domain_info_db.registrant.zip_code, registrant_country=domain_info_db.registrant.country, registrant_phone=domain_info_db.registrant.phone, registrant_fax=domain_info_db.registrant.fax, registrant_email=domain_info_db.registrant.email, registrant_address=domain_info_db.registrant.address, admin_name=domain_info_db.admin.name, admin_id=domain_info_db.admin.id_str, admin_organization=domain_info_db.admin.organization, admin_city=domain_info_db.admin.city, admin_state=domain_info_db.admin.state, admin_zip_code=domain_info_db.admin.zip_code, admin_country=domain_info_db.admin.country, admin_phone=domain_info_db.admin.phone, admin_fax=domain_info_db.admin.fax, admin_email=domain_info_db.admin.email, admin_address=domain_info_db.admin.address, tech_name=domain_info_db.tech.name, tech_id=domain_info_db.tech.id_str, tech_organization=domain_info_db.tech.organization, tech_city=domain_info_db.tech.city, tech_state=domain_info_db.tech.state, tech_zip_code=domain_info_db.tech.zip_code, tech_country=domain_info_db.tech.country, tech_phone=domain_info_db.tech.phone, tech_fax=domain_info_db.tech.fax, tech_email=domain_info_db.tech.email, tech_address=domain_info_db.tech.address, related_tlds=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_tlds, many=True).data], related_domains=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_domains, many=True).data], historical_ips=[ip for ip in HistoricalIPSerializer(domain_info_db.historical_ips, many=True).data], ) if domain_info_db.dns_records: a_records = [] txt_records = [] mx_records = [] dns_records = [{'name': dns['name'], 'type': dns['type']} for dns in DomainDNSRecordSerializer(domain_info_db.dns_records, many=True).data] for dns in dns_records: if dns['type'] == 'a': a_records.append(dns['name']) elif dns['type'] == 'txt': txt_records.append(dns['name']) elif dns['type'] == 'mx': mx_records.append(dns['name']) domain_info.a_records = a_records domain_info.txt_records = txt_records domain_info.mx_records = mx_records else: logger.info(f'Domain info for "{ip_domain}" not found in DB, querying whois') domain_info = DottedDict() # find domain historical ip try: historical_ips = get_domain_historical_ip_address(ip_domain) domain_info.historical_ips = historical_ips except Exception as e: logger.error(f'HistoricalIP for {ip_domain} not found!\nError: {str(e)}') historical_ips = [] # find associated domains using ip_domain try: related_domains = reverse_whois(ip_domain.split('.')[0]) except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] # find related tlds using TLSx try: related_tlds = [] output_path = '/tmp/ip_domain_tlsx.txt' tlsx_command = f'tlsx -san -cn -silent -ro -host {ip_domain} -o {output_path}' run_command( tlsx_command, shell=True, ) tlsx_output = [] with open(output_path) as f: tlsx_output = f.readlines() tldextract_target = tldextract.extract(ip_domain) for doms in tlsx_output: doms = doms.strip() tldextract_res = tldextract.extract(doms) if ip_domain != doms and tldextract_res.domain == tldextract_target.domain and tldextract_res.subdomain == '': related_tlds.append(doms) related_tlds = list(set(related_tlds)) domain_info.related_tlds = related_tlds except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] related_domains_list = [] if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_related = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_related) related_domains_list.append(_domain['name']) for _domain in related_tlds: domain_related = RelatedDomain.objects.get_or_create( name=_domain, )[0] db_domain_info.related_tlds.add(domain_related) for _ip in historical_ips: historical_ip = HistoricalIP.objects.get_or_create( ip=_ip['ip'], owner=_ip['owner'], location=_ip['location'], last_seen=_ip['last_seen'], )[0] db_domain_info.historical_ips.add(historical_ip) domain.domain_info = db_domain_info domain.save() command = f'netlas host {ip_domain} -f json' # check if netlas key is provided netlas_key = get_netlas_key() command += f' -a {netlas_key}' if netlas_key else '' result = subprocess.check_output(command.split()).decode('utf-8') if 'Failed to parse response data' in result: # do fallback return { 'status': False, 'ip_domain': ip_domain, 'result': "Netlas limit exceeded.", 'message': 'Netlas limit exceeded.' } try: result = json.loads(result) logger.info(result) whois = result.get('whois') if result.get('whois') else {} domain_info.created = whois.get('created_date') domain_info.expires = whois.get('expiration_date') domain_info.updated = whois.get('updated_date') domain_info.whois_server = whois.get('whois_server') if 'registrant' in whois: registrant = whois.get('registrant') domain_info.registrant_name = registrant.get('name') domain_info.registrant_country = registrant.get('country') domain_info.registrant_id = registrant.get('id') domain_info.registrant_state = registrant.get('province') domain_info.registrant_city = registrant.get('city') domain_info.registrant_phone = registrant.get('phone') domain_info.registrant_address = registrant.get('street') domain_info.registrant_organization = registrant.get('organization') domain_info.registrant_fax = registrant.get('fax') domain_info.registrant_zip_code = registrant.get('postal_code') email_search = EMAIL_REGEX.search(str(registrant.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.registrant_email = field_content if 'administrative' in whois: administrative = whois.get('administrative') domain_info.admin_name = administrative.get('name') domain_info.admin_country = administrative.get('country') domain_info.admin_id = administrative.get('id') domain_info.admin_state = administrative.get('province') domain_info.admin_city = administrative.get('city') domain_info.admin_phone = administrative.get('phone') domain_info.admin_address = administrative.get('street') domain_info.admin_organization = administrative.get('organization') domain_info.admin_fax = administrative.get('fax') domain_info.admin_zip_code = administrative.get('postal_code') mail_search = EMAIL_REGEX.search(str(administrative.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.admin_email = field_content if 'technical' in whois: technical = whois.get('technical') domain_info.tech_name = technical.get('name') domain_info.tech_country = technical.get('country') domain_info.tech_state = technical.get('province') domain_info.tech_id = technical.get('id') domain_info.tech_city = technical.get('city') domain_info.tech_phone = technical.get('phone') domain_info.tech_address = technical.get('street') domain_info.tech_organization = technical.get('organization') domain_info.tech_fax = technical.get('fax') domain_info.tech_zip_code = technical.get('postal_code') mail_search = EMAIL_REGEX.search(str(technical.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.tech_email = field_content if 'dns' in result: dns = result.get('dns') domain_info.mx_records = dns.get('mx') domain_info.txt_records = dns.get('txt') domain_info.a_records = dns.get('a') domain_info.ns_records = whois.get('name_servers') domain_info.dnssec = True if whois.get('dnssec') else False domain_info.status = whois.get('status') if 'registrar' in whois: registrar = whois.get('registrar') domain_info.registrar_name = registrar.get('name') domain_info.registrar_email = registrar.get('email') domain_info.registrar_phone = registrar.get('phone') domain_info.registrar_url = registrar.get('url') # find associated domains if registrant email is found related_domains = reverse_whois(domain_info.get('registrant_email')) if domain_info.get('registrant_email') else [] for _domain in related_domains: related_domains_list.append(_domain['name']) # remove duplicate domains from related domains list related_domains_list = list(set(related_domains_list)) domain_info.related_domains = related_domains_list # save to db if domain exists if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_rel = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_rel) db_domain_info.dnssec = domain_info.get('dnssec') #dates db_domain_info.created = domain_info.get('created') db_domain_info.updated = domain_info.get('updated') db_domain_info.expires = domain_info.get('expires') #registrar db_domain_info.registrar = Registrar.objects.get_or_create( name=domain_info.get('registrar_name'), email=domain_info.get('registrar_email'), phone=domain_info.get('registrar_phone'), url=domain_info.get('registrar_url'), )[0] db_domain_info.registrant = DomainRegistration.objects.get_or_create( name=domain_info.get('registrant_name'), organization=domain_info.get('registrant_organization'), address=domain_info.get('registrant_address'), city=domain_info.get('registrant_city'), state=domain_info.get('registrant_state'), zip_code=domain_info.get('registrant_zip_code'), country=domain_info.get('registrant_country'), email=domain_info.get('registrant_email'), phone=domain_info.get('registrant_phone'), fax=domain_info.get('registrant_fax'), id_str=domain_info.get('registrant_id'), )[0] db_domain_info.admin = DomainRegistration.objects.get_or_create( name=domain_info.get('admin_name'), organization=domain_info.get('admin_organization'), address=domain_info.get('admin_address'), city=domain_info.get('admin_city'), state=domain_info.get('admin_state'), zip_code=domain_info.get('admin_zip_code'), country=domain_info.get('admin_country'), email=domain_info.get('admin_email'), phone=domain_info.get('admin_phone'), fax=domain_info.get('admin_fax'), id_str=domain_info.get('admin_id'), )[0] db_domain_info.tech = DomainRegistration.objects.get_or_create( name=domain_info.get('tech_name'), organization=domain_info.get('tech_organization'), address=domain_info.get('tech_address'), city=domain_info.get('tech_city'), state=domain_info.get('tech_state'), zip_code=domain_info.get('tech_zip_code'), country=domain_info.get('tech_country'), email=domain_info.get('tech_email'), phone=domain_info.get('tech_phone'), fax=domain_info.get('tech_fax'), id_str=domain_info.get('tech_id'), )[0] for status in domain_info.get('status') or []: _status = WhoisStatus.objects.get_or_create( name=status )[0] _status.save() db_domain_info.status.add(_status) for ns in domain_info.get('ns_records') or []: _ns = NameServer.objects.get_or_create( name=ns )[0] _ns.save() db_domain_info.name_servers.add(_ns) for a in domain_info.get('a_records') or []: _a = DNSRecord.objects.get_or_create( name=a, type='a' )[0] _a.save() db_domain_info.dns_records.add(_a) for mx in domain_info.get('mx_records') or []: _mx = DNSRecord.objects.get_or_create( name=mx, type='mx' )[0] _mx.save() db_domain_info.dns_records.add(_mx) for txt in domain_info.get('txt_records') or []: _txt = DNSRecord.objects.get_or_create( name=txt, type='txt' )[0] _txt.save() db_domain_info.dns_records.add(_txt) db_domain_info.geolocation_iso = domain_info.get('registrant_country') db_domain_info.whois_server = domain_info.get('whois_server') db_domain_info.save() domain.domain_info = db_domain_info domain.save() except Exception as e: return { 'status': False, 'ip_domain': ip_domain, 'result': "unable to fetch records from WHOIS database.", 'message': str(e) } return { 'status': True, 'ip_domain': ip_domain, 'dnssec': domain_info.get('dnssec'), 'created': domain_info.get('created'), 'updated': domain_info.get('updated'), 'expires': domain_info.get('expires'), 'geolocation_iso': domain_info.get('registrant_country'), 'domain_statuses': domain_info.get('status'), 'whois_server': domain_info.get('whois_server'), 'dns': { 'a': domain_info.get('a_records'), 'mx': domain_info.get('mx_records'), 'txt': domain_info.get('txt_records'), }, 'registrar': { 'name': domain_info.get('registrar_name'), 'phone': domain_info.get('registrar_phone'), 'email': domain_info.get('registrar_email'), 'url': domain_info.get('registrar_url'), }, 'registrant': { 'name': domain_info.get('registrant_name'), 'id': domain_info.get('registrant_id'), 'organization': domain_info.get('registrant_organization'), 'address': domain_info.get('registrant_address'), 'city': domain_info.get('registrant_city'), 'state': domain_info.get('registrant_state'), 'zipcode': domain_info.get('registrant_zip_code'), 'country': domain_info.get('registrant_country'), 'phone': domain_info.get('registrant_phone'), 'fax': domain_info.get('registrant_fax'), 'email': domain_info.get('registrant_email'), }, 'admin': { 'name': domain_info.get('admin_name'), 'id': domain_info.get('admin_id'), 'organization': domain_info.get('admin_organization'), 'address':domain_info.get('admin_address'), 'city': domain_info.get('admin_city'), 'state': domain_info.get('admin_state'), 'zipcode': domain_info.get('admin_zip_code'), 'country': domain_info.get('admin_country'), 'phone': domain_info.get('admin_phone'), 'fax': domain_info.get('admin_fax'), 'email': domain_info.get('admin_email'), }, 'technical_contact': { 'name': domain_info.get('tech_name'), 'id': domain_info.get('tech_id'), 'organization': domain_info.get('tech_organization'), 'address': domain_info.get('tech_address'), 'city': domain_info.get('tech_city'), 'state': domain_info.get('tech_state'), 'zipcode': domain_info.get('tech_zip_code'), 'country': domain_info.get('tech_country'), 'phone': domain_info.get('tech_phone'), 'fax': domain_info.get('tech_fax'), 'email': domain_info.get('tech_email'), }, 'nameservers': domain_info.get('ns_records'), # 'similar_domains': domain_info.get('similar_domains'), 'related_domains': domain_info.get('related_domains'), 'related_tlds': domain_info.get('related_tlds'), 'historical_ips': domain_info.get('historical_ips'), } @app.task(name='remove_duplicate_endpoints', bind=False, queue='remove_duplicate_endpoints_queue') def remove_duplicate_endpoints( scan_history_id, domain_id, subdomain_id=None, filter_ids=[], filter_status=[200, 301, 404], duplicate_removal_fields=ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS ): """Remove duplicate endpoints. Check for implicit redirections by comparing endpoints: - [x] `content_length` similarities indicating redirections - [x] `page_title` (check for same page title) - [ ] Sign-in / login page (check for endpoints with the same words) Args: scan_history_id: ScanHistory id. domain_id (int): Domain id. subdomain_id (int, optional): Subdomain id. filter_ids (list): List of endpoint ids to filter on. filter_status (list): List of HTTP status codes to filter on. duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates """ logger.info(f'Removing duplicate endpoints based on {duplicate_removal_fields}') endpoints = ( EndPoint.objects .filter(scan_history__id=scan_history_id) .filter(target_domain__id=domain_id) ) if filter_status: endpoints = endpoints.filter(http_status__in=filter_status) if subdomain_id: endpoints = endpoints.filter(subdomain__id=subdomain_id) if filter_ids: endpoints = endpoints.filter(id__in=filter_ids) for field_name in duplicate_removal_fields: cl_query = ( endpoints .values_list(field_name) .annotate(mc=Count(field_name)) .order_by('-mc') ) for (field_value, count) in cl_query: if count > DELETE_DUPLICATES_THRESHOLD: eps_to_delete = ( endpoints .filter(**{field_name: field_value}) .order_by('discovered_date') .all()[1:] ) msg = f'Deleting {len(eps_to_delete)} endpoints [reason: same {field_name} {field_value}]' for ep in eps_to_delete: url = urlparse(ep.http_url) if url.path in ['', '/', '/login']: # try do not delete the original page that other pages redirect to continue msg += f'\n\t {ep.http_url} [{ep.http_status}] [{field_name}={field_value}]' ep.delete() logger.warning(msg) @app.task(name='run_command', bind=False, queue='run_command_queue') def run_command(cmd, cwd=None, shell=False, history_file=None, scan_id=None, activity_id=None): """Run a given command using subprocess module. Args: cmd (str): Command to run. cwd (str): Current working directory. echo (bool): Log command. shell (bool): Run within separate shell if True. history_file (str): Write command + output to history file. Returns: tuple: Tuple with return_code, output. """ logger.info(cmd) logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Run the command using subprocess popen = subprocess.Popen( cmd if shell else cmd.split(), shell=shell, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, cwd=cwd, universal_newlines=True) output = '' for stdout_line in iter(popen.stdout.readline, ""): item = stdout_line.strip() output += '\n' + item logger.debug(item) popen.stdout.close() popen.wait() return_code = popen.returncode command_obj.output = output command_obj.return_code = return_code command_obj.save() if history_file: mode = 'a' if not os.path.exists(history_file): mode = 'w' with open(history_file, mode) as f: f.write(f'\n{cmd}\n{return_code}\n{output}\n------------------\n') return return_code, output #-------------# # Other utils # #-------------# def stream_command(cmd, cwd=None, shell=False, history_file=None, encoding='utf-8', scan_id=None, activity_id=None, trunc_char=None): # Log cmd logger.info(cmd) # logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Sanitize the cmd command = cmd if shell else cmd.split() # Run the command using subprocess process = subprocess.Popen( command, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, universal_newlines=True, shell=shell) # Log the output in real-time to the database output = "" # Process the output for line in iter(lambda: process.stdout.readline(), b''): if not line: break line = line.strip() ansi_escape = re.compile(r'\x1B(?:[@-Z\\-_]|\[[0-?]*[ -/]*[@-~])') line = ansi_escape.sub('', line) line = line.replace('\\x0d\\x0a', '\n') if trunc_char and line.endswith(trunc_char): line = line[:-1] item = line # Try to parse the line as JSON try: item = json.loads(line) except json.JSONDecodeError: pass # Yield the line #logger.debug(item) yield item # Add the log line to the output output += line + "\n" # Update the command record in the database command_obj.output = output command_obj.save() # Retrieve the return code and output process.wait() return_code = process.returncode # Update the return code and final output in the database command_obj.return_code = return_code command_obj.save() # Append the command, return code and output to the history file if history_file is not None: with open(history_file, "a") as f: f.write(f"{cmd}\n{return_code}\n{output}\n") def process_httpx_response(line): """TODO: implement this""" def extract_httpx_url(line): """Extract final URL from httpx results. Always follow redirects to find the last URL. Args: line (dict): URL data output by httpx. Returns: tuple: (final_url, redirect_bool) tuple. """ status_code = line.get('status_code', 0) final_url = line.get('final_url') location = line.get('location') chain_status_codes = line.get('chain_status_codes', []) # Final URL is already looking nice, if it exists return it if final_url: return final_url, False http_url = line['url'] # fallback to url field # Handle redirects manually REDIRECT_STATUS_CODES = [301, 302] is_redirect = ( status_code in REDIRECT_STATUS_CODES or any(x in REDIRECT_STATUS_CODES for x in chain_status_codes) ) if is_redirect and location: if location.startswith(('http', 'https')): http_url = location else: http_url = f'{http_url}/{location.lstrip("/")}' # Sanitize URL http_url = sanitize_url(http_url) return http_url, is_redirect #-------------# # OSInt utils # #-------------# def get_and_save_dork_results(lookup_target, results_dir, type, lookup_keywords=None, lookup_extensions=None, delay=3, page_count=2, scan_history=None): """ Uses gofuzz to dork and store information Args: lookup_target (str): target to look into such as stackoverflow or even the target itself results_dir (str): Results directory type (str): Dork Type Title lookup_keywords (str): comma separated keywords or paths to look for lookup_extensions (str): comma separated extensions to look for delay (int): delay between each requests page_count (int): pages in google to extract information scan_history (startScan.ScanHistory): Scan History Object """ results = [] gofuzz_command = f'{GOFUZZ_EXEC_PATH} -t {lookup_target} -d {delay} -p {page_count}' if lookup_extensions: gofuzz_command += f' -e {lookup_extensions}' elif lookup_keywords: gofuzz_command += f' -w {lookup_keywords}' output_file = f'{results_dir}/gofuzz.txt' gofuzz_command += f' -o {output_file}' history_file = f'{results_dir}/commands.txt' try: run_command( gofuzz_command, shell=False, history_file=history_file, scan_id=scan_history.id, ) if not os.path.isfile(output_file): return with open(output_file) as f: for line in f.readlines(): url = line.strip() if url: results.append(url) dork, created = Dork.objects.get_or_create( type=type, url=url ) if scan_history: scan_history.dorks.add(dork) # remove output file os.remove(output_file) except Exception as e: logger.exception(e) return results def get_and_save_emails(scan_history, activity_id, results_dir): """Get and save emails from Google, Bing and Baidu. Args: scan_history (startScan.ScanHistory): Scan history object. activity_id: ScanActivity Object results_dir (str): Results directory. Returns: list: List of emails found. """ emails = [] # Proxy settings # get_random_proxy() # Gather emails from Google, Bing and Baidu output_file = f'{results_dir}/emails_tmp.txt' history_file = f'{results_dir}/commands.txt' command = f'python3 /usr/src/github/Infoga/infoga.py --domain {scan_history.domain.name} --source all --report {output_file}' try: run_command( command, shell=False, history_file=history_file, scan_id=scan_history.id, activity_id=activity_id) if not os.path.isfile(output_file): logger.info('No Email results') return [] with open(output_file) as f: for line in f.readlines(): if 'Email' in line: split_email = line.split(' ')[2] emails.append(split_email) output_path = f'{results_dir}/emails.txt' with open(output_path, 'w') as output_file: for email_address in emails: save_email(email_address, scan_history) output_file.write(f'{email_address}\n') except Exception as e: logger.exception(e) return emails def save_metadata_info(meta_dict): """Extract metadata from Google Search. Args: meta_dict (dict): Info dict. Returns: list: List of startScan.MetaFinderDocument objects. """ logger.warning(f'Getting metadata for {meta_dict.osint_target}') scan_history = ScanHistory.objects.get(id=meta_dict.scan_id) # Proxy settings get_random_proxy() # Get metadata result = extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if not result: logger.error(f'No metadata result from Google Search for {meta_dict.osint_target}.') return [] # Add metadata info to DB results = [] for metadata_name, data in result.get_metadata().items(): subdomain = Subdomain.objects.get( scan_history=meta_dict.scan_id, name=meta_dict.osint_target) metadata = DottedDict({k: v for k, v in data.items()}) meta_finder_document = MetaFinderDocument( subdomain=subdomain, target_domain=meta_dict.domain, scan_history=scan_history, url=metadata.url, doc_name=metadata_name, http_status=metadata.status_code, producer=metadata.metadata.get('Producer'), creator=metadata.metadata.get('Creator'), creation_date=metadata.metadata.get('CreationDate'), modified_date=metadata.metadata.get('ModDate'), author=metadata.metadata.get('Author'), title=metadata.metadata.get('Title'), os=metadata.metadata.get('OSInfo')) meta_finder_document.save() results.append(data) return results #-----------------# # Utils functions # #-----------------# def create_scan_activity(scan_history_id, message, status): scan_activity = ScanActivity() scan_activity.scan_of = ScanHistory.objects.get(pk=scan_history_id) scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id #--------------------# # Database functions # #--------------------# def save_vulnerability(**vuln_data): references = vuln_data.pop('references', []) cve_ids = vuln_data.pop('cve_ids', []) cwe_ids = vuln_data.pop('cwe_ids', []) tags = vuln_data.pop('tags', []) subscan = vuln_data.pop('subscan', None) # remove nulls vuln_data = replace_nulls(vuln_data) # Create vulnerability vuln, created = Vulnerability.objects.get_or_create(**vuln_data) if created: vuln.discovered_date = timezone.now() vuln.open_status = True vuln.save() # Save vuln tags for tag_name in tags or []: tag, created = VulnerabilityTags.objects.get_or_create(name=tag_name) if tag: vuln.tags.add(tag) vuln.save() # Save CVEs for cve_id in cve_ids or []: cve, created = CveId.objects.get_or_create(name=cve_id) if cve: vuln.cve_ids.add(cve) vuln.save() # Save CWEs for cve_id in cwe_ids or []: cwe, created = CweId.objects.get_or_create(name=cve_id) if cwe: vuln.cwe_ids.add(cwe) vuln.save() # Save vuln reference for url in references or []: ref, created = VulnerabilityReference.objects.get_or_create(url=url) if created: vuln.references.add(ref) vuln.save() # Save subscan id in vuln object if subscan: vuln.vuln_subscan_ids.add(subscan) vuln.save() return vuln, created def save_endpoint( http_url, ctx={}, crawl=False, is_default=False, **endpoint_data): """Get or create EndPoint object. If crawl is True, also crawl the endpoint HTTP URL with httpx. Args: http_url (str): Input HTTP URL. is_default (bool): If the url is a default url for SubDomains. scan_history (startScan.models.ScanHistory): ScanHistory object. domain (startScan.models.Domain): Domain object. subdomain (starScan.models.Subdomain): Subdomain object. results_dir (str, optional): Results directory. crawl (bool, optional): Run httpx on endpoint if True. Default: False. force (bool, optional): Force crawl even if ENABLE_HTTP_CRAWL mode is on. subscan (startScan.models.SubScan, optional): SubScan object. Returns: tuple: (startScan.models.EndPoint, created) where `created` is a boolean indicating if the object is new or already existed. """ # remove nulls endpoint_data = replace_nulls(endpoint_data) scheme = urlparse(http_url).scheme endpoint = None created = False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in http_url: logger.error(f"{http_url} is not a URL of domain {domain.name}. Skipping.") return None, False if crawl: ctx['track'] = False results = http_crawl( urls=[http_url], method='HEAD', ctx=ctx) if results: endpoint_data = results[0] endpoint_id = endpoint_data['endpoint_id'] created = endpoint_data['endpoint_created'] endpoint = EndPoint.objects.get(pk=endpoint_id) elif not scheme: return None, False else: # add dumb endpoint without probing it scan = ScanHistory.objects.filter(pk=ctx.get('scan_history_id')).first() domain = Domain.objects.filter(pk=ctx.get('domain_id')).first() if not validators.url(http_url): return None, False http_url = sanitize_url(http_url) endpoint, created = EndPoint.objects.get_or_create( scan_history=scan, target_domain=domain, http_url=http_url, **endpoint_data) if created: endpoint.is_default = is_default endpoint.discovered_date = timezone.now() endpoint.save() subscan_id = ctx.get('subscan_id') if subscan_id: endpoint.endpoint_subscan_ids.add(subscan_id) endpoint.save() return endpoint, created def save_subdomain(subdomain_name, ctx={}): """Get or create Subdomain object. Args: subdomain_name (str): Subdomain name. scan_history (startScan.models.ScanHistory): ScanHistory object. Returns: tuple: (startScan.models.Subdomain, created) where `created` is a boolean indicating if the object has been created in DB. """ scan_id = ctx.get('scan_history_id') subscan_id = ctx.get('subscan_id') out_of_scope_subdomains = ctx.get('out_of_scope_subdomains', []) valid_domain = ( validators.domain(subdomain_name) or validators.ipv4(subdomain_name) or validators.ipv6(subdomain_name) ) if not valid_domain: logger.error(f'{subdomain_name} is not an invalid domain. Skipping.') return None, False if subdomain_name in out_of_scope_subdomains: logger.error(f'{subdomain_name} is out-of-scope. Skipping.') return None, False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in subdomain_name: logger.error(f"{subdomain_name} is not a subdomain of domain {domain.name}. Skipping.") return None, False scan = ScanHistory.objects.filter(pk=scan_id).first() domain = scan.domain if scan else None subdomain, created = Subdomain.objects.get_or_create( scan_history=scan, target_domain=domain, name=subdomain_name) if created: # logger.warning(f'Found new subdomain {subdomain_name}') subdomain.discovered_date = timezone.now() if subscan_id: subdomain.subdomain_subscan_ids.add(subscan_id) subdomain.save() return subdomain, created def save_email(email_address, scan_history=None): if not validators.email(email_address): logger.info(f'Email {email_address} is invalid. Skipping.') return None, False email, created = Email.objects.get_or_create(address=email_address) # if created: # logger.warning(f'Found new email address {email_address}') # Add email to ScanHistory if scan_history: scan_history.emails.add(email) scan_history.save() return email, created def save_employee(name, designation, scan_history=None): employee, created = Employee.objects.get_or_create( name=name, designation=designation) # if created: # logger.warning(f'Found new employee {name}') # Add employee to ScanHistory if scan_history: scan_history.employees.add(employee) scan_history.save() return employee, created def save_ip_address(ip_address, subdomain=None, subscan=None, **kwargs): if not (validators.ipv4(ip_address) or validators.ipv6(ip_address)): logger.info(f'IP {ip_address} is not a valid IP. Skipping.') return None, False ip, created = IpAddress.objects.get_or_create(address=ip_address) # if created: # logger.warning(f'Found new IP {ip_address}') # Set extra attributes for key, value in kwargs.items(): setattr(ip, key, value) ip.save() # Add IP to subdomain if subdomain: subdomain.ip_addresses.add(ip) subdomain.save() # Add subscan to IP if subscan: ip.ip_subscan_ids.add(subscan) # Geo-localize IP asynchronously if created: geo_localize.delay(ip_address, ip.id) return ip, created def save_imported_subdomains(subdomains, ctx={}): """Take a list of subdomains imported and write them to from_imported.txt. Args: subdomains (list): List of subdomain names. scan_history (startScan.models.ScanHistory): ScanHistory instance. domain (startScan.models.Domain): Domain instance. results_dir (str): Results directory. """ domain_id = ctx['domain_id'] domain = Domain.objects.get(pk=domain_id) results_dir = ctx.get('results_dir', RENGINE_RESULTS) # Validate each subdomain and de-duplicate entries subdomains = list(set([ subdomain for subdomain in subdomains if validators.domain(subdomain) and domain.name == get_domain_from_subdomain(subdomain) ])) if not subdomains: return logger.warning(f'Found {len(subdomains)} imported subdomains.') with open(f'{results_dir}/from_imported.txt', 'w+') as output_file: for name in subdomains: subdomain_name = name.strip() subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain.is_imported_subdomain = True subdomain.save() output_file.write(f'{subdomain}\n') @app.task(name='query_reverse_whois', bind=False, queue='query_reverse_whois_queue') def query_reverse_whois(lookup_keyword): """Queries Reverse WHOIS information for an organization or email address. Args: lookup_keyword (str): Registrar Name or email Returns: dict: Reverse WHOIS information. """ return get_associated_domains(lookup_keyword) @app.task(name='query_ip_history', bind=False, queue='query_ip_history_queue') def query_ip_history(domain): """Queries the IP history for a domain Args: domain (str): domain_name Returns: list: list of historical ip addresses """ return get_domain_historical_ip_address(domain) @app.task(name='gpt_vulnerability_description', bind=False, queue='gpt_queue') def gpt_vulnerability_description(vulnerability_id): """Generate and store Vulnerability Description using GPT. Args: vulnerability_id (Vulnerability Model ID): Vulnerability ID to fetch Description. """ logger.info('Getting GPT Vulnerability Description') try: lookup_vulnerability = Vulnerability.objects.get(id=vulnerability_id) lookup_url = urlparse(lookup_vulnerability.http_url) path = lookup_url.path except Exception as e: return { 'status': False, 'error': str(e) } # check in db GPTVulnerabilityReport model if vulnerability description and path matches stored = GPTVulnerabilityReport.objects.filter(url_path=path).filter(title=lookup_vulnerability.name).first() if stored: response = { 'status': True, 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: vulnerability_description = get_gpt_vuln_input_description( lookup_vulnerability.name, path ) # one can add more description here later gpt_generator = GPTVulnerabilityReportGenerator() response = gpt_generator.get_vulnerability_description(vulnerability_description) add_gpt_description_db( lookup_vulnerability.name, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) # for all vulnerabilities with the same vulnerability name this description has to be stored. # also the consition is that the url must contain a part of this. for vuln in Vulnerability.objects.filter(name=lookup_vulnerability.name, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() return response

import csv import json import os import pprint import subprocess import time import validators import whatportis import xmltodict import yaml import tldextract import concurrent.futures from datetime import datetime from urllib.parse import urlparse from api.serializers import SubdomainSerializer from celery import chain, chord, group from celery.result import allow_join_result from celery.utils.log import get_task_logger from django.db.models import Count from dotted_dict import DottedDict from django.utils import timezone from pycvesearch import CVESearch from metafinder.extractor import extract_metadata_from_google_search from reNgine.celery import app from reNgine.gpt import GPTVulnerabilityReportGenerator from reNgine.celery_custom_task import RengineTask from reNgine.common_func import * from reNgine.definitions import * from reNgine.settings import * from reNgine.gpt import * from reNgine.utilities import * from scanEngine.models import (EngineType, InstalledExternalTool, Notification, Proxy) from startScan.models import * from startScan.models import EndPoint, Subdomain, Vulnerability from targetApp.models import Domain """ Celery tasks. """ logger = get_task_logger(__name__) #----------------------# # Scan / Subscan tasks # #----------------------# @app.task(name='initiate_scan', bind=False, queue='initiate_scan_queue') def initiate_scan( scan_history_id, domain_id, engine_id=None, scan_type=LIVE_SCAN, results_dir=RENGINE_RESULTS, imported_subdomains=[], out_of_scope_subdomains=[], url_filter=''): """Initiate a new scan. Args: scan_history_id (int): ScanHistory id. domain_id (int): Domain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. imported_subdomains (list): Imported subdomains. out_of_scope_subdomains (list): Out-of-scope subdomains. url_filter (str): URL path. Default: '' """ # Get scan history scan = ScanHistory.objects.get(pk=scan_history_id) # Get scan engine engine_id = engine_id or scan.scan_type.id # scan history engine_id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, []) # Get domain and set last_scan_date domain = Domain.objects.get(pk=domain_id) domain.last_scan_date = timezone.now() domain.save() # Get path filter url_filter = url_filter.rstrip('/') # Get or create ScanHistory() object if scan_type == LIVE_SCAN: # immediate scan = ScanHistory.objects.get(pk=scan_history_id) scan.scan_status = RUNNING_TASK elif scan_type == SCHEDULED_SCAN: # scheduled scan = ScanHistory() scan.scan_status = INITIATED_TASK scan.scan_type = engine scan.celery_ids = [initiate_scan.request.id] scan.domain = domain scan.start_scan_date = timezone.now() scan.tasks = engine.tasks scan.results_dir = f'{results_dir}/{domain.name}_{scan.id}' add_gf_patterns = gf_patterns and 'fetch_url' in engine.tasks if add_gf_patterns: scan.used_gf_patterns = ','.join(gf_patterns) scan.save() # Create scan results dir os.makedirs(scan.results_dir) # Build task context ctx = { 'scan_history_id': scan_history_id, 'engine_id': engine_id, 'domain_id': domain.id, 'results_dir': scan.results_dir, 'url_filter': url_filter, 'yaml_configuration': config, 'out_of_scope_subdomains': out_of_scope_subdomains } ctx_str = json.dumps(ctx, indent=2) # Send start notif logger.warning(f'Starting scan {scan_history_id} with context:\n{ctx_str}') send_scan_notif.delay( scan_history_id, subscan_id=None, engine_id=engine_id, status=CELERY_TASK_STATUS_MAP[scan.scan_status]) # Save imported subdomains in DB save_imported_subdomains(imported_subdomains, ctx=ctx) # Create initial subdomain in DB: make a copy of domain as a subdomain so # that other tasks using subdomains can use it. subdomain_name = domain.name subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) # If enable_http_crawl is set, create an initial root HTTP endpoint so that # HTTP crawling can start somewhere http_url = f'{domain.name}{url_filter}' if url_filter else domain.name endpoint, _ = save_endpoint( http_url, ctx=ctx, crawl=enable_http_crawl, is_default=True, subdomain=subdomain ) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build Celery tasks, crafted according to the dependency graph below: # subdomain_discovery --> port_scan --> fetch_url --> dir_file_fuzz # osint vulnerability_scan # osint dalfox xss scan # screenshot # waf_detection workflow = chain( group( subdomain_discovery.si(ctx=ctx, description='Subdomain discovery'), osint.si(ctx=ctx, description='OS Intelligence') ), port_scan.si(ctx=ctx, description='Port scan'), fetch_url.si(ctx=ctx, description='Fetch URL'), group( dir_file_fuzz.si(ctx=ctx, description='Directories & files fuzz'), vulnerability_scan.si(ctx=ctx, description='Vulnerability scan'), screenshot.si(ctx=ctx, description='Screenshot'), waf_detection.si(ctx=ctx, description='WAF detection') ) ) # Build callback callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery chord logger.info(f'Running Celery workflow with {len(workflow.tasks) + 1} tasks') task = chain(workflow, callback).on_error(callback).delay() scan.celery_ids.append(task.id) scan.save() return { 'success': True, 'task_id': task.id } @app.task(name='initiate_subscan', bind=False, queue='subscan_queue') def initiate_subscan( scan_history_id, subdomain_id, engine_id=None, scan_type=None, results_dir=RENGINE_RESULTS, url_filter=''): """Initiate a new subscan. Args: scan_history_id (int): ScanHistory id. subdomain_id (int): Subdomain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. url_filter (str): URL path. Default: '' """ # Get Subdomain, Domain and ScanHistory subdomain = Subdomain.objects.get(pk=subdomain_id) scan = ScanHistory.objects.get(pk=subdomain.scan_history.id) domain = Domain.objects.get(pk=subdomain.target_domain.id) # Get EngineType engine_id = engine_id or scan.scan_type.id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Create scan activity of SubScan Model subscan = SubScan( start_scan_date=timezone.now(), celery_ids=[initiate_subscan.request.id], scan_history=scan, subdomain=subdomain, type=scan_type, status=RUNNING_TASK, engine=engine) subscan.save() # Get YAML configuration config = yaml.safe_load(engine.yaml_configuration) # Create results directory results_dir = f'{scan.results_dir}/subscans/{subscan.id}' os.makedirs(results_dir, exist_ok=True) # Run task method = globals().get(scan_type) if not method: logger.warning(f'Task {scan_type} is not supported by reNgine. Skipping') return scan.tasks.append(scan_type) scan.save() # Send start notif send_scan_notif.delay( scan.id, subscan_id=subscan.id, engine_id=engine_id, status='RUNNING') # Build context ctx = { 'scan_history_id': scan.id, 'subscan_id': subscan.id, 'engine_id': engine_id, 'domain_id': domain.id, 'subdomain_id': subdomain.id, 'yaml_configuration': config, 'results_dir': results_dir, 'url_filter': url_filter } # Create initial endpoints in DB: find domain HTTP endpoint so that HTTP # crawling can start somewhere base_url = f'{subdomain.name}{url_filter}' if url_filter else subdomain.name endpoint, _ = save_endpoint( base_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build header + callback workflow = method.si(ctx=ctx) callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery tasks task = chain(workflow, callback).on_error(callback).delay() subscan.celery_ids.append(task.id) subscan.save() return { 'success': True, 'task_id': task.id } @app.task(name='report', bind=False, queue='report_queue') def report(ctx={}, description=None): """Report task running after all other tasks. Mark ScanHistory or SubScan object as completed and update with final status, log run details and send notification. Args: description (str, optional): Task description shown in UI. """ # Get objects subscan_id = ctx.get('subscan_id') scan_id = ctx.get('scan_history_id') engine_id = ctx.get('engine_id') scan = ScanHistory.objects.filter(pk=scan_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() # Get failed tasks tasks = ScanActivity.objects.filter(scan_of=scan).all() if subscan: tasks = tasks.filter(celery_id__in=subscan.celery_ids) failed_tasks = tasks.filter(status=FAILED_TASK) # Get task status failed_count = failed_tasks.count() status = SUCCESS_TASK if failed_count == 0 else FAILED_TASK status_h = 'SUCCESS' if failed_count == 0 else 'FAILED' # Update scan / subscan status if subscan: subscan.stop_scan_date = timezone.now() subscan.status = status subscan.save() else: scan.scan_status = status scan.stop_scan_date = timezone.now() scan.save() # Send scan status notif send_scan_notif.delay( scan_history_id=scan_id, subscan_id=subscan_id, engine_id=engine_id, status=status_h) #------------------------- # # Tracked reNgine tasks # #--------------------------# @app.task(name='subdomain_discovery', queue='main_scan_queue', base=RengineTask, bind=True) def subdomain_discovery( self, host=None, ctx=None, description=None): """Uses a set of tools (see SUBDOMAIN_SCAN_DEFAULT_TOOLS) to scan all subdomains associated with a domain. Args: host (str): Hostname to scan. Returns: subdomains (list): List of subdomain names. """ if not host: host = self.subdomain.name if self.subdomain else self.domain.name if self.url_filter: logger.warning(f'Ignoring subdomains scan as an URL path filter was passed ({self.url_filter}).') return # Config config = self.yaml_configuration.get(SUBDOMAIN_DISCOVERY) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL) or self.yaml_configuration.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) tools = config.get(USES_TOOLS, SUBDOMAIN_SCAN_DEFAULT_TOOLS) default_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=True).filter(is_subdomain_gathering=True)] custom_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=False).filter(is_subdomain_gathering=True)] send_subdomain_changes, send_interesting = False, False notif = Notification.objects.first() if notif: send_subdomain_changes = notif.send_subdomain_changes_notif send_interesting = notif.send_interesting_notif # Gather tools to run for subdomain scan if ALL in tools: tools = SUBDOMAIN_SCAN_DEFAULT_TOOLS + custom_subdomain_tools tools = [t.lower() for t in tools] # Make exception for amass since tool name is amass, but command is amass-active/passive default_subdomain_tools.append('amass-passive') default_subdomain_tools.append('amass-active') # Run tools for tool in tools: cmd = None logger.info(f'Scanning subdomains for {host} with {tool}') proxy = get_random_proxy() if tool in default_subdomain_tools: if tool == 'amass-passive': cmd = f'amass enum -passive -d {host} -o {self.results_dir}/subdomains_amass.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' elif tool == 'amass-active': use_amass_config = config.get(USE_AMASS_CONFIG, False) amass_wordlist_name = config.get(AMASS_WORDLIST, 'deepmagic.com-prefixes-top50000') wordlist_path = f'/usr/src/wordlist/{amass_wordlist_name}.txt' cmd = f'amass enum -active -d {host} -o {self.results_dir}/subdomains_amass_active.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' cmd += f' -brute -w {wordlist_path}' elif tool == 'sublist3r': cmd = f'python3 /usr/src/github/Sublist3r/sublist3r.py -d {host} -t {threads} -o {self.results_dir}/subdomains_sublister.txt' elif tool == 'subfinder': cmd = f'subfinder -d {host} -o {self.results_dir}/subdomains_subfinder.txt' use_subfinder_config = config.get(USE_SUBFINDER_CONFIG, False) cmd += ' -config /root/.config/subfinder/config.yaml' if use_subfinder_config else '' cmd += f' -proxy {proxy}' if proxy else '' cmd += f' -timeout {timeout}' if timeout else '' cmd += f' -t {threads}' if threads else '' cmd += f' -silent' elif tool == 'oneforall': cmd = f'python3 /usr/src/github/OneForAll/oneforall.py --target {host} run' cmd_extract = f'cut -d\',\' -f6 /usr/src/github/OneForAll/results/{host}.csv > {self.results_dir}/subdomains_oneforall.txt' cmd_rm = f'rm -rf /usr/src/github/OneForAll/results/{host}.csv' cmd += f' && {cmd_extract} && {cmd_rm}' elif tool == 'ctfr': results_file = self.results_dir + '/subdomains_ctfr.txt' cmd = f'python3 /usr/src/github/ctfr/ctfr.py -d {host} -o {results_file}' cmd_extract = f"cat {results_file} | sed 's/\*.//g' | tail -n +12 | uniq | sort > {results_file}" cmd += f' && {cmd_extract}' elif tool == 'tlsx': results_file = self.results_dir + '/subdomains_tlsx.txt' cmd = f'tlsx -san -cn -silent -ro -host {host}' cmd += f" | sed -n '/^\([a-zA-Z0-9]\([-a-zA-Z0-9]*[a-zA-Z0-9]\)\?\.\)\+{host}$/p' | uniq | sort" cmd += f' > {results_file}' elif tool == 'netlas': results_file = self.results_dir + '/subdomains_netlas.txt' cmd = f'netlas search -d domain -i domain domain:"*.{host}" -f json' netlas_key = get_netlas_key() cmd += f' -a {netlas_key}' if netlas_key else '' cmd_extract = f"grep -oE '([a-zA-Z0-9]([-a-zA-Z0-9]*[a-zA-Z0-9])?\.)+{host}'" cmd += f' | {cmd_extract} > {results_file}' elif tool in custom_subdomain_tools: tool_query = InstalledExternalTool.objects.filter(name__icontains=tool.lower()) if not tool_query.exists(): logger.error(f'Missing {{TARGET}} and {{OUTPUT}} placeholders in {tool} configuration. Skipping.') continue custom_tool = tool_query.first() cmd = custom_tool.subdomain_gathering_command if '{TARGET}' in cmd and '{OUTPUT}' in cmd: cmd = cmd.replace('{TARGET}', host) cmd = cmd.replace('{OUTPUT}', f'{self.results_dir}/subdomains_{tool}.txt') cmd = cmd.replace('{PATH}', custom_tool.github_clone_path) if '{PATH}' in cmd else cmd else: logger.warning( f'Subdomain discovery tool "{tool}" is not supported by reNgine. Skipping.') continue # Run tool try: run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) except Exception as e: logger.error( f'Subdomain discovery tool "{tool}" raised an exception') logger.exception(e) # Gather all the tools' results in one single file. Write subdomains into # separate files, and sort all subdomains. run_command( f'cat {self.results_dir}/subdomains_*.txt > {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {self.output_path} -o {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) with open(self.output_path) as f: lines = f.readlines() # Parse the output_file file and store Subdomain and EndPoint objects found # in db. subdomain_count = 0 subdomains = [] urls = [] for line in lines: subdomain_name = line.strip() valid_url = bool(validators.url(subdomain_name)) valid_domain = ( bool(validators.domain(subdomain_name)) or bool(validators.ipv4(subdomain_name)) or bool(validators.ipv6(subdomain_name)) or valid_url ) if not valid_domain: logger.error(f'Subdomain {subdomain_name} is not a valid domain, IP or URL. Skipping.') continue if valid_url: subdomain_name = urlparse(subdomain_name).netloc if subdomain_name in self.out_of_scope_subdomains: logger.error(f'Subdomain {subdomain_name} is out of scope. Skipping.') continue # Add subdomain subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain_count += 1 subdomains.append(subdomain) urls.append(subdomain.name) # Bulk crawl subdomains if enable_http_crawl: ctx['track'] = True http_crawl(urls, ctx=ctx, is_ran_from_subdomain_scan=True) # Find root subdomain endpoints for subdomain in subdomains: pass # Send notifications subdomains_str = '\n'.join([f'• `{subdomain.name}`' for subdomain in subdomains]) self.notify(fields={ 'Subdomain count': len(subdomains), 'Subdomains': subdomains_str, }) if send_subdomain_changes and self.scan_id and self.domain_id: added = get_new_added_subdomain(self.scan_id, self.domain_id) removed = get_removed_subdomain(self.scan_id, self.domain_id) if added: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in added]) self.notify(fields={'Added subdomains': subdomains_str}) if removed: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in removed]) self.notify(fields={'Removed subdomains': subdomains_str}) if send_interesting and self.scan_id and self.domain_id: interesting_subdomains = get_interesting_subdomains(self.scan_id, self.domain_id) if interesting_subdomains: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in interesting_subdomains]) self.notify(fields={'Interesting subdomains': subdomains_str}) return SubdomainSerializer(subdomains, many=True).data @app.task(name='osint', queue='main_scan_queue', base=RengineTask, bind=True) def osint(self, host=None, ctx={}, description=None): """Run Open-Source Intelligence tools on selected domain. Args: host (str): Hostname to scan. Returns: dict: Results from osint discovery and dorking. """ config = self.yaml_configuration.get(OSINT) or OSINT_DEFAULT_CONFIG results = {} grouped_tasks = [] if 'discover' in config: ctx['track'] = False # results = osint_discovery(host=host, ctx=ctx) _task = osint_discovery.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, activity_id=self.activity_id, results_dir=self.results_dir, ctx=ctx ) grouped_tasks.append(_task) if OSINT_DORK in config or OSINT_CUSTOM_DORK in config: _task = dorking.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, results_dir=self.results_dir ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('OSINT Tasks finished...') # with open(self.output_path, 'w') as f: # json.dump(results, f, indent=4) # # return results @app.task(name='osint_discovery', queue='osint_discovery_queue', bind=False) def osint_discovery(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run OSINT discovery. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: dict: osint metadat and theHarvester and h8mail results. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) osint_lookup = config.get(OSINT_DISCOVER, []) osint_intensity = config.get(INTENSITY, 'normal') documents_limit = config.get(OSINT_DOCUMENTS_LIMIT, 50) results = {} meta_info = [] emails = [] creds = [] # Get and save meta info if 'metainfo' in osint_lookup: if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': host, 'domain': host, 'scan_id': scan_history_id, 'documents_limit': documents_limit }) meta_info.append(save_metadata_info(meta_dict)) # TODO: disabled for now # elif osint_intensity == 'deep': # subdomains = Subdomain.objects # if self.scan: # subdomains = subdomains.filter(scan_history=self.scan) # for subdomain in subdomains: # meta_dict = DottedDict({ # 'osint_target': subdomain.name, # 'domain': self.domain, # 'scan_id': self.scan_id, # 'documents_limit': documents_limit # }) # meta_info.append(save_metadata_info(meta_dict)) grouped_tasks = [] if 'emails' in osint_lookup: emails = get_and_save_emails(scan_history, activity_id, results_dir) emails_str = '\n'.join([f'• `{email}`' for email in emails]) # self.notify(fields={'Emails': emails_str}) # ctx['track'] = False _task = h8mail.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) if 'employees' in osint_lookup: ctx['track'] = False _task = theHarvester.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) # results['emails'] = results.get('emails', []) + emails # results['creds'] = creds # results['meta_info'] = meta_info return results @app.task(name='dorking', bind=False, queue='dorking_queue') def dorking(config, host, scan_history_id, results_dir): """Run Google dorks. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: list: Dorking results for each dork ran. """ # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh scan_history = ScanHistory.objects.get(pk=scan_history_id) dorks = config.get(OSINT_DORK, []) custom_dorks = config.get(OSINT_CUSTOM_DORK, []) results = [] # custom dorking has higher priority try: for custom_dork in custom_dorks: lookup_target = custom_dork.get('lookup_site') # replace with original host if _target_ lookup_target = host if lookup_target == '_target_' else lookup_target if 'lookup_extensions' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_extensions=custom_dork.get('lookup_extensions'), scan_history=scan_history ) elif 'lookup_keywords' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_keywords=custom_dork.get('lookup_keywords'), scan_history=scan_history ) except Exception as e: logger.exception(e) # default dorking try: for dork in dorks: logger.info(f'Getting dork information for {dork}') if dork == 'stackoverflow': results = get_and_save_dork_results( lookup_target='stackoverflow.com', results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'login_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/login/,login.html', page_count=5, scan_history=scan_history ) elif dork == 'admin_panels': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/admin/,admin.html', page_count=5, scan_history=scan_history ) elif dork == 'dashboard_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/dashboard/,dashboard.html', page_count=5, scan_history=scan_history ) elif dork == 'social_media' : social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'reddit.com' ] for site in social_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'project_management' : project_websites = [ 'trello.com', 'atlassian.net' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'code_sharing' : project_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'config_files' : config_file_exts = [ 'env', 'xml', 'conf', 'toml', 'yml', 'yaml', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(config_file_exts), page_count=4, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keyword = 'Jenkins' results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=lookup_keyword, page_count=1, scan_history=scan_history ) elif dork == 'wordpress_files' : lookup_keywords = [ '/wp-content/', '/wp-includes/' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'php_error' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'exposed_documents' : docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(docs_file_ext), page_count=7, scan_history=scan_history ) elif dork == 'db_files' : file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) elif dork == 'git_exposed' : file_ext = [ 'git', ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) except Exception as e: logger.exception(e) return results @app.task(name='theHarvester', queue='theHarvester_queue', bind=False) def theHarvester(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run theHarvester to get save emails, hosts, employees found in domain. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: dict: Dict of emails, employees, hosts and ips found during crawling. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) output_path_json = f'{results_dir}/theHarvester.json' theHarvester_dir = '/usr/src/github/theHarvester' history_file = f'{results_dir}/commands.txt' cmd = f'python3 {theHarvester_dir}/theHarvester.py -d {host} -b all -f {output_path_json}' # Update proxies.yaml proxy_query = Proxy.objects.all() if proxy_query.exists(): proxy = proxy_query.first() if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(f'{theHarvester_dir}/proxies.yaml', 'w') as file: yaml.dump(yaml_data, file) # Run cmd run_command( cmd, shell=False, cwd=theHarvester_dir, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) # Get file location if not os.path.isfile(output_path_json): logger.error(f'Could not open {output_path_json}') return {} # Load theHarvester results with open(output_path_json, 'r') as f: data = json.load(f) # Re-indent theHarvester JSON with open(output_path_json, 'w') as f: json.dump(data, f, indent=4) emails = data.get('emails', []) for email_address in emails: email, _ = save_email(email_address, scan_history=scan_history) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) linkedin_people = data.get('linkedin_people', []) for people in linkedin_people: employee, _ = save_employee( people, designation='linkedin', scan_history=scan_history) # if employee: # self.notify(fields={'LinkedIn people': f'• {employee.name}'}) twitter_people = data.get('twitter_people', []) for people in twitter_people: employee, _ = save_employee( people, designation='twitter', scan_history=scan_history) # if employee: # self.notify(fields={'Twitter people': f'• {employee.name}'}) hosts = data.get('hosts', []) urls = [] for host in hosts: split = tuple(host.split(':')) http_url = split[0] subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) endpoint, _ = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain) # if endpoint: # urls.append(endpoint.http_url) # self.notify(fields={'Hosts': f'• {endpoint.http_url}'}) # if enable_http_crawl: # ctx['track'] = False # http_crawl(urls, ctx=ctx) # TODO: Lots of ips unrelated with our domain are found, disabling # this for now. # ips = data.get('ips', []) # for ip_address in ips: # ip, created = save_ip_address( # ip_address, # subscan=subscan) # if ip: # send_task_notif.delay( # 'osint', # scan_history_id=scan_history_id, # subscan_id=subscan_id, # severity='success', # update_fields={'IPs': f'{ip.address}'}) return data @app.task(name='h8mail', queue='h8mail_queue', bind=False) def h8mail(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run h8mail. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: list[dict]: List of credentials info. """ logger.warning('Getting leaked credentials') scan_history = ScanHistory.objects.get(pk=scan_history_id) input_path = f'{results_dir}/emails.txt' output_file = f'{results_dir}/h8mail.json' cmd = f'h8mail -t {input_path} --json {output_file}' history_file = f'{results_dir}/commands.txt' run_command( cmd, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) with open(output_file) as f: data = json.load(f) creds = data.get('targets', []) # TODO: go through h8mail output and save emails to DB for cred in creds: logger.warning(cred) email_address = cred['target'] pwn_num = cred['pwn_num'] pwn_data = cred.get('data', []) email, created = save_email(email_address, scan_history=scan) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) return creds @app.task(name='screenshot', queue='main_scan_queue', base=RengineTask, bind=True) def screenshot(self, ctx={}, description=None): """Uses EyeWitness to gather screenshot of a domain and/or url. Args: description (str, optional): Task description shown in UI. """ # Config screenshots_path = f'{self.results_dir}/screenshots' output_path = f'{self.results_dir}/screenshots/{self.filename}' alive_endpoints_file = f'{self.results_dir}/endpoints_alive.txt' config = self.yaml_configuration.get(SCREENSHOT) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT + 5) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) # If intensity is normal, grab only the root endpoints of each subdomain strict = True if intensity == 'normal' else False # Get URLs to take screenshot of get_http_urls( is_alive=enable_http_crawl, strict=strict, write_filepath=alive_endpoints_file, get_only_default_urls=True, ctx=ctx ) # Send start notif notification = Notification.objects.first() send_output_file = notification.send_scan_output_file if notification else False # Run cmd cmd = f'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py -f {alive_endpoints_file} -d {screenshots_path} --no-prompt' cmd += f' --timeout {timeout}' if timeout > 0 else '' cmd += f' --threads {threads}' if threads > 0 else '' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(output_path): logger.error(f'Could not load EyeWitness results at {output_path} for {self.domain.name}.') return # Loop through results and save objects in DB screenshot_paths = [] with open(output_path, 'r') as file: reader = csv.reader(file) for row in reader: "Protocol,Port,Domain,Request Status,Screenshot Path, Source Path" protocol, port, subdomain_name, status, screenshot_path, source_path = tuple(row) logger.info(f'{protocol}:{port}:{subdomain_name}:{status}') subdomain_query = Subdomain.objects.filter(name=subdomain_name) if self.scan: subdomain_query = subdomain_query.filter(scan_history=self.scan) if status == 'Successful' and subdomain_query.exists(): subdomain = subdomain_query.first() screenshot_paths.append(screenshot_path) subdomain.screenshot_path = screenshot_path.replace('/usr/src/scan_results/', '') subdomain.save() logger.warning(f'Added screenshot for {subdomain.name} to DB') # Remove all db, html extra files in screenshot results run_command( 'rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format(screenshots_path), shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'rm -rf {screenshots_path}/source', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Send finish notifs screenshots_str = '• ' + '\n• '.join([f'`{path}`' for path in screenshot_paths]) self.notify(fields={'Screenshots': screenshots_str}) if send_output_file: for path in screenshot_paths: title = get_output_file_name( self.scan_id, self.subscan_id, self.filename) send_file_to_discord.delay(path, title) @app.task(name='port_scan', queue='main_scan_queue', base=RengineTask, bind=True) def port_scan(self, hosts=[], ctx={}, description=None): """Run port scan. Args: hosts (list, optional): Hosts to run port scan on. description (str, optional): Task description shown in UI. Returns: list: List of open ports (dict). """ input_file = f'{self.results_dir}/input_subdomains_port_scan.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(PORT_SCAN) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) exclude_ports = config.get(NAABU_EXCLUDE_PORTS, []) exclude_subdomains = config.get(NAABU_EXCLUDE_SUBDOMAINS, False) ports = config.get(PORTS, NAABU_DEFAULT_PORTS) ports = [str(port) for port in ports] rate_limit = config.get(NAABU_RATE) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) passive = config.get(NAABU_PASSIVE, False) use_naabu_config = config.get(USE_NAABU_CONFIG, False) exclude_ports_str = ','.join(return_iterable(exclude_ports)) # nmap args nmap_enabled = config.get(ENABLE_NMAP, False) nmap_cmd = config.get(NMAP_COMMAND, '') nmap_script = config.get(NMAP_SCRIPT, '') nmap_script = ','.join(return_iterable(nmap_script)) nmap_script_args = config.get(NMAP_SCRIPT_ARGS) if hosts: with open(input_file, 'w') as f: f.write('\n'.join(hosts)) else: hosts = get_subdomains( write_filepath=input_file, exclude_subdomains=exclude_subdomains, ctx=ctx) # Build cmd cmd = 'naabu -json -exclude-cdn' cmd += f' -list {input_file}' if len(hosts) > 0 else f' -host {hosts[0]}' if 'full' in ports or 'all' in ports: ports_str = ' -p "-"' elif 'top-100' in ports: ports_str = ' -top-ports 100' elif 'top-1000' in ports: ports_str = ' -top-ports 1000' else: ports_str = ','.join(ports) ports_str = f' -p {ports_str}' cmd += ports_str cmd += ' -config /root/.config/naabu/config.yaml' if use_naabu_config else '' cmd += f' -proxy "{proxy}"' if proxy else '' cmd += f' -c {threads}' if threads else '' cmd += f' -rate {rate_limit}' if rate_limit > 0 else '' cmd += f' -timeout {timeout*1000}' if timeout > 0 else '' cmd += f' -passive' if passive else '' cmd += f' -exclude-ports {exclude_ports_str}' if exclude_ports else '' cmd += f' -silent' # Execute cmd and gather results results = [] urls = [] ports_data = {} for line in stream_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) port_number = line['port'] ip_address = line['ip'] host = line.get('host') or ip_address if port_number == 0: continue # Grab subdomain subdomain = Subdomain.objects.filter( name=host, target_domain=self.domain, scan_history=self.scan ).first() # Add IP DB ip, _ = save_ip_address(ip_address, subdomain, subscan=self.subscan) if self.subscan: ip.ip_subscan_ids.add(self.subscan) ip.save() # Add endpoint to DB # port 80 and 443 not needed as http crawl already does that. if port_number not in [80, 443]: http_url = f'{host}:{port_number}' endpoint, _ = save_endpoint( http_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint: http_url = endpoint.http_url urls.append(http_url) # Add Port in DB port_details = whatportis.get_ports(str(port_number)) service_name = port_details[0].name if len(port_details) > 0 else 'unknown' description = port_details[0].description if len(port_details) > 0 else '' # get or create port port, created = Port.objects.get_or_create( number=port_number, service_name=service_name, description=description ) if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port.save() ip.ports.add(port) ip.save() if host in ports_data: ports_data[host].append(port_number) else: ports_data[host] = [port_number] # Send notification logger.warning(f'Found opened port {port_number} on {ip_address} ({host})') if len(ports_data) == 0: logger.info('Finished running naabu port scan - No open ports found.') if nmap_enabled: logger.info('Nmap scans skipped') return ports_data # Send notification fields_str = '' for host, ports in ports_data.items(): ports_str = ', '.join([f'`{port}`' for port in ports]) fields_str += f'• `{host}`: {ports_str}\n' self.notify(fields={'Ports discovered': fields_str}) # Save output to file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) logger.info('Finished running naabu port scan.') # Process nmap results: 1 process per host sigs = [] if nmap_enabled: logger.warning(f'Starting nmap scans ...') logger.warning(ports_data) for host, port_list in ports_data.items(): ports_str = '_'.join([str(p) for p in port_list]) ctx_nmap = ctx.copy() ctx_nmap['description'] = get_task_title(f'nmap_{host}', self.scan_id, self.subscan_id) ctx_nmap['track'] = False sig = nmap.si( cmd=nmap_cmd, ports=port_list, host=host, script=nmap_script, script_args=nmap_script_args, max_rate=rate_limit, ctx=ctx_nmap) sigs.append(sig) task = group(sigs).apply_async() with allow_join_result(): results = task.get() return ports_data @app.task(name='nmap', queue='main_scan_queue', base=RengineTask, bind=True) def nmap( self, cmd=None, ports=[], host=None, input_file=None, script=None, script_args=None, max_rate=None, ctx={}, description=None): """Run nmap on a host. Args: cmd (str, optional): Existing nmap command to complete. ports (list, optional): List of ports to scan. host (str, optional): Host to scan. input_file (str, optional): Input hosts file. script (str, optional): NSE script to run. script_args (str, optional): NSE script args. max_rate (int): Max rate. description (str, optional): Task description shown in UI. """ notif = Notification.objects.first() ports_str = ','.join(str(port) for port in ports) self.filename = self.filename.replace('.txt', '.xml') filename_vulns = self.filename.replace('.xml', '_vulns.json') output_file = self.output_path output_file_xml = f'{self.results_dir}/{host}_{self.filename}' vulns_file = f'{self.results_dir}/{host}_{filename_vulns}' logger.warning(f'Running nmap on {host}:{ports}') # Build cmd nmap_cmd = get_nmap_cmd( cmd=cmd, ports=ports_str, script=script, script_args=script_args, max_rate=max_rate, host=host, input_file=input_file, output_file=output_file_xml) # Run cmd run_command( nmap_cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Get nmap XML results and convert to JSON vulns = parse_nmap_results(output_file_xml, output_file) with open(vulns_file, 'w') as f: json.dump(vulns, f, indent=4) # Save vulnerabilities found by nmap vulns_str = '' for vuln_data in vulns: # URL is not necessarily an HTTP URL when running nmap (can be any # other vulnerable protocols). Look for existing endpoint and use its # URL as vulnerability.http_url if it exists. url = vuln_data['http_url'] endpoint = EndPoint.objects.filter(http_url__contains=url).first() if endpoint: vuln_data['http_url'] = endpoint.http_url vuln, created = save_vulnerability( target_domain=self.domain, subdomain=self.subdomain, scan_history=self.scan, subscan=self.subscan, endpoint=endpoint, **vuln_data) vulns_str += f'• {str(vuln)}\n' if created: logger.warning(str(vuln)) # Send only 1 notif for all vulns to reduce number of notifs if notif and notif.send_vuln_notif and vulns_str: logger.warning(vulns_str) self.notify(fields={'CVEs': vulns_str}) return vulns @app.task(name='waf_detection', queue='main_scan_queue', base=RengineTask, bind=True) def waf_detection(self, ctx={}, description=None): """ Uses wafw00f to check for the presence of a WAF. Args: description (str, optional): Task description shown in UI. Returns: list: List of startScan.models.Waf objects. """ input_path = f'{self.results_dir}/input_endpoints_waf_detection.txt' config = self.yaml_configuration.get(WAF_DETECTION) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Get alive endpoints from DB get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) cmd = f'wafw00f -i {input_path} -o {self.output_path}' run_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(self.output_path): logger.error(f'Could not find {self.output_path}') return with open(self.output_path) as file: wafs = file.readlines() for line in wafs: line = " ".join(line.split()) splitted = line.split(' ', 1) waf_info = splitted[1].strip() waf_name = waf_info[:waf_info.find('(')].strip() waf_manufacturer = waf_info[waf_info.find('(')+1:waf_info.find(')')].strip().replace('.', '') http_url = sanitize_url(splitted[0].strip()) if not waf_name or waf_name == 'None': continue # Add waf to db waf, _ = Waf.objects.get_or_create( name=waf_name, manufacturer=waf_manufacturer ) # Add waf info to Subdomain in DB subdomain = get_subdomain_from_url(http_url) logger.info(f'Wafw00f Subdomain : {subdomain}') subdomain_query, _ = Subdomain.objects.get_or_create(scan_history=self.scan, name=subdomain) subdomain_query.waf.add(waf) subdomain_query.save() return wafs @app.task(name='dir_file_fuzz', queue='main_scan_queue', base=RengineTask, bind=True) def dir_file_fuzz(self, ctx={}, description=None): """Perform directory scan, and currently uses `ffuf` as a default tool. Args: description (str, optional): Task description shown in UI. Returns: list: List of URLs discovered. """ # Config cmd = 'ffuf' config = self.yaml_configuration.get(DIR_FILE_FUZZ) or {} custom_header = self.yaml_configuration.get(CUSTOM_HEADER) auto_calibration = config.get(AUTO_CALIBRATION, True) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) extensions = config.get(EXTENSIONS, DEFAULT_DIR_FILE_FUZZ_EXTENSIONS) # prepend . on extensions extensions = [ext if ext.startswith('.') else '.' + ext for ext in extensions] extensions_str = ','.join(map(str, extensions)) follow_redirect = config.get(FOLLOW_REDIRECT, FFUF_DEFAULT_FOLLOW_REDIRECT) max_time = config.get(MAX_TIME, 0) match_http_status = config.get(MATCH_HTTP_STATUS, FFUF_DEFAULT_MATCH_HTTP_STATUS) mc = ','.join([str(c) for c in match_http_status]) recursive_level = config.get(RECURSIVE_LEVEL, FFUF_DEFAULT_RECURSIVE_LEVEL) stop_on_error = config.get(STOP_ON_ERROR, False) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) wordlist_name = config.get(WORDLIST, 'dicc') delay = rate_limit / (threads * 100) # calculate request pause delay from rate_limit and number of threads input_path = f'{self.results_dir}/input_dir_file_fuzz.txt' # Get wordlist wordlist_name = 'dicc' if wordlist_name == 'default' else wordlist_name wordlist_path = f'/usr/src/wordlist/{wordlist_name}.txt' # Build command cmd += f' -w {wordlist_path}' cmd += f' -e {extensions_str}' if extensions else '' cmd += f' -maxtime {max_time}' if max_time > 0 else '' cmd += f' -p {delay}' if delay > 0 else '' cmd += f' -recursion -recursion-depth {recursive_level} ' if recursive_level > 0 else '' cmd += f' -t {threads}' if threads and threads > 0 else '' cmd += f' -timeout {timeout}' if timeout and timeout > 0 else '' cmd += ' -se' if stop_on_error else '' cmd += ' -fr' if follow_redirect else '' cmd += ' -ac' if auto_calibration else '' cmd += f' -mc {mc}' if mc else '' cmd += f' -H "{custom_header}"' if custom_header else '' # Grab URLs to fuzz urls = get_http_urls( is_alive=True, ignore_files=False, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) logger.warning(urls) # Loop through URLs and run command results = [] for url in urls: ''' Above while fetching urls, we are not ignoring files, because some default urls may redirect to https://example.com/login.php so, ignore_files is set to False but, during fuzzing, we will only need part of the path, in above example it is still a good idea to ffuf base url https://example.com so files from base url ''' url_parse = urlparse(url) url = url_parse.scheme + '://' + url_parse.netloc url += '/FUZZ' # TODO: fuzz not only URL but also POST / PUT / headers proxy = get_random_proxy() # Build final cmd fcmd = cmd fcmd += f' -x {proxy}' if proxy else '' fcmd += f' -u {url} -json' # Initialize DirectoryScan object dirscan = DirectoryScan() dirscan.scanned_date = timezone.now() dirscan.command_line = fcmd dirscan.save() # Loop through results and populate EndPoint and DirectoryFile in DB results = [] for line in stream_command( fcmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) name = line['input'].get('FUZZ') length = line['length'] status = line['status'] words = line['words'] url = line['url'] lines = line['lines'] content_type = line['content-type'] duration = line['duration'] if not name: logger.error(f'FUZZ not found for "{url}"') continue endpoint, created = save_endpoint(url, crawl=False, ctx=ctx) endpoint.http_status = status endpoint.content_length = length endpoint.response_time = duration / 1000000000 endpoint.save() if created: urls.append(endpoint.http_url) endpoint.status = status endpoint.content_type = content_type endpoint.content_length = length dfile, created = DirectoryFile.objects.get_or_create( name=name, length=length, words=words, lines=lines, content_type=content_type, url=url) dfile.http_status = status dfile.save() # if created: # logger.warning(f'Found new directory or file {url}') dirscan.directory_files.add(dfile) dirscan.save() if self.subscan: dirscan.dir_subscan_ids.add(self.subscan) subdomain_name = get_subdomain_from_url(endpoint.http_url) subdomain = Subdomain.objects.get(name=subdomain_name, scan_history=self.scan) subdomain.directories.add(dirscan) subdomain.save() # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl(urls, ctx=ctx) return results @app.task(name='fetch_url', queue='main_scan_queue', base=RengineTask, bind=True) def fetch_url(self, urls=[], ctx={}, description=None): """Fetch URLs using different tools like gauplus, gau, gospider, waybackurls ... Args: urls (list): List of URLs to start from. description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/input_endpoints_fetch_url.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(FETCH_URL) or {} should_remove_duplicate_endpoints = config.get(REMOVE_DUPLICATE_ENDPOINTS, True) duplicate_removal_fields = config.get(DUPLICATE_REMOVAL_FIELDS, ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, DEFAULT_GF_PATTERNS) ignore_file_extension = config.get(IGNORE_FILE_EXTENSION, DEFAULT_IGNORE_FILE_EXTENSIONS) tools = config.get(USES_TOOLS, ENDPOINT_SCAN_DEFAULT_TOOLS) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) domain_request_headers = self.domain.request_headers if self.domain else None custom_header = domain_request_headers or self.yaml_configuration.get(CUSTOM_HEADER) exclude_subdomains = config.get(EXCLUDED_SUBDOMAINS, False) # Get URLs to scan and save to input file if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, exclude_subdomains=exclude_subdomains, get_only_default_urls=True, ctx=ctx ) # Domain regex host = self.domain.name if self.domain else urlparse(urls[0]).netloc host_regex = f"\'https?://([a-z0-9]+[.])*{host}.*\'" # Tools cmds cmd_map = { 'gau': f'gau', 'gauplus': f'gauplus -random-agent', 'hakrawler': 'hakrawler -subs -u', 'waybackurls': 'waybackurls', 'gospider': f'gospider -S {input_path} --js -d 2 --sitemap --robots -w -r', 'katana': f'katana -list {input_path} -silent -jc -kf all -d 3 -fs rdn', } if proxy: cmd_map['gau'] += f' --proxy "{proxy}"' cmd_map['gauplus'] += f' -p "{proxy}"' cmd_map['gospider'] += f' -p {proxy}' cmd_map['hakrawler'] += f' -proxy {proxy}' cmd_map['katana'] += f' -proxy {proxy}' if threads > 0: cmd_map['gau'] += f' --threads {threads}' cmd_map['gauplus'] += f' -t {threads}' cmd_map['gospider'] += f' -t {threads}' cmd_map['katana'] += f' -c {threads}' if custom_header: header_string = ';;'.join([ f'{key}: {value}' for key, value in custom_header.items() ]) cmd_map['hakrawler'] += f' -h {header_string}' cmd_map['katana'] += f' -H {header_string}' header_flags = [':'.join(h) for h in header_string.split(';;')] for flag in header_flags: cmd_map['gospider'] += f' -H {flag}' cat_input = f'cat {input_path}' grep_output = f'grep -Eo {host_regex}' cmd_map = { tool: f'{cat_input} | {cmd} | {grep_output} > {self.results_dir}/urls_{tool}.txt' for tool, cmd in cmd_map.items() } tasks = group( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for tool, cmd in cmd_map.items() if tool in tools ) # Cleanup task sort_output = [ f'cat {self.results_dir}/urls_* > {self.output_path}', f'cat {input_path} >> {self.output_path}', f'sort -u {self.output_path} -o {self.output_path}', ] if ignore_file_extension: ignore_exts = '|'.join(ignore_file_extension) grep_ext_filtered_output = [ f'cat {self.output_path} | grep -Eiv "\\.({ignore_exts}).*" > {self.results_dir}/urls_filtered.txt', f'mv {self.results_dir}/urls_filtered.txt {self.output_path}' ] sort_output.extend(grep_ext_filtered_output) cleanup = chain( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for cmd in sort_output ) # Run all commands task = chord(tasks)(cleanup) with allow_join_result(): task.get() # Store all the endpoints and run httpx with open(self.output_path) as f: discovered_urls = f.readlines() self.notify(fields={'Discovered URLs': len(discovered_urls)}) # Some tools can have an URL in the format <URL>] - <PATH> or <URL> - <PATH>, add them # to the final URL list all_urls = [] for url in discovered_urls: url = url.strip() urlpath = None base_url = None if '] ' in url: # found JS scraped endpoint e.g from gospider split = tuple(url.split('] ')) if not len(split) == 2: logger.warning(f'URL format not recognized for "{url}". Skipping.') continue base_url, urlpath = split urlpath = urlpath.lstrip('- ') elif ' - ' in url: # found JS scraped endpoint e.g from gospider base_url, urlpath = tuple(url.split(' - ')) if base_url and urlpath: subdomain = urlparse(base_url) url = f'{subdomain.scheme}://{subdomain.netloc}{self.url_filter}' if not validators.url(url): logger.warning(f'Invalid URL "{url}". Skipping.') if url not in all_urls: all_urls.append(url) # Filter out URLs if a path filter was passed if self.url_filter: all_urls = [url for url in all_urls if self.url_filter in url] # Write result to output path with open(self.output_path, 'w') as f: f.write('\n'.join(all_urls)) logger.warning(f'Found {len(all_urls)} usable URLs') # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl( all_urls, ctx=ctx, should_remove_duplicate_endpoints=should_remove_duplicate_endpoints, duplicate_removal_fields=duplicate_removal_fields ) #-------------------# # GF PATTERNS MATCH # #-------------------# # Combine old gf patterns with new ones if gf_patterns: self.scan.used_gf_patterns = ','.join(gf_patterns) self.scan.save() # Run gf patterns on saved endpoints # TODO: refactor to Celery task for gf_pattern in gf_patterns: # TODO: js var is causing issues, removing for now if gf_pattern == 'jsvar': logger.info('Ignoring jsvar as it is causing issues.') continue # Run gf on current pattern logger.warning(f'Running gf on pattern "{gf_pattern}"') gf_output_file = f'{self.results_dir}/gf_patterns_{gf_pattern}.txt' cmd = f'cat {self.output_path} | gf {gf_pattern} | grep -Eo {host_regex} >> {gf_output_file}' run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Check output file if not os.path.exists(gf_output_file): logger.error(f'Could not find GF output file {gf_output_file}. Skipping GF pattern "{gf_pattern}"') continue # Read output file line by line and with open(gf_output_file, 'r') as f: lines = f.readlines() # Add endpoints / subdomains to DB for url in lines: http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue endpoint, created = save_endpoint( http_url, crawl=False, subdomain=subdomain, ctx=ctx) if not endpoint: continue earlier_pattern = None if not created: earlier_pattern = endpoint.matched_gf_patterns pattern = f'{earlier_pattern},{gf_pattern}' if earlier_pattern else gf_pattern endpoint.matched_gf_patterns = pattern endpoint.save() return all_urls def parse_curl_output(response): # TODO: Enrich from other cURL fields. CURL_REGEX_HTTP_STATUS = f'HTTP\/(?:(?:\d\.?)+)\s(\d+)\s(?:\w+)' http_status = 0 if response: failed = False regex = re.compile(CURL_REGEX_HTTP_STATUS, re.MULTILINE) try: http_status = int(regex.findall(response)[0]) except (KeyError, TypeError, IndexError): pass return { 'http_status': http_status, } @app.task(name='vulnerability_scan', queue='main_scan_queue', bind=True, base=RengineTask) def vulnerability_scan(self, urls=[], ctx={}, description=None): """ This function will serve as an entrypoint to vulnerability scan. All other vulnerability scan will be run from here including nuclei, crlfuzz, etc """ logger.info('Running Vulnerability Scan Queue') config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_run_nuclei = config.get(RUN_NUCLEI, True) should_run_crlfuzz = config.get(RUN_CRLFUZZ, False) should_run_dalfox = config.get(RUN_DALFOX, False) should_run_s3scanner = config.get(RUN_S3SCANNER, True) grouped_tasks = [] if should_run_nuclei: _task = nuclei_scan.si( urls=urls, ctx=ctx, description=f'Nuclei Scan' ) grouped_tasks.append(_task) if should_run_crlfuzz: _task = crlfuzz_scan.si( urls=urls, ctx=ctx, description=f'CRLFuzz Scan' ) grouped_tasks.append(_task) if should_run_dalfox: _task = dalfox_xss_scan.si( urls=urls, ctx=ctx, description=f'Dalfox XSS Scan' ) grouped_tasks.append(_task) if should_run_s3scanner: _task = s3scanner.si( ctx=ctx, description=f'Misconfigured S3 Buckets Scanner' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan completed...') # return results return None @app.task(name='nuclei_individual_severity_module', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_individual_severity_module(self, cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx={}, description=None): ''' This celery task will run vulnerability scan in parallel. All severities supplied should run in parallel as grouped tasks. ''' results = [] logger.info(f'Running vulnerability scan with severity: {severity}') cmd += f' -severity {severity}' # Send start notification notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) # Gather nuclei results vuln_data = parse_nuclei_result(line) # Get corresponding subdomain http_url = sanitize_url(line.get('matched-at')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) # Look for duplicate vulnerabilities by excluding records that might change but are irrelevant. object_comparison_exclude = ['response', 'curl_command', 'tags', 'references', 'cve_ids', 'cwe_ids'] # Add subdomain and target domain to the duplicate check vuln_data_copy = vuln_data.copy() vuln_data_copy['subdomain'] = subdomain vuln_data_copy['target_domain'] = self.domain # Check if record exists, if exists do not save it if record_exists(Vulnerability, data=vuln_data_copy, exclude_keys=object_comparison_exclude): logger.warning(f'Nuclei vulnerability of severity {severity} : {vuln_data_copy["name"]} for {subdomain_name} already exists') continue # Get or create EndPoint object response = line.get('response') httpx_crawl = False if response else enable_http_crawl # avoid yet another httpx crawl endpoint, _ = save_endpoint( http_url, crawl=httpx_crawl, subdomain=subdomain, ctx=ctx) if endpoint: http_url = endpoint.http_url if not httpx_crawl: output = parse_curl_output(response) endpoint.http_status = output['http_status'] endpoint.save() # Get or create Vulnerability object vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, subdomain=subdomain, **vuln_data) if not vuln: continue # Print vuln severity = line['info'].get('severity', 'unknown') logger.warning(str(vuln)) # Send notification for all vulnerabilities except info url = vuln.http_url or vuln.subdomain send_vuln = ( notif and notif.send_vuln_notif and vuln and severity in ['low', 'medium', 'high', 'critical']) if send_vuln: fields = { 'Severity': f'**{severity.upper()}**', 'URL': http_url, 'Subdomain': subdomain_name, 'Name': vuln.name, 'Type': vuln.type, 'Description': vuln.description, 'Template': vuln.template_url, 'Tags': vuln.get_tags_str(), 'CVEs': vuln.get_cve_str(), 'CWEs': vuln.get_cwe_str(), 'References': vuln.get_refs_str() } severity_map = { 'low': 'info', 'medium': 'warning', 'high': 'error', 'critical': 'error' } self.notify( f'vulnerability_scan_#{vuln.id}', severity_map[severity], fields, add_meta_info=False) # Send report to hackerone hackerone_query = Hackerone.objects.all() send_report = ( hackerone_query.exists() and severity not in ('info', 'low') and vuln.target_domain.h1_team_handle ) if send_report: hackerone = hackerone_query.first() if hackerone.send_critical and severity == 'critical': send_hackerone_report.delay(vuln.id) elif hackerone.send_high and severity == 'high': send_hackerone_report.delay(vuln.id) elif hackerone.send_medium and severity == 'medium': send_hackerone_report.delay(vuln.id) # Write results to JSON file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) # Send finish notif if send_status: vulns = Vulnerability.objects.filter(scan_history__id=self.scan_id) info_count = vulns.filter(severity=0).count() low_count = vulns.filter(severity=1).count() medium_count = vulns.filter(severity=2).count() high_count = vulns.filter(severity=3).count() critical_count = vulns.filter(severity=4).count() unknown_count = vulns.filter(severity=-1).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count + unknown_count fields = { 'Total': vulnerability_count, 'Critical': critical_count, 'High': high_count, 'Medium': medium_count, 'Low': low_count, 'Info': info_count, 'Unknown': unknown_count } self.notify(fields=fields) # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=NUCLEI ).exclude( severity=0 ) # find all unique vulnerabilities based on path and title # all unique vulnerability will go thru gpt function and get report # once report is got, it will be matched with other vulnerabilities and saved unique_vulns = set() for vuln in vulns: unique_vulns.add((vuln.name, vuln.get_path())) unique_vulns = list(unique_vulns) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in unique_vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return None def get_vulnerability_gpt_report(vuln): title = vuln[0] path = vuln[1] logger.info(f'Getting GPT Report for {title}, PATH: {path}') # check if in db already exists stored = GPTVulnerabilityReport.objects.filter( url_path=path ).filter( title=title ).first() if stored: response = { 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: report = GPTVulnerabilityReportGenerator() vulnerability_description = get_gpt_vuln_input_description( title, path ) response = report.get_vulnerability_description(vulnerability_description) add_gpt_description_db( title, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) for vuln in Vulnerability.objects.filter(name=title, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() def add_gpt_description_db(title, path, description, impact, remediation, references): gpt_report = GPTVulnerabilityReport() gpt_report.url_path = path gpt_report.title = title gpt_report.description = description gpt_report.impact = impact gpt_report.remediation = remediation gpt_report.save() for url in references: ref, created = VulnerabilityReference.objects.get_or_create(url=url) gpt_report.references.add(ref) gpt_report.save() @app.task(name='nuclei_scan', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_scan(self, urls=[], ctx={}, description=None): """HTTP vulnerability scan using Nuclei Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. Notes: Unfurl the urls to keep only domain and path, will be sent to vuln scan and ignore certain file extensions. Thanks: https://github.com/six2dez/reconftw """ # Config config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} input_path = f'{self.results_dir}/input_endpoints_vulnerability_scan.txt' enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) concurrency = config.get(NUCLEI_CONCURRENCY) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) retries = config.get(RETRIES) or self.yaml_configuration.get(RETRIES, DEFAULT_RETRIES) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) custom_header = config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) should_fetch_gpt_report = config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) proxy = get_random_proxy() nuclei_specific_config = config.get('nuclei', {}) use_nuclei_conf = nuclei_specific_config.get(USE_CONFIG, False) severities = nuclei_specific_config.get(NUCLEI_SEVERITY, NUCLEI_DEFAULT_SEVERITIES) tags = nuclei_specific_config.get(NUCLEI_TAGS, []) tags = ','.join(tags) nuclei_templates = nuclei_specific_config.get(NUCLEI_TEMPLATE) custom_nuclei_templates = nuclei_specific_config.get(NUCLEI_CUSTOM_TEMPLATE) # severities_str = ','.join(severities) # Get alive endpoints if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=enable_http_crawl, ignore_files=True, write_filepath=input_path, ctx=ctx ) if intensity == 'normal': # reduce number of endpoints to scan unfurl_filter = f'{self.results_dir}/urls_unfurled.txt' run_command( f"cat {input_path} | unfurl -u format %s://%d%p |uro > {unfurl_filter}", shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {unfurl_filter} -o {unfurl_filter}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) input_path = unfurl_filter # Build templates # logger.info('Updating Nuclei templates ...') run_command( 'nuclei -update-templates', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) templates = [] if not (nuclei_templates or custom_nuclei_templates): templates.append(NUCLEI_DEFAULT_TEMPLATES_PATH) if nuclei_templates: if ALL in nuclei_templates: template = NUCLEI_DEFAULT_TEMPLATES_PATH templates.append(template) else: templates.extend(nuclei_templates) if custom_nuclei_templates: custom_nuclei_template_paths = [f'{str(elem)}.yaml' for elem in custom_nuclei_templates] template = templates.extend(custom_nuclei_template_paths) # Build CMD cmd = 'nuclei -j' cmd += ' -config /root/.config/nuclei/config.yaml' if use_nuclei_conf else '' cmd += f' -irr' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -l {input_path}' cmd += f' -c {str(concurrency)}' if concurrency > 0 else '' cmd += f' -proxy {proxy} ' if proxy else '' cmd += f' -retries {retries}' if retries > 0 else '' cmd += f' -rl {rate_limit}' if rate_limit > 0 else '' # cmd += f' -severity {severities_str}' cmd += f' -timeout {str(timeout)}' if timeout and timeout > 0 else '' cmd += f' -tags {tags}' if tags else '' cmd += f' -silent' for tpl in templates: cmd += f' -t {tpl}' grouped_tasks = [] custom_ctx = ctx for severity in severities: custom_ctx['track'] = True _task = nuclei_individual_severity_module.si( cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx=custom_ctx, description=f'Nuclei Scan with severity {severity}' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan with all severities completed...') return None @app.task(name='dalfox_xss_scan', queue='main_scan_queue', base=RengineTask, bind=True) def dalfox_xss_scan(self, urls=[], ctx={}, description=None): """XSS Scan using dalfox Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) dalfox_config = vuln_config.get(DALFOX) or {} custom_header = dalfox_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() is_waf_evasion = dalfox_config.get(WAF_EVASION, False) blind_xss_server = dalfox_config.get(BLIND_XSS_SERVER) user_agent = dalfox_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) timeout = dalfox_config.get(TIMEOUT) delay = dalfox_config.get(DELAY) threads = dalfox_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_dalfox_xss.txt' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=False, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'dalfox --silence --no-color --no-spinner' cmd += f' --only-poc r ' cmd += f' --ignore-return 302,404,403' cmd += f' --skip-bav' cmd += f' file {input_path}' cmd += f' --proxy {proxy}' if proxy else '' cmd += f' --waf-evasion' if is_waf_evasion else '' cmd += f' -b {blind_xss_server}' if blind_xss_server else '' cmd += f' --delay {delay}' if delay else '' cmd += f' --timeout {timeout}' if timeout else '' cmd += f' --user-agent {user_agent}' if user_agent else '' cmd += f' --header {custom_header}' if custom_header else '' cmd += f' --worker {threads}' if threads else '' cmd += f' --format json' results = [] for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id, trunc_char=',' ): if not isinstance(line, dict): continue results.append(line) vuln_data = parse_dalfox_result(line) http_url = sanitize_url(line.get('data')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Dalfox Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=DALFOX ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='crlfuzz_scan', queue='main_scan_queue', base=RengineTask, bind=True) def crlfuzz_scan(self, urls=[], ctx={}, description=None): """CRLF Fuzzing with CRLFuzz Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) custom_header = vuln_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() user_agent = vuln_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) threads = vuln_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_crlf.txt' output_path = f'{self.results_dir}/{self.filename}' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=True, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'crlfuzz -s' cmd += f' -l {input_path}' cmd += f' -x {proxy}' if proxy else '' cmd += f' --H {custom_header}' if custom_header else '' cmd += f' -o {output_path}' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id ) if not os.path.isfile(output_path): logger.info('No Results from CRLFuzz') return crlfs = [] results = [] with open(output_path, 'r') as file: crlfs = file.readlines() for crlf in crlfs: url = crlf.strip() vuln_data = parse_crlfuzz_result(url) http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting CRLFuzz Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=CRLFUZZ ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='s3scanner', queue='main_scan_queue', base=RengineTask, bind=True) def s3scanner(self, ctx={}, description=None): """Bucket Scanner Args: ctx (dict): Context description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/#{self.scan_id}_subdomain_discovery.txt' vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} s3_config = vuln_config.get(S3SCANNER) or {} threads = s3_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) providers = s3_config.get(PROVIDERS, S3SCANNER_DEFAULT_PROVIDERS) scan_history = ScanHistory.objects.filter(pk=self.scan_id).first() for provider in providers: cmd = f's3scanner -bucket-file {input_path} -enumerate -provider {provider} -threads {threads} -json' for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue if line.get('bucket', {}).get('exists', 0) == 1: result = parse_s3scanner_result(line) s3bucket, created = S3Bucket.objects.get_or_create(**result) scan_history.buckets.add(s3bucket) logger.info(f"s3 bucket added {result['provider']}-{result['name']}-{result['region']}") @app.task(name='http_crawl', queue='main_scan_queue', base=RengineTask, bind=True) def http_crawl( self, urls=[], method=None, recrawl=False, ctx={}, track=True, description=None, is_ran_from_subdomain_scan=False, should_remove_duplicate_endpoints=True, duplicate_removal_fields=[]): """Use httpx to query HTTP URLs for important info like page titles, http status, etc... Args: urls (list, optional): A set of URLs to check. Overrides default behavior which queries all endpoints related to this scan. method (str): HTTP method to use (GET, HEAD, POST, PUT, DELETE). recrawl (bool, optional): If False, filter out URLs that have already been crawled. should_remove_duplicate_endpoints (bool): Whether to remove duplicate endpoints duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates Returns: list: httpx results. """ logger.info('Initiating HTTP Crawl') if is_ran_from_subdomain_scan: logger.info('Running From Subdomain Scan...') cmd = '/go/bin/httpx' cfg = self.yaml_configuration.get(HTTP_CRAWL) or {} custom_header = cfg.get(CUSTOM_HEADER, '') threads = cfg.get(THREADS, DEFAULT_THREADS) follow_redirect = cfg.get(FOLLOW_REDIRECT, True) self.output_path = None input_path = f'{self.results_dir}/httpx_input.txt' history_file = f'{self.results_dir}/commands.txt' if urls: # direct passing URLs to check if self.url_filter: urls = [u for u in urls if self.url_filter in u] with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_uncrawled=not recrawl, write_filepath=input_path, ctx=ctx ) # logger.debug(urls) # If no URLs found, skip it if not urls: return # Re-adjust thread number if few URLs to avoid spinning up a monster to # kill a fly. if len(urls) < threads: threads = len(urls) # Get random proxy proxy = get_random_proxy() # Run command cmd += f' -cl -ct -rt -location -td -websocket -cname -asn -cdn -probe -random-agent' cmd += f' -t {threads}' if threads > 0 else '' cmd += f' --http-proxy {proxy}' if proxy else '' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -json' cmd += f' -u {urls[0]}' if len(urls) == 1 else f' -l {input_path}' cmd += f' -x {method}' if method else '' cmd += f' -silent' if follow_redirect: cmd += ' -fr' results = [] endpoint_ids = [] for line in stream_command( cmd, history_file=history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not line or not isinstance(line, dict): continue logger.debug(line) # No response from endpoint if line.get('failed', False): continue # Parse httpx output host = line.get('host', '') content_length = line.get('content_length', 0) http_status = line.get('status_code') http_url, is_redirect = extract_httpx_url(line) page_title = line.get('title') webserver = line.get('webserver') cdn = line.get('cdn', False) rt = line.get('time') techs = line.get('tech', []) cname = line.get('cname', '') content_type = line.get('content_type', '') response_time = -1 if rt: response_time = float(''.join(ch for ch in rt if not ch.isalpha())) if rt[-2:] == 'ms': response_time = response_time / 1000 # Create Subdomain object in DB subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue # Save default HTTP URL to endpoint object in DB endpoint, created = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain, is_default=is_ran_from_subdomain_scan ) if not endpoint: continue endpoint.http_status = http_status endpoint.page_title = page_title endpoint.content_length = content_length endpoint.webserver = webserver endpoint.response_time = response_time endpoint.content_type = content_type endpoint.save() endpoint_str = f'{http_url} [{http_status}] `{content_length}B` `{webserver}` `{rt}`' logger.warning(endpoint_str) if endpoint and endpoint.is_alive and endpoint.http_status != 403: self.notify( fields={'Alive endpoint': f'• {endpoint_str}'}, add_meta_info=False) # Add endpoint to results line['_cmd'] = cmd line['final_url'] = http_url line['endpoint_id'] = endpoint.id line['endpoint_created'] = created line['is_redirect'] = is_redirect results.append(line) # Add technology objects to DB for technology in techs: tech, _ = Technology.objects.get_or_create(name=technology) endpoint.techs.add(tech) if is_ran_from_subdomain_scan: subdomain.technologies.add(tech) subdomain.save() endpoint.save() techs_str = ', '.join([f'`{tech}`' for tech in techs]) self.notify( fields={'Technologies': techs_str}, add_meta_info=False) # Add IP objects for 'a' records to DB a_records = line.get('a', []) for ip_address in a_records: ip, created = save_ip_address( ip_address, subdomain, subscan=self.subscan, cdn=cdn) ips_str = '• ' + '\n• '.join([f'`{ip}`' for ip in a_records]) self.notify( fields={'IPs': ips_str}, add_meta_info=False) # Add IP object for host in DB if host: ip, created = save_ip_address( host, subdomain, subscan=self.subscan, cdn=cdn) self.notify( fields={'IPs': f'• `{ip.address}`'}, add_meta_info=False) # Save subdomain and endpoint if is_ran_from_subdomain_scan: # save subdomain stuffs subdomain.http_url = http_url subdomain.http_status = http_status subdomain.page_title = page_title subdomain.content_length = content_length subdomain.webserver = webserver subdomain.response_time = response_time subdomain.content_type = content_type subdomain.cname = ','.join(cname) subdomain.is_cdn = cdn if cdn: subdomain.cdn_name = line.get('cdn_name') subdomain.save() endpoint.save() endpoint_ids.append(endpoint.id) if should_remove_duplicate_endpoints: # Remove 'fake' alive endpoints that are just redirects to the same page remove_duplicate_endpoints( self.scan_id, self.domain_id, self.subdomain_id, filter_ids=endpoint_ids ) # Remove input file run_command( f'rm {input_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) return results #---------------------# # Notifications tasks # #---------------------# @app.task(name='send_notif', bind=False, queue='send_notif_queue') def send_notif( message, scan_history_id=None, subscan_id=None, **options): if not 'title' in options: message = enrich_notification(message, scan_history_id, subscan_id) send_discord_message(message, **options) send_slack_message(message) send_telegram_message(message) @app.task(name='send_scan_notif', bind=False, queue='send_scan_notif_queue') def send_scan_notif( scan_history_id, subscan_id=None, engine_id=None, status='RUNNING'): """Send scan status notification. Works for scan or a subscan if subscan_id is passed. Args: scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Get domain, engine, scan_history objects engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() tasks = ScanActivity.objects.filter(scan_of=scan) if scan else 0 # Build notif options url = get_scan_url(scan_history_id, subscan_id) title = get_scan_title(scan_history_id, subscan_id) fields = get_scan_fields(engine, scan, subscan, status, tasks) severity = None msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) if status: severity = STATUS_TO_SEVERITIES.get(status) opts = { 'title': title, 'url': url, 'fields': fields, 'severity': severity } logger.warning(f'Sending notification "{title}" [{severity}]') # Send notification send_notif( msg, scan_history_id, subscan_id, **opts) @app.task(name='send_task_notif', bind=False, queue='send_task_notif_queue') def send_task_notif( task_name, status=None, result=None, output_path=None, traceback=None, scan_history_id=None, engine_id=None, subscan_id=None, severity=None, add_meta_info=True, update_fields={}): """Send task status notification. Args: task_name (str): Task name. status (str, optional): Task status. result (str, optional): Task result. output_path (str, optional): Task output path. traceback (str, optional): Task traceback. scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. severity (str, optional): Severity (will be mapped to notif colors) add_meta_info (bool, optional): Wheter to add scan / subscan info to notif. update_fields (dict, optional): Fields key / value to update. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Build fields url = None fields = {} if add_meta_info: engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() url = get_scan_url(scan_history_id) if status: fields['Status'] = f'**{status}**' if engine: fields['Engine'] = engine.engine_name if scan: fields['Scan ID'] = f'[#{scan.id}]({url})' if subscan: url = get_scan_url(scan_history_id, subscan_id) fields['Subscan ID'] = f'[#{subscan.id}]({url})' title = get_task_title(task_name, scan_history_id, subscan_id) if status: severity = STATUS_TO_SEVERITIES.get(status) msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) # Add fields to update for k, v in update_fields.items(): fields[k] = v # Add traceback to notif if traceback and notif.send_scan_tracebacks: fields['Traceback'] = f'```\n{traceback}\n```' # Add files to notif files = [] attach_file = ( notif.send_scan_output_file and output_path and result and not traceback ) if attach_file: output_title = output_path.split('/')[-1] files = [(output_path, output_title)] # Send notif opts = { 'title': title, 'url': url, 'files': files, 'severity': severity, 'fields': fields, 'fields_append': update_fields.keys() } send_notif( msg, scan_history_id=scan_history_id, subscan_id=subscan_id, **opts) @app.task(name='send_file_to_discord', bind=False, queue='send_file_to_discord_queue') def send_file_to_discord(file_path, title=None): notif = Notification.objects.first() do_send = notif and notif.send_to_discord and notif.discord_hook_url if not do_send: return False webhook = DiscordWebhook( url=notif.discord_hook_url, rate_limit_retry=True, username=title or "reNgine Discord Plugin" ) with open(file_path, "rb") as f: head, tail = os.path.split(file_path) webhook.add_file(file=f.read(), filename=tail) webhook.execute() @app.task(name='send_hackerone_report', bind=False, queue='send_hackerone_report_queue') def send_hackerone_report(vulnerability_id): """Send HackerOne vulnerability report. Args: vulnerability_id (int): Vulnerability id. Returns: int: HTTP response status code. """ vulnerability = Vulnerability.objects.get(id=vulnerability_id) severities = {v: k for k,v in NUCLEI_SEVERITY_MAP.items()} headers = { 'Content-Type': 'application/json', 'Accept': 'application/json' } # can only send vulnerability report if team_handle exists if len(vulnerability.target_domain.h1_team_handle) !=0: hackerone_query = Hackerone.objects.all() if hackerone_query.exists(): hackerone = Hackerone.objects.first() severity_value = severities[vulnerability.severity] tpl = hackerone.report_template # Replace syntax of report template with actual content tpl = tpl.replace('{vulnerability_name}', vulnerability.name) tpl = tpl.replace('{vulnerable_url}', vulnerability.http_url) tpl = tpl.replace('{vulnerability_severity}', severity_value) tpl = tpl.replace('{vulnerability_description}', vulnerability.description if vulnerability.description else '') tpl = tpl.replace('{vulnerability_extracted_results}', vulnerability.extracted_results if vulnerability.extracted_results else '') tpl = tpl.replace('{vulnerability_reference}', vulnerability.reference if vulnerability.reference else '') data = { "data": { "type": "report", "attributes": { "team_handle": vulnerability.target_domain.h1_team_handle, "title": '{} found in {}'.format(vulnerability.name, vulnerability.http_url), "vulnerability_information": tpl, "severity_rating": severity_value, "impact": "More information about the impact and vulnerability can be found here: \n" + vulnerability.reference if vulnerability.reference else "NA", } } } r = requests.post( 'https://api.hackerone.com/v1/hackers/reports', auth=(hackerone.username, hackerone.api_key), json=data, headers=headers ) response = r.json() status_code = r.status_code if status_code == 201: vulnerability.hackerone_report_id = response['data']["id"] vulnerability.open_status = False vulnerability.save() return status_code else: logger.error('No team handle found.') status_code = 111 return status_code #-------------# # Utils tasks # #-------------# @app.task(name='parse_nmap_results', bind=False, queue='parse_nmap_results_queue') def parse_nmap_results(xml_file, output_file=None): """Parse results from nmap output file. Args: xml_file (str): nmap XML report file path. Returns: list: List of vulnerabilities found from nmap results. """ with open(xml_file, encoding='utf8') as f: content = f.read() try: nmap_results = xmltodict.parse(content) # parse XML to dict except Exception as e: logger.exception(e) logger.error(f'Cannot parse {xml_file} to valid JSON. Skipping.') return [] # Write JSON to output file if output_file: with open(output_file, 'w') as f: json.dump(nmap_results, f, indent=4) logger.warning(json.dumps(nmap_results, indent=4)) hosts = ( nmap_results .get('nmaprun', {}) .get('host', {}) ) all_vulns = [] if isinstance(hosts, dict): hosts = [hosts] for host in hosts: # Grab hostname / IP from output hostnames_dict = host.get('hostnames', {}) if hostnames_dict: # Ensure that hostnames['hostname'] is a list for consistency hostnames_list = hostnames_dict['hostname'] if isinstance(hostnames_dict['hostname'], list) else [hostnames_dict['hostname']] # Extract all the @name values from the list of dictionaries hostnames = [entry.get('@name') for entry in hostnames_list] else: hostnames = [host.get('address')['@addr']] # Iterate over each hostname for each port for hostname in hostnames: # Grab ports from output ports = host.get('ports', {}).get('port', []) if isinstance(ports, dict): ports = [ports] for port in ports: url_vulns = [] port_number = port['@portid'] url = sanitize_url(f'{hostname}:{port_number}') logger.info(f'Parsing nmap results for {hostname}:{port_number} ...') if not port_number or not port_number.isdigit(): continue port_protocol = port['@protocol'] scripts = port.get('script', []) if isinstance(scripts, dict): scripts = [scripts] for script in scripts: script_id = script['@id'] script_output = script['@output'] script_output_table = script.get('table', []) logger.debug(f'Ran nmap script "{script_id}" on {port_number}/{port_protocol}:\n{script_output}\n') if script_id == 'vulscan': vulns = parse_nmap_vulscan_output(script_output) url_vulns.extend(vulns) elif script_id == 'vulners': vulns = parse_nmap_vulners_output(script_output) url_vulns.extend(vulns) # elif script_id == 'http-server-header': # TODO: nmap can help find technologies as well using the http-server-header script # regex = r'(\w+)/([\d.]+)\s?(?:\((\w+)\))?' # tech_name, tech_version, tech_os = re.match(regex, test_string).groups() # Technology.objects.get_or_create(...) # elif script_id == 'http_csrf': # vulns = parse_nmap_http_csrf_output(script_output) # url_vulns.extend(vulns) else: logger.warning(f'Script output parsing for script "{script_id}" is not supported yet.') # Add URL to vuln for vuln in url_vulns: # TODO: This should extend to any URL, not just HTTP vuln['http_url'] = url if 'http_path' in vuln: vuln['http_url'] += vuln['http_path'] all_vulns.append(vuln) return all_vulns def parse_nmap_http_csrf_output(script_output): pass def parse_nmap_vulscan_output(script_output): """Parse nmap vulscan script output. Args: script_output (str): Vulscan script output. Returns: list: List of Vulnerability dicts. """ data = {} vulns = [] provider_name = '' # Sort all vulns found by provider so that we can match each provider with # a function that pulls from its API to get more info about the # vulnerability. for line in script_output.splitlines(): if not line: continue if not line.startswith('['): # provider line if "No findings" in line: logger.info(f"No findings: {line}") continue elif ' - ' in line: provider_name, provider_url = tuple(line.split(' - ')) data[provider_name] = {'url': provider_url.rstrip(':'), 'entries': []} continue else: # Log a warning logger.warning(f"Unexpected line format: {line}") continue reg = r'\[(.*)\] (.*)' matches = re.match(reg, line) id, title = matches.groups() entry = {'id': id, 'title': title} data[provider_name]['entries'].append(entry) logger.warning('Vulscan parsed output:') logger.warning(pprint.pformat(data)) for provider_name in data: if provider_name == 'Exploit-DB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'IBM X-Force': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'MITRE CVE': logger.error(f'Provider {provider_name} is not supported YET.') for entry in data[provider_name]['entries']: cve_id = entry['id'] vuln = cve_to_vuln(cve_id) vulns.append(vuln) elif provider_name == 'OSVDB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'OpenVAS (Nessus)': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'SecurityFocus': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'VulDB': logger.error(f'Provider {provider_name} is not supported YET.') pass else: logger.error(f'Provider {provider_name} is not supported.') return vulns def parse_nmap_vulners_output(script_output, url=''): """Parse nmap vulners script output. TODO: Rework this as it's currently matching all CVEs no matter the confidence. Args: script_output (str): Script output. Returns: list: List of found vulnerabilities. """ vulns = [] # Check for CVE in script output CVE_REGEX = re.compile(r'.*(CVE-\d\d\d\d-\d+).*') matches = CVE_REGEX.findall(script_output) matches = list(dict.fromkeys(matches)) for cve_id in matches: # get CVE info vuln = cve_to_vuln(cve_id, vuln_type='nmap-vulners-nse') if vuln: vulns.append(vuln) return vulns def cve_to_vuln(cve_id, vuln_type=''): """Search for a CVE using CVESearch and return Vulnerability data. Args: cve_id (str): CVE ID in the form CVE-* Returns: dict: Vulnerability dict. """ cve_info = CVESearch('https://cve.circl.lu').id(cve_id) if not cve_info: logger.error(f'Could not fetch CVE info for cve {cve_id}. Skipping.') return None vuln_cve_id = cve_info['id'] vuln_name = vuln_cve_id vuln_description = cve_info.get('summary', 'none').replace(vuln_cve_id, '').strip() try: vuln_cvss = float(cve_info.get('cvss', -1)) except (ValueError, TypeError): vuln_cvss = -1 vuln_cwe_id = cve_info.get('cwe', '') exploit_ids = cve_info.get('refmap', {}).get('exploit-db', []) osvdb_ids = cve_info.get('refmap', {}).get('osvdb', []) references = cve_info.get('references', []) capec_objects = cve_info.get('capec', []) # Parse ovals for a better vuln name / type ovals = cve_info.get('oval', []) if ovals: vuln_name = ovals[0]['title'] vuln_type = ovals[0]['family'] # Set vulnerability severity based on CVSS score vuln_severity = 'info' if vuln_cvss < 4: vuln_severity = 'low' elif vuln_cvss < 7: vuln_severity = 'medium' elif vuln_cvss < 9: vuln_severity = 'high' else: vuln_severity = 'critical' # Build console warning message msg = f'{vuln_name} | {vuln_severity.upper()} | {vuln_cve_id} | {vuln_cwe_id} | {vuln_cvss}' for id in osvdb_ids: msg += f'\n\tOSVDB: {id}' for exploit_id in exploit_ids: msg += f'\n\tEXPLOITDB: {exploit_id}' logger.warning(msg) vuln = { 'name': vuln_name, 'type': vuln_type, 'severity': NUCLEI_SEVERITY_MAP[vuln_severity], 'description': vuln_description, 'cvss_score': vuln_cvss, 'references': references, 'cve_ids': [vuln_cve_id], 'cwe_ids': [vuln_cwe_id] } return vuln def parse_s3scanner_result(line): ''' Parses and returns s3Scanner Data ''' bucket = line['bucket'] return { 'name': bucket['name'], 'region': bucket['region'], 'provider': bucket['provider'], 'owner_display_name': bucket['owner_display_name'], 'owner_id': bucket['owner_id'], 'perm_auth_users_read': bucket['perm_auth_users_read'], 'perm_auth_users_write': bucket['perm_auth_users_write'], 'perm_auth_users_read_acl': bucket['perm_auth_users_read_acl'], 'perm_auth_users_write_acl': bucket['perm_auth_users_write_acl'], 'perm_auth_users_full_control': bucket['perm_auth_users_full_control'], 'perm_all_users_read': bucket['perm_all_users_read'], 'perm_all_users_write': bucket['perm_all_users_write'], 'perm_all_users_read_acl': bucket['perm_all_users_read_acl'], 'perm_all_users_write_acl': bucket['perm_all_users_write_acl'], 'perm_all_users_full_control': bucket['perm_all_users_full_control'], 'num_objects': bucket['num_objects'], 'size': bucket['bucket_size'] } def parse_nuclei_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ return { 'name': line['info'].get('name', ''), 'type': line['type'], 'severity': NUCLEI_SEVERITY_MAP[line['info'].get('severity', 'unknown')], 'template': line['template'], 'template_url': line['template-url'], 'template_id': line['template-id'], 'description': line['info'].get('description', ''), 'matcher_name': line.get('matcher-name', ''), 'curl_command': line.get('curl-command'), 'request': line.get('request'), 'response': line.get('response'), 'extracted_results': line.get('extracted-results', []), 'cvss_metrics': line['info'].get('classification', {}).get('cvss-metrics', ''), 'cvss_score': line['info'].get('classification', {}).get('cvss-score'), 'cve_ids': line['info'].get('classification', {}).get('cve_id', []) or [], 'cwe_ids': line['info'].get('classification', {}).get('cwe_id', []) or [], 'references': line['info'].get('reference', []) or [], 'tags': line['info'].get('tags', []), 'source': NUCLEI, } def parse_dalfox_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ description = '' description += f" Evidence: {line.get('evidence')} <br>" if line.get('evidence') else '' description += f" Message: {line.get('message')} <br>" if line.get('message') else '' description += f" Payload: {line.get('message_str')} <br>" if line.get('message_str') else '' description += f" Vulnerable Parameter: {line.get('param')} <br>" if line.get('param') else '' return { 'name': 'XSS (Cross Site Scripting)', 'type': 'XSS', 'severity': DALFOX_SEVERITY_MAP[line.get('severity', 'unknown')], 'description': description, 'source': DALFOX, 'cwe_ids': [line.get('cwe')] } def parse_crlfuzz_result(url): """Parse CRLF results Args: url (str): CRLF Vulnerable URL Returns: dict: Vulnerability data. """ return { 'name': 'CRLF (HTTP Response Splitting)', 'type': 'CRLF', 'severity': 2, 'description': 'A CRLF (HTTP Response Splitting) vulnerability has been discovered.', 'source': CRLFUZZ, } def record_exists(model, data, exclude_keys=[]): """ Check if a record already exists in the database based on the given data. Args: model (django.db.models.Model): The Django model to check against. data (dict): Data dictionary containing fields and values. exclude_keys (list): List of keys to exclude from the lookup. Returns: bool: True if the record exists, False otherwise. """ # Extract the keys that will be used for the lookup lookup_fields = {key: data[key] for key in data if key not in exclude_keys} # Return True if a record exists based on the lookup fields, False otherwise return model.objects.filter(**lookup_fields).exists() @app.task(name='geo_localize', bind=False, queue='geo_localize_queue') def geo_localize(host, ip_id=None): """Uses geoiplookup to find location associated with host. Args: host (str): Hostname. ip_id (int): IpAddress object id. Returns: startScan.models.CountryISO: CountryISO object from DB or None. """ if validators.ipv6(host): logger.info(f'Ipv6 "{host}" is not supported by geoiplookup. Skipping.') return None cmd = f'geoiplookup {host}' _, out = run_command(cmd) if 'IP Address not found' not in out and "can't resolve hostname" not in out: country_iso = out.split(':')[1].strip().split(',')[0] country_name = out.split(':')[1].strip().split(',')[1].strip() geo_object, _ = CountryISO.objects.get_or_create( iso=country_iso, name=country_name ) geo_json = { 'iso': country_iso, 'name': country_name } if ip_id: ip = IpAddress.objects.get(pk=ip_id) ip.geo_iso = geo_object ip.save() return geo_json logger.info(f'Geo IP lookup failed for host "{host}"') return None @app.task(name='query_whois', bind=False, queue='query_whois_queue') def query_whois(ip_domain, force_reload_whois=False): """Query WHOIS information for an IP or a domain name. Args: ip_domain (str): IP address or domain name. save_domain (bool): Whether to save domain or not, default False Returns: dict: WHOIS information. """ if not force_reload_whois and Domain.objects.filter(name=ip_domain).exists() and Domain.objects.get(name=ip_domain).domain_info: domain = Domain.objects.get(name=ip_domain) if not domain.insert_date: domain.insert_date = timezone.now() domain.save() domain_info_db = domain.domain_info domain_info = DottedDict( dnssec=domain_info_db.dnssec, created=domain_info_db.created, updated=domain_info_db.updated, expires=domain_info_db.expires, geolocation_iso=domain_info_db.geolocation_iso, status=[status['name'] for status in DomainWhoisStatusSerializer(domain_info_db.status, many=True).data], whois_server=domain_info_db.whois_server, ns_records=[ns['name'] for ns in NameServersSerializer(domain_info_db.name_servers, many=True).data], registrar_name=domain_info_db.registrar.name, registrar_phone=domain_info_db.registrar.phone, registrar_email=domain_info_db.registrar.email, registrar_url=domain_info_db.registrar.url, registrant_name=domain_info_db.registrant.name, registrant_id=domain_info_db.registrant.id_str, registrant_organization=domain_info_db.registrant.organization, registrant_city=domain_info_db.registrant.city, registrant_state=domain_info_db.registrant.state, registrant_zip_code=domain_info_db.registrant.zip_code, registrant_country=domain_info_db.registrant.country, registrant_phone=domain_info_db.registrant.phone, registrant_fax=domain_info_db.registrant.fax, registrant_email=domain_info_db.registrant.email, registrant_address=domain_info_db.registrant.address, admin_name=domain_info_db.admin.name, admin_id=domain_info_db.admin.id_str, admin_organization=domain_info_db.admin.organization, admin_city=domain_info_db.admin.city, admin_state=domain_info_db.admin.state, admin_zip_code=domain_info_db.admin.zip_code, admin_country=domain_info_db.admin.country, admin_phone=domain_info_db.admin.phone, admin_fax=domain_info_db.admin.fax, admin_email=domain_info_db.admin.email, admin_address=domain_info_db.admin.address, tech_name=domain_info_db.tech.name, tech_id=domain_info_db.tech.id_str, tech_organization=domain_info_db.tech.organization, tech_city=domain_info_db.tech.city, tech_state=domain_info_db.tech.state, tech_zip_code=domain_info_db.tech.zip_code, tech_country=domain_info_db.tech.country, tech_phone=domain_info_db.tech.phone, tech_fax=domain_info_db.tech.fax, tech_email=domain_info_db.tech.email, tech_address=domain_info_db.tech.address, related_tlds=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_tlds, many=True).data], related_domains=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_domains, many=True).data], historical_ips=[ip for ip in HistoricalIPSerializer(domain_info_db.historical_ips, many=True).data], ) if domain_info_db.dns_records: a_records = [] txt_records = [] mx_records = [] dns_records = [{'name': dns['name'], 'type': dns['type']} for dns in DomainDNSRecordSerializer(domain_info_db.dns_records, many=True).data] for dns in dns_records: if dns['type'] == 'a': a_records.append(dns['name']) elif dns['type'] == 'txt': txt_records.append(dns['name']) elif dns['type'] == 'mx': mx_records.append(dns['name']) domain_info.a_records = a_records domain_info.txt_records = txt_records domain_info.mx_records = mx_records else: logger.info(f'Domain info for "{ip_domain}" not found in DB, querying whois') domain_info = DottedDict() # find domain historical ip try: historical_ips = get_domain_historical_ip_address(ip_domain) domain_info.historical_ips = historical_ips except Exception as e: logger.error(f'HistoricalIP for {ip_domain} not found!\nError: {str(e)}') historical_ips = [] # find associated domains using ip_domain try: related_domains = reverse_whois(ip_domain.split('.')[0]) except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] # find related tlds using TLSx try: related_tlds = [] output_path = '/tmp/ip_domain_tlsx.txt' tlsx_command = f'tlsx -san -cn -silent -ro -host {ip_domain} -o {output_path}' run_command( tlsx_command, shell=True, ) tlsx_output = [] with open(output_path) as f: tlsx_output = f.readlines() tldextract_target = tldextract.extract(ip_domain) for doms in tlsx_output: doms = doms.strip() tldextract_res = tldextract.extract(doms) if ip_domain != doms and tldextract_res.domain == tldextract_target.domain and tldextract_res.subdomain == '': related_tlds.append(doms) related_tlds = list(set(related_tlds)) domain_info.related_tlds = related_tlds except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] related_domains_list = [] if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_related = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_related) related_domains_list.append(_domain['name']) for _domain in related_tlds: domain_related = RelatedDomain.objects.get_or_create( name=_domain, )[0] db_domain_info.related_tlds.add(domain_related) for _ip in historical_ips: historical_ip = HistoricalIP.objects.get_or_create( ip=_ip['ip'], owner=_ip['owner'], location=_ip['location'], last_seen=_ip['last_seen'], )[0] db_domain_info.historical_ips.add(historical_ip) domain.domain_info = db_domain_info domain.save() command = f'netlas host {ip_domain} -f json' # check if netlas key is provided netlas_key = get_netlas_key() command += f' -a {netlas_key}' if netlas_key else '' result = subprocess.check_output(command.split()).decode('utf-8') if 'Failed to parse response data' in result: # do fallback return { 'status': False, 'ip_domain': ip_domain, 'result': "Netlas limit exceeded.", 'message': 'Netlas limit exceeded.' } try: result = json.loads(result) logger.info(result) whois = result.get('whois') if result.get('whois') else {} domain_info.created = whois.get('created_date') domain_info.expires = whois.get('expiration_date') domain_info.updated = whois.get('updated_date') domain_info.whois_server = whois.get('whois_server') if 'registrant' in whois: registrant = whois.get('registrant') domain_info.registrant_name = registrant.get('name') domain_info.registrant_country = registrant.get('country') domain_info.registrant_id = registrant.get('id') domain_info.registrant_state = registrant.get('province') domain_info.registrant_city = registrant.get('city') domain_info.registrant_phone = registrant.get('phone') domain_info.registrant_address = registrant.get('street') domain_info.registrant_organization = registrant.get('organization') domain_info.registrant_fax = registrant.get('fax') domain_info.registrant_zip_code = registrant.get('postal_code') email_search = EMAIL_REGEX.search(str(registrant.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.registrant_email = field_content if 'administrative' in whois: administrative = whois.get('administrative') domain_info.admin_name = administrative.get('name') domain_info.admin_country = administrative.get('country') domain_info.admin_id = administrative.get('id') domain_info.admin_state = administrative.get('province') domain_info.admin_city = administrative.get('city') domain_info.admin_phone = administrative.get('phone') domain_info.admin_address = administrative.get('street') domain_info.admin_organization = administrative.get('organization') domain_info.admin_fax = administrative.get('fax') domain_info.admin_zip_code = administrative.get('postal_code') mail_search = EMAIL_REGEX.search(str(administrative.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.admin_email = field_content if 'technical' in whois: technical = whois.get('technical') domain_info.tech_name = technical.get('name') domain_info.tech_country = technical.get('country') domain_info.tech_state = technical.get('province') domain_info.tech_id = technical.get('id') domain_info.tech_city = technical.get('city') domain_info.tech_phone = technical.get('phone') domain_info.tech_address = technical.get('street') domain_info.tech_organization = technical.get('organization') domain_info.tech_fax = technical.get('fax') domain_info.tech_zip_code = technical.get('postal_code') mail_search = EMAIL_REGEX.search(str(technical.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.tech_email = field_content if 'dns' in result: dns = result.get('dns') domain_info.mx_records = dns.get('mx') domain_info.txt_records = dns.get('txt') domain_info.a_records = dns.get('a') domain_info.ns_records = whois.get('name_servers') domain_info.dnssec = True if whois.get('dnssec') else False domain_info.status = whois.get('status') if 'registrar' in whois: registrar = whois.get('registrar') domain_info.registrar_name = registrar.get('name') domain_info.registrar_email = registrar.get('email') domain_info.registrar_phone = registrar.get('phone') domain_info.registrar_url = registrar.get('url') # find associated domains if registrant email is found related_domains = reverse_whois(domain_info.get('registrant_email')) if domain_info.get('registrant_email') else [] for _domain in related_domains: related_domains_list.append(_domain['name']) # remove duplicate domains from related domains list related_domains_list = list(set(related_domains_list)) domain_info.related_domains = related_domains_list # save to db if domain exists if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_rel = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_rel) db_domain_info.dnssec = domain_info.get('dnssec') #dates db_domain_info.created = domain_info.get('created') db_domain_info.updated = domain_info.get('updated') db_domain_info.expires = domain_info.get('expires') #registrar db_domain_info.registrar = Registrar.objects.get_or_create( name=domain_info.get('registrar_name'), email=domain_info.get('registrar_email'), phone=domain_info.get('registrar_phone'), url=domain_info.get('registrar_url'), )[0] db_domain_info.registrant = DomainRegistration.objects.get_or_create( name=domain_info.get('registrant_name'), organization=domain_info.get('registrant_organization'), address=domain_info.get('registrant_address'), city=domain_info.get('registrant_city'), state=domain_info.get('registrant_state'), zip_code=domain_info.get('registrant_zip_code'), country=domain_info.get('registrant_country'), email=domain_info.get('registrant_email'), phone=domain_info.get('registrant_phone'), fax=domain_info.get('registrant_fax'), id_str=domain_info.get('registrant_id'), )[0] db_domain_info.admin = DomainRegistration.objects.get_or_create( name=domain_info.get('admin_name'), organization=domain_info.get('admin_organization'), address=domain_info.get('admin_address'), city=domain_info.get('admin_city'), state=domain_info.get('admin_state'), zip_code=domain_info.get('admin_zip_code'), country=domain_info.get('admin_country'), email=domain_info.get('admin_email'), phone=domain_info.get('admin_phone'), fax=domain_info.get('admin_fax'), id_str=domain_info.get('admin_id'), )[0] db_domain_info.tech = DomainRegistration.objects.get_or_create( name=domain_info.get('tech_name'), organization=domain_info.get('tech_organization'), address=domain_info.get('tech_address'), city=domain_info.get('tech_city'), state=domain_info.get('tech_state'), zip_code=domain_info.get('tech_zip_code'), country=domain_info.get('tech_country'), email=domain_info.get('tech_email'), phone=domain_info.get('tech_phone'), fax=domain_info.get('tech_fax'), id_str=domain_info.get('tech_id'), )[0] for status in domain_info.get('status') or []: _status = WhoisStatus.objects.get_or_create( name=status )[0] _status.save() db_domain_info.status.add(_status) for ns in domain_info.get('ns_records') or []: _ns = NameServer.objects.get_or_create( name=ns )[0] _ns.save() db_domain_info.name_servers.add(_ns) for a in domain_info.get('a_records') or []: _a = DNSRecord.objects.get_or_create( name=a, type='a' )[0] _a.save() db_domain_info.dns_records.add(_a) for mx in domain_info.get('mx_records') or []: _mx = DNSRecord.objects.get_or_create( name=mx, type='mx' )[0] _mx.save() db_domain_info.dns_records.add(_mx) for txt in domain_info.get('txt_records') or []: _txt = DNSRecord.objects.get_or_create( name=txt, type='txt' )[0] _txt.save() db_domain_info.dns_records.add(_txt) db_domain_info.geolocation_iso = domain_info.get('registrant_country') db_domain_info.whois_server = domain_info.get('whois_server') db_domain_info.save() domain.domain_info = db_domain_info domain.save() except Exception as e: return { 'status': False, 'ip_domain': ip_domain, 'result': "unable to fetch records from WHOIS database.", 'message': str(e) } return { 'status': True, 'ip_domain': ip_domain, 'dnssec': domain_info.get('dnssec'), 'created': domain_info.get('created'), 'updated': domain_info.get('updated'), 'expires': domain_info.get('expires'), 'geolocation_iso': domain_info.get('registrant_country'), 'domain_statuses': domain_info.get('status'), 'whois_server': domain_info.get('whois_server'), 'dns': { 'a': domain_info.get('a_records'), 'mx': domain_info.get('mx_records'), 'txt': domain_info.get('txt_records'), }, 'registrar': { 'name': domain_info.get('registrar_name'), 'phone': domain_info.get('registrar_phone'), 'email': domain_info.get('registrar_email'), 'url': domain_info.get('registrar_url'), }, 'registrant': { 'name': domain_info.get('registrant_name'), 'id': domain_info.get('registrant_id'), 'organization': domain_info.get('registrant_organization'), 'address': domain_info.get('registrant_address'), 'city': domain_info.get('registrant_city'), 'state': domain_info.get('registrant_state'), 'zipcode': domain_info.get('registrant_zip_code'), 'country': domain_info.get('registrant_country'), 'phone': domain_info.get('registrant_phone'), 'fax': domain_info.get('registrant_fax'), 'email': domain_info.get('registrant_email'), }, 'admin': { 'name': domain_info.get('admin_name'), 'id': domain_info.get('admin_id'), 'organization': domain_info.get('admin_organization'), 'address':domain_info.get('admin_address'), 'city': domain_info.get('admin_city'), 'state': domain_info.get('admin_state'), 'zipcode': domain_info.get('admin_zip_code'), 'country': domain_info.get('admin_country'), 'phone': domain_info.get('admin_phone'), 'fax': domain_info.get('admin_fax'), 'email': domain_info.get('admin_email'), }, 'technical_contact': { 'name': domain_info.get('tech_name'), 'id': domain_info.get('tech_id'), 'organization': domain_info.get('tech_organization'), 'address': domain_info.get('tech_address'), 'city': domain_info.get('tech_city'), 'state': domain_info.get('tech_state'), 'zipcode': domain_info.get('tech_zip_code'), 'country': domain_info.get('tech_country'), 'phone': domain_info.get('tech_phone'), 'fax': domain_info.get('tech_fax'), 'email': domain_info.get('tech_email'), }, 'nameservers': domain_info.get('ns_records'), # 'similar_domains': domain_info.get('similar_domains'), 'related_domains': domain_info.get('related_domains'), 'related_tlds': domain_info.get('related_tlds'), 'historical_ips': domain_info.get('historical_ips'), } @app.task(name='remove_duplicate_endpoints', bind=False, queue='remove_duplicate_endpoints_queue') def remove_duplicate_endpoints( scan_history_id, domain_id, subdomain_id=None, filter_ids=[], filter_status=[200, 301, 404], duplicate_removal_fields=ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS ): """Remove duplicate endpoints. Check for implicit redirections by comparing endpoints: - [x] `content_length` similarities indicating redirections - [x] `page_title` (check for same page title) - [ ] Sign-in / login page (check for endpoints with the same words) Args: scan_history_id: ScanHistory id. domain_id (int): Domain id. subdomain_id (int, optional): Subdomain id. filter_ids (list): List of endpoint ids to filter on. filter_status (list): List of HTTP status codes to filter on. duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates """ logger.info(f'Removing duplicate endpoints based on {duplicate_removal_fields}') endpoints = ( EndPoint.objects .filter(scan_history__id=scan_history_id) .filter(target_domain__id=domain_id) ) if filter_status: endpoints = endpoints.filter(http_status__in=filter_status) if subdomain_id: endpoints = endpoints.filter(subdomain__id=subdomain_id) if filter_ids: endpoints = endpoints.filter(id__in=filter_ids) for field_name in duplicate_removal_fields: cl_query = ( endpoints .values_list(field_name) .annotate(mc=Count(field_name)) .order_by('-mc') ) for (field_value, count) in cl_query: if count > DELETE_DUPLICATES_THRESHOLD: eps_to_delete = ( endpoints .filter(**{field_name: field_value}) .order_by('discovered_date') .all()[1:] ) msg = f'Deleting {len(eps_to_delete)} endpoints [reason: same {field_name} {field_value}]' for ep in eps_to_delete: url = urlparse(ep.http_url) if url.path in ['', '/', '/login']: # try do not delete the original page that other pages redirect to continue msg += f'\n\t {ep.http_url} [{ep.http_status}] [{field_name}={field_value}]' ep.delete() logger.warning(msg) @app.task(name='run_command', bind=False, queue='run_command_queue') def run_command(cmd, cwd=None, shell=False, history_file=None, scan_id=None, activity_id=None): """Run a given command using subprocess module. Args: cmd (str): Command to run. cwd (str): Current working directory. echo (bool): Log command. shell (bool): Run within separate shell if True. history_file (str): Write command + output to history file. Returns: tuple: Tuple with return_code, output. """ logger.info(cmd) logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Run the command using subprocess popen = subprocess.Popen( cmd if shell else cmd.split(), shell=shell, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, cwd=cwd, universal_newlines=True) output = '' for stdout_line in iter(popen.stdout.readline, ""): item = stdout_line.strip() output += '\n' + item logger.debug(item) popen.stdout.close() popen.wait() return_code = popen.returncode command_obj.output = output command_obj.return_code = return_code command_obj.save() if history_file: mode = 'a' if not os.path.exists(history_file): mode = 'w' with open(history_file, mode) as f: f.write(f'\n{cmd}\n{return_code}\n{output}\n------------------\n') return return_code, output #-------------# # Other utils # #-------------# def stream_command(cmd, cwd=None, shell=False, history_file=None, encoding='utf-8', scan_id=None, activity_id=None, trunc_char=None): # Log cmd logger.info(cmd) # logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Sanitize the cmd command = cmd if shell else cmd.split() # Run the command using subprocess process = subprocess.Popen( command, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, universal_newlines=True, shell=shell) # Log the output in real-time to the database output = "" # Process the output for line in iter(lambda: process.stdout.readline(), b''): if not line: break line = line.strip() ansi_escape = re.compile(r'\x1B(?:[@-Z\\-_]|\[[0-?]*[ -/]*[@-~])') line = ansi_escape.sub('', line) line = line.replace('\\x0d\\x0a', '\n') if trunc_char and line.endswith(trunc_char): line = line[:-1] item = line # Try to parse the line as JSON try: item = json.loads(line) except json.JSONDecodeError: pass # Yield the line #logger.debug(item) yield item # Add the log line to the output output += line + "\n" # Update the command record in the database command_obj.output = output command_obj.save() # Retrieve the return code and output process.wait() return_code = process.returncode # Update the return code and final output in the database command_obj.return_code = return_code command_obj.save() # Append the command, return code and output to the history file if history_file is not None: with open(history_file, "a") as f: f.write(f"{cmd}\n{return_code}\n{output}\n") def process_httpx_response(line): """TODO: implement this""" def extract_httpx_url(line): """Extract final URL from httpx results. Always follow redirects to find the last URL. Args: line (dict): URL data output by httpx. Returns: tuple: (final_url, redirect_bool) tuple. """ status_code = line.get('status_code', 0) final_url = line.get('final_url') location = line.get('location') chain_status_codes = line.get('chain_status_codes', []) # Final URL is already looking nice, if it exists return it if final_url: return final_url, False http_url = line['url'] # fallback to url field # Handle redirects manually REDIRECT_STATUS_CODES = [301, 302] is_redirect = ( status_code in REDIRECT_STATUS_CODES or any(x in REDIRECT_STATUS_CODES for x in chain_status_codes) ) if is_redirect and location: if location.startswith(('http', 'https')): http_url = location else: http_url = f'{http_url}/{location.lstrip("/")}' # Sanitize URL http_url = sanitize_url(http_url) return http_url, is_redirect #-------------# # OSInt utils # #-------------# def get_and_save_dork_results(lookup_target, results_dir, type, lookup_keywords=None, lookup_extensions=None, delay=3, page_count=2, scan_history=None): """ Uses gofuzz to dork and store information Args: lookup_target (str): target to look into such as stackoverflow or even the target itself results_dir (str): Results directory type (str): Dork Type Title lookup_keywords (str): comma separated keywords or paths to look for lookup_extensions (str): comma separated extensions to look for delay (int): delay between each requests page_count (int): pages in google to extract information scan_history (startScan.ScanHistory): Scan History Object """ results = [] gofuzz_command = f'{GOFUZZ_EXEC_PATH} -t {lookup_target} -d {delay} -p {page_count}' if lookup_extensions: gofuzz_command += f' -e {lookup_extensions}' elif lookup_keywords: gofuzz_command += f' -w {lookup_keywords}' output_file = f'{results_dir}/gofuzz.txt' gofuzz_command += f' -o {output_file}' history_file = f'{results_dir}/commands.txt' try: run_command( gofuzz_command, shell=False, history_file=history_file, scan_id=scan_history.id, ) if not os.path.isfile(output_file): return with open(output_file) as f: for line in f.readlines(): url = line.strip() if url: results.append(url) dork, created = Dork.objects.get_or_create( type=type, url=url ) if scan_history: scan_history.dorks.add(dork) # remove output file os.remove(output_file) except Exception as e: logger.exception(e) return results def get_and_save_emails(scan_history, activity_id, results_dir): """Get and save emails from Google, Bing and Baidu. Args: scan_history (startScan.ScanHistory): Scan history object. activity_id: ScanActivity Object results_dir (str): Results directory. Returns: list: List of emails found. """ emails = [] # Proxy settings # get_random_proxy() # Gather emails from Google, Bing and Baidu output_file = f'{results_dir}/emails_tmp.txt' history_file = f'{results_dir}/commands.txt' command = f'python3 /usr/src/github/Infoga/infoga.py --domain {scan_history.domain.name} --source all --report {output_file}' try: run_command( command, shell=False, history_file=history_file, scan_id=scan_history.id, activity_id=activity_id) if not os.path.isfile(output_file): logger.info('No Email results') return [] with open(output_file) as f: for line in f.readlines(): if 'Email' in line: split_email = line.split(' ')[2] emails.append(split_email) output_path = f'{results_dir}/emails.txt' with open(output_path, 'w') as output_file: for email_address in emails: save_email(email_address, scan_history) output_file.write(f'{email_address}\n') except Exception as e: logger.exception(e) return emails def save_metadata_info(meta_dict): """Extract metadata from Google Search. Args: meta_dict (dict): Info dict. Returns: list: List of startScan.MetaFinderDocument objects. """ logger.warning(f'Getting metadata for {meta_dict.osint_target}') scan_history = ScanHistory.objects.get(id=meta_dict.scan_id) # Proxy settings get_random_proxy() # Get metadata result = extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if not result: logger.error(f'No metadata result from Google Search for {meta_dict.osint_target}.') return [] # Add metadata info to DB results = [] for metadata_name, data in result.get_metadata().items(): subdomain = Subdomain.objects.get( scan_history=meta_dict.scan_id, name=meta_dict.osint_target) metadata = DottedDict({k: v for k, v in data.items()}) meta_finder_document = MetaFinderDocument( subdomain=subdomain, target_domain=meta_dict.domain, scan_history=scan_history, url=metadata.url, doc_name=metadata_name, http_status=metadata.status_code, producer=metadata.metadata.get('Producer'), creator=metadata.metadata.get('Creator'), creation_date=metadata.metadata.get('CreationDate'), modified_date=metadata.metadata.get('ModDate'), author=metadata.metadata.get('Author'), title=metadata.metadata.get('Title'), os=metadata.metadata.get('OSInfo')) meta_finder_document.save() results.append(data) return results #-----------------# # Utils functions # #-----------------# def create_scan_activity(scan_history_id, message, status): scan_activity = ScanActivity() scan_activity.scan_of = ScanHistory.objects.get(pk=scan_history_id) scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id #--------------------# # Database functions # #--------------------# def save_vulnerability(**vuln_data): references = vuln_data.pop('references', []) cve_ids = vuln_data.pop('cve_ids', []) cwe_ids = vuln_data.pop('cwe_ids', []) tags = vuln_data.pop('tags', []) subscan = vuln_data.pop('subscan', None) # remove nulls vuln_data = replace_nulls(vuln_data) # Create vulnerability vuln, created = Vulnerability.objects.get_or_create(**vuln_data) if created: vuln.discovered_date = timezone.now() vuln.open_status = True vuln.save() # Save vuln tags for tag_name in tags or []: tag, created = VulnerabilityTags.objects.get_or_create(name=tag_name) if tag: vuln.tags.add(tag) vuln.save() # Save CVEs for cve_id in cve_ids or []: cve, created = CveId.objects.get_or_create(name=cve_id) if cve: vuln.cve_ids.add(cve) vuln.save() # Save CWEs for cve_id in cwe_ids or []: cwe, created = CweId.objects.get_or_create(name=cve_id) if cwe: vuln.cwe_ids.add(cwe) vuln.save() # Save vuln reference for url in references or []: ref, created = VulnerabilityReference.objects.get_or_create(url=url) if created: vuln.references.add(ref) vuln.save() # Save subscan id in vuln object if subscan: vuln.vuln_subscan_ids.add(subscan) vuln.save() return vuln, created def save_endpoint( http_url, ctx={}, crawl=False, is_default=False, **endpoint_data): """Get or create EndPoint object. If crawl is True, also crawl the endpoint HTTP URL with httpx. Args: http_url (str): Input HTTP URL. is_default (bool): If the url is a default url for SubDomains. scan_history (startScan.models.ScanHistory): ScanHistory object. domain (startScan.models.Domain): Domain object. subdomain (starScan.models.Subdomain): Subdomain object. results_dir (str, optional): Results directory. crawl (bool, optional): Run httpx on endpoint if True. Default: False. force (bool, optional): Force crawl even if ENABLE_HTTP_CRAWL mode is on. subscan (startScan.models.SubScan, optional): SubScan object. Returns: tuple: (startScan.models.EndPoint, created) where `created` is a boolean indicating if the object is new or already existed. """ # remove nulls endpoint_data = replace_nulls(endpoint_data) scheme = urlparse(http_url).scheme endpoint = None created = False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in http_url: logger.error(f"{http_url} is not a URL of domain {domain.name}. Skipping.") return None, False if crawl: ctx['track'] = False results = http_crawl( urls=[http_url], method='HEAD', ctx=ctx) if results: endpoint_data = results[0] endpoint_id = endpoint_data['endpoint_id'] created = endpoint_data['endpoint_created'] endpoint = EndPoint.objects.get(pk=endpoint_id) elif not scheme: return None, False else: # add dumb endpoint without probing it scan = ScanHistory.objects.filter(pk=ctx.get('scan_history_id')).first() domain = Domain.objects.filter(pk=ctx.get('domain_id')).first() if not validators.url(http_url): return None, False http_url = sanitize_url(http_url) # Try to get the first matching record (prevent duplicate error) endpoints = EndPoint.objects.filter( scan_history=scan, target_domain=domain, http_url=http_url, **endpoint_data ) if endpoints.exists(): endpoint = endpoints.first() created = False else: # No existing record, create a new one endpoint = EndPoint.objects.create( scan_history=scan, target_domain=domain, http_url=http_url, **endpoint_data ) created = True if created: endpoint.is_default = is_default endpoint.discovered_date = timezone.now() endpoint.save() subscan_id = ctx.get('subscan_id') if subscan_id: endpoint.endpoint_subscan_ids.add(subscan_id) endpoint.save() return endpoint, created def save_subdomain(subdomain_name, ctx={}): """Get or create Subdomain object. Args: subdomain_name (str): Subdomain name. scan_history (startScan.models.ScanHistory): ScanHistory object. Returns: tuple: (startScan.models.Subdomain, created) where `created` is a boolean indicating if the object has been created in DB. """ scan_id = ctx.get('scan_history_id') subscan_id = ctx.get('subscan_id') out_of_scope_subdomains = ctx.get('out_of_scope_subdomains', []) valid_domain = ( validators.domain(subdomain_name) or validators.ipv4(subdomain_name) or validators.ipv6(subdomain_name) ) if not valid_domain: logger.error(f'{subdomain_name} is not an invalid domain. Skipping.') return None, False if subdomain_name in out_of_scope_subdomains: logger.error(f'{subdomain_name} is out-of-scope. Skipping.') return None, False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in subdomain_name: logger.error(f"{subdomain_name} is not a subdomain of domain {domain.name}. Skipping.") return None, False scan = ScanHistory.objects.filter(pk=scan_id).first() domain = scan.domain if scan else None subdomain, created = Subdomain.objects.get_or_create( scan_history=scan, target_domain=domain, name=subdomain_name) if created: # logger.warning(f'Found new subdomain {subdomain_name}') subdomain.discovered_date = timezone.now() if subscan_id: subdomain.subdomain_subscan_ids.add(subscan_id) subdomain.save() return subdomain, created def save_email(email_address, scan_history=None): if not validators.email(email_address): logger.info(f'Email {email_address} is invalid. Skipping.') return None, False email, created = Email.objects.get_or_create(address=email_address) # if created: # logger.warning(f'Found new email address {email_address}') # Add email to ScanHistory if scan_history: scan_history.emails.add(email) scan_history.save() return email, created def save_employee(name, designation, scan_history=None): employee, created = Employee.objects.get_or_create( name=name, designation=designation) # if created: # logger.warning(f'Found new employee {name}') # Add employee to ScanHistory if scan_history: scan_history.employees.add(employee) scan_history.save() return employee, created def save_ip_address(ip_address, subdomain=None, subscan=None, **kwargs): if not (validators.ipv4(ip_address) or validators.ipv6(ip_address)): logger.info(f'IP {ip_address} is not a valid IP. Skipping.') return None, False ip, created = IpAddress.objects.get_or_create(address=ip_address) # if created: # logger.warning(f'Found new IP {ip_address}') # Set extra attributes for key, value in kwargs.items(): setattr(ip, key, value) ip.save() # Add IP to subdomain if subdomain: subdomain.ip_addresses.add(ip) subdomain.save() # Add subscan to IP if subscan: ip.ip_subscan_ids.add(subscan) # Geo-localize IP asynchronously if created: geo_localize.delay(ip_address, ip.id) return ip, created def save_imported_subdomains(subdomains, ctx={}): """Take a list of subdomains imported and write them to from_imported.txt. Args: subdomains (list): List of subdomain names. scan_history (startScan.models.ScanHistory): ScanHistory instance. domain (startScan.models.Domain): Domain instance. results_dir (str): Results directory. """ domain_id = ctx['domain_id'] domain = Domain.objects.get(pk=domain_id) results_dir = ctx.get('results_dir', RENGINE_RESULTS) # Validate each subdomain and de-duplicate entries subdomains = list(set([ subdomain for subdomain in subdomains if validators.domain(subdomain) and domain.name == get_domain_from_subdomain(subdomain) ])) if not subdomains: return logger.warning(f'Found {len(subdomains)} imported subdomains.') with open(f'{results_dir}/from_imported.txt', 'w+') as output_file: for name in subdomains: subdomain_name = name.strip() subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain.is_imported_subdomain = True subdomain.save() output_file.write(f'{subdomain}\n') @app.task(name='query_reverse_whois', bind=False, queue='query_reverse_whois_queue') def query_reverse_whois(lookup_keyword): """Queries Reverse WHOIS information for an organization or email address. Args: lookup_keyword (str): Registrar Name or email Returns: dict: Reverse WHOIS information. """ return get_associated_domains(lookup_keyword) @app.task(name='query_ip_history', bind=False, queue='query_ip_history_queue') def query_ip_history(domain): """Queries the IP history for a domain Args: domain (str): domain_name Returns: list: list of historical ip addresses """ return get_domain_historical_ip_address(domain) @app.task(name='gpt_vulnerability_description', bind=False, queue='gpt_queue') def gpt_vulnerability_description(vulnerability_id): """Generate and store Vulnerability Description using GPT. Args: vulnerability_id (Vulnerability Model ID): Vulnerability ID to fetch Description. """ logger.info('Getting GPT Vulnerability Description') try: lookup_vulnerability = Vulnerability.objects.get(id=vulnerability_id) lookup_url = urlparse(lookup_vulnerability.http_url) path = lookup_url.path except Exception as e: return { 'status': False, 'error': str(e) } # check in db GPTVulnerabilityReport model if vulnerability description and path matches stored = GPTVulnerabilityReport.objects.filter(url_path=path).filter(title=lookup_vulnerability.name).first() if stored: response = { 'status': True, 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: vulnerability_description = get_gpt_vuln_input_description( lookup_vulnerability.name, path ) # one can add more description here later gpt_generator = GPTVulnerabilityReportGenerator() response = gpt_generator.get_vulnerability_description(vulnerability_description) add_gpt_description_db( lookup_vulnerability.name, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) # for all vulnerabilities with the same vulnerability name this description has to be stored. # also the consition is that the url must contain a part of this. for vuln in Vulnerability.objects.filter(name=lookup_vulnerability.name, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() return response

psyray

7c01a46cea370e74385682ba7c28eaf4e58f5d69

2e089dc62f1bd64aa481750da10fa750e3aa232d

Not needed, feel free to remove

yogeshojha

yogeshojha/rengine

Fix crash on saving endpoint (FFUF related only)

Fix #1006 I've added : - a **try except** block to catch error on duplicate record returned by **get_or_create** in **saving_endpoint** method - a **check** on endpoint existence in **dir_file_fuzz** method Errors are logged to the console with the URL.  Tested and working Now we need to find why there are duplicates endpoints in the db But it's another issue

2023-11-22 02:57:45+00:00

2023-11-27 12:37:27+00:00

web/reNgine/tasks.py

import csv import json import os import pprint import subprocess import time import validators import whatportis import xmltodict import yaml import tldextract import concurrent.futures from datetime import datetime from urllib.parse import urlparse from api.serializers import SubdomainSerializer from celery import chain, chord, group from celery.result import allow_join_result from celery.utils.log import get_task_logger from django.db.models import Count from dotted_dict import DottedDict from django.utils import timezone from pycvesearch import CVESearch from metafinder.extractor import extract_metadata_from_google_search from reNgine.celery import app from reNgine.gpt import GPTVulnerabilityReportGenerator from reNgine.celery_custom_task import RengineTask from reNgine.common_func import * from reNgine.definitions import * from reNgine.settings import * from reNgine.gpt import * from reNgine.utilities import * from scanEngine.models import (EngineType, InstalledExternalTool, Notification, Proxy) from startScan.models import * from startScan.models import EndPoint, Subdomain, Vulnerability from targetApp.models import Domain """ Celery tasks. """ logger = get_task_logger(__name__) #----------------------# # Scan / Subscan tasks # #----------------------# @app.task(name='initiate_scan', bind=False, queue='initiate_scan_queue') def initiate_scan( scan_history_id, domain_id, engine_id=None, scan_type=LIVE_SCAN, results_dir=RENGINE_RESULTS, imported_subdomains=[], out_of_scope_subdomains=[], url_filter=''): """Initiate a new scan. Args: scan_history_id (int): ScanHistory id. domain_id (int): Domain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. imported_subdomains (list): Imported subdomains. out_of_scope_subdomains (list): Out-of-scope subdomains. url_filter (str): URL path. Default: '' """ # Get scan history scan = ScanHistory.objects.get(pk=scan_history_id) # Get scan engine engine_id = engine_id or scan.scan_type.id # scan history engine_id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, []) # Get domain and set last_scan_date domain = Domain.objects.get(pk=domain_id) domain.last_scan_date = timezone.now() domain.save() # Get path filter url_filter = url_filter.rstrip('/') # Get or create ScanHistory() object if scan_type == LIVE_SCAN: # immediate scan = ScanHistory.objects.get(pk=scan_history_id) scan.scan_status = RUNNING_TASK elif scan_type == SCHEDULED_SCAN: # scheduled scan = ScanHistory() scan.scan_status = INITIATED_TASK scan.scan_type = engine scan.celery_ids = [initiate_scan.request.id] scan.domain = domain scan.start_scan_date = timezone.now() scan.tasks = engine.tasks scan.results_dir = f'{results_dir}/{domain.name}_{scan.id}' add_gf_patterns = gf_patterns and 'fetch_url' in engine.tasks if add_gf_patterns: scan.used_gf_patterns = ','.join(gf_patterns) scan.save() # Create scan results dir os.makedirs(scan.results_dir) # Build task context ctx = { 'scan_history_id': scan_history_id, 'engine_id': engine_id, 'domain_id': domain.id, 'results_dir': scan.results_dir, 'url_filter': url_filter, 'yaml_configuration': config, 'out_of_scope_subdomains': out_of_scope_subdomains } ctx_str = json.dumps(ctx, indent=2) # Send start notif logger.warning(f'Starting scan {scan_history_id} with context:\n{ctx_str}') send_scan_notif.delay( scan_history_id, subscan_id=None, engine_id=engine_id, status=CELERY_TASK_STATUS_MAP[scan.scan_status]) # Save imported subdomains in DB save_imported_subdomains(imported_subdomains, ctx=ctx) # Create initial subdomain in DB: make a copy of domain as a subdomain so # that other tasks using subdomains can use it. subdomain_name = domain.name subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) # If enable_http_crawl is set, create an initial root HTTP endpoint so that # HTTP crawling can start somewhere http_url = f'{domain.name}{url_filter}' if url_filter else domain.name endpoint, _ = save_endpoint( http_url, ctx=ctx, crawl=enable_http_crawl, is_default=True, subdomain=subdomain ) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build Celery tasks, crafted according to the dependency graph below: # subdomain_discovery --> port_scan --> fetch_url --> dir_file_fuzz # osint vulnerability_scan # osint dalfox xss scan # screenshot # waf_detection workflow = chain( group( subdomain_discovery.si(ctx=ctx, description='Subdomain discovery'), osint.si(ctx=ctx, description='OS Intelligence') ), port_scan.si(ctx=ctx, description='Port scan'), fetch_url.si(ctx=ctx, description='Fetch URL'), group( dir_file_fuzz.si(ctx=ctx, description='Directories & files fuzz'), vulnerability_scan.si(ctx=ctx, description='Vulnerability scan'), screenshot.si(ctx=ctx, description='Screenshot'), waf_detection.si(ctx=ctx, description='WAF detection') ) ) # Build callback callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery chord logger.info(f'Running Celery workflow with {len(workflow.tasks) + 1} tasks') task = chain(workflow, callback).on_error(callback).delay() scan.celery_ids.append(task.id) scan.save() return { 'success': True, 'task_id': task.id } @app.task(name='initiate_subscan', bind=False, queue='subscan_queue') def initiate_subscan( scan_history_id, subdomain_id, engine_id=None, scan_type=None, results_dir=RENGINE_RESULTS, url_filter=''): """Initiate a new subscan. Args: scan_history_id (int): ScanHistory id. subdomain_id (int): Subdomain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. url_filter (str): URL path. Default: '' """ # Get Subdomain, Domain and ScanHistory subdomain = Subdomain.objects.get(pk=subdomain_id) scan = ScanHistory.objects.get(pk=subdomain.scan_history.id) domain = Domain.objects.get(pk=subdomain.target_domain.id) # Get EngineType engine_id = engine_id or scan.scan_type.id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Create scan activity of SubScan Model subscan = SubScan( start_scan_date=timezone.now(), celery_ids=[initiate_subscan.request.id], scan_history=scan, subdomain=subdomain, type=scan_type, status=RUNNING_TASK, engine=engine) subscan.save() # Get YAML configuration config = yaml.safe_load(engine.yaml_configuration) # Create results directory results_dir = f'{scan.results_dir}/subscans/{subscan.id}' os.makedirs(results_dir, exist_ok=True) # Run task method = globals().get(scan_type) if not method: logger.warning(f'Task {scan_type} is not supported by reNgine. Skipping') return scan.tasks.append(scan_type) scan.save() # Send start notif send_scan_notif.delay( scan.id, subscan_id=subscan.id, engine_id=engine_id, status='RUNNING') # Build context ctx = { 'scan_history_id': scan.id, 'subscan_id': subscan.id, 'engine_id': engine_id, 'domain_id': domain.id, 'subdomain_id': subdomain.id, 'yaml_configuration': config, 'results_dir': results_dir, 'url_filter': url_filter } # Create initial endpoints in DB: find domain HTTP endpoint so that HTTP # crawling can start somewhere base_url = f'{subdomain.name}{url_filter}' if url_filter else subdomain.name endpoint, _ = save_endpoint( base_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build header + callback workflow = method.si(ctx=ctx) callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery tasks task = chain(workflow, callback).on_error(callback).delay() subscan.celery_ids.append(task.id) subscan.save() return { 'success': True, 'task_id': task.id } @app.task(name='report', bind=False, queue='report_queue') def report(ctx={}, description=None): """Report task running after all other tasks. Mark ScanHistory or SubScan object as completed and update with final status, log run details and send notification. Args: description (str, optional): Task description shown in UI. """ # Get objects subscan_id = ctx.get('subscan_id') scan_id = ctx.get('scan_history_id') engine_id = ctx.get('engine_id') scan = ScanHistory.objects.filter(pk=scan_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() # Get failed tasks tasks = ScanActivity.objects.filter(scan_of=scan).all() if subscan: tasks = tasks.filter(celery_id__in=subscan.celery_ids) failed_tasks = tasks.filter(status=FAILED_TASK) # Get task status failed_count = failed_tasks.count() status = SUCCESS_TASK if failed_count == 0 else FAILED_TASK status_h = 'SUCCESS' if failed_count == 0 else 'FAILED' # Update scan / subscan status if subscan: subscan.stop_scan_date = timezone.now() subscan.status = status subscan.save() else: scan.scan_status = status scan.stop_scan_date = timezone.now() scan.save() # Send scan status notif send_scan_notif.delay( scan_history_id=scan_id, subscan_id=subscan_id, engine_id=engine_id, status=status_h) #------------------------- # # Tracked reNgine tasks # #--------------------------# @app.task(name='subdomain_discovery', queue='main_scan_queue', base=RengineTask, bind=True) def subdomain_discovery( self, host=None, ctx=None, description=None): """Uses a set of tools (see SUBDOMAIN_SCAN_DEFAULT_TOOLS) to scan all subdomains associated with a domain. Args: host (str): Hostname to scan. Returns: subdomains (list): List of subdomain names. """ if not host: host = self.subdomain.name if self.subdomain else self.domain.name if self.url_filter: logger.warning(f'Ignoring subdomains scan as an URL path filter was passed ({self.url_filter}).') return # Config config = self.yaml_configuration.get(SUBDOMAIN_DISCOVERY) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL) or self.yaml_configuration.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) tools = config.get(USES_TOOLS, SUBDOMAIN_SCAN_DEFAULT_TOOLS) default_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=True).filter(is_subdomain_gathering=True)] custom_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=False).filter(is_subdomain_gathering=True)] send_subdomain_changes, send_interesting = False, False notif = Notification.objects.first() if notif: send_subdomain_changes = notif.send_subdomain_changes_notif send_interesting = notif.send_interesting_notif # Gather tools to run for subdomain scan if ALL in tools: tools = SUBDOMAIN_SCAN_DEFAULT_TOOLS + custom_subdomain_tools tools = [t.lower() for t in tools] # Make exception for amass since tool name is amass, but command is amass-active/passive default_subdomain_tools.append('amass-passive') default_subdomain_tools.append('amass-active') # Run tools for tool in tools: cmd = None logger.info(f'Scanning subdomains for {host} with {tool}') proxy = get_random_proxy() if tool in default_subdomain_tools: if tool == 'amass-passive': cmd = f'amass enum -passive -d {host} -o {self.results_dir}/subdomains_amass.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' elif tool == 'amass-active': use_amass_config = config.get(USE_AMASS_CONFIG, False) amass_wordlist_name = config.get(AMASS_WORDLIST, 'deepmagic.com-prefixes-top50000') wordlist_path = f'/usr/src/wordlist/{amass_wordlist_name}.txt' cmd = f'amass enum -active -d {host} -o {self.results_dir}/subdomains_amass_active.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' cmd += f' -brute -w {wordlist_path}' elif tool == 'sublist3r': cmd = f'python3 /usr/src/github/Sublist3r/sublist3r.py -d {host} -t {threads} -o {self.results_dir}/subdomains_sublister.txt' elif tool == 'subfinder': cmd = f'subfinder -d {host} -o {self.results_dir}/subdomains_subfinder.txt' use_subfinder_config = config.get(USE_SUBFINDER_CONFIG, False) cmd += ' -config /root/.config/subfinder/config.yaml' if use_subfinder_config else '' cmd += f' -proxy {proxy}' if proxy else '' cmd += f' -timeout {timeout}' if timeout else '' cmd += f' -t {threads}' if threads else '' cmd += f' -silent' elif tool == 'oneforall': cmd = f'python3 /usr/src/github/OneForAll/oneforall.py --target {host} run' cmd_extract = f'cut -d\',\' -f6 /usr/src/github/OneForAll/results/{host}.csv > {self.results_dir}/subdomains_oneforall.txt' cmd_rm = f'rm -rf /usr/src/github/OneForAll/results/{host}.csv' cmd += f' && {cmd_extract} && {cmd_rm}' elif tool == 'ctfr': results_file = self.results_dir + '/subdomains_ctfr.txt' cmd = f'python3 /usr/src/github/ctfr/ctfr.py -d {host} -o {results_file}' cmd_extract = f"cat {results_file} | sed 's/\*.//g' | tail -n +12 | uniq | sort > {results_file}" cmd += f' && {cmd_extract}' elif tool == 'tlsx': results_file = self.results_dir + '/subdomains_tlsx.txt' cmd = f'tlsx -san -cn -silent -ro -host {host}' cmd += f" | sed -n '/^\([a-zA-Z0-9]\([-a-zA-Z0-9]*[a-zA-Z0-9]\)\?\.\)\+{host}$/p' | uniq | sort" cmd += f' > {results_file}' elif tool == 'netlas': results_file = self.results_dir + '/subdomains_netlas.txt' cmd = f'netlas search -d domain -i domain domain:"*.{host}" -f json' netlas_key = get_netlas_key() cmd += f' -a {netlas_key}' if netlas_key else '' cmd_extract = f"grep -oE '([a-zA-Z0-9]([-a-zA-Z0-9]*[a-zA-Z0-9])?\.)+{host}'" cmd += f' | {cmd_extract} > {results_file}' elif tool in custom_subdomain_tools: tool_query = InstalledExternalTool.objects.filter(name__icontains=tool.lower()) if not tool_query.exists(): logger.error(f'Missing {{TARGET}} and {{OUTPUT}} placeholders in {tool} configuration. Skipping.') continue custom_tool = tool_query.first() cmd = custom_tool.subdomain_gathering_command if '{TARGET}' in cmd and '{OUTPUT}' in cmd: cmd = cmd.replace('{TARGET}', host) cmd = cmd.replace('{OUTPUT}', f'{self.results_dir}/subdomains_{tool}.txt') cmd = cmd.replace('{PATH}', custom_tool.github_clone_path) if '{PATH}' in cmd else cmd else: logger.warning( f'Subdomain discovery tool "{tool}" is not supported by reNgine. Skipping.') continue # Run tool try: run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) except Exception as e: logger.error( f'Subdomain discovery tool "{tool}" raised an exception') logger.exception(e) # Gather all the tools' results in one single file. Write subdomains into # separate files, and sort all subdomains. run_command( f'cat {self.results_dir}/subdomains_*.txt > {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {self.output_path} -o {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) with open(self.output_path) as f: lines = f.readlines() # Parse the output_file file and store Subdomain and EndPoint objects found # in db. subdomain_count = 0 subdomains = [] urls = [] for line in lines: subdomain_name = line.strip() valid_url = bool(validators.url(subdomain_name)) valid_domain = ( bool(validators.domain(subdomain_name)) or bool(validators.ipv4(subdomain_name)) or bool(validators.ipv6(subdomain_name)) or valid_url ) if not valid_domain: logger.error(f'Subdomain {subdomain_name} is not a valid domain, IP or URL. Skipping.') continue if valid_url: subdomain_name = urlparse(subdomain_name).netloc if subdomain_name in self.out_of_scope_subdomains: logger.error(f'Subdomain {subdomain_name} is out of scope. Skipping.') continue # Add subdomain subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain_count += 1 subdomains.append(subdomain) urls.append(subdomain.name) # Bulk crawl subdomains if enable_http_crawl: ctx['track'] = True http_crawl(urls, ctx=ctx, is_ran_from_subdomain_scan=True) # Find root subdomain endpoints for subdomain in subdomains: pass # Send notifications subdomains_str = '\n'.join([f'• `{subdomain.name}`' for subdomain in subdomains]) self.notify(fields={ 'Subdomain count': len(subdomains), 'Subdomains': subdomains_str, }) if send_subdomain_changes and self.scan_id and self.domain_id: added = get_new_added_subdomain(self.scan_id, self.domain_id) removed = get_removed_subdomain(self.scan_id, self.domain_id) if added: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in added]) self.notify(fields={'Added subdomains': subdomains_str}) if removed: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in removed]) self.notify(fields={'Removed subdomains': subdomains_str}) if send_interesting and self.scan_id and self.domain_id: interesting_subdomains = get_interesting_subdomains(self.scan_id, self.domain_id) if interesting_subdomains: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in interesting_subdomains]) self.notify(fields={'Interesting subdomains': subdomains_str}) return SubdomainSerializer(subdomains, many=True).data @app.task(name='osint', queue='main_scan_queue', base=RengineTask, bind=True) def osint(self, host=None, ctx={}, description=None): """Run Open-Source Intelligence tools on selected domain. Args: host (str): Hostname to scan. Returns: dict: Results from osint discovery and dorking. """ config = self.yaml_configuration.get(OSINT) or OSINT_DEFAULT_CONFIG results = {} grouped_tasks = [] if 'discover' in config: ctx['track'] = False # results = osint_discovery(host=host, ctx=ctx) _task = osint_discovery.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, activity_id=self.activity_id, results_dir=self.results_dir, ctx=ctx ) grouped_tasks.append(_task) if OSINT_DORK in config or OSINT_CUSTOM_DORK in config: _task = dorking.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, results_dir=self.results_dir ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('OSINT Tasks finished...') # with open(self.output_path, 'w') as f: # json.dump(results, f, indent=4) # # return results @app.task(name='osint_discovery', queue='osint_discovery_queue', bind=False) def osint_discovery(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run OSINT discovery. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: dict: osint metadat and theHarvester and h8mail results. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) osint_lookup = config.get(OSINT_DISCOVER, []) osint_intensity = config.get(INTENSITY, 'normal') documents_limit = config.get(OSINT_DOCUMENTS_LIMIT, 50) results = {} meta_info = [] emails = [] creds = [] # Get and save meta info if 'metainfo' in osint_lookup: if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': host, 'domain': host, 'scan_id': scan_history_id, 'documents_limit': documents_limit }) meta_info.append(save_metadata_info(meta_dict)) # TODO: disabled for now # elif osint_intensity == 'deep': # subdomains = Subdomain.objects # if self.scan: # subdomains = subdomains.filter(scan_history=self.scan) # for subdomain in subdomains: # meta_dict = DottedDict({ # 'osint_target': subdomain.name, # 'domain': self.domain, # 'scan_id': self.scan_id, # 'documents_limit': documents_limit # }) # meta_info.append(save_metadata_info(meta_dict)) grouped_tasks = [] if 'emails' in osint_lookup: emails = get_and_save_emails(scan_history, activity_id, results_dir) emails_str = '\n'.join([f'• `{email}`' for email in emails]) # self.notify(fields={'Emails': emails_str}) # ctx['track'] = False _task = h8mail.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) if 'employees' in osint_lookup: ctx['track'] = False _task = theHarvester.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) # results['emails'] = results.get('emails', []) + emails # results['creds'] = creds # results['meta_info'] = meta_info return results @app.task(name='dorking', bind=False, queue='dorking_queue') def dorking(config, host, scan_history_id, results_dir): """Run Google dorks. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: list: Dorking results for each dork ran. """ # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh scan_history = ScanHistory.objects.get(pk=scan_history_id) dorks = config.get(OSINT_DORK, []) custom_dorks = config.get(OSINT_CUSTOM_DORK, []) results = [] # custom dorking has higher priority try: for custom_dork in custom_dorks: lookup_target = custom_dork.get('lookup_site') # replace with original host if _target_ lookup_target = host if lookup_target == '_target_' else lookup_target if 'lookup_extensions' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_extensions=custom_dork.get('lookup_extensions'), scan_history=scan_history ) elif 'lookup_keywords' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_keywords=custom_dork.get('lookup_keywords'), scan_history=scan_history ) except Exception as e: logger.exception(e) # default dorking try: for dork in dorks: logger.info(f'Getting dork information for {dork}') if dork == 'stackoverflow': results = get_and_save_dork_results( lookup_target='stackoverflow.com', results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'login_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/login/,login.html', page_count=5, scan_history=scan_history ) elif dork == 'admin_panels': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/admin/,admin.html', page_count=5, scan_history=scan_history ) elif dork == 'dashboard_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/dashboard/,dashboard.html', page_count=5, scan_history=scan_history ) elif dork == 'social_media' : social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'reddit.com' ] for site in social_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'project_management' : project_websites = [ 'trello.com', 'atlassian.net' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'code_sharing' : project_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'config_files' : config_file_exts = [ 'env', 'xml', 'conf', 'toml', 'yml', 'yaml', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(config_file_exts), page_count=4, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keyword = 'Jenkins' results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=lookup_keyword, page_count=1, scan_history=scan_history ) elif dork == 'wordpress_files' : lookup_keywords = [ '/wp-content/', '/wp-includes/' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'php_error' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'exposed_documents' : docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(docs_file_ext), page_count=7, scan_history=scan_history ) elif dork == 'db_files' : file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) elif dork == 'git_exposed' : file_ext = [ 'git', ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) except Exception as e: logger.exception(e) return results @app.task(name='theHarvester', queue='theHarvester_queue', bind=False) def theHarvester(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run theHarvester to get save emails, hosts, employees found in domain. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: dict: Dict of emails, employees, hosts and ips found during crawling. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) output_path_json = f'{results_dir}/theHarvester.json' theHarvester_dir = '/usr/src/github/theHarvester' history_file = f'{results_dir}/commands.txt' cmd = f'python3 {theHarvester_dir}/theHarvester.py -d {host} -b all -f {output_path_json}' # Update proxies.yaml proxy_query = Proxy.objects.all() if proxy_query.exists(): proxy = proxy_query.first() if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(f'{theHarvester_dir}/proxies.yaml', 'w') as file: yaml.dump(yaml_data, file) # Run cmd run_command( cmd, shell=False, cwd=theHarvester_dir, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) # Get file location if not os.path.isfile(output_path_json): logger.error(f'Could not open {output_path_json}') return {} # Load theHarvester results with open(output_path_json, 'r') as f: data = json.load(f) # Re-indent theHarvester JSON with open(output_path_json, 'w') as f: json.dump(data, f, indent=4) emails = data.get('emails', []) for email_address in emails: email, _ = save_email(email_address, scan_history=scan_history) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) linkedin_people = data.get('linkedin_people', []) for people in linkedin_people: employee, _ = save_employee( people, designation='linkedin', scan_history=scan_history) # if employee: # self.notify(fields={'LinkedIn people': f'• {employee.name}'}) twitter_people = data.get('twitter_people', []) for people in twitter_people: employee, _ = save_employee( people, designation='twitter', scan_history=scan_history) # if employee: # self.notify(fields={'Twitter people': f'• {employee.name}'}) hosts = data.get('hosts', []) urls = [] for host in hosts: split = tuple(host.split(':')) http_url = split[0] subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) endpoint, _ = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain) # if endpoint: # urls.append(endpoint.http_url) # self.notify(fields={'Hosts': f'• {endpoint.http_url}'}) # if enable_http_crawl: # ctx['track'] = False # http_crawl(urls, ctx=ctx) # TODO: Lots of ips unrelated with our domain are found, disabling # this for now. # ips = data.get('ips', []) # for ip_address in ips: # ip, created = save_ip_address( # ip_address, # subscan=subscan) # if ip: # send_task_notif.delay( # 'osint', # scan_history_id=scan_history_id, # subscan_id=subscan_id, # severity='success', # update_fields={'IPs': f'{ip.address}'}) return data @app.task(name='h8mail', queue='h8mail_queue', bind=False) def h8mail(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run h8mail. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: list[dict]: List of credentials info. """ logger.warning('Getting leaked credentials') scan_history = ScanHistory.objects.get(pk=scan_history_id) input_path = f'{results_dir}/emails.txt' output_file = f'{results_dir}/h8mail.json' cmd = f'h8mail -t {input_path} --json {output_file}' history_file = f'{results_dir}/commands.txt' run_command( cmd, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) with open(output_file) as f: data = json.load(f) creds = data.get('targets', []) # TODO: go through h8mail output and save emails to DB for cred in creds: logger.warning(cred) email_address = cred['target'] pwn_num = cred['pwn_num'] pwn_data = cred.get('data', []) email, created = save_email(email_address, scan_history=scan) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) return creds @app.task(name='screenshot', queue='main_scan_queue', base=RengineTask, bind=True) def screenshot(self, ctx={}, description=None): """Uses EyeWitness to gather screenshot of a domain and/or url. Args: description (str, optional): Task description shown in UI. """ # Config screenshots_path = f'{self.results_dir}/screenshots' output_path = f'{self.results_dir}/screenshots/{self.filename}' alive_endpoints_file = f'{self.results_dir}/endpoints_alive.txt' config = self.yaml_configuration.get(SCREENSHOT) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT + 5) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) # If intensity is normal, grab only the root endpoints of each subdomain strict = True if intensity == 'normal' else False # Get URLs to take screenshot of get_http_urls( is_alive=enable_http_crawl, strict=strict, write_filepath=alive_endpoints_file, get_only_default_urls=True, ctx=ctx ) # Send start notif notification = Notification.objects.first() send_output_file = notification.send_scan_output_file if notification else False # Run cmd cmd = f'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py -f {alive_endpoints_file} -d {screenshots_path} --no-prompt' cmd += f' --timeout {timeout}' if timeout > 0 else '' cmd += f' --threads {threads}' if threads > 0 else '' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(output_path): logger.error(f'Could not load EyeWitness results at {output_path} for {self.domain.name}.') return # Loop through results and save objects in DB screenshot_paths = [] with open(output_path, 'r') as file: reader = csv.reader(file) for row in reader: "Protocol,Port,Domain,Request Status,Screenshot Path, Source Path" protocol, port, subdomain_name, status, screenshot_path, source_path = tuple(row) logger.info(f'{protocol}:{port}:{subdomain_name}:{status}') subdomain_query = Subdomain.objects.filter(name=subdomain_name) if self.scan: subdomain_query = subdomain_query.filter(scan_history=self.scan) if status == 'Successful' and subdomain_query.exists(): subdomain = subdomain_query.first() screenshot_paths.append(screenshot_path) subdomain.screenshot_path = screenshot_path.replace('/usr/src/scan_results/', '') subdomain.save() logger.warning(f'Added screenshot for {subdomain.name} to DB') # Remove all db, html extra files in screenshot results run_command( 'rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format(screenshots_path), shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'rm -rf {screenshots_path}/source', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Send finish notifs screenshots_str = '• ' + '\n• '.join([f'`{path}`' for path in screenshot_paths]) self.notify(fields={'Screenshots': screenshots_str}) if send_output_file: for path in screenshot_paths: title = get_output_file_name( self.scan_id, self.subscan_id, self.filename) send_file_to_discord.delay(path, title) @app.task(name='port_scan', queue='main_scan_queue', base=RengineTask, bind=True) def port_scan(self, hosts=[], ctx={}, description=None): """Run port scan. Args: hosts (list, optional): Hosts to run port scan on. description (str, optional): Task description shown in UI. Returns: list: List of open ports (dict). """ input_file = f'{self.results_dir}/input_subdomains_port_scan.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(PORT_SCAN) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) exclude_ports = config.get(NAABU_EXCLUDE_PORTS, []) exclude_subdomains = config.get(NAABU_EXCLUDE_SUBDOMAINS, False) ports = config.get(PORTS, NAABU_DEFAULT_PORTS) ports = [str(port) for port in ports] rate_limit = config.get(NAABU_RATE) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) passive = config.get(NAABU_PASSIVE, False) use_naabu_config = config.get(USE_NAABU_CONFIG, False) exclude_ports_str = ','.join(return_iterable(exclude_ports)) # nmap args nmap_enabled = config.get(ENABLE_NMAP, False) nmap_cmd = config.get(NMAP_COMMAND, '') nmap_script = config.get(NMAP_SCRIPT, '') nmap_script = ','.join(return_iterable(nmap_script)) nmap_script_args = config.get(NMAP_SCRIPT_ARGS) if hosts: with open(input_file, 'w') as f: f.write('\n'.join(hosts)) else: hosts = get_subdomains( write_filepath=input_file, exclude_subdomains=exclude_subdomains, ctx=ctx) # Build cmd cmd = 'naabu -json -exclude-cdn' cmd += f' -list {input_file}' if len(hosts) > 0 else f' -host {hosts[0]}' if 'full' in ports or 'all' in ports: ports_str = ' -p "-"' elif 'top-100' in ports: ports_str = ' -top-ports 100' elif 'top-1000' in ports: ports_str = ' -top-ports 1000' else: ports_str = ','.join(ports) ports_str = f' -p {ports_str}' cmd += ports_str cmd += ' -config /root/.config/naabu/config.yaml' if use_naabu_config else '' cmd += f' -proxy "{proxy}"' if proxy else '' cmd += f' -c {threads}' if threads else '' cmd += f' -rate {rate_limit}' if rate_limit > 0 else '' cmd += f' -timeout {timeout*1000}' if timeout > 0 else '' cmd += f' -passive' if passive else '' cmd += f' -exclude-ports {exclude_ports_str}' if exclude_ports else '' cmd += f' -silent' # Execute cmd and gather results results = [] urls = [] ports_data = {} for line in stream_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) port_number = line['port'] ip_address = line['ip'] host = line.get('host') or ip_address if port_number == 0: continue # Grab subdomain subdomain = Subdomain.objects.filter( name=host, target_domain=self.domain, scan_history=self.scan ).first() # Add IP DB ip, _ = save_ip_address(ip_address, subdomain, subscan=self.subscan) if self.subscan: ip.ip_subscan_ids.add(self.subscan) ip.save() # Add endpoint to DB # port 80 and 443 not needed as http crawl already does that. if port_number not in [80, 443]: http_url = f'{host}:{port_number}' endpoint, _ = save_endpoint( http_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint: http_url = endpoint.http_url urls.append(http_url) # Add Port in DB port_details = whatportis.get_ports(str(port_number)) service_name = port_details[0].name if len(port_details) > 0 else 'unknown' description = port_details[0].description if len(port_details) > 0 else '' # get or create port port, created = Port.objects.get_or_create( number=port_number, service_name=service_name, description=description ) if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port.save() ip.ports.add(port) ip.save() if host in ports_data: ports_data[host].append(port_number) else: ports_data[host] = [port_number] # Send notification logger.warning(f'Found opened port {port_number} on {ip_address} ({host})') if len(ports_data) == 0: logger.info('Finished running naabu port scan - No open ports found.') if nmap_enabled: logger.info('Nmap scans skipped') return ports_data # Send notification fields_str = '' for host, ports in ports_data.items(): ports_str = ', '.join([f'`{port}`' for port in ports]) fields_str += f'• `{host}`: {ports_str}\n' self.notify(fields={'Ports discovered': fields_str}) # Save output to file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) logger.info('Finished running naabu port scan.') # Process nmap results: 1 process per host sigs = [] if nmap_enabled: logger.warning(f'Starting nmap scans ...') logger.warning(ports_data) for host, port_list in ports_data.items(): ports_str = '_'.join([str(p) for p in port_list]) ctx_nmap = ctx.copy() ctx_nmap['description'] = get_task_title(f'nmap_{host}', self.scan_id, self.subscan_id) ctx_nmap['track'] = False sig = nmap.si( cmd=nmap_cmd, ports=port_list, host=host, script=nmap_script, script_args=nmap_script_args, max_rate=rate_limit, ctx=ctx_nmap) sigs.append(sig) task = group(sigs).apply_async() with allow_join_result(): results = task.get() return ports_data @app.task(name='nmap', queue='main_scan_queue', base=RengineTask, bind=True) def nmap( self, cmd=None, ports=[], host=None, input_file=None, script=None, script_args=None, max_rate=None, ctx={}, description=None): """Run nmap on a host. Args: cmd (str, optional): Existing nmap command to complete. ports (list, optional): List of ports to scan. host (str, optional): Host to scan. input_file (str, optional): Input hosts file. script (str, optional): NSE script to run. script_args (str, optional): NSE script args. max_rate (int): Max rate. description (str, optional): Task description shown in UI. """ notif = Notification.objects.first() ports_str = ','.join(str(port) for port in ports) self.filename = self.filename.replace('.txt', '.xml') filename_vulns = self.filename.replace('.xml', '_vulns.json') output_file = self.output_path output_file_xml = f'{self.results_dir}/{host}_{self.filename}' vulns_file = f'{self.results_dir}/{host}_{filename_vulns}' logger.warning(f'Running nmap on {host}:{ports}') # Build cmd nmap_cmd = get_nmap_cmd( cmd=cmd, ports=ports_str, script=script, script_args=script_args, max_rate=max_rate, host=host, input_file=input_file, output_file=output_file_xml) # Run cmd run_command( nmap_cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Get nmap XML results and convert to JSON vulns = parse_nmap_results(output_file_xml, output_file) with open(vulns_file, 'w') as f: json.dump(vulns, f, indent=4) # Save vulnerabilities found by nmap vulns_str = '' for vuln_data in vulns: # URL is not necessarily an HTTP URL when running nmap (can be any # other vulnerable protocols). Look for existing endpoint and use its # URL as vulnerability.http_url if it exists. url = vuln_data['http_url'] endpoint = EndPoint.objects.filter(http_url__contains=url).first() if endpoint: vuln_data['http_url'] = endpoint.http_url vuln, created = save_vulnerability( target_domain=self.domain, subdomain=self.subdomain, scan_history=self.scan, subscan=self.subscan, endpoint=endpoint, **vuln_data) vulns_str += f'• {str(vuln)}\n' if created: logger.warning(str(vuln)) # Send only 1 notif for all vulns to reduce number of notifs if notif and notif.send_vuln_notif and vulns_str: logger.warning(vulns_str) self.notify(fields={'CVEs': vulns_str}) return vulns @app.task(name='waf_detection', queue='main_scan_queue', base=RengineTask, bind=True) def waf_detection(self, ctx={}, description=None): """ Uses wafw00f to check for the presence of a WAF. Args: description (str, optional): Task description shown in UI. Returns: list: List of startScan.models.Waf objects. """ input_path = f'{self.results_dir}/input_endpoints_waf_detection.txt' config = self.yaml_configuration.get(WAF_DETECTION) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Get alive endpoints from DB get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) cmd = f'wafw00f -i {input_path} -o {self.output_path}' run_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(self.output_path): logger.error(f'Could not find {self.output_path}') return with open(self.output_path) as file: wafs = file.readlines() for line in wafs: line = " ".join(line.split()) splitted = line.split(' ', 1) waf_info = splitted[1].strip() waf_name = waf_info[:waf_info.find('(')].strip() waf_manufacturer = waf_info[waf_info.find('(')+1:waf_info.find(')')].strip().replace('.', '') http_url = sanitize_url(splitted[0].strip()) if not waf_name or waf_name == 'None': continue # Add waf to db waf, _ = Waf.objects.get_or_create( name=waf_name, manufacturer=waf_manufacturer ) # Add waf info to Subdomain in DB subdomain = get_subdomain_from_url(http_url) logger.info(f'Wafw00f Subdomain : {subdomain}') subdomain_query, _ = Subdomain.objects.get_or_create(scan_history=self.scan, name=subdomain) subdomain_query.waf.add(waf) subdomain_query.save() return wafs @app.task(name='dir_file_fuzz', queue='main_scan_queue', base=RengineTask, bind=True) def dir_file_fuzz(self, ctx={}, description=None): """Perform directory scan, and currently uses `ffuf` as a default tool. Args: description (str, optional): Task description shown in UI. Returns: list: List of URLs discovered. """ # Config cmd = 'ffuf' config = self.yaml_configuration.get(DIR_FILE_FUZZ) or {} custom_header = self.yaml_configuration.get(CUSTOM_HEADER) auto_calibration = config.get(AUTO_CALIBRATION, True) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) extensions = config.get(EXTENSIONS, DEFAULT_DIR_FILE_FUZZ_EXTENSIONS) # prepend . on extensions extensions = [ext if ext.startswith('.') else '.' + ext for ext in extensions] extensions_str = ','.join(map(str, extensions)) follow_redirect = config.get(FOLLOW_REDIRECT, FFUF_DEFAULT_FOLLOW_REDIRECT) max_time = config.get(MAX_TIME, 0) match_http_status = config.get(MATCH_HTTP_STATUS, FFUF_DEFAULT_MATCH_HTTP_STATUS) mc = ','.join([str(c) for c in match_http_status]) recursive_level = config.get(RECURSIVE_LEVEL, FFUF_DEFAULT_RECURSIVE_LEVEL) stop_on_error = config.get(STOP_ON_ERROR, False) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) wordlist_name = config.get(WORDLIST, 'dicc') delay = rate_limit / (threads * 100) # calculate request pause delay from rate_limit and number of threads input_path = f'{self.results_dir}/input_dir_file_fuzz.txt' # Get wordlist wordlist_name = 'dicc' if wordlist_name == 'default' else wordlist_name wordlist_path = f'/usr/src/wordlist/{wordlist_name}.txt' # Build command cmd += f' -w {wordlist_path}' cmd += f' -e {extensions_str}' if extensions else '' cmd += f' -maxtime {max_time}' if max_time > 0 else '' cmd += f' -p {delay}' if delay > 0 else '' cmd += f' -recursion -recursion-depth {recursive_level} ' if recursive_level > 0 else '' cmd += f' -t {threads}' if threads and threads > 0 else '' cmd += f' -timeout {timeout}' if timeout and timeout > 0 else '' cmd += ' -se' if stop_on_error else '' cmd += ' -fr' if follow_redirect else '' cmd += ' -ac' if auto_calibration else '' cmd += f' -mc {mc}' if mc else '' cmd += f' -H "{custom_header}"' if custom_header else '' # Grab URLs to fuzz urls = get_http_urls( is_alive=True, ignore_files=False, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) logger.warning(urls) # Loop through URLs and run command results = [] for url in urls: ''' Above while fetching urls, we are not ignoring files, because some default urls may redirect to https://example.com/login.php so, ignore_files is set to False but, during fuzzing, we will only need part of the path, in above example it is still a good idea to ffuf base url https://example.com so files from base url ''' url_parse = urlparse(url) url = url_parse.scheme + '://' + url_parse.netloc url += '/FUZZ' # TODO: fuzz not only URL but also POST / PUT / headers proxy = get_random_proxy() # Build final cmd fcmd = cmd fcmd += f' -x {proxy}' if proxy else '' fcmd += f' -u {url} -json' # Initialize DirectoryScan object dirscan = DirectoryScan() dirscan.scanned_date = timezone.now() dirscan.command_line = fcmd dirscan.save() # Loop through results and populate EndPoint and DirectoryFile in DB results = [] for line in stream_command( fcmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) name = line['input'].get('FUZZ') length = line['length'] status = line['status'] words = line['words'] url = line['url'] lines = line['lines'] content_type = line['content-type'] duration = line['duration'] if not name: logger.error(f'FUZZ not found for "{url}"') continue endpoint, created = save_endpoint(url, crawl=False, ctx=ctx) # endpoint.is_default = False endpoint.http_status = status endpoint.content_length = length endpoint.response_time = duration / 1000000000 endpoint.save() if created: urls.append(endpoint.http_url) endpoint.status = status endpoint.content_type = content_type endpoint.content_length = length dfile, created = DirectoryFile.objects.get_or_create( name=name, length=length, words=words, lines=lines, content_type=content_type, url=url) dfile.http_status = status dfile.save() # if created: # logger.warning(f'Found new directory or file {url}') dirscan.directory_files.add(dfile) dirscan.save() if self.subscan: dirscan.dir_subscan_ids.add(self.subscan) subdomain_name = get_subdomain_from_url(endpoint.http_url) subdomain = Subdomain.objects.get(name=subdomain_name, scan_history=self.scan) subdomain.directories.add(dirscan) subdomain.save() # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl(urls, ctx=ctx) return results @app.task(name='fetch_url', queue='main_scan_queue', base=RengineTask, bind=True) def fetch_url(self, urls=[], ctx={}, description=None): """Fetch URLs using different tools like gauplus, gau, gospider, waybackurls ... Args: urls (list): List of URLs to start from. description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/input_endpoints_fetch_url.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(FETCH_URL) or {} should_remove_duplicate_endpoints = config.get(REMOVE_DUPLICATE_ENDPOINTS, True) duplicate_removal_fields = config.get(DUPLICATE_REMOVAL_FIELDS, ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, DEFAULT_GF_PATTERNS) ignore_file_extension = config.get(IGNORE_FILE_EXTENSION, DEFAULT_IGNORE_FILE_EXTENSIONS) tools = config.get(USES_TOOLS, ENDPOINT_SCAN_DEFAULT_TOOLS) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) domain_request_headers = self.domain.request_headers if self.domain else None custom_header = domain_request_headers or self.yaml_configuration.get(CUSTOM_HEADER) exclude_subdomains = config.get(EXCLUDED_SUBDOMAINS, False) # Get URLs to scan and save to input file if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, exclude_subdomains=exclude_subdomains, get_only_default_urls=True, ctx=ctx ) # Domain regex host = self.domain.name if self.domain else urlparse(urls[0]).netloc host_regex = f"\'https?://([a-z0-9]+[.])*{host}.*\'" # Tools cmds cmd_map = { 'gau': f'gau', 'gauplus': f'gauplus -random-agent', 'hakrawler': 'hakrawler -subs -u', 'waybackurls': 'waybackurls', 'gospider': f'gospider -S {input_path} --js -d 2 --sitemap --robots -w -r', 'katana': f'katana -list {input_path} -silent -jc -kf all -d 3 -fs rdn', } if proxy: cmd_map['gau'] += f' --proxy "{proxy}"' cmd_map['gauplus'] += f' -p "{proxy}"' cmd_map['gospider'] += f' -p {proxy}' cmd_map['hakrawler'] += f' -proxy {proxy}' cmd_map['katana'] += f' -proxy {proxy}' if threads > 0: cmd_map['gau'] += f' --threads {threads}' cmd_map['gauplus'] += f' -t {threads}' cmd_map['gospider'] += f' -t {threads}' cmd_map['katana'] += f' -c {threads}' if custom_header: header_string = ';;'.join([ f'{key}: {value}' for key, value in custom_header.items() ]) cmd_map['hakrawler'] += f' -h {header_string}' cmd_map['katana'] += f' -H {header_string}' header_flags = [':'.join(h) for h in header_string.split(';;')] for flag in header_flags: cmd_map['gospider'] += f' -H {flag}' cat_input = f'cat {input_path}' grep_output = f'grep -Eo {host_regex}' cmd_map = { tool: f'{cat_input} | {cmd} | {grep_output} > {self.results_dir}/urls_{tool}.txt' for tool, cmd in cmd_map.items() } tasks = group( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for tool, cmd in cmd_map.items() if tool in tools ) # Cleanup task sort_output = [ f'cat {self.results_dir}/urls_* > {self.output_path}', f'cat {input_path} >> {self.output_path}', f'sort -u {self.output_path} -o {self.output_path}', ] if ignore_file_extension: ignore_exts = '|'.join(ignore_file_extension) grep_ext_filtered_output = [ f'cat {self.output_path} | grep -Eiv "\\.({ignore_exts}).*" > {self.results_dir}/urls_filtered.txt', f'mv {self.results_dir}/urls_filtered.txt {self.output_path}' ] sort_output.extend(grep_ext_filtered_output) cleanup = chain( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for cmd in sort_output ) # Run all commands task = chord(tasks)(cleanup) with allow_join_result(): task.get() # Store all the endpoints and run httpx with open(self.output_path) as f: discovered_urls = f.readlines() self.notify(fields={'Discovered URLs': len(discovered_urls)}) # Some tools can have an URL in the format <URL>] - <PATH> or <URL> - <PATH>, add them # to the final URL list all_urls = [] for url in discovered_urls: url = url.strip() urlpath = None base_url = None if '] ' in url: # found JS scraped endpoint e.g from gospider split = tuple(url.split('] ')) if not len(split) == 2: logger.warning(f'URL format not recognized for "{url}". Skipping.') continue base_url, urlpath = split urlpath = urlpath.lstrip('- ') elif ' - ' in url: # found JS scraped endpoint e.g from gospider base_url, urlpath = tuple(url.split(' - ')) if base_url and urlpath: subdomain = urlparse(base_url) url = f'{subdomain.scheme}://{subdomain.netloc}{self.url_filter}' if not validators.url(url): logger.warning(f'Invalid URL "{url}". Skipping.') if url not in all_urls: all_urls.append(url) # Filter out URLs if a path filter was passed if self.url_filter: all_urls = [url for url in all_urls if self.url_filter in url] # Write result to output path with open(self.output_path, 'w') as f: f.write('\n'.join(all_urls)) logger.warning(f'Found {len(all_urls)} usable URLs') # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl( all_urls, ctx=ctx, should_remove_duplicate_endpoints=should_remove_duplicate_endpoints, duplicate_removal_fields=duplicate_removal_fields ) #-------------------# # GF PATTERNS MATCH # #-------------------# # Combine old gf patterns with new ones if gf_patterns: self.scan.used_gf_patterns = ','.join(gf_patterns) self.scan.save() # Run gf patterns on saved endpoints # TODO: refactor to Celery task for gf_pattern in gf_patterns: # TODO: js var is causing issues, removing for now if gf_pattern == 'jsvar': logger.info('Ignoring jsvar as it is causing issues.') continue # Run gf on current pattern logger.warning(f'Running gf on pattern "{gf_pattern}"') gf_output_file = f'{self.results_dir}/gf_patterns_{gf_pattern}.txt' cmd = f'cat {self.output_path} | gf {gf_pattern} | grep -Eo {host_regex} >> {gf_output_file}' run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Check output file if not os.path.exists(gf_output_file): logger.error(f'Could not find GF output file {gf_output_file}. Skipping GF pattern "{gf_pattern}"') continue # Read output file line by line and with open(gf_output_file, 'r') as f: lines = f.readlines() # Add endpoints / subdomains to DB for url in lines: http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue endpoint, created = save_endpoint( http_url, crawl=False, subdomain=subdomain, ctx=ctx) if not endpoint: continue earlier_pattern = None if not created: earlier_pattern = endpoint.matched_gf_patterns pattern = f'{earlier_pattern},{gf_pattern}' if earlier_pattern else gf_pattern endpoint.matched_gf_patterns = pattern endpoint.save() return all_urls def parse_curl_output(response): # TODO: Enrich from other cURL fields. CURL_REGEX_HTTP_STATUS = f'HTTP\/(?:(?:\d\.?)+)\s(\d+)\s(?:\w+)' http_status = 0 if response: failed = False regex = re.compile(CURL_REGEX_HTTP_STATUS, re.MULTILINE) try: http_status = int(regex.findall(response)[0]) except (KeyError, TypeError, IndexError): pass return { 'http_status': http_status, } @app.task(name='vulnerability_scan', queue='main_scan_queue', bind=True, base=RengineTask) def vulnerability_scan(self, urls=[], ctx={}, description=None): """ This function will serve as an entrypoint to vulnerability scan. All other vulnerability scan will be run from here including nuclei, crlfuzz, etc """ logger.info('Running Vulnerability Scan Queue') config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_run_nuclei = config.get(RUN_NUCLEI, True) should_run_crlfuzz = config.get(RUN_CRLFUZZ, False) should_run_dalfox = config.get(RUN_DALFOX, False) should_run_s3scanner = config.get(RUN_S3SCANNER, True) grouped_tasks = [] if should_run_nuclei: _task = nuclei_scan.si( urls=urls, ctx=ctx, description=f'Nuclei Scan' ) grouped_tasks.append(_task) if should_run_crlfuzz: _task = crlfuzz_scan.si( urls=urls, ctx=ctx, description=f'CRLFuzz Scan' ) grouped_tasks.append(_task) if should_run_dalfox: _task = dalfox_xss_scan.si( urls=urls, ctx=ctx, description=f'Dalfox XSS Scan' ) grouped_tasks.append(_task) if should_run_s3scanner: _task = s3scanner.si( ctx=ctx, description=f'Misconfigured S3 Buckets Scanner' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan completed...') # return results return None @app.task(name='nuclei_individual_severity_module', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_individual_severity_module(self, cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx={}, description=None): ''' This celery task will run vulnerability scan in parallel. All severities supplied should run in parallel as grouped tasks. ''' results = [] logger.info(f'Running vulnerability scan with severity: {severity}') cmd += f' -severity {severity}' # Send start notification notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) # Gather nuclei results vuln_data = parse_nuclei_result(line) # Get corresponding subdomain http_url = sanitize_url(line.get('matched-at')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) # Look for duplicate vulnerabilities by excluding records that might change but are irrelevant. object_comparison_exclude = ['response', 'curl_command', 'tags', 'references', 'cve_ids', 'cwe_ids'] # Add subdomain and target domain to the duplicate check vuln_data_copy = vuln_data.copy() vuln_data_copy['subdomain'] = subdomain vuln_data_copy['target_domain'] = self.domain # Check if record exists, if exists do not save it if record_exists(Vulnerability, data=vuln_data_copy, exclude_keys=object_comparison_exclude): logger.warning(f'Nuclei vulnerability of severity {severity} : {vuln_data_copy["name"]} for {subdomain_name} already exists') continue # Get or create EndPoint object response = line.get('response') httpx_crawl = False if response else enable_http_crawl # avoid yet another httpx crawl endpoint, _ = save_endpoint( http_url, crawl=httpx_crawl, subdomain=subdomain, ctx=ctx) if endpoint: http_url = endpoint.http_url if not httpx_crawl: output = parse_curl_output(response) endpoint.http_status = output['http_status'] endpoint.save() # Get or create Vulnerability object vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, subdomain=subdomain, **vuln_data) if not vuln: continue # Print vuln severity = line['info'].get('severity', 'unknown') logger.warning(str(vuln)) # Send notification for all vulnerabilities except info url = vuln.http_url or vuln.subdomain send_vuln = ( notif and notif.send_vuln_notif and vuln and severity in ['low', 'medium', 'high', 'critical']) if send_vuln: fields = { 'Severity': f'**{severity.upper()}**', 'URL': http_url, 'Subdomain': subdomain_name, 'Name': vuln.name, 'Type': vuln.type, 'Description': vuln.description, 'Template': vuln.template_url, 'Tags': vuln.get_tags_str(), 'CVEs': vuln.get_cve_str(), 'CWEs': vuln.get_cwe_str(), 'References': vuln.get_refs_str() } severity_map = { 'low': 'info', 'medium': 'warning', 'high': 'error', 'critical': 'error' } self.notify( f'vulnerability_scan_#{vuln.id}', severity_map[severity], fields, add_meta_info=False) # Send report to hackerone hackerone_query = Hackerone.objects.all() send_report = ( hackerone_query.exists() and severity not in ('info', 'low') and vuln.target_domain.h1_team_handle ) if send_report: hackerone = hackerone_query.first() if hackerone.send_critical and severity == 'critical': send_hackerone_report.delay(vuln.id) elif hackerone.send_high and severity == 'high': send_hackerone_report.delay(vuln.id) elif hackerone.send_medium and severity == 'medium': send_hackerone_report.delay(vuln.id) # Write results to JSON file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) # Send finish notif if send_status: vulns = Vulnerability.objects.filter(scan_history__id=self.scan_id) info_count = vulns.filter(severity=0).count() low_count = vulns.filter(severity=1).count() medium_count = vulns.filter(severity=2).count() high_count = vulns.filter(severity=3).count() critical_count = vulns.filter(severity=4).count() unknown_count = vulns.filter(severity=-1).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count + unknown_count fields = { 'Total': vulnerability_count, 'Critical': critical_count, 'High': high_count, 'Medium': medium_count, 'Low': low_count, 'Info': info_count, 'Unknown': unknown_count } self.notify(fields=fields) # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=NUCLEI ).exclude( severity=0 ) # find all unique vulnerabilities based on path and title # all unique vulnerability will go thru gpt function and get report # once report is got, it will be matched with other vulnerabilities and saved unique_vulns = set() for vuln in vulns: unique_vulns.add((vuln.name, vuln.get_path())) unique_vulns = list(unique_vulns) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in unique_vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return None def get_vulnerability_gpt_report(vuln): title = vuln[0] path = vuln[1] logger.info(f'Getting GPT Report for {title}, PATH: {path}') # check if in db already exists stored = GPTVulnerabilityReport.objects.filter( url_path=path ).filter( title=title ).first() if stored: response = { 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: report = GPTVulnerabilityReportGenerator() vulnerability_description = get_gpt_vuln_input_description( title, path ) response = report.get_vulnerability_description(vulnerability_description) add_gpt_description_db( title, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) for vuln in Vulnerability.objects.filter(name=title, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() def add_gpt_description_db(title, path, description, impact, remediation, references): gpt_report = GPTVulnerabilityReport() gpt_report.url_path = path gpt_report.title = title gpt_report.description = description gpt_report.impact = impact gpt_report.remediation = remediation gpt_report.save() for url in references: ref, created = VulnerabilityReference.objects.get_or_create(url=url) gpt_report.references.add(ref) gpt_report.save() @app.task(name='nuclei_scan', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_scan(self, urls=[], ctx={}, description=None): """HTTP vulnerability scan using Nuclei Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. Notes: Unfurl the urls to keep only domain and path, will be sent to vuln scan and ignore certain file extensions. Thanks: https://github.com/six2dez/reconftw """ # Config config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} input_path = f'{self.results_dir}/input_endpoints_vulnerability_scan.txt' enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) concurrency = config.get(NUCLEI_CONCURRENCY) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) retries = config.get(RETRIES) or self.yaml_configuration.get(RETRIES, DEFAULT_RETRIES) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) custom_header = config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) should_fetch_gpt_report = config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) proxy = get_random_proxy() nuclei_specific_config = config.get('nuclei', {}) use_nuclei_conf = nuclei_specific_config.get(USE_CONFIG, False) severities = nuclei_specific_config.get(NUCLEI_SEVERITY, NUCLEI_DEFAULT_SEVERITIES) tags = nuclei_specific_config.get(NUCLEI_TAGS, []) tags = ','.join(tags) nuclei_templates = nuclei_specific_config.get(NUCLEI_TEMPLATE) custom_nuclei_templates = nuclei_specific_config.get(NUCLEI_CUSTOM_TEMPLATE) # severities_str = ','.join(severities) # Get alive endpoints if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=enable_http_crawl, ignore_files=True, write_filepath=input_path, ctx=ctx ) if intensity == 'normal': # reduce number of endpoints to scan unfurl_filter = f'{self.results_dir}/urls_unfurled.txt' run_command( f"cat {input_path} | unfurl -u format %s://%d%p |uro > {unfurl_filter}", shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {unfurl_filter} -o {unfurl_filter}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) input_path = unfurl_filter # Build templates # logger.info('Updating Nuclei templates ...') run_command( 'nuclei -update-templates', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) templates = [] if not (nuclei_templates or custom_nuclei_templates): templates.append(NUCLEI_DEFAULT_TEMPLATES_PATH) if nuclei_templates: if ALL in nuclei_templates: template = NUCLEI_DEFAULT_TEMPLATES_PATH templates.append(template) else: templates.extend(nuclei_templates) if custom_nuclei_templates: custom_nuclei_template_paths = [f'{str(elem)}.yaml' for elem in custom_nuclei_templates] template = templates.extend(custom_nuclei_template_paths) # Build CMD cmd = 'nuclei -j' cmd += ' -config /root/.config/nuclei/config.yaml' if use_nuclei_conf else '' cmd += f' -irr' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -l {input_path}' cmd += f' -c {str(concurrency)}' if concurrency > 0 else '' cmd += f' -proxy {proxy} ' if proxy else '' cmd += f' -retries {retries}' if retries > 0 else '' cmd += f' -rl {rate_limit}' if rate_limit > 0 else '' # cmd += f' -severity {severities_str}' cmd += f' -timeout {str(timeout)}' if timeout and timeout > 0 else '' cmd += f' -tags {tags}' if tags else '' cmd += f' -silent' for tpl in templates: cmd += f' -t {tpl}' grouped_tasks = [] custom_ctx = ctx for severity in severities: custom_ctx['track'] = True _task = nuclei_individual_severity_module.si( cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx=custom_ctx, description=f'Nuclei Scan with severity {severity}' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan with all severities completed...') return None @app.task(name='dalfox_xss_scan', queue='main_scan_queue', base=RengineTask, bind=True) def dalfox_xss_scan(self, urls=[], ctx={}, description=None): """XSS Scan using dalfox Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) dalfox_config = vuln_config.get(DALFOX) or {} custom_header = dalfox_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() is_waf_evasion = dalfox_config.get(WAF_EVASION, False) blind_xss_server = dalfox_config.get(BLIND_XSS_SERVER) user_agent = dalfox_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) timeout = dalfox_config.get(TIMEOUT) delay = dalfox_config.get(DELAY) threads = dalfox_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_dalfox_xss.txt' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=False, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'dalfox --silence --no-color --no-spinner' cmd += f' --only-poc r ' cmd += f' --ignore-return 302,404,403' cmd += f' --skip-bav' cmd += f' file {input_path}' cmd += f' --proxy {proxy}' if proxy else '' cmd += f' --waf-evasion' if is_waf_evasion else '' cmd += f' -b {blind_xss_server}' if blind_xss_server else '' cmd += f' --delay {delay}' if delay else '' cmd += f' --timeout {timeout}' if timeout else '' cmd += f' --user-agent {user_agent}' if user_agent else '' cmd += f' --header {custom_header}' if custom_header else '' cmd += f' --worker {threads}' if threads else '' cmd += f' --format json' results = [] for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id, trunc_char=',' ): if not isinstance(line, dict): continue results.append(line) vuln_data = parse_dalfox_result(line) http_url = sanitize_url(line.get('data')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Dalfox Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=DALFOX ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='crlfuzz_scan', queue='main_scan_queue', base=RengineTask, bind=True) def crlfuzz_scan(self, urls=[], ctx={}, description=None): """CRLF Fuzzing with CRLFuzz Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) custom_header = vuln_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() user_agent = vuln_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) threads = vuln_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_crlf.txt' output_path = f'{self.results_dir}/{self.filename}' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=True, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'crlfuzz -s' cmd += f' -l {input_path}' cmd += f' -x {proxy}' if proxy else '' cmd += f' --H {custom_header}' if custom_header else '' cmd += f' -o {output_path}' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id ) if not os.path.isfile(output_path): logger.info('No Results from CRLFuzz') return crlfs = [] results = [] with open(output_path, 'r') as file: crlfs = file.readlines() for crlf in crlfs: url = crlf.strip() vuln_data = parse_crlfuzz_result(url) http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting CRLFuzz Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=CRLFUZZ ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='s3scanner', queue='main_scan_queue', base=RengineTask, bind=True) def s3scanner(self, ctx={}, description=None): """Bucket Scanner Args: ctx (dict): Context description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/#{self.scan_id}_subdomain_discovery.txt' vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} s3_config = vuln_config.get(S3SCANNER) or {} threads = s3_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) providers = s3_config.get(PROVIDERS, S3SCANNER_DEFAULT_PROVIDERS) scan_history = ScanHistory.objects.filter(pk=self.scan_id).first() for provider in providers: cmd = f's3scanner -bucket-file {input_path} -enumerate -provider {provider} -threads {threads} -json' for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue if line.get('bucket', {}).get('exists', 0) == 1: result = parse_s3scanner_result(line) s3bucket, created = S3Bucket.objects.get_or_create(**result) scan_history.buckets.add(s3bucket) logger.info(f"s3 bucket added {result['provider']}-{result['name']}-{result['region']}") @app.task(name='http_crawl', queue='main_scan_queue', base=RengineTask, bind=True) def http_crawl( self, urls=[], method=None, recrawl=False, ctx={}, track=True, description=None, is_ran_from_subdomain_scan=False, should_remove_duplicate_endpoints=True, duplicate_removal_fields=[]): """Use httpx to query HTTP URLs for important info like page titles, http status, etc... Args: urls (list, optional): A set of URLs to check. Overrides default behavior which queries all endpoints related to this scan. method (str): HTTP method to use (GET, HEAD, POST, PUT, DELETE). recrawl (bool, optional): If False, filter out URLs that have already been crawled. should_remove_duplicate_endpoints (bool): Whether to remove duplicate endpoints duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates Returns: list: httpx results. """ logger.info('Initiating HTTP Crawl') if is_ran_from_subdomain_scan: logger.info('Running From Subdomain Scan...') cmd = '/go/bin/httpx' cfg = self.yaml_configuration.get(HTTP_CRAWL) or {} custom_header = cfg.get(CUSTOM_HEADER, '') threads = cfg.get(THREADS, DEFAULT_THREADS) follow_redirect = cfg.get(FOLLOW_REDIRECT, True) self.output_path = None input_path = f'{self.results_dir}/httpx_input.txt' history_file = f'{self.results_dir}/commands.txt' if urls: # direct passing URLs to check if self.url_filter: urls = [u for u in urls if self.url_filter in u] with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_uncrawled=not recrawl, write_filepath=input_path, ctx=ctx ) # logger.debug(urls) # If no URLs found, skip it if not urls: return # Re-adjust thread number if few URLs to avoid spinning up a monster to # kill a fly. if len(urls) < threads: threads = len(urls) # Get random proxy proxy = get_random_proxy() # Run command cmd += f' -cl -ct -rt -location -td -websocket -cname -asn -cdn -probe -random-agent' cmd += f' -t {threads}' if threads > 0 else '' cmd += f' --http-proxy {proxy}' if proxy else '' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -json' cmd += f' -u {urls[0]}' if len(urls) == 1 else f' -l {input_path}' cmd += f' -x {method}' if method else '' cmd += f' -silent' if follow_redirect: cmd += ' -fr' results = [] endpoint_ids = [] for line in stream_command( cmd, history_file=history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not line or not isinstance(line, dict): continue logger.debug(line) # No response from endpoint if line.get('failed', False): continue # Parse httpx output host = line.get('host', '') content_length = line.get('content_length', 0) http_status = line.get('status_code') http_url, is_redirect = extract_httpx_url(line) page_title = line.get('title') webserver = line.get('webserver') cdn = line.get('cdn', False) rt = line.get('time') techs = line.get('tech', []) cname = line.get('cname', '') content_type = line.get('content_type', '') response_time = -1 if rt: response_time = float(''.join(ch for ch in rt if not ch.isalpha())) if rt[-2:] == 'ms': response_time = response_time / 1000 # Create Subdomain object in DB subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue # Save default HTTP URL to endpoint object in DB endpoint, created = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain, is_default=is_ran_from_subdomain_scan ) if not endpoint: continue endpoint.http_status = http_status endpoint.page_title = page_title endpoint.content_length = content_length endpoint.webserver = webserver endpoint.response_time = response_time endpoint.content_type = content_type endpoint.save() endpoint_str = f'{http_url} [{http_status}] `{content_length}B` `{webserver}` `{rt}`' logger.warning(endpoint_str) if endpoint and endpoint.is_alive and endpoint.http_status != 403: self.notify( fields={'Alive endpoint': f'• {endpoint_str}'}, add_meta_info=False) # Add endpoint to results line['_cmd'] = cmd line['final_url'] = http_url line['endpoint_id'] = endpoint.id line['endpoint_created'] = created line['is_redirect'] = is_redirect results.append(line) # Add technology objects to DB for technology in techs: tech, _ = Technology.objects.get_or_create(name=technology) endpoint.techs.add(tech) if is_ran_from_subdomain_scan: subdomain.technologies.add(tech) subdomain.save() endpoint.save() techs_str = ', '.join([f'`{tech}`' for tech in techs]) self.notify( fields={'Technologies': techs_str}, add_meta_info=False) # Add IP objects for 'a' records to DB a_records = line.get('a', []) for ip_address in a_records: ip, created = save_ip_address( ip_address, subdomain, subscan=self.subscan, cdn=cdn) ips_str = '• ' + '\n• '.join([f'`{ip}`' for ip in a_records]) self.notify( fields={'IPs': ips_str}, add_meta_info=False) # Add IP object for host in DB if host: ip, created = save_ip_address( host, subdomain, subscan=self.subscan, cdn=cdn) self.notify( fields={'IPs': f'• `{ip.address}`'}, add_meta_info=False) # Save subdomain and endpoint if is_ran_from_subdomain_scan: # save subdomain stuffs subdomain.http_url = http_url subdomain.http_status = http_status subdomain.page_title = page_title subdomain.content_length = content_length subdomain.webserver = webserver subdomain.response_time = response_time subdomain.content_type = content_type subdomain.cname = ','.join(cname) subdomain.is_cdn = cdn if cdn: subdomain.cdn_name = line.get('cdn_name') subdomain.save() endpoint.save() endpoint_ids.append(endpoint.id) if should_remove_duplicate_endpoints: # Remove 'fake' alive endpoints that are just redirects to the same page remove_duplicate_endpoints( self.scan_id, self.domain_id, self.subdomain_id, filter_ids=endpoint_ids ) # Remove input file run_command( f'rm {input_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) return results #---------------------# # Notifications tasks # #---------------------# @app.task(name='send_notif', bind=False, queue='send_notif_queue') def send_notif( message, scan_history_id=None, subscan_id=None, **options): if not 'title' in options: message = enrich_notification(message, scan_history_id, subscan_id) send_discord_message(message, **options) send_slack_message(message) send_telegram_message(message) @app.task(name='send_scan_notif', bind=False, queue='send_scan_notif_queue') def send_scan_notif( scan_history_id, subscan_id=None, engine_id=None, status='RUNNING'): """Send scan status notification. Works for scan or a subscan if subscan_id is passed. Args: scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Get domain, engine, scan_history objects engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() tasks = ScanActivity.objects.filter(scan_of=scan) if scan else 0 # Build notif options url = get_scan_url(scan_history_id, subscan_id) title = get_scan_title(scan_history_id, subscan_id) fields = get_scan_fields(engine, scan, subscan, status, tasks) severity = None msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) if status: severity = STATUS_TO_SEVERITIES.get(status) opts = { 'title': title, 'url': url, 'fields': fields, 'severity': severity } logger.warning(f'Sending notification "{title}" [{severity}]') # Send notification send_notif( msg, scan_history_id, subscan_id, **opts) @app.task(name='send_task_notif', bind=False, queue='send_task_notif_queue') def send_task_notif( task_name, status=None, result=None, output_path=None, traceback=None, scan_history_id=None, engine_id=None, subscan_id=None, severity=None, add_meta_info=True, update_fields={}): """Send task status notification. Args: task_name (str): Task name. status (str, optional): Task status. result (str, optional): Task result. output_path (str, optional): Task output path. traceback (str, optional): Task traceback. scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. severity (str, optional): Severity (will be mapped to notif colors) add_meta_info (bool, optional): Wheter to add scan / subscan info to notif. update_fields (dict, optional): Fields key / value to update. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Build fields url = None fields = {} if add_meta_info: engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() url = get_scan_url(scan_history_id) if status: fields['Status'] = f'**{status}**' if engine: fields['Engine'] = engine.engine_name if scan: fields['Scan ID'] = f'[#{scan.id}]({url})' if subscan: url = get_scan_url(scan_history_id, subscan_id) fields['Subscan ID'] = f'[#{subscan.id}]({url})' title = get_task_title(task_name, scan_history_id, subscan_id) if status: severity = STATUS_TO_SEVERITIES.get(status) msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) # Add fields to update for k, v in update_fields.items(): fields[k] = v # Add traceback to notif if traceback and notif.send_scan_tracebacks: fields['Traceback'] = f'```\n{traceback}\n```' # Add files to notif files = [] attach_file = ( notif.send_scan_output_file and output_path and result and not traceback ) if attach_file: output_title = output_path.split('/')[-1] files = [(output_path, output_title)] # Send notif opts = { 'title': title, 'url': url, 'files': files, 'severity': severity, 'fields': fields, 'fields_append': update_fields.keys() } send_notif( msg, scan_history_id=scan_history_id, subscan_id=subscan_id, **opts) @app.task(name='send_file_to_discord', bind=False, queue='send_file_to_discord_queue') def send_file_to_discord(file_path, title=None): notif = Notification.objects.first() do_send = notif and notif.send_to_discord and notif.discord_hook_url if not do_send: return False webhook = DiscordWebhook( url=notif.discord_hook_url, rate_limit_retry=True, username=title or "reNgine Discord Plugin" ) with open(file_path, "rb") as f: head, tail = os.path.split(file_path) webhook.add_file(file=f.read(), filename=tail) webhook.execute() @app.task(name='send_hackerone_report', bind=False, queue='send_hackerone_report_queue') def send_hackerone_report(vulnerability_id): """Send HackerOne vulnerability report. Args: vulnerability_id (int): Vulnerability id. Returns: int: HTTP response status code. """ vulnerability = Vulnerability.objects.get(id=vulnerability_id) severities = {v: k for k,v in NUCLEI_SEVERITY_MAP.items()} headers = { 'Content-Type': 'application/json', 'Accept': 'application/json' } # can only send vulnerability report if team_handle exists if len(vulnerability.target_domain.h1_team_handle) !=0: hackerone_query = Hackerone.objects.all() if hackerone_query.exists(): hackerone = Hackerone.objects.first() severity_value = severities[vulnerability.severity] tpl = hackerone.report_template # Replace syntax of report template with actual content tpl = tpl.replace('{vulnerability_name}', vulnerability.name) tpl = tpl.replace('{vulnerable_url}', vulnerability.http_url) tpl = tpl.replace('{vulnerability_severity}', severity_value) tpl = tpl.replace('{vulnerability_description}', vulnerability.description if vulnerability.description else '') tpl = tpl.replace('{vulnerability_extracted_results}', vulnerability.extracted_results if vulnerability.extracted_results else '') tpl = tpl.replace('{vulnerability_reference}', vulnerability.reference if vulnerability.reference else '') data = { "data": { "type": "report", "attributes": { "team_handle": vulnerability.target_domain.h1_team_handle, "title": '{} found in {}'.format(vulnerability.name, vulnerability.http_url), "vulnerability_information": tpl, "severity_rating": severity_value, "impact": "More information about the impact and vulnerability can be found here: \n" + vulnerability.reference if vulnerability.reference else "NA", } } } r = requests.post( 'https://api.hackerone.com/v1/hackers/reports', auth=(hackerone.username, hackerone.api_key), json=data, headers=headers ) response = r.json() status_code = r.status_code if status_code == 201: vulnerability.hackerone_report_id = response['data']["id"] vulnerability.open_status = False vulnerability.save() return status_code else: logger.error('No team handle found.') status_code = 111 return status_code #-------------# # Utils tasks # #-------------# @app.task(name='parse_nmap_results', bind=False, queue='parse_nmap_results_queue') def parse_nmap_results(xml_file, output_file=None): """Parse results from nmap output file. Args: xml_file (str): nmap XML report file path. Returns: list: List of vulnerabilities found from nmap results. """ with open(xml_file, encoding='utf8') as f: content = f.read() try: nmap_results = xmltodict.parse(content) # parse XML to dict except Exception as e: logger.exception(e) logger.error(f'Cannot parse {xml_file} to valid JSON. Skipping.') return [] # Write JSON to output file if output_file: with open(output_file, 'w') as f: json.dump(nmap_results, f, indent=4) logger.warning(json.dumps(nmap_results, indent=4)) hosts = ( nmap_results .get('nmaprun', {}) .get('host', {}) ) all_vulns = [] if isinstance(hosts, dict): hosts = [hosts] for host in hosts: # Grab hostname / IP from output hostnames_dict = host.get('hostnames', {}) if hostnames_dict: # Ensure that hostnames['hostname'] is a list for consistency hostnames_list = hostnames_dict['hostname'] if isinstance(hostnames_dict['hostname'], list) else [hostnames_dict['hostname']] # Extract all the @name values from the list of dictionaries hostnames = [entry.get('@name') for entry in hostnames_list] else: hostnames = [host.get('address')['@addr']] # Iterate over each hostname for each port for hostname in hostnames: # Grab ports from output ports = host.get('ports', {}).get('port', []) if isinstance(ports, dict): ports = [ports] for port in ports: url_vulns = [] port_number = port['@portid'] url = sanitize_url(f'{hostname}:{port_number}') logger.info(f'Parsing nmap results for {hostname}:{port_number} ...') if not port_number or not port_number.isdigit(): continue port_protocol = port['@protocol'] scripts = port.get('script', []) if isinstance(scripts, dict): scripts = [scripts] for script in scripts: script_id = script['@id'] script_output = script['@output'] script_output_table = script.get('table', []) logger.debug(f'Ran nmap script "{script_id}" on {port_number}/{port_protocol}:\n{script_output}\n') if script_id == 'vulscan': vulns = parse_nmap_vulscan_output(script_output) url_vulns.extend(vulns) elif script_id == 'vulners': vulns = parse_nmap_vulners_output(script_output) url_vulns.extend(vulns) # elif script_id == 'http-server-header': # TODO: nmap can help find technologies as well using the http-server-header script # regex = r'(\w+)/([\d.]+)\s?(?:\((\w+)\))?' # tech_name, tech_version, tech_os = re.match(regex, test_string).groups() # Technology.objects.get_or_create(...) # elif script_id == 'http_csrf': # vulns = parse_nmap_http_csrf_output(script_output) # url_vulns.extend(vulns) else: logger.warning(f'Script output parsing for script "{script_id}" is not supported yet.') # Add URL to vuln for vuln in url_vulns: # TODO: This should extend to any URL, not just HTTP vuln['http_url'] = url if 'http_path' in vuln: vuln['http_url'] += vuln['http_path'] all_vulns.append(vuln) return all_vulns def parse_nmap_http_csrf_output(script_output): pass def parse_nmap_vulscan_output(script_output): """Parse nmap vulscan script output. Args: script_output (str): Vulscan script output. Returns: list: List of Vulnerability dicts. """ data = {} vulns = [] provider_name = '' # Sort all vulns found by provider so that we can match each provider with # a function that pulls from its API to get more info about the # vulnerability. for line in script_output.splitlines(): if not line: continue if not line.startswith('['): # provider line if "No findings" in line: logger.info(f"No findings: {line}") continue elif ' - ' in line: provider_name, provider_url = tuple(line.split(' - ')) data[provider_name] = {'url': provider_url.rstrip(':'), 'entries': []} continue else: # Log a warning logger.warning(f"Unexpected line format: {line}") continue reg = r'\[(.*)\] (.*)' matches = re.match(reg, line) id, title = matches.groups() entry = {'id': id, 'title': title} data[provider_name]['entries'].append(entry) logger.warning('Vulscan parsed output:') logger.warning(pprint.pformat(data)) for provider_name in data: if provider_name == 'Exploit-DB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'IBM X-Force': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'MITRE CVE': logger.error(f'Provider {provider_name} is not supported YET.') for entry in data[provider_name]['entries']: cve_id = entry['id'] vuln = cve_to_vuln(cve_id) vulns.append(vuln) elif provider_name == 'OSVDB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'OpenVAS (Nessus)': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'SecurityFocus': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'VulDB': logger.error(f'Provider {provider_name} is not supported YET.') pass else: logger.error(f'Provider {provider_name} is not supported.') return vulns def parse_nmap_vulners_output(script_output, url=''): """Parse nmap vulners script output. TODO: Rework this as it's currently matching all CVEs no matter the confidence. Args: script_output (str): Script output. Returns: list: List of found vulnerabilities. """ vulns = [] # Check for CVE in script output CVE_REGEX = re.compile(r'.*(CVE-\d\d\d\d-\d+).*') matches = CVE_REGEX.findall(script_output) matches = list(dict.fromkeys(matches)) for cve_id in matches: # get CVE info vuln = cve_to_vuln(cve_id, vuln_type='nmap-vulners-nse') if vuln: vulns.append(vuln) return vulns def cve_to_vuln(cve_id, vuln_type=''): """Search for a CVE using CVESearch and return Vulnerability data. Args: cve_id (str): CVE ID in the form CVE-* Returns: dict: Vulnerability dict. """ cve_info = CVESearch('https://cve.circl.lu').id(cve_id) if not cve_info: logger.error(f'Could not fetch CVE info for cve {cve_id}. Skipping.') return None vuln_cve_id = cve_info['id'] vuln_name = vuln_cve_id vuln_description = cve_info.get('summary', 'none').replace(vuln_cve_id, '').strip() try: vuln_cvss = float(cve_info.get('cvss', -1)) except (ValueError, TypeError): vuln_cvss = -1 vuln_cwe_id = cve_info.get('cwe', '') exploit_ids = cve_info.get('refmap', {}).get('exploit-db', []) osvdb_ids = cve_info.get('refmap', {}).get('osvdb', []) references = cve_info.get('references', []) capec_objects = cve_info.get('capec', []) # Parse ovals for a better vuln name / type ovals = cve_info.get('oval', []) if ovals: vuln_name = ovals[0]['title'] vuln_type = ovals[0]['family'] # Set vulnerability severity based on CVSS score vuln_severity = 'info' if vuln_cvss < 4: vuln_severity = 'low' elif vuln_cvss < 7: vuln_severity = 'medium' elif vuln_cvss < 9: vuln_severity = 'high' else: vuln_severity = 'critical' # Build console warning message msg = f'{vuln_name} | {vuln_severity.upper()} | {vuln_cve_id} | {vuln_cwe_id} | {vuln_cvss}' for id in osvdb_ids: msg += f'\n\tOSVDB: {id}' for exploit_id in exploit_ids: msg += f'\n\tEXPLOITDB: {exploit_id}' logger.warning(msg) vuln = { 'name': vuln_name, 'type': vuln_type, 'severity': NUCLEI_SEVERITY_MAP[vuln_severity], 'description': vuln_description, 'cvss_score': vuln_cvss, 'references': references, 'cve_ids': [vuln_cve_id], 'cwe_ids': [vuln_cwe_id] } return vuln def parse_s3scanner_result(line): ''' Parses and returns s3Scanner Data ''' bucket = line['bucket'] return { 'name': bucket['name'], 'region': bucket['region'], 'provider': bucket['provider'], 'owner_display_name': bucket['owner_display_name'], 'owner_id': bucket['owner_id'], 'perm_auth_users_read': bucket['perm_auth_users_read'], 'perm_auth_users_write': bucket['perm_auth_users_write'], 'perm_auth_users_read_acl': bucket['perm_auth_users_read_acl'], 'perm_auth_users_write_acl': bucket['perm_auth_users_write_acl'], 'perm_auth_users_full_control': bucket['perm_auth_users_full_control'], 'perm_all_users_read': bucket['perm_all_users_read'], 'perm_all_users_write': bucket['perm_all_users_write'], 'perm_all_users_read_acl': bucket['perm_all_users_read_acl'], 'perm_all_users_write_acl': bucket['perm_all_users_write_acl'], 'perm_all_users_full_control': bucket['perm_all_users_full_control'], 'num_objects': bucket['num_objects'], 'size': bucket['bucket_size'] } def parse_nuclei_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ return { 'name': line['info'].get('name', ''), 'type': line['type'], 'severity': NUCLEI_SEVERITY_MAP[line['info'].get('severity', 'unknown')], 'template': line['template'], 'template_url': line['template-url'], 'template_id': line['template-id'], 'description': line['info'].get('description', ''), 'matcher_name': line.get('matcher-name', ''), 'curl_command': line.get('curl-command'), 'request': line.get('request'), 'response': line.get('response'), 'extracted_results': line.get('extracted-results', []), 'cvss_metrics': line['info'].get('classification', {}).get('cvss-metrics', ''), 'cvss_score': line['info'].get('classification', {}).get('cvss-score'), 'cve_ids': line['info'].get('classification', {}).get('cve_id', []) or [], 'cwe_ids': line['info'].get('classification', {}).get('cwe_id', []) or [], 'references': line['info'].get('reference', []) or [], 'tags': line['info'].get('tags', []), 'source': NUCLEI, } def parse_dalfox_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ description = '' description += f" Evidence: {line.get('evidence')} <br>" if line.get('evidence') else '' description += f" Message: {line.get('message')} <br>" if line.get('message') else '' description += f" Payload: {line.get('message_str')} <br>" if line.get('message_str') else '' description += f" Vulnerable Parameter: {line.get('param')} <br>" if line.get('param') else '' return { 'name': 'XSS (Cross Site Scripting)', 'type': 'XSS', 'severity': DALFOX_SEVERITY_MAP[line.get('severity', 'unknown')], 'description': description, 'source': DALFOX, 'cwe_ids': [line.get('cwe')] } def parse_crlfuzz_result(url): """Parse CRLF results Args: url (str): CRLF Vulnerable URL Returns: dict: Vulnerability data. """ return { 'name': 'CRLF (HTTP Response Splitting)', 'type': 'CRLF', 'severity': 2, 'description': 'A CRLF (HTTP Response Splitting) vulnerability has been discovered.', 'source': CRLFUZZ, } def record_exists(model, data, exclude_keys=[]): """ Check if a record already exists in the database based on the given data. Args: model (django.db.models.Model): The Django model to check against. data (dict): Data dictionary containing fields and values. exclude_keys (list): List of keys to exclude from the lookup. Returns: bool: True if the record exists, False otherwise. """ # Extract the keys that will be used for the lookup lookup_fields = {key: data[key] for key in data if key not in exclude_keys} # Return True if a record exists based on the lookup fields, False otherwise return model.objects.filter(**lookup_fields).exists() @app.task(name='geo_localize', bind=False, queue='geo_localize_queue') def geo_localize(host, ip_id=None): """Uses geoiplookup to find location associated with host. Args: host (str): Hostname. ip_id (int): IpAddress object id. Returns: startScan.models.CountryISO: CountryISO object from DB or None. """ if validators.ipv6(host): logger.info(f'Ipv6 "{host}" is not supported by geoiplookup. Skipping.') return None cmd = f'geoiplookup {host}' _, out = run_command(cmd) if 'IP Address not found' not in out and "can't resolve hostname" not in out: country_iso = out.split(':')[1].strip().split(',')[0] country_name = out.split(':')[1].strip().split(',')[1].strip() geo_object, _ = CountryISO.objects.get_or_create( iso=country_iso, name=country_name ) geo_json = { 'iso': country_iso, 'name': country_name } if ip_id: ip = IpAddress.objects.get(pk=ip_id) ip.geo_iso = geo_object ip.save() return geo_json logger.info(f'Geo IP lookup failed for host "{host}"') return None @app.task(name='query_whois', bind=False, queue='query_whois_queue') def query_whois(ip_domain, force_reload_whois=False): """Query WHOIS information for an IP or a domain name. Args: ip_domain (str): IP address or domain name. save_domain (bool): Whether to save domain or not, default False Returns: dict: WHOIS information. """ if not force_reload_whois and Domain.objects.filter(name=ip_domain).exists() and Domain.objects.get(name=ip_domain).domain_info: domain = Domain.objects.get(name=ip_domain) if not domain.insert_date: domain.insert_date = timezone.now() domain.save() domain_info_db = domain.domain_info domain_info = DottedDict( dnssec=domain_info_db.dnssec, created=domain_info_db.created, updated=domain_info_db.updated, expires=domain_info_db.expires, geolocation_iso=domain_info_db.geolocation_iso, status=[status['name'] for status in DomainWhoisStatusSerializer(domain_info_db.status, many=True).data], whois_server=domain_info_db.whois_server, ns_records=[ns['name'] for ns in NameServersSerializer(domain_info_db.name_servers, many=True).data], registrar_name=domain_info_db.registrar.name, registrar_phone=domain_info_db.registrar.phone, registrar_email=domain_info_db.registrar.email, registrar_url=domain_info_db.registrar.url, registrant_name=domain_info_db.registrant.name, registrant_id=domain_info_db.registrant.id_str, registrant_organization=domain_info_db.registrant.organization, registrant_city=domain_info_db.registrant.city, registrant_state=domain_info_db.registrant.state, registrant_zip_code=domain_info_db.registrant.zip_code, registrant_country=domain_info_db.registrant.country, registrant_phone=domain_info_db.registrant.phone, registrant_fax=domain_info_db.registrant.fax, registrant_email=domain_info_db.registrant.email, registrant_address=domain_info_db.registrant.address, admin_name=domain_info_db.admin.name, admin_id=domain_info_db.admin.id_str, admin_organization=domain_info_db.admin.organization, admin_city=domain_info_db.admin.city, admin_state=domain_info_db.admin.state, admin_zip_code=domain_info_db.admin.zip_code, admin_country=domain_info_db.admin.country, admin_phone=domain_info_db.admin.phone, admin_fax=domain_info_db.admin.fax, admin_email=domain_info_db.admin.email, admin_address=domain_info_db.admin.address, tech_name=domain_info_db.tech.name, tech_id=domain_info_db.tech.id_str, tech_organization=domain_info_db.tech.organization, tech_city=domain_info_db.tech.city, tech_state=domain_info_db.tech.state, tech_zip_code=domain_info_db.tech.zip_code, tech_country=domain_info_db.tech.country, tech_phone=domain_info_db.tech.phone, tech_fax=domain_info_db.tech.fax, tech_email=domain_info_db.tech.email, tech_address=domain_info_db.tech.address, related_tlds=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_tlds, many=True).data], related_domains=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_domains, many=True).data], historical_ips=[ip for ip in HistoricalIPSerializer(domain_info_db.historical_ips, many=True).data], ) if domain_info_db.dns_records: a_records = [] txt_records = [] mx_records = [] dns_records = [{'name': dns['name'], 'type': dns['type']} for dns in DomainDNSRecordSerializer(domain_info_db.dns_records, many=True).data] for dns in dns_records: if dns['type'] == 'a': a_records.append(dns['name']) elif dns['type'] == 'txt': txt_records.append(dns['name']) elif dns['type'] == 'mx': mx_records.append(dns['name']) domain_info.a_records = a_records domain_info.txt_records = txt_records domain_info.mx_records = mx_records else: logger.info(f'Domain info for "{ip_domain}" not found in DB, querying whois') domain_info = DottedDict() # find domain historical ip try: historical_ips = get_domain_historical_ip_address(ip_domain) domain_info.historical_ips = historical_ips except Exception as e: logger.error(f'HistoricalIP for {ip_domain} not found!\nError: {str(e)}') historical_ips = [] # find associated domains using ip_domain try: related_domains = reverse_whois(ip_domain.split('.')[0]) except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] # find related tlds using TLSx try: related_tlds = [] output_path = '/tmp/ip_domain_tlsx.txt' tlsx_command = f'tlsx -san -cn -silent -ro -host {ip_domain} -o {output_path}' run_command( tlsx_command, shell=True, ) tlsx_output = [] with open(output_path) as f: tlsx_output = f.readlines() tldextract_target = tldextract.extract(ip_domain) for doms in tlsx_output: doms = doms.strip() tldextract_res = tldextract.extract(doms) if ip_domain != doms and tldextract_res.domain == tldextract_target.domain and tldextract_res.subdomain == '': related_tlds.append(doms) related_tlds = list(set(related_tlds)) domain_info.related_tlds = related_tlds except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] related_domains_list = [] if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_related = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_related) related_domains_list.append(_domain['name']) for _domain in related_tlds: domain_related = RelatedDomain.objects.get_or_create( name=_domain, )[0] db_domain_info.related_tlds.add(domain_related) for _ip in historical_ips: historical_ip = HistoricalIP.objects.get_or_create( ip=_ip['ip'], owner=_ip['owner'], location=_ip['location'], last_seen=_ip['last_seen'], )[0] db_domain_info.historical_ips.add(historical_ip) domain.domain_info = db_domain_info domain.save() command = f'netlas host {ip_domain} -f json' # check if netlas key is provided netlas_key = get_netlas_key() command += f' -a {netlas_key}' if netlas_key else '' result = subprocess.check_output(command.split()).decode('utf-8') if 'Failed to parse response data' in result: # do fallback return { 'status': False, 'ip_domain': ip_domain, 'result': "Netlas limit exceeded.", 'message': 'Netlas limit exceeded.' } try: result = json.loads(result) logger.info(result) whois = result.get('whois') if result.get('whois') else {} domain_info.created = whois.get('created_date') domain_info.expires = whois.get('expiration_date') domain_info.updated = whois.get('updated_date') domain_info.whois_server = whois.get('whois_server') if 'registrant' in whois: registrant = whois.get('registrant') domain_info.registrant_name = registrant.get('name') domain_info.registrant_country = registrant.get('country') domain_info.registrant_id = registrant.get('id') domain_info.registrant_state = registrant.get('province') domain_info.registrant_city = registrant.get('city') domain_info.registrant_phone = registrant.get('phone') domain_info.registrant_address = registrant.get('street') domain_info.registrant_organization = registrant.get('organization') domain_info.registrant_fax = registrant.get('fax') domain_info.registrant_zip_code = registrant.get('postal_code') email_search = EMAIL_REGEX.search(str(registrant.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.registrant_email = field_content if 'administrative' in whois: administrative = whois.get('administrative') domain_info.admin_name = administrative.get('name') domain_info.admin_country = administrative.get('country') domain_info.admin_id = administrative.get('id') domain_info.admin_state = administrative.get('province') domain_info.admin_city = administrative.get('city') domain_info.admin_phone = administrative.get('phone') domain_info.admin_address = administrative.get('street') domain_info.admin_organization = administrative.get('organization') domain_info.admin_fax = administrative.get('fax') domain_info.admin_zip_code = administrative.get('postal_code') mail_search = EMAIL_REGEX.search(str(administrative.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.admin_email = field_content if 'technical' in whois: technical = whois.get('technical') domain_info.tech_name = technical.get('name') domain_info.tech_country = technical.get('country') domain_info.tech_state = technical.get('province') domain_info.tech_id = technical.get('id') domain_info.tech_city = technical.get('city') domain_info.tech_phone = technical.get('phone') domain_info.tech_address = technical.get('street') domain_info.tech_organization = technical.get('organization') domain_info.tech_fax = technical.get('fax') domain_info.tech_zip_code = technical.get('postal_code') mail_search = EMAIL_REGEX.search(str(technical.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.tech_email = field_content if 'dns' in result: dns = result.get('dns') domain_info.mx_records = dns.get('mx') domain_info.txt_records = dns.get('txt') domain_info.a_records = dns.get('a') domain_info.ns_records = whois.get('name_servers') domain_info.dnssec = True if whois.get('dnssec') else False domain_info.status = whois.get('status') if 'registrar' in whois: registrar = whois.get('registrar') domain_info.registrar_name = registrar.get('name') domain_info.registrar_email = registrar.get('email') domain_info.registrar_phone = registrar.get('phone') domain_info.registrar_url = registrar.get('url') # find associated domains if registrant email is found related_domains = reverse_whois(domain_info.get('registrant_email')) if domain_info.get('registrant_email') else [] for _domain in related_domains: related_domains_list.append(_domain['name']) # remove duplicate domains from related domains list related_domains_list = list(set(related_domains_list)) domain_info.related_domains = related_domains_list # save to db if domain exists if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_rel = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_rel) db_domain_info.dnssec = domain_info.get('dnssec') #dates db_domain_info.created = domain_info.get('created') db_domain_info.updated = domain_info.get('updated') db_domain_info.expires = domain_info.get('expires') #registrar db_domain_info.registrar = Registrar.objects.get_or_create( name=domain_info.get('registrar_name'), email=domain_info.get('registrar_email'), phone=domain_info.get('registrar_phone'), url=domain_info.get('registrar_url'), )[0] db_domain_info.registrant = DomainRegistration.objects.get_or_create( name=domain_info.get('registrant_name'), organization=domain_info.get('registrant_organization'), address=domain_info.get('registrant_address'), city=domain_info.get('registrant_city'), state=domain_info.get('registrant_state'), zip_code=domain_info.get('registrant_zip_code'), country=domain_info.get('registrant_country'), email=domain_info.get('registrant_email'), phone=domain_info.get('registrant_phone'), fax=domain_info.get('registrant_fax'), id_str=domain_info.get('registrant_id'), )[0] db_domain_info.admin = DomainRegistration.objects.get_or_create( name=domain_info.get('admin_name'), organization=domain_info.get('admin_organization'), address=domain_info.get('admin_address'), city=domain_info.get('admin_city'), state=domain_info.get('admin_state'), zip_code=domain_info.get('admin_zip_code'), country=domain_info.get('admin_country'), email=domain_info.get('admin_email'), phone=domain_info.get('admin_phone'), fax=domain_info.get('admin_fax'), id_str=domain_info.get('admin_id'), )[0] db_domain_info.tech = DomainRegistration.objects.get_or_create( name=domain_info.get('tech_name'), organization=domain_info.get('tech_organization'), address=domain_info.get('tech_address'), city=domain_info.get('tech_city'), state=domain_info.get('tech_state'), zip_code=domain_info.get('tech_zip_code'), country=domain_info.get('tech_country'), email=domain_info.get('tech_email'), phone=domain_info.get('tech_phone'), fax=domain_info.get('tech_fax'), id_str=domain_info.get('tech_id'), )[0] for status in domain_info.get('status') or []: _status = WhoisStatus.objects.get_or_create( name=status )[0] _status.save() db_domain_info.status.add(_status) for ns in domain_info.get('ns_records') or []: _ns = NameServer.objects.get_or_create( name=ns )[0] _ns.save() db_domain_info.name_servers.add(_ns) for a in domain_info.get('a_records') or []: _a = DNSRecord.objects.get_or_create( name=a, type='a' )[0] _a.save() db_domain_info.dns_records.add(_a) for mx in domain_info.get('mx_records') or []: _mx = DNSRecord.objects.get_or_create( name=mx, type='mx' )[0] _mx.save() db_domain_info.dns_records.add(_mx) for txt in domain_info.get('txt_records') or []: _txt = DNSRecord.objects.get_or_create( name=txt, type='txt' )[0] _txt.save() db_domain_info.dns_records.add(_txt) db_domain_info.geolocation_iso = domain_info.get('registrant_country') db_domain_info.whois_server = domain_info.get('whois_server') db_domain_info.save() domain.domain_info = db_domain_info domain.save() except Exception as e: return { 'status': False, 'ip_domain': ip_domain, 'result': "unable to fetch records from WHOIS database.", 'message': str(e) } return { 'status': True, 'ip_domain': ip_domain, 'dnssec': domain_info.get('dnssec'), 'created': domain_info.get('created'), 'updated': domain_info.get('updated'), 'expires': domain_info.get('expires'), 'geolocation_iso': domain_info.get('registrant_country'), 'domain_statuses': domain_info.get('status'), 'whois_server': domain_info.get('whois_server'), 'dns': { 'a': domain_info.get('a_records'), 'mx': domain_info.get('mx_records'), 'txt': domain_info.get('txt_records'), }, 'registrar': { 'name': domain_info.get('registrar_name'), 'phone': domain_info.get('registrar_phone'), 'email': domain_info.get('registrar_email'), 'url': domain_info.get('registrar_url'), }, 'registrant': { 'name': domain_info.get('registrant_name'), 'id': domain_info.get('registrant_id'), 'organization': domain_info.get('registrant_organization'), 'address': domain_info.get('registrant_address'), 'city': domain_info.get('registrant_city'), 'state': domain_info.get('registrant_state'), 'zipcode': domain_info.get('registrant_zip_code'), 'country': domain_info.get('registrant_country'), 'phone': domain_info.get('registrant_phone'), 'fax': domain_info.get('registrant_fax'), 'email': domain_info.get('registrant_email'), }, 'admin': { 'name': domain_info.get('admin_name'), 'id': domain_info.get('admin_id'), 'organization': domain_info.get('admin_organization'), 'address':domain_info.get('admin_address'), 'city': domain_info.get('admin_city'), 'state': domain_info.get('admin_state'), 'zipcode': domain_info.get('admin_zip_code'), 'country': domain_info.get('admin_country'), 'phone': domain_info.get('admin_phone'), 'fax': domain_info.get('admin_fax'), 'email': domain_info.get('admin_email'), }, 'technical_contact': { 'name': domain_info.get('tech_name'), 'id': domain_info.get('tech_id'), 'organization': domain_info.get('tech_organization'), 'address': domain_info.get('tech_address'), 'city': domain_info.get('tech_city'), 'state': domain_info.get('tech_state'), 'zipcode': domain_info.get('tech_zip_code'), 'country': domain_info.get('tech_country'), 'phone': domain_info.get('tech_phone'), 'fax': domain_info.get('tech_fax'), 'email': domain_info.get('tech_email'), }, 'nameservers': domain_info.get('ns_records'), # 'similar_domains': domain_info.get('similar_domains'), 'related_domains': domain_info.get('related_domains'), 'related_tlds': domain_info.get('related_tlds'), 'historical_ips': domain_info.get('historical_ips'), } @app.task(name='remove_duplicate_endpoints', bind=False, queue='remove_duplicate_endpoints_queue') def remove_duplicate_endpoints( scan_history_id, domain_id, subdomain_id=None, filter_ids=[], filter_status=[200, 301, 404], duplicate_removal_fields=ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS ): """Remove duplicate endpoints. Check for implicit redirections by comparing endpoints: - [x] `content_length` similarities indicating redirections - [x] `page_title` (check for same page title) - [ ] Sign-in / login page (check for endpoints with the same words) Args: scan_history_id: ScanHistory id. domain_id (int): Domain id. subdomain_id (int, optional): Subdomain id. filter_ids (list): List of endpoint ids to filter on. filter_status (list): List of HTTP status codes to filter on. duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates """ logger.info(f'Removing duplicate endpoints based on {duplicate_removal_fields}') endpoints = ( EndPoint.objects .filter(scan_history__id=scan_history_id) .filter(target_domain__id=domain_id) ) if filter_status: endpoints = endpoints.filter(http_status__in=filter_status) if subdomain_id: endpoints = endpoints.filter(subdomain__id=subdomain_id) if filter_ids: endpoints = endpoints.filter(id__in=filter_ids) for field_name in duplicate_removal_fields: cl_query = ( endpoints .values_list(field_name) .annotate(mc=Count(field_name)) .order_by('-mc') ) for (field_value, count) in cl_query: if count > DELETE_DUPLICATES_THRESHOLD: eps_to_delete = ( endpoints .filter(**{field_name: field_value}) .order_by('discovered_date') .all()[1:] ) msg = f'Deleting {len(eps_to_delete)} endpoints [reason: same {field_name} {field_value}]' for ep in eps_to_delete: url = urlparse(ep.http_url) if url.path in ['', '/', '/login']: # try do not delete the original page that other pages redirect to continue msg += f'\n\t {ep.http_url} [{ep.http_status}] [{field_name}={field_value}]' ep.delete() logger.warning(msg) @app.task(name='run_command', bind=False, queue='run_command_queue') def run_command(cmd, cwd=None, shell=False, history_file=None, scan_id=None, activity_id=None): """Run a given command using subprocess module. Args: cmd (str): Command to run. cwd (str): Current working directory. echo (bool): Log command. shell (bool): Run within separate shell if True. history_file (str): Write command + output to history file. Returns: tuple: Tuple with return_code, output. """ logger.info(cmd) logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Run the command using subprocess popen = subprocess.Popen( cmd if shell else cmd.split(), shell=shell, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, cwd=cwd, universal_newlines=True) output = '' for stdout_line in iter(popen.stdout.readline, ""): item = stdout_line.strip() output += '\n' + item logger.debug(item) popen.stdout.close() popen.wait() return_code = popen.returncode command_obj.output = output command_obj.return_code = return_code command_obj.save() if history_file: mode = 'a' if not os.path.exists(history_file): mode = 'w' with open(history_file, mode) as f: f.write(f'\n{cmd}\n{return_code}\n{output}\n------------------\n') return return_code, output #-------------# # Other utils # #-------------# def stream_command(cmd, cwd=None, shell=False, history_file=None, encoding='utf-8', scan_id=None, activity_id=None, trunc_char=None): # Log cmd logger.info(cmd) # logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Sanitize the cmd command = cmd if shell else cmd.split() # Run the command using subprocess process = subprocess.Popen( command, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, universal_newlines=True, shell=shell) # Log the output in real-time to the database output = "" # Process the output for line in iter(lambda: process.stdout.readline(), b''): if not line: break line = line.strip() ansi_escape = re.compile(r'\x1B(?:[@-Z\\-_]|\[[0-?]*[ -/]*[@-~])') line = ansi_escape.sub('', line) line = line.replace('\\x0d\\x0a', '\n') if trunc_char and line.endswith(trunc_char): line = line[:-1] item = line # Try to parse the line as JSON try: item = json.loads(line) except json.JSONDecodeError: pass # Yield the line #logger.debug(item) yield item # Add the log line to the output output += line + "\n" # Update the command record in the database command_obj.output = output command_obj.save() # Retrieve the return code and output process.wait() return_code = process.returncode # Update the return code and final output in the database command_obj.return_code = return_code command_obj.save() # Append the command, return code and output to the history file if history_file is not None: with open(history_file, "a") as f: f.write(f"{cmd}\n{return_code}\n{output}\n") def process_httpx_response(line): """TODO: implement this""" def extract_httpx_url(line): """Extract final URL from httpx results. Always follow redirects to find the last URL. Args: line (dict): URL data output by httpx. Returns: tuple: (final_url, redirect_bool) tuple. """ status_code = line.get('status_code', 0) final_url = line.get('final_url') location = line.get('location') chain_status_codes = line.get('chain_status_codes', []) # Final URL is already looking nice, if it exists return it if final_url: return final_url, False http_url = line['url'] # fallback to url field # Handle redirects manually REDIRECT_STATUS_CODES = [301, 302] is_redirect = ( status_code in REDIRECT_STATUS_CODES or any(x in REDIRECT_STATUS_CODES for x in chain_status_codes) ) if is_redirect and location: if location.startswith(('http', 'https')): http_url = location else: http_url = f'{http_url}/{location.lstrip("/")}' # Sanitize URL http_url = sanitize_url(http_url) return http_url, is_redirect #-------------# # OSInt utils # #-------------# def get_and_save_dork_results(lookup_target, results_dir, type, lookup_keywords=None, lookup_extensions=None, delay=3, page_count=2, scan_history=None): """ Uses gofuzz to dork and store information Args: lookup_target (str): target to look into such as stackoverflow or even the target itself results_dir (str): Results directory type (str): Dork Type Title lookup_keywords (str): comma separated keywords or paths to look for lookup_extensions (str): comma separated extensions to look for delay (int): delay between each requests page_count (int): pages in google to extract information scan_history (startScan.ScanHistory): Scan History Object """ results = [] gofuzz_command = f'{GOFUZZ_EXEC_PATH} -t {lookup_target} -d {delay} -p {page_count}' if lookup_extensions: gofuzz_command += f' -e {lookup_extensions}' elif lookup_keywords: gofuzz_command += f' -w {lookup_keywords}' output_file = f'{results_dir}/gofuzz.txt' gofuzz_command += f' -o {output_file}' history_file = f'{results_dir}/commands.txt' try: run_command( gofuzz_command, shell=False, history_file=history_file, scan_id=scan_history.id, ) if not os.path.isfile(output_file): return with open(output_file) as f: for line in f.readlines(): url = line.strip() if url: results.append(url) dork, created = Dork.objects.get_or_create( type=type, url=url ) if scan_history: scan_history.dorks.add(dork) # remove output file os.remove(output_file) except Exception as e: logger.exception(e) return results def get_and_save_emails(scan_history, activity_id, results_dir): """Get and save emails from Google, Bing and Baidu. Args: scan_history (startScan.ScanHistory): Scan history object. activity_id: ScanActivity Object results_dir (str): Results directory. Returns: list: List of emails found. """ emails = [] # Proxy settings # get_random_proxy() # Gather emails from Google, Bing and Baidu output_file = f'{results_dir}/emails_tmp.txt' history_file = f'{results_dir}/commands.txt' command = f'python3 /usr/src/github/Infoga/infoga.py --domain {scan_history.domain.name} --source all --report {output_file}' try: run_command( command, shell=False, history_file=history_file, scan_id=scan_history.id, activity_id=activity_id) if not os.path.isfile(output_file): logger.info('No Email results') return [] with open(output_file) as f: for line in f.readlines(): if 'Email' in line: split_email = line.split(' ')[2] emails.append(split_email) output_path = f'{results_dir}/emails.txt' with open(output_path, 'w') as output_file: for email_address in emails: save_email(email_address, scan_history) output_file.write(f'{email_address}\n') except Exception as e: logger.exception(e) return emails def save_metadata_info(meta_dict): """Extract metadata from Google Search. Args: meta_dict (dict): Info dict. Returns: list: List of startScan.MetaFinderDocument objects. """ logger.warning(f'Getting metadata for {meta_dict.osint_target}') scan_history = ScanHistory.objects.get(id=meta_dict.scan_id) # Proxy settings get_random_proxy() # Get metadata result = extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if not result: logger.error(f'No metadata result from Google Search for {meta_dict.osint_target}.') return [] # Add metadata info to DB results = [] for metadata_name, data in result.get_metadata().items(): subdomain = Subdomain.objects.get( scan_history=meta_dict.scan_id, name=meta_dict.osint_target) metadata = DottedDict({k: v for k, v in data.items()}) meta_finder_document = MetaFinderDocument( subdomain=subdomain, target_domain=meta_dict.domain, scan_history=scan_history, url=metadata.url, doc_name=metadata_name, http_status=metadata.status_code, producer=metadata.metadata.get('Producer'), creator=metadata.metadata.get('Creator'), creation_date=metadata.metadata.get('CreationDate'), modified_date=metadata.metadata.get('ModDate'), author=metadata.metadata.get('Author'), title=metadata.metadata.get('Title'), os=metadata.metadata.get('OSInfo')) meta_finder_document.save() results.append(data) return results #-----------------# # Utils functions # #-----------------# def create_scan_activity(scan_history_id, message, status): scan_activity = ScanActivity() scan_activity.scan_of = ScanHistory.objects.get(pk=scan_history_id) scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id #--------------------# # Database functions # #--------------------# def save_vulnerability(**vuln_data): references = vuln_data.pop('references', []) cve_ids = vuln_data.pop('cve_ids', []) cwe_ids = vuln_data.pop('cwe_ids', []) tags = vuln_data.pop('tags', []) subscan = vuln_data.pop('subscan', None) # remove nulls vuln_data = replace_nulls(vuln_data) # Create vulnerability vuln, created = Vulnerability.objects.get_or_create(**vuln_data) if created: vuln.discovered_date = timezone.now() vuln.open_status = True vuln.save() # Save vuln tags for tag_name in tags or []: tag, created = VulnerabilityTags.objects.get_or_create(name=tag_name) if tag: vuln.tags.add(tag) vuln.save() # Save CVEs for cve_id in cve_ids or []: cve, created = CveId.objects.get_or_create(name=cve_id) if cve: vuln.cve_ids.add(cve) vuln.save() # Save CWEs for cve_id in cwe_ids or []: cwe, created = CweId.objects.get_or_create(name=cve_id) if cwe: vuln.cwe_ids.add(cwe) vuln.save() # Save vuln reference for url in references or []: ref, created = VulnerabilityReference.objects.get_or_create(url=url) if created: vuln.references.add(ref) vuln.save() # Save subscan id in vuln object if subscan: vuln.vuln_subscan_ids.add(subscan) vuln.save() return vuln, created def save_endpoint( http_url, ctx={}, crawl=False, is_default=False, **endpoint_data): """Get or create EndPoint object. If crawl is True, also crawl the endpoint HTTP URL with httpx. Args: http_url (str): Input HTTP URL. is_default (bool): If the url is a default url for SubDomains. scan_history (startScan.models.ScanHistory): ScanHistory object. domain (startScan.models.Domain): Domain object. subdomain (starScan.models.Subdomain): Subdomain object. results_dir (str, optional): Results directory. crawl (bool, optional): Run httpx on endpoint if True. Default: False. force (bool, optional): Force crawl even if ENABLE_HTTP_CRAWL mode is on. subscan (startScan.models.SubScan, optional): SubScan object. Returns: tuple: (startScan.models.EndPoint, created) where `created` is a boolean indicating if the object is new or already existed. """ # remove nulls endpoint_data = replace_nulls(endpoint_data) scheme = urlparse(http_url).scheme endpoint = None created = False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in http_url: logger.error(f"{http_url} is not a URL of domain {domain.name}. Skipping.") return None, False if crawl: ctx['track'] = False results = http_crawl( urls=[http_url], method='HEAD', ctx=ctx) if results: endpoint_data = results[0] endpoint_id = endpoint_data['endpoint_id'] created = endpoint_data['endpoint_created'] endpoint = EndPoint.objects.get(pk=endpoint_id) elif not scheme: return None, False else: # add dumb endpoint without probing it scan = ScanHistory.objects.filter(pk=ctx.get('scan_history_id')).first() domain = Domain.objects.filter(pk=ctx.get('domain_id')).first() if not validators.url(http_url): return None, False http_url = sanitize_url(http_url) endpoint, created = EndPoint.objects.get_or_create( scan_history=scan, target_domain=domain, http_url=http_url, **endpoint_data) if created: endpoint.is_default = is_default endpoint.discovered_date = timezone.now() endpoint.save() subscan_id = ctx.get('subscan_id') if subscan_id: endpoint.endpoint_subscan_ids.add(subscan_id) endpoint.save() return endpoint, created def save_subdomain(subdomain_name, ctx={}): """Get or create Subdomain object. Args: subdomain_name (str): Subdomain name. scan_history (startScan.models.ScanHistory): ScanHistory object. Returns: tuple: (startScan.models.Subdomain, created) where `created` is a boolean indicating if the object has been created in DB. """ scan_id = ctx.get('scan_history_id') subscan_id = ctx.get('subscan_id') out_of_scope_subdomains = ctx.get('out_of_scope_subdomains', []) valid_domain = ( validators.domain(subdomain_name) or validators.ipv4(subdomain_name) or validators.ipv6(subdomain_name) ) if not valid_domain: logger.error(f'{subdomain_name} is not an invalid domain. Skipping.') return None, False if subdomain_name in out_of_scope_subdomains: logger.error(f'{subdomain_name} is out-of-scope. Skipping.') return None, False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in subdomain_name: logger.error(f"{subdomain_name} is not a subdomain of domain {domain.name}. Skipping.") return None, False scan = ScanHistory.objects.filter(pk=scan_id).first() domain = scan.domain if scan else None subdomain, created = Subdomain.objects.get_or_create( scan_history=scan, target_domain=domain, name=subdomain_name) if created: # logger.warning(f'Found new subdomain {subdomain_name}') subdomain.discovered_date = timezone.now() if subscan_id: subdomain.subdomain_subscan_ids.add(subscan_id) subdomain.save() return subdomain, created def save_email(email_address, scan_history=None): if not validators.email(email_address): logger.info(f'Email {email_address} is invalid. Skipping.') return None, False email, created = Email.objects.get_or_create(address=email_address) # if created: # logger.warning(f'Found new email address {email_address}') # Add email to ScanHistory if scan_history: scan_history.emails.add(email) scan_history.save() return email, created def save_employee(name, designation, scan_history=None): employee, created = Employee.objects.get_or_create( name=name, designation=designation) # if created: # logger.warning(f'Found new employee {name}') # Add employee to ScanHistory if scan_history: scan_history.employees.add(employee) scan_history.save() return employee, created def save_ip_address(ip_address, subdomain=None, subscan=None, **kwargs): if not (validators.ipv4(ip_address) or validators.ipv6(ip_address)): logger.info(f'IP {ip_address} is not a valid IP. Skipping.') return None, False ip, created = IpAddress.objects.get_or_create(address=ip_address) # if created: # logger.warning(f'Found new IP {ip_address}') # Set extra attributes for key, value in kwargs.items(): setattr(ip, key, value) ip.save() # Add IP to subdomain if subdomain: subdomain.ip_addresses.add(ip) subdomain.save() # Add subscan to IP if subscan: ip.ip_subscan_ids.add(subscan) # Geo-localize IP asynchronously if created: geo_localize.delay(ip_address, ip.id) return ip, created def save_imported_subdomains(subdomains, ctx={}): """Take a list of subdomains imported and write them to from_imported.txt. Args: subdomains (list): List of subdomain names. scan_history (startScan.models.ScanHistory): ScanHistory instance. domain (startScan.models.Domain): Domain instance. results_dir (str): Results directory. """ domain_id = ctx['domain_id'] domain = Domain.objects.get(pk=domain_id) results_dir = ctx.get('results_dir', RENGINE_RESULTS) # Validate each subdomain and de-duplicate entries subdomains = list(set([ subdomain for subdomain in subdomains if validators.domain(subdomain) and domain.name == get_domain_from_subdomain(subdomain) ])) if not subdomains: return logger.warning(f'Found {len(subdomains)} imported subdomains.') with open(f'{results_dir}/from_imported.txt', 'w+') as output_file: for name in subdomains: subdomain_name = name.strip() subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain.is_imported_subdomain = True subdomain.save() output_file.write(f'{subdomain}\n') @app.task(name='query_reverse_whois', bind=False, queue='query_reverse_whois_queue') def query_reverse_whois(lookup_keyword): """Queries Reverse WHOIS information for an organization or email address. Args: lookup_keyword (str): Registrar Name or email Returns: dict: Reverse WHOIS information. """ return get_associated_domains(lookup_keyword) @app.task(name='query_ip_history', bind=False, queue='query_ip_history_queue') def query_ip_history(domain): """Queries the IP history for a domain Args: domain (str): domain_name Returns: list: list of historical ip addresses """ return get_domain_historical_ip_address(domain) @app.task(name='gpt_vulnerability_description', bind=False, queue='gpt_queue') def gpt_vulnerability_description(vulnerability_id): """Generate and store Vulnerability Description using GPT. Args: vulnerability_id (Vulnerability Model ID): Vulnerability ID to fetch Description. """ logger.info('Getting GPT Vulnerability Description') try: lookup_vulnerability = Vulnerability.objects.get(id=vulnerability_id) lookup_url = urlparse(lookup_vulnerability.http_url) path = lookup_url.path except Exception as e: return { 'status': False, 'error': str(e) } # check in db GPTVulnerabilityReport model if vulnerability description and path matches stored = GPTVulnerabilityReport.objects.filter(url_path=path).filter(title=lookup_vulnerability.name).first() if stored: response = { 'status': True, 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: vulnerability_description = get_gpt_vuln_input_description( lookup_vulnerability.name, path ) # one can add more description here later gpt_generator = GPTVulnerabilityReportGenerator() response = gpt_generator.get_vulnerability_description(vulnerability_description) add_gpt_description_db( lookup_vulnerability.name, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) # for all vulnerabilities with the same vulnerability name this description has to be stored. # also the consition is that the url must contain a part of this. for vuln in Vulnerability.objects.filter(name=lookup_vulnerability.name, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() return response

import csv import json import os import pprint import subprocess import time import validators import whatportis import xmltodict import yaml import tldextract import concurrent.futures from datetime import datetime from urllib.parse import urlparse from api.serializers import SubdomainSerializer from celery import chain, chord, group from celery.result import allow_join_result from celery.utils.log import get_task_logger from django.db.models import Count from dotted_dict import DottedDict from django.utils import timezone from pycvesearch import CVESearch from metafinder.extractor import extract_metadata_from_google_search from reNgine.celery import app from reNgine.gpt import GPTVulnerabilityReportGenerator from reNgine.celery_custom_task import RengineTask from reNgine.common_func import * from reNgine.definitions import * from reNgine.settings import * from reNgine.gpt import * from reNgine.utilities import * from scanEngine.models import (EngineType, InstalledExternalTool, Notification, Proxy) from startScan.models import * from startScan.models import EndPoint, Subdomain, Vulnerability from targetApp.models import Domain """ Celery tasks. """ logger = get_task_logger(__name__) #----------------------# # Scan / Subscan tasks # #----------------------# @app.task(name='initiate_scan', bind=False, queue='initiate_scan_queue') def initiate_scan( scan_history_id, domain_id, engine_id=None, scan_type=LIVE_SCAN, results_dir=RENGINE_RESULTS, imported_subdomains=[], out_of_scope_subdomains=[], url_filter=''): """Initiate a new scan. Args: scan_history_id (int): ScanHistory id. domain_id (int): Domain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. imported_subdomains (list): Imported subdomains. out_of_scope_subdomains (list): Out-of-scope subdomains. url_filter (str): URL path. Default: '' """ # Get scan history scan = ScanHistory.objects.get(pk=scan_history_id) # Get scan engine engine_id = engine_id or scan.scan_type.id # scan history engine_id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, []) # Get domain and set last_scan_date domain = Domain.objects.get(pk=domain_id) domain.last_scan_date = timezone.now() domain.save() # Get path filter url_filter = url_filter.rstrip('/') # Get or create ScanHistory() object if scan_type == LIVE_SCAN: # immediate scan = ScanHistory.objects.get(pk=scan_history_id) scan.scan_status = RUNNING_TASK elif scan_type == SCHEDULED_SCAN: # scheduled scan = ScanHistory() scan.scan_status = INITIATED_TASK scan.scan_type = engine scan.celery_ids = [initiate_scan.request.id] scan.domain = domain scan.start_scan_date = timezone.now() scan.tasks = engine.tasks scan.results_dir = f'{results_dir}/{domain.name}_{scan.id}' add_gf_patterns = gf_patterns and 'fetch_url' in engine.tasks if add_gf_patterns: scan.used_gf_patterns = ','.join(gf_patterns) scan.save() # Create scan results dir os.makedirs(scan.results_dir) # Build task context ctx = { 'scan_history_id': scan_history_id, 'engine_id': engine_id, 'domain_id': domain.id, 'results_dir': scan.results_dir, 'url_filter': url_filter, 'yaml_configuration': config, 'out_of_scope_subdomains': out_of_scope_subdomains } ctx_str = json.dumps(ctx, indent=2) # Send start notif logger.warning(f'Starting scan {scan_history_id} with context:\n{ctx_str}') send_scan_notif.delay( scan_history_id, subscan_id=None, engine_id=engine_id, status=CELERY_TASK_STATUS_MAP[scan.scan_status]) # Save imported subdomains in DB save_imported_subdomains(imported_subdomains, ctx=ctx) # Create initial subdomain in DB: make a copy of domain as a subdomain so # that other tasks using subdomains can use it. subdomain_name = domain.name subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) # If enable_http_crawl is set, create an initial root HTTP endpoint so that # HTTP crawling can start somewhere http_url = f'{domain.name}{url_filter}' if url_filter else domain.name endpoint, _ = save_endpoint( http_url, ctx=ctx, crawl=enable_http_crawl, is_default=True, subdomain=subdomain ) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build Celery tasks, crafted according to the dependency graph below: # subdomain_discovery --> port_scan --> fetch_url --> dir_file_fuzz # osint vulnerability_scan # osint dalfox xss scan # screenshot # waf_detection workflow = chain( group( subdomain_discovery.si(ctx=ctx, description='Subdomain discovery'), osint.si(ctx=ctx, description='OS Intelligence') ), port_scan.si(ctx=ctx, description='Port scan'), fetch_url.si(ctx=ctx, description='Fetch URL'), group( dir_file_fuzz.si(ctx=ctx, description='Directories & files fuzz'), vulnerability_scan.si(ctx=ctx, description='Vulnerability scan'), screenshot.si(ctx=ctx, description='Screenshot'), waf_detection.si(ctx=ctx, description='WAF detection') ) ) # Build callback callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery chord logger.info(f'Running Celery workflow with {len(workflow.tasks) + 1} tasks') task = chain(workflow, callback).on_error(callback).delay() scan.celery_ids.append(task.id) scan.save() return { 'success': True, 'task_id': task.id } @app.task(name='initiate_subscan', bind=False, queue='subscan_queue') def initiate_subscan( scan_history_id, subdomain_id, engine_id=None, scan_type=None, results_dir=RENGINE_RESULTS, url_filter=''): """Initiate a new subscan. Args: scan_history_id (int): ScanHistory id. subdomain_id (int): Subdomain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. url_filter (str): URL path. Default: '' """ # Get Subdomain, Domain and ScanHistory subdomain = Subdomain.objects.get(pk=subdomain_id) scan = ScanHistory.objects.get(pk=subdomain.scan_history.id) domain = Domain.objects.get(pk=subdomain.target_domain.id) # Get EngineType engine_id = engine_id or scan.scan_type.id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Create scan activity of SubScan Model subscan = SubScan( start_scan_date=timezone.now(), celery_ids=[initiate_subscan.request.id], scan_history=scan, subdomain=subdomain, type=scan_type, status=RUNNING_TASK, engine=engine) subscan.save() # Get YAML configuration config = yaml.safe_load(engine.yaml_configuration) # Create results directory results_dir = f'{scan.results_dir}/subscans/{subscan.id}' os.makedirs(results_dir, exist_ok=True) # Run task method = globals().get(scan_type) if not method: logger.warning(f'Task {scan_type} is not supported by reNgine. Skipping') return scan.tasks.append(scan_type) scan.save() # Send start notif send_scan_notif.delay( scan.id, subscan_id=subscan.id, engine_id=engine_id, status='RUNNING') # Build context ctx = { 'scan_history_id': scan.id, 'subscan_id': subscan.id, 'engine_id': engine_id, 'domain_id': domain.id, 'subdomain_id': subdomain.id, 'yaml_configuration': config, 'results_dir': results_dir, 'url_filter': url_filter } # Create initial endpoints in DB: find domain HTTP endpoint so that HTTP # crawling can start somewhere base_url = f'{subdomain.name}{url_filter}' if url_filter else subdomain.name endpoint, _ = save_endpoint( base_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build header + callback workflow = method.si(ctx=ctx) callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery tasks task = chain(workflow, callback).on_error(callback).delay() subscan.celery_ids.append(task.id) subscan.save() return { 'success': True, 'task_id': task.id } @app.task(name='report', bind=False, queue='report_queue') def report(ctx={}, description=None): """Report task running after all other tasks. Mark ScanHistory or SubScan object as completed and update with final status, log run details and send notification. Args: description (str, optional): Task description shown in UI. """ # Get objects subscan_id = ctx.get('subscan_id') scan_id = ctx.get('scan_history_id') engine_id = ctx.get('engine_id') scan = ScanHistory.objects.filter(pk=scan_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() # Get failed tasks tasks = ScanActivity.objects.filter(scan_of=scan).all() if subscan: tasks = tasks.filter(celery_id__in=subscan.celery_ids) failed_tasks = tasks.filter(status=FAILED_TASK) # Get task status failed_count = failed_tasks.count() status = SUCCESS_TASK if failed_count == 0 else FAILED_TASK status_h = 'SUCCESS' if failed_count == 0 else 'FAILED' # Update scan / subscan status if subscan: subscan.stop_scan_date = timezone.now() subscan.status = status subscan.save() else: scan.scan_status = status scan.stop_scan_date = timezone.now() scan.save() # Send scan status notif send_scan_notif.delay( scan_history_id=scan_id, subscan_id=subscan_id, engine_id=engine_id, status=status_h) #------------------------- # # Tracked reNgine tasks # #--------------------------# @app.task(name='subdomain_discovery', queue='main_scan_queue', base=RengineTask, bind=True) def subdomain_discovery( self, host=None, ctx=None, description=None): """Uses a set of tools (see SUBDOMAIN_SCAN_DEFAULT_TOOLS) to scan all subdomains associated with a domain. Args: host (str): Hostname to scan. Returns: subdomains (list): List of subdomain names. """ if not host: host = self.subdomain.name if self.subdomain else self.domain.name if self.url_filter: logger.warning(f'Ignoring subdomains scan as an URL path filter was passed ({self.url_filter}).') return # Config config = self.yaml_configuration.get(SUBDOMAIN_DISCOVERY) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL) or self.yaml_configuration.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) tools = config.get(USES_TOOLS, SUBDOMAIN_SCAN_DEFAULT_TOOLS) default_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=True).filter(is_subdomain_gathering=True)] custom_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=False).filter(is_subdomain_gathering=True)] send_subdomain_changes, send_interesting = False, False notif = Notification.objects.first() if notif: send_subdomain_changes = notif.send_subdomain_changes_notif send_interesting = notif.send_interesting_notif # Gather tools to run for subdomain scan if ALL in tools: tools = SUBDOMAIN_SCAN_DEFAULT_TOOLS + custom_subdomain_tools tools = [t.lower() for t in tools] # Make exception for amass since tool name is amass, but command is amass-active/passive default_subdomain_tools.append('amass-passive') default_subdomain_tools.append('amass-active') # Run tools for tool in tools: cmd = None logger.info(f'Scanning subdomains for {host} with {tool}') proxy = get_random_proxy() if tool in default_subdomain_tools: if tool == 'amass-passive': cmd = f'amass enum -passive -d {host} -o {self.results_dir}/subdomains_amass.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' elif tool == 'amass-active': use_amass_config = config.get(USE_AMASS_CONFIG, False) amass_wordlist_name = config.get(AMASS_WORDLIST, 'deepmagic.com-prefixes-top50000') wordlist_path = f'/usr/src/wordlist/{amass_wordlist_name}.txt' cmd = f'amass enum -active -d {host} -o {self.results_dir}/subdomains_amass_active.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' cmd += f' -brute -w {wordlist_path}' elif tool == 'sublist3r': cmd = f'python3 /usr/src/github/Sublist3r/sublist3r.py -d {host} -t {threads} -o {self.results_dir}/subdomains_sublister.txt' elif tool == 'subfinder': cmd = f'subfinder -d {host} -o {self.results_dir}/subdomains_subfinder.txt' use_subfinder_config = config.get(USE_SUBFINDER_CONFIG, False) cmd += ' -config /root/.config/subfinder/config.yaml' if use_subfinder_config else '' cmd += f' -proxy {proxy}' if proxy else '' cmd += f' -timeout {timeout}' if timeout else '' cmd += f' -t {threads}' if threads else '' cmd += f' -silent' elif tool == 'oneforall': cmd = f'python3 /usr/src/github/OneForAll/oneforall.py --target {host} run' cmd_extract = f'cut -d\',\' -f6 /usr/src/github/OneForAll/results/{host}.csv > {self.results_dir}/subdomains_oneforall.txt' cmd_rm = f'rm -rf /usr/src/github/OneForAll/results/{host}.csv' cmd += f' && {cmd_extract} && {cmd_rm}' elif tool == 'ctfr': results_file = self.results_dir + '/subdomains_ctfr.txt' cmd = f'python3 /usr/src/github/ctfr/ctfr.py -d {host} -o {results_file}' cmd_extract = f"cat {results_file} | sed 's/\*.//g' | tail -n +12 | uniq | sort > {results_file}" cmd += f' && {cmd_extract}' elif tool == 'tlsx': results_file = self.results_dir + '/subdomains_tlsx.txt' cmd = f'tlsx -san -cn -silent -ro -host {host}' cmd += f" | sed -n '/^\([a-zA-Z0-9]\([-a-zA-Z0-9]*[a-zA-Z0-9]\)\?\.\)\+{host}$/p' | uniq | sort" cmd += f' > {results_file}' elif tool == 'netlas': results_file = self.results_dir + '/subdomains_netlas.txt' cmd = f'netlas search -d domain -i domain domain:"*.{host}" -f json' netlas_key = get_netlas_key() cmd += f' -a {netlas_key}' if netlas_key else '' cmd_extract = f"grep -oE '([a-zA-Z0-9]([-a-zA-Z0-9]*[a-zA-Z0-9])?\.)+{host}'" cmd += f' | {cmd_extract} > {results_file}' elif tool in custom_subdomain_tools: tool_query = InstalledExternalTool.objects.filter(name__icontains=tool.lower()) if not tool_query.exists(): logger.error(f'Missing {{TARGET}} and {{OUTPUT}} placeholders in {tool} configuration. Skipping.') continue custom_tool = tool_query.first() cmd = custom_tool.subdomain_gathering_command if '{TARGET}' in cmd and '{OUTPUT}' in cmd: cmd = cmd.replace('{TARGET}', host) cmd = cmd.replace('{OUTPUT}', f'{self.results_dir}/subdomains_{tool}.txt') cmd = cmd.replace('{PATH}', custom_tool.github_clone_path) if '{PATH}' in cmd else cmd else: logger.warning( f'Subdomain discovery tool "{tool}" is not supported by reNgine. Skipping.') continue # Run tool try: run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) except Exception as e: logger.error( f'Subdomain discovery tool "{tool}" raised an exception') logger.exception(e) # Gather all the tools' results in one single file. Write subdomains into # separate files, and sort all subdomains. run_command( f'cat {self.results_dir}/subdomains_*.txt > {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {self.output_path} -o {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) with open(self.output_path) as f: lines = f.readlines() # Parse the output_file file and store Subdomain and EndPoint objects found # in db. subdomain_count = 0 subdomains = [] urls = [] for line in lines: subdomain_name = line.strip() valid_url = bool(validators.url(subdomain_name)) valid_domain = ( bool(validators.domain(subdomain_name)) or bool(validators.ipv4(subdomain_name)) or bool(validators.ipv6(subdomain_name)) or valid_url ) if not valid_domain: logger.error(f'Subdomain {subdomain_name} is not a valid domain, IP or URL. Skipping.') continue if valid_url: subdomain_name = urlparse(subdomain_name).netloc if subdomain_name in self.out_of_scope_subdomains: logger.error(f'Subdomain {subdomain_name} is out of scope. Skipping.') continue # Add subdomain subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain_count += 1 subdomains.append(subdomain) urls.append(subdomain.name) # Bulk crawl subdomains if enable_http_crawl: ctx['track'] = True http_crawl(urls, ctx=ctx, is_ran_from_subdomain_scan=True) # Find root subdomain endpoints for subdomain in subdomains: pass # Send notifications subdomains_str = '\n'.join([f'• `{subdomain.name}`' for subdomain in subdomains]) self.notify(fields={ 'Subdomain count': len(subdomains), 'Subdomains': subdomains_str, }) if send_subdomain_changes and self.scan_id and self.domain_id: added = get_new_added_subdomain(self.scan_id, self.domain_id) removed = get_removed_subdomain(self.scan_id, self.domain_id) if added: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in added]) self.notify(fields={'Added subdomains': subdomains_str}) if removed: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in removed]) self.notify(fields={'Removed subdomains': subdomains_str}) if send_interesting and self.scan_id and self.domain_id: interesting_subdomains = get_interesting_subdomains(self.scan_id, self.domain_id) if interesting_subdomains: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in interesting_subdomains]) self.notify(fields={'Interesting subdomains': subdomains_str}) return SubdomainSerializer(subdomains, many=True).data @app.task(name='osint', queue='main_scan_queue', base=RengineTask, bind=True) def osint(self, host=None, ctx={}, description=None): """Run Open-Source Intelligence tools on selected domain. Args: host (str): Hostname to scan. Returns: dict: Results from osint discovery and dorking. """ config = self.yaml_configuration.get(OSINT) or OSINT_DEFAULT_CONFIG results = {} grouped_tasks = [] if 'discover' in config: ctx['track'] = False # results = osint_discovery(host=host, ctx=ctx) _task = osint_discovery.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, activity_id=self.activity_id, results_dir=self.results_dir, ctx=ctx ) grouped_tasks.append(_task) if OSINT_DORK in config or OSINT_CUSTOM_DORK in config: _task = dorking.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, results_dir=self.results_dir ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('OSINT Tasks finished...') # with open(self.output_path, 'w') as f: # json.dump(results, f, indent=4) # # return results @app.task(name='osint_discovery', queue='osint_discovery_queue', bind=False) def osint_discovery(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run OSINT discovery. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: dict: osint metadat and theHarvester and h8mail results. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) osint_lookup = config.get(OSINT_DISCOVER, []) osint_intensity = config.get(INTENSITY, 'normal') documents_limit = config.get(OSINT_DOCUMENTS_LIMIT, 50) results = {} meta_info = [] emails = [] creds = [] # Get and save meta info if 'metainfo' in osint_lookup: if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': host, 'domain': host, 'scan_id': scan_history_id, 'documents_limit': documents_limit }) meta_info.append(save_metadata_info(meta_dict)) # TODO: disabled for now # elif osint_intensity == 'deep': # subdomains = Subdomain.objects # if self.scan: # subdomains = subdomains.filter(scan_history=self.scan) # for subdomain in subdomains: # meta_dict = DottedDict({ # 'osint_target': subdomain.name, # 'domain': self.domain, # 'scan_id': self.scan_id, # 'documents_limit': documents_limit # }) # meta_info.append(save_metadata_info(meta_dict)) grouped_tasks = [] if 'emails' in osint_lookup: emails = get_and_save_emails(scan_history, activity_id, results_dir) emails_str = '\n'.join([f'• `{email}`' for email in emails]) # self.notify(fields={'Emails': emails_str}) # ctx['track'] = False _task = h8mail.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) if 'employees' in osint_lookup: ctx['track'] = False _task = theHarvester.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) # results['emails'] = results.get('emails', []) + emails # results['creds'] = creds # results['meta_info'] = meta_info return results @app.task(name='dorking', bind=False, queue='dorking_queue') def dorking(config, host, scan_history_id, results_dir): """Run Google dorks. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: list: Dorking results for each dork ran. """ # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh scan_history = ScanHistory.objects.get(pk=scan_history_id) dorks = config.get(OSINT_DORK, []) custom_dorks = config.get(OSINT_CUSTOM_DORK, []) results = [] # custom dorking has higher priority try: for custom_dork in custom_dorks: lookup_target = custom_dork.get('lookup_site') # replace with original host if _target_ lookup_target = host if lookup_target == '_target_' else lookup_target if 'lookup_extensions' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_extensions=custom_dork.get('lookup_extensions'), scan_history=scan_history ) elif 'lookup_keywords' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_keywords=custom_dork.get('lookup_keywords'), scan_history=scan_history ) except Exception as e: logger.exception(e) # default dorking try: for dork in dorks: logger.info(f'Getting dork information for {dork}') if dork == 'stackoverflow': results = get_and_save_dork_results( lookup_target='stackoverflow.com', results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'login_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/login/,login.html', page_count=5, scan_history=scan_history ) elif dork == 'admin_panels': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/admin/,admin.html', page_count=5, scan_history=scan_history ) elif dork == 'dashboard_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/dashboard/,dashboard.html', page_count=5, scan_history=scan_history ) elif dork == 'social_media' : social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'reddit.com' ] for site in social_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'project_management' : project_websites = [ 'trello.com', 'atlassian.net' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'code_sharing' : project_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'config_files' : config_file_exts = [ 'env', 'xml', 'conf', 'toml', 'yml', 'yaml', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(config_file_exts), page_count=4, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keyword = 'Jenkins' results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=lookup_keyword, page_count=1, scan_history=scan_history ) elif dork == 'wordpress_files' : lookup_keywords = [ '/wp-content/', '/wp-includes/' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'php_error' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'exposed_documents' : docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(docs_file_ext), page_count=7, scan_history=scan_history ) elif dork == 'db_files' : file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) elif dork == 'git_exposed' : file_ext = [ 'git', ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) except Exception as e: logger.exception(e) return results @app.task(name='theHarvester', queue='theHarvester_queue', bind=False) def theHarvester(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run theHarvester to get save emails, hosts, employees found in domain. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: dict: Dict of emails, employees, hosts and ips found during crawling. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) output_path_json = f'{results_dir}/theHarvester.json' theHarvester_dir = '/usr/src/github/theHarvester' history_file = f'{results_dir}/commands.txt' cmd = f'python3 {theHarvester_dir}/theHarvester.py -d {host} -b all -f {output_path_json}' # Update proxies.yaml proxy_query = Proxy.objects.all() if proxy_query.exists(): proxy = proxy_query.first() if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(f'{theHarvester_dir}/proxies.yaml', 'w') as file: yaml.dump(yaml_data, file) # Run cmd run_command( cmd, shell=False, cwd=theHarvester_dir, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) # Get file location if not os.path.isfile(output_path_json): logger.error(f'Could not open {output_path_json}') return {} # Load theHarvester results with open(output_path_json, 'r') as f: data = json.load(f) # Re-indent theHarvester JSON with open(output_path_json, 'w') as f: json.dump(data, f, indent=4) emails = data.get('emails', []) for email_address in emails: email, _ = save_email(email_address, scan_history=scan_history) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) linkedin_people = data.get('linkedin_people', []) for people in linkedin_people: employee, _ = save_employee( people, designation='linkedin', scan_history=scan_history) # if employee: # self.notify(fields={'LinkedIn people': f'• {employee.name}'}) twitter_people = data.get('twitter_people', []) for people in twitter_people: employee, _ = save_employee( people, designation='twitter', scan_history=scan_history) # if employee: # self.notify(fields={'Twitter people': f'• {employee.name}'}) hosts = data.get('hosts', []) urls = [] for host in hosts: split = tuple(host.split(':')) http_url = split[0] subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) endpoint, _ = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain) # if endpoint: # urls.append(endpoint.http_url) # self.notify(fields={'Hosts': f'• {endpoint.http_url}'}) # if enable_http_crawl: # ctx['track'] = False # http_crawl(urls, ctx=ctx) # TODO: Lots of ips unrelated with our domain are found, disabling # this for now. # ips = data.get('ips', []) # for ip_address in ips: # ip, created = save_ip_address( # ip_address, # subscan=subscan) # if ip: # send_task_notif.delay( # 'osint', # scan_history_id=scan_history_id, # subscan_id=subscan_id, # severity='success', # update_fields={'IPs': f'{ip.address}'}) return data @app.task(name='h8mail', queue='h8mail_queue', bind=False) def h8mail(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run h8mail. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: list[dict]: List of credentials info. """ logger.warning('Getting leaked credentials') scan_history = ScanHistory.objects.get(pk=scan_history_id) input_path = f'{results_dir}/emails.txt' output_file = f'{results_dir}/h8mail.json' cmd = f'h8mail -t {input_path} --json {output_file}' history_file = f'{results_dir}/commands.txt' run_command( cmd, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) with open(output_file) as f: data = json.load(f) creds = data.get('targets', []) # TODO: go through h8mail output and save emails to DB for cred in creds: logger.warning(cred) email_address = cred['target'] pwn_num = cred['pwn_num'] pwn_data = cred.get('data', []) email, created = save_email(email_address, scan_history=scan) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) return creds @app.task(name='screenshot', queue='main_scan_queue', base=RengineTask, bind=True) def screenshot(self, ctx={}, description=None): """Uses EyeWitness to gather screenshot of a domain and/or url. Args: description (str, optional): Task description shown in UI. """ # Config screenshots_path = f'{self.results_dir}/screenshots' output_path = f'{self.results_dir}/screenshots/{self.filename}' alive_endpoints_file = f'{self.results_dir}/endpoints_alive.txt' config = self.yaml_configuration.get(SCREENSHOT) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT + 5) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) # If intensity is normal, grab only the root endpoints of each subdomain strict = True if intensity == 'normal' else False # Get URLs to take screenshot of get_http_urls( is_alive=enable_http_crawl, strict=strict, write_filepath=alive_endpoints_file, get_only_default_urls=True, ctx=ctx ) # Send start notif notification = Notification.objects.first() send_output_file = notification.send_scan_output_file if notification else False # Run cmd cmd = f'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py -f {alive_endpoints_file} -d {screenshots_path} --no-prompt' cmd += f' --timeout {timeout}' if timeout > 0 else '' cmd += f' --threads {threads}' if threads > 0 else '' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(output_path): logger.error(f'Could not load EyeWitness results at {output_path} for {self.domain.name}.') return # Loop through results and save objects in DB screenshot_paths = [] with open(output_path, 'r') as file: reader = csv.reader(file) for row in reader: "Protocol,Port,Domain,Request Status,Screenshot Path, Source Path" protocol, port, subdomain_name, status, screenshot_path, source_path = tuple(row) logger.info(f'{protocol}:{port}:{subdomain_name}:{status}') subdomain_query = Subdomain.objects.filter(name=subdomain_name) if self.scan: subdomain_query = subdomain_query.filter(scan_history=self.scan) if status == 'Successful' and subdomain_query.exists(): subdomain = subdomain_query.first() screenshot_paths.append(screenshot_path) subdomain.screenshot_path = screenshot_path.replace('/usr/src/scan_results/', '') subdomain.save() logger.warning(f'Added screenshot for {subdomain.name} to DB') # Remove all db, html extra files in screenshot results run_command( 'rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format(screenshots_path), shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'rm -rf {screenshots_path}/source', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Send finish notifs screenshots_str = '• ' + '\n• '.join([f'`{path}`' for path in screenshot_paths]) self.notify(fields={'Screenshots': screenshots_str}) if send_output_file: for path in screenshot_paths: title = get_output_file_name( self.scan_id, self.subscan_id, self.filename) send_file_to_discord.delay(path, title) @app.task(name='port_scan', queue='main_scan_queue', base=RengineTask, bind=True) def port_scan(self, hosts=[], ctx={}, description=None): """Run port scan. Args: hosts (list, optional): Hosts to run port scan on. description (str, optional): Task description shown in UI. Returns: list: List of open ports (dict). """ input_file = f'{self.results_dir}/input_subdomains_port_scan.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(PORT_SCAN) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) exclude_ports = config.get(NAABU_EXCLUDE_PORTS, []) exclude_subdomains = config.get(NAABU_EXCLUDE_SUBDOMAINS, False) ports = config.get(PORTS, NAABU_DEFAULT_PORTS) ports = [str(port) for port in ports] rate_limit = config.get(NAABU_RATE) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) passive = config.get(NAABU_PASSIVE, False) use_naabu_config = config.get(USE_NAABU_CONFIG, False) exclude_ports_str = ','.join(return_iterable(exclude_ports)) # nmap args nmap_enabled = config.get(ENABLE_NMAP, False) nmap_cmd = config.get(NMAP_COMMAND, '') nmap_script = config.get(NMAP_SCRIPT, '') nmap_script = ','.join(return_iterable(nmap_script)) nmap_script_args = config.get(NMAP_SCRIPT_ARGS) if hosts: with open(input_file, 'w') as f: f.write('\n'.join(hosts)) else: hosts = get_subdomains( write_filepath=input_file, exclude_subdomains=exclude_subdomains, ctx=ctx) # Build cmd cmd = 'naabu -json -exclude-cdn' cmd += f' -list {input_file}' if len(hosts) > 0 else f' -host {hosts[0]}' if 'full' in ports or 'all' in ports: ports_str = ' -p "-"' elif 'top-100' in ports: ports_str = ' -top-ports 100' elif 'top-1000' in ports: ports_str = ' -top-ports 1000' else: ports_str = ','.join(ports) ports_str = f' -p {ports_str}' cmd += ports_str cmd += ' -config /root/.config/naabu/config.yaml' if use_naabu_config else '' cmd += f' -proxy "{proxy}"' if proxy else '' cmd += f' -c {threads}' if threads else '' cmd += f' -rate {rate_limit}' if rate_limit > 0 else '' cmd += f' -timeout {timeout*1000}' if timeout > 0 else '' cmd += f' -passive' if passive else '' cmd += f' -exclude-ports {exclude_ports_str}' if exclude_ports else '' cmd += f' -silent' # Execute cmd and gather results results = [] urls = [] ports_data = {} for line in stream_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) port_number = line['port'] ip_address = line['ip'] host = line.get('host') or ip_address if port_number == 0: continue # Grab subdomain subdomain = Subdomain.objects.filter( name=host, target_domain=self.domain, scan_history=self.scan ).first() # Add IP DB ip, _ = save_ip_address(ip_address, subdomain, subscan=self.subscan) if self.subscan: ip.ip_subscan_ids.add(self.subscan) ip.save() # Add endpoint to DB # port 80 and 443 not needed as http crawl already does that. if port_number not in [80, 443]: http_url = f'{host}:{port_number}' endpoint, _ = save_endpoint( http_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint: http_url = endpoint.http_url urls.append(http_url) # Add Port in DB port_details = whatportis.get_ports(str(port_number)) service_name = port_details[0].name if len(port_details) > 0 else 'unknown' description = port_details[0].description if len(port_details) > 0 else '' # get or create port port, created = Port.objects.get_or_create( number=port_number, service_name=service_name, description=description ) if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port.save() ip.ports.add(port) ip.save() if host in ports_data: ports_data[host].append(port_number) else: ports_data[host] = [port_number] # Send notification logger.warning(f'Found opened port {port_number} on {ip_address} ({host})') if len(ports_data) == 0: logger.info('Finished running naabu port scan - No open ports found.') if nmap_enabled: logger.info('Nmap scans skipped') return ports_data # Send notification fields_str = '' for host, ports in ports_data.items(): ports_str = ', '.join([f'`{port}`' for port in ports]) fields_str += f'• `{host}`: {ports_str}\n' self.notify(fields={'Ports discovered': fields_str}) # Save output to file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) logger.info('Finished running naabu port scan.') # Process nmap results: 1 process per host sigs = [] if nmap_enabled: logger.warning(f'Starting nmap scans ...') logger.warning(ports_data) for host, port_list in ports_data.items(): ports_str = '_'.join([str(p) for p in port_list]) ctx_nmap = ctx.copy() ctx_nmap['description'] = get_task_title(f'nmap_{host}', self.scan_id, self.subscan_id) ctx_nmap['track'] = False sig = nmap.si( cmd=nmap_cmd, ports=port_list, host=host, script=nmap_script, script_args=nmap_script_args, max_rate=rate_limit, ctx=ctx_nmap) sigs.append(sig) task = group(sigs).apply_async() with allow_join_result(): results = task.get() return ports_data @app.task(name='nmap', queue='main_scan_queue', base=RengineTask, bind=True) def nmap( self, cmd=None, ports=[], host=None, input_file=None, script=None, script_args=None, max_rate=None, ctx={}, description=None): """Run nmap on a host. Args: cmd (str, optional): Existing nmap command to complete. ports (list, optional): List of ports to scan. host (str, optional): Host to scan. input_file (str, optional): Input hosts file. script (str, optional): NSE script to run. script_args (str, optional): NSE script args. max_rate (int): Max rate. description (str, optional): Task description shown in UI. """ notif = Notification.objects.first() ports_str = ','.join(str(port) for port in ports) self.filename = self.filename.replace('.txt', '.xml') filename_vulns = self.filename.replace('.xml', '_vulns.json') output_file = self.output_path output_file_xml = f'{self.results_dir}/{host}_{self.filename}' vulns_file = f'{self.results_dir}/{host}_{filename_vulns}' logger.warning(f'Running nmap on {host}:{ports}') # Build cmd nmap_cmd = get_nmap_cmd( cmd=cmd, ports=ports_str, script=script, script_args=script_args, max_rate=max_rate, host=host, input_file=input_file, output_file=output_file_xml) # Run cmd run_command( nmap_cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Get nmap XML results and convert to JSON vulns = parse_nmap_results(output_file_xml, output_file) with open(vulns_file, 'w') as f: json.dump(vulns, f, indent=4) # Save vulnerabilities found by nmap vulns_str = '' for vuln_data in vulns: # URL is not necessarily an HTTP URL when running nmap (can be any # other vulnerable protocols). Look for existing endpoint and use its # URL as vulnerability.http_url if it exists. url = vuln_data['http_url'] endpoint = EndPoint.objects.filter(http_url__contains=url).first() if endpoint: vuln_data['http_url'] = endpoint.http_url vuln, created = save_vulnerability( target_domain=self.domain, subdomain=self.subdomain, scan_history=self.scan, subscan=self.subscan, endpoint=endpoint, **vuln_data) vulns_str += f'• {str(vuln)}\n' if created: logger.warning(str(vuln)) # Send only 1 notif for all vulns to reduce number of notifs if notif and notif.send_vuln_notif and vulns_str: logger.warning(vulns_str) self.notify(fields={'CVEs': vulns_str}) return vulns @app.task(name='waf_detection', queue='main_scan_queue', base=RengineTask, bind=True) def waf_detection(self, ctx={}, description=None): """ Uses wafw00f to check for the presence of a WAF. Args: description (str, optional): Task description shown in UI. Returns: list: List of startScan.models.Waf objects. """ input_path = f'{self.results_dir}/input_endpoints_waf_detection.txt' config = self.yaml_configuration.get(WAF_DETECTION) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Get alive endpoints from DB get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) cmd = f'wafw00f -i {input_path} -o {self.output_path}' run_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(self.output_path): logger.error(f'Could not find {self.output_path}') return with open(self.output_path) as file: wafs = file.readlines() for line in wafs: line = " ".join(line.split()) splitted = line.split(' ', 1) waf_info = splitted[1].strip() waf_name = waf_info[:waf_info.find('(')].strip() waf_manufacturer = waf_info[waf_info.find('(')+1:waf_info.find(')')].strip().replace('.', '') http_url = sanitize_url(splitted[0].strip()) if not waf_name or waf_name == 'None': continue # Add waf to db waf, _ = Waf.objects.get_or_create( name=waf_name, manufacturer=waf_manufacturer ) # Add waf info to Subdomain in DB subdomain = get_subdomain_from_url(http_url) logger.info(f'Wafw00f Subdomain : {subdomain}') subdomain_query, _ = Subdomain.objects.get_or_create(scan_history=self.scan, name=subdomain) subdomain_query.waf.add(waf) subdomain_query.save() return wafs @app.task(name='dir_file_fuzz', queue='main_scan_queue', base=RengineTask, bind=True) def dir_file_fuzz(self, ctx={}, description=None): """Perform directory scan, and currently uses `ffuf` as a default tool. Args: description (str, optional): Task description shown in UI. Returns: list: List of URLs discovered. """ # Config cmd = 'ffuf' config = self.yaml_configuration.get(DIR_FILE_FUZZ) or {} custom_header = self.yaml_configuration.get(CUSTOM_HEADER) auto_calibration = config.get(AUTO_CALIBRATION, True) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) extensions = config.get(EXTENSIONS, DEFAULT_DIR_FILE_FUZZ_EXTENSIONS) # prepend . on extensions extensions = [ext if ext.startswith('.') else '.' + ext for ext in extensions] extensions_str = ','.join(map(str, extensions)) follow_redirect = config.get(FOLLOW_REDIRECT, FFUF_DEFAULT_FOLLOW_REDIRECT) max_time = config.get(MAX_TIME, 0) match_http_status = config.get(MATCH_HTTP_STATUS, FFUF_DEFAULT_MATCH_HTTP_STATUS) mc = ','.join([str(c) for c in match_http_status]) recursive_level = config.get(RECURSIVE_LEVEL, FFUF_DEFAULT_RECURSIVE_LEVEL) stop_on_error = config.get(STOP_ON_ERROR, False) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) wordlist_name = config.get(WORDLIST, 'dicc') delay = rate_limit / (threads * 100) # calculate request pause delay from rate_limit and number of threads input_path = f'{self.results_dir}/input_dir_file_fuzz.txt' # Get wordlist wordlist_name = 'dicc' if wordlist_name == 'default' else wordlist_name wordlist_path = f'/usr/src/wordlist/{wordlist_name}.txt' # Build command cmd += f' -w {wordlist_path}' cmd += f' -e {extensions_str}' if extensions else '' cmd += f' -maxtime {max_time}' if max_time > 0 else '' cmd += f' -p {delay}' if delay > 0 else '' cmd += f' -recursion -recursion-depth {recursive_level} ' if recursive_level > 0 else '' cmd += f' -t {threads}' if threads and threads > 0 else '' cmd += f' -timeout {timeout}' if timeout and timeout > 0 else '' cmd += ' -se' if stop_on_error else '' cmd += ' -fr' if follow_redirect else '' cmd += ' -ac' if auto_calibration else '' cmd += f' -mc {mc}' if mc else '' cmd += f' -H "{custom_header}"' if custom_header else '' # Grab URLs to fuzz urls = get_http_urls( is_alive=True, ignore_files=False, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) logger.warning(urls) # Loop through URLs and run command results = [] for url in urls: ''' Above while fetching urls, we are not ignoring files, because some default urls may redirect to https://example.com/login.php so, ignore_files is set to False but, during fuzzing, we will only need part of the path, in above example it is still a good idea to ffuf base url https://example.com so files from base url ''' url_parse = urlparse(url) url = url_parse.scheme + '://' + url_parse.netloc url += '/FUZZ' # TODO: fuzz not only URL but also POST / PUT / headers proxy = get_random_proxy() # Build final cmd fcmd = cmd fcmd += f' -x {proxy}' if proxy else '' fcmd += f' -u {url} -json' # Initialize DirectoryScan object dirscan = DirectoryScan() dirscan.scanned_date = timezone.now() dirscan.command_line = fcmd dirscan.save() # Loop through results and populate EndPoint and DirectoryFile in DB results = [] for line in stream_command( fcmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) name = line['input'].get('FUZZ') length = line['length'] status = line['status'] words = line['words'] url = line['url'] lines = line['lines'] content_type = line['content-type'] duration = line['duration'] if not name: logger.error(f'FUZZ not found for "{url}"') continue endpoint, created = save_endpoint(url, crawl=False, ctx=ctx) endpoint.http_status = status endpoint.content_length = length endpoint.response_time = duration / 1000000000 endpoint.save() if created: urls.append(endpoint.http_url) endpoint.status = status endpoint.content_type = content_type endpoint.content_length = length dfile, created = DirectoryFile.objects.get_or_create( name=name, length=length, words=words, lines=lines, content_type=content_type, url=url) dfile.http_status = status dfile.save() # if created: # logger.warning(f'Found new directory or file {url}') dirscan.directory_files.add(dfile) dirscan.save() if self.subscan: dirscan.dir_subscan_ids.add(self.subscan) subdomain_name = get_subdomain_from_url(endpoint.http_url) subdomain = Subdomain.objects.get(name=subdomain_name, scan_history=self.scan) subdomain.directories.add(dirscan) subdomain.save() # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl(urls, ctx=ctx) return results @app.task(name='fetch_url', queue='main_scan_queue', base=RengineTask, bind=True) def fetch_url(self, urls=[], ctx={}, description=None): """Fetch URLs using different tools like gauplus, gau, gospider, waybackurls ... Args: urls (list): List of URLs to start from. description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/input_endpoints_fetch_url.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(FETCH_URL) or {} should_remove_duplicate_endpoints = config.get(REMOVE_DUPLICATE_ENDPOINTS, True) duplicate_removal_fields = config.get(DUPLICATE_REMOVAL_FIELDS, ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, DEFAULT_GF_PATTERNS) ignore_file_extension = config.get(IGNORE_FILE_EXTENSION, DEFAULT_IGNORE_FILE_EXTENSIONS) tools = config.get(USES_TOOLS, ENDPOINT_SCAN_DEFAULT_TOOLS) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) domain_request_headers = self.domain.request_headers if self.domain else None custom_header = domain_request_headers or self.yaml_configuration.get(CUSTOM_HEADER) exclude_subdomains = config.get(EXCLUDED_SUBDOMAINS, False) # Get URLs to scan and save to input file if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, exclude_subdomains=exclude_subdomains, get_only_default_urls=True, ctx=ctx ) # Domain regex host = self.domain.name if self.domain else urlparse(urls[0]).netloc host_regex = f"\'https?://([a-z0-9]+[.])*{host}.*\'" # Tools cmds cmd_map = { 'gau': f'gau', 'gauplus': f'gauplus -random-agent', 'hakrawler': 'hakrawler -subs -u', 'waybackurls': 'waybackurls', 'gospider': f'gospider -S {input_path} --js -d 2 --sitemap --robots -w -r', 'katana': f'katana -list {input_path} -silent -jc -kf all -d 3 -fs rdn', } if proxy: cmd_map['gau'] += f' --proxy "{proxy}"' cmd_map['gauplus'] += f' -p "{proxy}"' cmd_map['gospider'] += f' -p {proxy}' cmd_map['hakrawler'] += f' -proxy {proxy}' cmd_map['katana'] += f' -proxy {proxy}' if threads > 0: cmd_map['gau'] += f' --threads {threads}' cmd_map['gauplus'] += f' -t {threads}' cmd_map['gospider'] += f' -t {threads}' cmd_map['katana'] += f' -c {threads}' if custom_header: header_string = ';;'.join([ f'{key}: {value}' for key, value in custom_header.items() ]) cmd_map['hakrawler'] += f' -h {header_string}' cmd_map['katana'] += f' -H {header_string}' header_flags = [':'.join(h) for h in header_string.split(';;')] for flag in header_flags: cmd_map['gospider'] += f' -H {flag}' cat_input = f'cat {input_path}' grep_output = f'grep -Eo {host_regex}' cmd_map = { tool: f'{cat_input} | {cmd} | {grep_output} > {self.results_dir}/urls_{tool}.txt' for tool, cmd in cmd_map.items() } tasks = group( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for tool, cmd in cmd_map.items() if tool in tools ) # Cleanup task sort_output = [ f'cat {self.results_dir}/urls_* > {self.output_path}', f'cat {input_path} >> {self.output_path}', f'sort -u {self.output_path} -o {self.output_path}', ] if ignore_file_extension: ignore_exts = '|'.join(ignore_file_extension) grep_ext_filtered_output = [ f'cat {self.output_path} | grep -Eiv "\\.({ignore_exts}).*" > {self.results_dir}/urls_filtered.txt', f'mv {self.results_dir}/urls_filtered.txt {self.output_path}' ] sort_output.extend(grep_ext_filtered_output) cleanup = chain( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for cmd in sort_output ) # Run all commands task = chord(tasks)(cleanup) with allow_join_result(): task.get() # Store all the endpoints and run httpx with open(self.output_path) as f: discovered_urls = f.readlines() self.notify(fields={'Discovered URLs': len(discovered_urls)}) # Some tools can have an URL in the format <URL>] - <PATH> or <URL> - <PATH>, add them # to the final URL list all_urls = [] for url in discovered_urls: url = url.strip() urlpath = None base_url = None if '] ' in url: # found JS scraped endpoint e.g from gospider split = tuple(url.split('] ')) if not len(split) == 2: logger.warning(f'URL format not recognized for "{url}". Skipping.') continue base_url, urlpath = split urlpath = urlpath.lstrip('- ') elif ' - ' in url: # found JS scraped endpoint e.g from gospider base_url, urlpath = tuple(url.split(' - ')) if base_url and urlpath: subdomain = urlparse(base_url) url = f'{subdomain.scheme}://{subdomain.netloc}{self.url_filter}' if not validators.url(url): logger.warning(f'Invalid URL "{url}". Skipping.') if url not in all_urls: all_urls.append(url) # Filter out URLs if a path filter was passed if self.url_filter: all_urls = [url for url in all_urls if self.url_filter in url] # Write result to output path with open(self.output_path, 'w') as f: f.write('\n'.join(all_urls)) logger.warning(f'Found {len(all_urls)} usable URLs') # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl( all_urls, ctx=ctx, should_remove_duplicate_endpoints=should_remove_duplicate_endpoints, duplicate_removal_fields=duplicate_removal_fields ) #-------------------# # GF PATTERNS MATCH # #-------------------# # Combine old gf patterns with new ones if gf_patterns: self.scan.used_gf_patterns = ','.join(gf_patterns) self.scan.save() # Run gf patterns on saved endpoints # TODO: refactor to Celery task for gf_pattern in gf_patterns: # TODO: js var is causing issues, removing for now if gf_pattern == 'jsvar': logger.info('Ignoring jsvar as it is causing issues.') continue # Run gf on current pattern logger.warning(f'Running gf on pattern "{gf_pattern}"') gf_output_file = f'{self.results_dir}/gf_patterns_{gf_pattern}.txt' cmd = f'cat {self.output_path} | gf {gf_pattern} | grep -Eo {host_regex} >> {gf_output_file}' run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Check output file if not os.path.exists(gf_output_file): logger.error(f'Could not find GF output file {gf_output_file}. Skipping GF pattern "{gf_pattern}"') continue # Read output file line by line and with open(gf_output_file, 'r') as f: lines = f.readlines() # Add endpoints / subdomains to DB for url in lines: http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue endpoint, created = save_endpoint( http_url, crawl=False, subdomain=subdomain, ctx=ctx) if not endpoint: continue earlier_pattern = None if not created: earlier_pattern = endpoint.matched_gf_patterns pattern = f'{earlier_pattern},{gf_pattern}' if earlier_pattern else gf_pattern endpoint.matched_gf_patterns = pattern endpoint.save() return all_urls def parse_curl_output(response): # TODO: Enrich from other cURL fields. CURL_REGEX_HTTP_STATUS = f'HTTP\/(?:(?:\d\.?)+)\s(\d+)\s(?:\w+)' http_status = 0 if response: failed = False regex = re.compile(CURL_REGEX_HTTP_STATUS, re.MULTILINE) try: http_status = int(regex.findall(response)[0]) except (KeyError, TypeError, IndexError): pass return { 'http_status': http_status, } @app.task(name='vulnerability_scan', queue='main_scan_queue', bind=True, base=RengineTask) def vulnerability_scan(self, urls=[], ctx={}, description=None): """ This function will serve as an entrypoint to vulnerability scan. All other vulnerability scan will be run from here including nuclei, crlfuzz, etc """ logger.info('Running Vulnerability Scan Queue') config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_run_nuclei = config.get(RUN_NUCLEI, True) should_run_crlfuzz = config.get(RUN_CRLFUZZ, False) should_run_dalfox = config.get(RUN_DALFOX, False) should_run_s3scanner = config.get(RUN_S3SCANNER, True) grouped_tasks = [] if should_run_nuclei: _task = nuclei_scan.si( urls=urls, ctx=ctx, description=f'Nuclei Scan' ) grouped_tasks.append(_task) if should_run_crlfuzz: _task = crlfuzz_scan.si( urls=urls, ctx=ctx, description=f'CRLFuzz Scan' ) grouped_tasks.append(_task) if should_run_dalfox: _task = dalfox_xss_scan.si( urls=urls, ctx=ctx, description=f'Dalfox XSS Scan' ) grouped_tasks.append(_task) if should_run_s3scanner: _task = s3scanner.si( ctx=ctx, description=f'Misconfigured S3 Buckets Scanner' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan completed...') # return results return None @app.task(name='nuclei_individual_severity_module', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_individual_severity_module(self, cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx={}, description=None): ''' This celery task will run vulnerability scan in parallel. All severities supplied should run in parallel as grouped tasks. ''' results = [] logger.info(f'Running vulnerability scan with severity: {severity}') cmd += f' -severity {severity}' # Send start notification notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) # Gather nuclei results vuln_data = parse_nuclei_result(line) # Get corresponding subdomain http_url = sanitize_url(line.get('matched-at')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) # Look for duplicate vulnerabilities by excluding records that might change but are irrelevant. object_comparison_exclude = ['response', 'curl_command', 'tags', 'references', 'cve_ids', 'cwe_ids'] # Add subdomain and target domain to the duplicate check vuln_data_copy = vuln_data.copy() vuln_data_copy['subdomain'] = subdomain vuln_data_copy['target_domain'] = self.domain # Check if record exists, if exists do not save it if record_exists(Vulnerability, data=vuln_data_copy, exclude_keys=object_comparison_exclude): logger.warning(f'Nuclei vulnerability of severity {severity} : {vuln_data_copy["name"]} for {subdomain_name} already exists') continue # Get or create EndPoint object response = line.get('response') httpx_crawl = False if response else enable_http_crawl # avoid yet another httpx crawl endpoint, _ = save_endpoint( http_url, crawl=httpx_crawl, subdomain=subdomain, ctx=ctx) if endpoint: http_url = endpoint.http_url if not httpx_crawl: output = parse_curl_output(response) endpoint.http_status = output['http_status'] endpoint.save() # Get or create Vulnerability object vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, subdomain=subdomain, **vuln_data) if not vuln: continue # Print vuln severity = line['info'].get('severity', 'unknown') logger.warning(str(vuln)) # Send notification for all vulnerabilities except info url = vuln.http_url or vuln.subdomain send_vuln = ( notif and notif.send_vuln_notif and vuln and severity in ['low', 'medium', 'high', 'critical']) if send_vuln: fields = { 'Severity': f'**{severity.upper()}**', 'URL': http_url, 'Subdomain': subdomain_name, 'Name': vuln.name, 'Type': vuln.type, 'Description': vuln.description, 'Template': vuln.template_url, 'Tags': vuln.get_tags_str(), 'CVEs': vuln.get_cve_str(), 'CWEs': vuln.get_cwe_str(), 'References': vuln.get_refs_str() } severity_map = { 'low': 'info', 'medium': 'warning', 'high': 'error', 'critical': 'error' } self.notify( f'vulnerability_scan_#{vuln.id}', severity_map[severity], fields, add_meta_info=False) # Send report to hackerone hackerone_query = Hackerone.objects.all() send_report = ( hackerone_query.exists() and severity not in ('info', 'low') and vuln.target_domain.h1_team_handle ) if send_report: hackerone = hackerone_query.first() if hackerone.send_critical and severity == 'critical': send_hackerone_report.delay(vuln.id) elif hackerone.send_high and severity == 'high': send_hackerone_report.delay(vuln.id) elif hackerone.send_medium and severity == 'medium': send_hackerone_report.delay(vuln.id) # Write results to JSON file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) # Send finish notif if send_status: vulns = Vulnerability.objects.filter(scan_history__id=self.scan_id) info_count = vulns.filter(severity=0).count() low_count = vulns.filter(severity=1).count() medium_count = vulns.filter(severity=2).count() high_count = vulns.filter(severity=3).count() critical_count = vulns.filter(severity=4).count() unknown_count = vulns.filter(severity=-1).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count + unknown_count fields = { 'Total': vulnerability_count, 'Critical': critical_count, 'High': high_count, 'Medium': medium_count, 'Low': low_count, 'Info': info_count, 'Unknown': unknown_count } self.notify(fields=fields) # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=NUCLEI ).exclude( severity=0 ) # find all unique vulnerabilities based on path and title # all unique vulnerability will go thru gpt function and get report # once report is got, it will be matched with other vulnerabilities and saved unique_vulns = set() for vuln in vulns: unique_vulns.add((vuln.name, vuln.get_path())) unique_vulns = list(unique_vulns) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in unique_vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return None def get_vulnerability_gpt_report(vuln): title = vuln[0] path = vuln[1] logger.info(f'Getting GPT Report for {title}, PATH: {path}') # check if in db already exists stored = GPTVulnerabilityReport.objects.filter( url_path=path ).filter( title=title ).first() if stored: response = { 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: report = GPTVulnerabilityReportGenerator() vulnerability_description = get_gpt_vuln_input_description( title, path ) response = report.get_vulnerability_description(vulnerability_description) add_gpt_description_db( title, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) for vuln in Vulnerability.objects.filter(name=title, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() def add_gpt_description_db(title, path, description, impact, remediation, references): gpt_report = GPTVulnerabilityReport() gpt_report.url_path = path gpt_report.title = title gpt_report.description = description gpt_report.impact = impact gpt_report.remediation = remediation gpt_report.save() for url in references: ref, created = VulnerabilityReference.objects.get_or_create(url=url) gpt_report.references.add(ref) gpt_report.save() @app.task(name='nuclei_scan', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_scan(self, urls=[], ctx={}, description=None): """HTTP vulnerability scan using Nuclei Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. Notes: Unfurl the urls to keep only domain and path, will be sent to vuln scan and ignore certain file extensions. Thanks: https://github.com/six2dez/reconftw """ # Config config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} input_path = f'{self.results_dir}/input_endpoints_vulnerability_scan.txt' enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) concurrency = config.get(NUCLEI_CONCURRENCY) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) retries = config.get(RETRIES) or self.yaml_configuration.get(RETRIES, DEFAULT_RETRIES) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) custom_header = config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) should_fetch_gpt_report = config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) proxy = get_random_proxy() nuclei_specific_config = config.get('nuclei', {}) use_nuclei_conf = nuclei_specific_config.get(USE_CONFIG, False) severities = nuclei_specific_config.get(NUCLEI_SEVERITY, NUCLEI_DEFAULT_SEVERITIES) tags = nuclei_specific_config.get(NUCLEI_TAGS, []) tags = ','.join(tags) nuclei_templates = nuclei_specific_config.get(NUCLEI_TEMPLATE) custom_nuclei_templates = nuclei_specific_config.get(NUCLEI_CUSTOM_TEMPLATE) # severities_str = ','.join(severities) # Get alive endpoints if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=enable_http_crawl, ignore_files=True, write_filepath=input_path, ctx=ctx ) if intensity == 'normal': # reduce number of endpoints to scan unfurl_filter = f'{self.results_dir}/urls_unfurled.txt' run_command( f"cat {input_path} | unfurl -u format %s://%d%p |uro > {unfurl_filter}", shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {unfurl_filter} -o {unfurl_filter}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) input_path = unfurl_filter # Build templates # logger.info('Updating Nuclei templates ...') run_command( 'nuclei -update-templates', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) templates = [] if not (nuclei_templates or custom_nuclei_templates): templates.append(NUCLEI_DEFAULT_TEMPLATES_PATH) if nuclei_templates: if ALL in nuclei_templates: template = NUCLEI_DEFAULT_TEMPLATES_PATH templates.append(template) else: templates.extend(nuclei_templates) if custom_nuclei_templates: custom_nuclei_template_paths = [f'{str(elem)}.yaml' for elem in custom_nuclei_templates] template = templates.extend(custom_nuclei_template_paths) # Build CMD cmd = 'nuclei -j' cmd += ' -config /root/.config/nuclei/config.yaml' if use_nuclei_conf else '' cmd += f' -irr' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -l {input_path}' cmd += f' -c {str(concurrency)}' if concurrency > 0 else '' cmd += f' -proxy {proxy} ' if proxy else '' cmd += f' -retries {retries}' if retries > 0 else '' cmd += f' -rl {rate_limit}' if rate_limit > 0 else '' # cmd += f' -severity {severities_str}' cmd += f' -timeout {str(timeout)}' if timeout and timeout > 0 else '' cmd += f' -tags {tags}' if tags else '' cmd += f' -silent' for tpl in templates: cmd += f' -t {tpl}' grouped_tasks = [] custom_ctx = ctx for severity in severities: custom_ctx['track'] = True _task = nuclei_individual_severity_module.si( cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx=custom_ctx, description=f'Nuclei Scan with severity {severity}' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan with all severities completed...') return None @app.task(name='dalfox_xss_scan', queue='main_scan_queue', base=RengineTask, bind=True) def dalfox_xss_scan(self, urls=[], ctx={}, description=None): """XSS Scan using dalfox Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) dalfox_config = vuln_config.get(DALFOX) or {} custom_header = dalfox_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() is_waf_evasion = dalfox_config.get(WAF_EVASION, False) blind_xss_server = dalfox_config.get(BLIND_XSS_SERVER) user_agent = dalfox_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) timeout = dalfox_config.get(TIMEOUT) delay = dalfox_config.get(DELAY) threads = dalfox_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_dalfox_xss.txt' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=False, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'dalfox --silence --no-color --no-spinner' cmd += f' --only-poc r ' cmd += f' --ignore-return 302,404,403' cmd += f' --skip-bav' cmd += f' file {input_path}' cmd += f' --proxy {proxy}' if proxy else '' cmd += f' --waf-evasion' if is_waf_evasion else '' cmd += f' -b {blind_xss_server}' if blind_xss_server else '' cmd += f' --delay {delay}' if delay else '' cmd += f' --timeout {timeout}' if timeout else '' cmd += f' --user-agent {user_agent}' if user_agent else '' cmd += f' --header {custom_header}' if custom_header else '' cmd += f' --worker {threads}' if threads else '' cmd += f' --format json' results = [] for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id, trunc_char=',' ): if not isinstance(line, dict): continue results.append(line) vuln_data = parse_dalfox_result(line) http_url = sanitize_url(line.get('data')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Dalfox Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=DALFOX ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='crlfuzz_scan', queue='main_scan_queue', base=RengineTask, bind=True) def crlfuzz_scan(self, urls=[], ctx={}, description=None): """CRLF Fuzzing with CRLFuzz Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) custom_header = vuln_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() user_agent = vuln_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) threads = vuln_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_crlf.txt' output_path = f'{self.results_dir}/{self.filename}' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=True, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'crlfuzz -s' cmd += f' -l {input_path}' cmd += f' -x {proxy}' if proxy else '' cmd += f' --H {custom_header}' if custom_header else '' cmd += f' -o {output_path}' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id ) if not os.path.isfile(output_path): logger.info('No Results from CRLFuzz') return crlfs = [] results = [] with open(output_path, 'r') as file: crlfs = file.readlines() for crlf in crlfs: url = crlf.strip() vuln_data = parse_crlfuzz_result(url) http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting CRLFuzz Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=CRLFUZZ ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='s3scanner', queue='main_scan_queue', base=RengineTask, bind=True) def s3scanner(self, ctx={}, description=None): """Bucket Scanner Args: ctx (dict): Context description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/#{self.scan_id}_subdomain_discovery.txt' vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} s3_config = vuln_config.get(S3SCANNER) or {} threads = s3_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) providers = s3_config.get(PROVIDERS, S3SCANNER_DEFAULT_PROVIDERS) scan_history = ScanHistory.objects.filter(pk=self.scan_id).first() for provider in providers: cmd = f's3scanner -bucket-file {input_path} -enumerate -provider {provider} -threads {threads} -json' for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue if line.get('bucket', {}).get('exists', 0) == 1: result = parse_s3scanner_result(line) s3bucket, created = S3Bucket.objects.get_or_create(**result) scan_history.buckets.add(s3bucket) logger.info(f"s3 bucket added {result['provider']}-{result['name']}-{result['region']}") @app.task(name='http_crawl', queue='main_scan_queue', base=RengineTask, bind=True) def http_crawl( self, urls=[], method=None, recrawl=False, ctx={}, track=True, description=None, is_ran_from_subdomain_scan=False, should_remove_duplicate_endpoints=True, duplicate_removal_fields=[]): """Use httpx to query HTTP URLs for important info like page titles, http status, etc... Args: urls (list, optional): A set of URLs to check. Overrides default behavior which queries all endpoints related to this scan. method (str): HTTP method to use (GET, HEAD, POST, PUT, DELETE). recrawl (bool, optional): If False, filter out URLs that have already been crawled. should_remove_duplicate_endpoints (bool): Whether to remove duplicate endpoints duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates Returns: list: httpx results. """ logger.info('Initiating HTTP Crawl') if is_ran_from_subdomain_scan: logger.info('Running From Subdomain Scan...') cmd = '/go/bin/httpx' cfg = self.yaml_configuration.get(HTTP_CRAWL) or {} custom_header = cfg.get(CUSTOM_HEADER, '') threads = cfg.get(THREADS, DEFAULT_THREADS) follow_redirect = cfg.get(FOLLOW_REDIRECT, True) self.output_path = None input_path = f'{self.results_dir}/httpx_input.txt' history_file = f'{self.results_dir}/commands.txt' if urls: # direct passing URLs to check if self.url_filter: urls = [u for u in urls if self.url_filter in u] with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_uncrawled=not recrawl, write_filepath=input_path, ctx=ctx ) # logger.debug(urls) # If no URLs found, skip it if not urls: return # Re-adjust thread number if few URLs to avoid spinning up a monster to # kill a fly. if len(urls) < threads: threads = len(urls) # Get random proxy proxy = get_random_proxy() # Run command cmd += f' -cl -ct -rt -location -td -websocket -cname -asn -cdn -probe -random-agent' cmd += f' -t {threads}' if threads > 0 else '' cmd += f' --http-proxy {proxy}' if proxy else '' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -json' cmd += f' -u {urls[0]}' if len(urls) == 1 else f' -l {input_path}' cmd += f' -x {method}' if method else '' cmd += f' -silent' if follow_redirect: cmd += ' -fr' results = [] endpoint_ids = [] for line in stream_command( cmd, history_file=history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not line or not isinstance(line, dict): continue logger.debug(line) # No response from endpoint if line.get('failed', False): continue # Parse httpx output host = line.get('host', '') content_length = line.get('content_length', 0) http_status = line.get('status_code') http_url, is_redirect = extract_httpx_url(line) page_title = line.get('title') webserver = line.get('webserver') cdn = line.get('cdn', False) rt = line.get('time') techs = line.get('tech', []) cname = line.get('cname', '') content_type = line.get('content_type', '') response_time = -1 if rt: response_time = float(''.join(ch for ch in rt if not ch.isalpha())) if rt[-2:] == 'ms': response_time = response_time / 1000 # Create Subdomain object in DB subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue # Save default HTTP URL to endpoint object in DB endpoint, created = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain, is_default=is_ran_from_subdomain_scan ) if not endpoint: continue endpoint.http_status = http_status endpoint.page_title = page_title endpoint.content_length = content_length endpoint.webserver = webserver endpoint.response_time = response_time endpoint.content_type = content_type endpoint.save() endpoint_str = f'{http_url} [{http_status}] `{content_length}B` `{webserver}` `{rt}`' logger.warning(endpoint_str) if endpoint and endpoint.is_alive and endpoint.http_status != 403: self.notify( fields={'Alive endpoint': f'• {endpoint_str}'}, add_meta_info=False) # Add endpoint to results line['_cmd'] = cmd line['final_url'] = http_url line['endpoint_id'] = endpoint.id line['endpoint_created'] = created line['is_redirect'] = is_redirect results.append(line) # Add technology objects to DB for technology in techs: tech, _ = Technology.objects.get_or_create(name=technology) endpoint.techs.add(tech) if is_ran_from_subdomain_scan: subdomain.technologies.add(tech) subdomain.save() endpoint.save() techs_str = ', '.join([f'`{tech}`' for tech in techs]) self.notify( fields={'Technologies': techs_str}, add_meta_info=False) # Add IP objects for 'a' records to DB a_records = line.get('a', []) for ip_address in a_records: ip, created = save_ip_address( ip_address, subdomain, subscan=self.subscan, cdn=cdn) ips_str = '• ' + '\n• '.join([f'`{ip}`' for ip in a_records]) self.notify( fields={'IPs': ips_str}, add_meta_info=False) # Add IP object for host in DB if host: ip, created = save_ip_address( host, subdomain, subscan=self.subscan, cdn=cdn) self.notify( fields={'IPs': f'• `{ip.address}`'}, add_meta_info=False) # Save subdomain and endpoint if is_ran_from_subdomain_scan: # save subdomain stuffs subdomain.http_url = http_url subdomain.http_status = http_status subdomain.page_title = page_title subdomain.content_length = content_length subdomain.webserver = webserver subdomain.response_time = response_time subdomain.content_type = content_type subdomain.cname = ','.join(cname) subdomain.is_cdn = cdn if cdn: subdomain.cdn_name = line.get('cdn_name') subdomain.save() endpoint.save() endpoint_ids.append(endpoint.id) if should_remove_duplicate_endpoints: # Remove 'fake' alive endpoints that are just redirects to the same page remove_duplicate_endpoints( self.scan_id, self.domain_id, self.subdomain_id, filter_ids=endpoint_ids ) # Remove input file run_command( f'rm {input_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) return results #---------------------# # Notifications tasks # #---------------------# @app.task(name='send_notif', bind=False, queue='send_notif_queue') def send_notif( message, scan_history_id=None, subscan_id=None, **options): if not 'title' in options: message = enrich_notification(message, scan_history_id, subscan_id) send_discord_message(message, **options) send_slack_message(message) send_telegram_message(message) @app.task(name='send_scan_notif', bind=False, queue='send_scan_notif_queue') def send_scan_notif( scan_history_id, subscan_id=None, engine_id=None, status='RUNNING'): """Send scan status notification. Works for scan or a subscan if subscan_id is passed. Args: scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Get domain, engine, scan_history objects engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() tasks = ScanActivity.objects.filter(scan_of=scan) if scan else 0 # Build notif options url = get_scan_url(scan_history_id, subscan_id) title = get_scan_title(scan_history_id, subscan_id) fields = get_scan_fields(engine, scan, subscan, status, tasks) severity = None msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) if status: severity = STATUS_TO_SEVERITIES.get(status) opts = { 'title': title, 'url': url, 'fields': fields, 'severity': severity } logger.warning(f'Sending notification "{title}" [{severity}]') # Send notification send_notif( msg, scan_history_id, subscan_id, **opts) @app.task(name='send_task_notif', bind=False, queue='send_task_notif_queue') def send_task_notif( task_name, status=None, result=None, output_path=None, traceback=None, scan_history_id=None, engine_id=None, subscan_id=None, severity=None, add_meta_info=True, update_fields={}): """Send task status notification. Args: task_name (str): Task name. status (str, optional): Task status. result (str, optional): Task result. output_path (str, optional): Task output path. traceback (str, optional): Task traceback. scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. severity (str, optional): Severity (will be mapped to notif colors) add_meta_info (bool, optional): Wheter to add scan / subscan info to notif. update_fields (dict, optional): Fields key / value to update. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Build fields url = None fields = {} if add_meta_info: engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() url = get_scan_url(scan_history_id) if status: fields['Status'] = f'**{status}**' if engine: fields['Engine'] = engine.engine_name if scan: fields['Scan ID'] = f'[#{scan.id}]({url})' if subscan: url = get_scan_url(scan_history_id, subscan_id) fields['Subscan ID'] = f'[#{subscan.id}]({url})' title = get_task_title(task_name, scan_history_id, subscan_id) if status: severity = STATUS_TO_SEVERITIES.get(status) msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) # Add fields to update for k, v in update_fields.items(): fields[k] = v # Add traceback to notif if traceback and notif.send_scan_tracebacks: fields['Traceback'] = f'```\n{traceback}\n```' # Add files to notif files = [] attach_file = ( notif.send_scan_output_file and output_path and result and not traceback ) if attach_file: output_title = output_path.split('/')[-1] files = [(output_path, output_title)] # Send notif opts = { 'title': title, 'url': url, 'files': files, 'severity': severity, 'fields': fields, 'fields_append': update_fields.keys() } send_notif( msg, scan_history_id=scan_history_id, subscan_id=subscan_id, **opts) @app.task(name='send_file_to_discord', bind=False, queue='send_file_to_discord_queue') def send_file_to_discord(file_path, title=None): notif = Notification.objects.first() do_send = notif and notif.send_to_discord and notif.discord_hook_url if not do_send: return False webhook = DiscordWebhook( url=notif.discord_hook_url, rate_limit_retry=True, username=title or "reNgine Discord Plugin" ) with open(file_path, "rb") as f: head, tail = os.path.split(file_path) webhook.add_file(file=f.read(), filename=tail) webhook.execute() @app.task(name='send_hackerone_report', bind=False, queue='send_hackerone_report_queue') def send_hackerone_report(vulnerability_id): """Send HackerOne vulnerability report. Args: vulnerability_id (int): Vulnerability id. Returns: int: HTTP response status code. """ vulnerability = Vulnerability.objects.get(id=vulnerability_id) severities = {v: k for k,v in NUCLEI_SEVERITY_MAP.items()} headers = { 'Content-Type': 'application/json', 'Accept': 'application/json' } # can only send vulnerability report if team_handle exists if len(vulnerability.target_domain.h1_team_handle) !=0: hackerone_query = Hackerone.objects.all() if hackerone_query.exists(): hackerone = Hackerone.objects.first() severity_value = severities[vulnerability.severity] tpl = hackerone.report_template # Replace syntax of report template with actual content tpl = tpl.replace('{vulnerability_name}', vulnerability.name) tpl = tpl.replace('{vulnerable_url}', vulnerability.http_url) tpl = tpl.replace('{vulnerability_severity}', severity_value) tpl = tpl.replace('{vulnerability_description}', vulnerability.description if vulnerability.description else '') tpl = tpl.replace('{vulnerability_extracted_results}', vulnerability.extracted_results if vulnerability.extracted_results else '') tpl = tpl.replace('{vulnerability_reference}', vulnerability.reference if vulnerability.reference else '') data = { "data": { "type": "report", "attributes": { "team_handle": vulnerability.target_domain.h1_team_handle, "title": '{} found in {}'.format(vulnerability.name, vulnerability.http_url), "vulnerability_information": tpl, "severity_rating": severity_value, "impact": "More information about the impact and vulnerability can be found here: \n" + vulnerability.reference if vulnerability.reference else "NA", } } } r = requests.post( 'https://api.hackerone.com/v1/hackers/reports', auth=(hackerone.username, hackerone.api_key), json=data, headers=headers ) response = r.json() status_code = r.status_code if status_code == 201: vulnerability.hackerone_report_id = response['data']["id"] vulnerability.open_status = False vulnerability.save() return status_code else: logger.error('No team handle found.') status_code = 111 return status_code #-------------# # Utils tasks # #-------------# @app.task(name='parse_nmap_results', bind=False, queue='parse_nmap_results_queue') def parse_nmap_results(xml_file, output_file=None): """Parse results from nmap output file. Args: xml_file (str): nmap XML report file path. Returns: list: List of vulnerabilities found from nmap results. """ with open(xml_file, encoding='utf8') as f: content = f.read() try: nmap_results = xmltodict.parse(content) # parse XML to dict except Exception as e: logger.exception(e) logger.error(f'Cannot parse {xml_file} to valid JSON. Skipping.') return [] # Write JSON to output file if output_file: with open(output_file, 'w') as f: json.dump(nmap_results, f, indent=4) logger.warning(json.dumps(nmap_results, indent=4)) hosts = ( nmap_results .get('nmaprun', {}) .get('host', {}) ) all_vulns = [] if isinstance(hosts, dict): hosts = [hosts] for host in hosts: # Grab hostname / IP from output hostnames_dict = host.get('hostnames', {}) if hostnames_dict: # Ensure that hostnames['hostname'] is a list for consistency hostnames_list = hostnames_dict['hostname'] if isinstance(hostnames_dict['hostname'], list) else [hostnames_dict['hostname']] # Extract all the @name values from the list of dictionaries hostnames = [entry.get('@name') for entry in hostnames_list] else: hostnames = [host.get('address')['@addr']] # Iterate over each hostname for each port for hostname in hostnames: # Grab ports from output ports = host.get('ports', {}).get('port', []) if isinstance(ports, dict): ports = [ports] for port in ports: url_vulns = [] port_number = port['@portid'] url = sanitize_url(f'{hostname}:{port_number}') logger.info(f'Parsing nmap results for {hostname}:{port_number} ...') if not port_number or not port_number.isdigit(): continue port_protocol = port['@protocol'] scripts = port.get('script', []) if isinstance(scripts, dict): scripts = [scripts] for script in scripts: script_id = script['@id'] script_output = script['@output'] script_output_table = script.get('table', []) logger.debug(f'Ran nmap script "{script_id}" on {port_number}/{port_protocol}:\n{script_output}\n') if script_id == 'vulscan': vulns = parse_nmap_vulscan_output(script_output) url_vulns.extend(vulns) elif script_id == 'vulners': vulns = parse_nmap_vulners_output(script_output) url_vulns.extend(vulns) # elif script_id == 'http-server-header': # TODO: nmap can help find technologies as well using the http-server-header script # regex = r'(\w+)/([\d.]+)\s?(?:\((\w+)\))?' # tech_name, tech_version, tech_os = re.match(regex, test_string).groups() # Technology.objects.get_or_create(...) # elif script_id == 'http_csrf': # vulns = parse_nmap_http_csrf_output(script_output) # url_vulns.extend(vulns) else: logger.warning(f'Script output parsing for script "{script_id}" is not supported yet.') # Add URL to vuln for vuln in url_vulns: # TODO: This should extend to any URL, not just HTTP vuln['http_url'] = url if 'http_path' in vuln: vuln['http_url'] += vuln['http_path'] all_vulns.append(vuln) return all_vulns def parse_nmap_http_csrf_output(script_output): pass def parse_nmap_vulscan_output(script_output): """Parse nmap vulscan script output. Args: script_output (str): Vulscan script output. Returns: list: List of Vulnerability dicts. """ data = {} vulns = [] provider_name = '' # Sort all vulns found by provider so that we can match each provider with # a function that pulls from its API to get more info about the # vulnerability. for line in script_output.splitlines(): if not line: continue if not line.startswith('['): # provider line if "No findings" in line: logger.info(f"No findings: {line}") continue elif ' - ' in line: provider_name, provider_url = tuple(line.split(' - ')) data[provider_name] = {'url': provider_url.rstrip(':'), 'entries': []} continue else: # Log a warning logger.warning(f"Unexpected line format: {line}") continue reg = r'\[(.*)\] (.*)' matches = re.match(reg, line) id, title = matches.groups() entry = {'id': id, 'title': title} data[provider_name]['entries'].append(entry) logger.warning('Vulscan parsed output:') logger.warning(pprint.pformat(data)) for provider_name in data: if provider_name == 'Exploit-DB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'IBM X-Force': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'MITRE CVE': logger.error(f'Provider {provider_name} is not supported YET.') for entry in data[provider_name]['entries']: cve_id = entry['id'] vuln = cve_to_vuln(cve_id) vulns.append(vuln) elif provider_name == 'OSVDB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'OpenVAS (Nessus)': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'SecurityFocus': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'VulDB': logger.error(f'Provider {provider_name} is not supported YET.') pass else: logger.error(f'Provider {provider_name} is not supported.') return vulns def parse_nmap_vulners_output(script_output, url=''): """Parse nmap vulners script output. TODO: Rework this as it's currently matching all CVEs no matter the confidence. Args: script_output (str): Script output. Returns: list: List of found vulnerabilities. """ vulns = [] # Check for CVE in script output CVE_REGEX = re.compile(r'.*(CVE-\d\d\d\d-\d+).*') matches = CVE_REGEX.findall(script_output) matches = list(dict.fromkeys(matches)) for cve_id in matches: # get CVE info vuln = cve_to_vuln(cve_id, vuln_type='nmap-vulners-nse') if vuln: vulns.append(vuln) return vulns def cve_to_vuln(cve_id, vuln_type=''): """Search for a CVE using CVESearch and return Vulnerability data. Args: cve_id (str): CVE ID in the form CVE-* Returns: dict: Vulnerability dict. """ cve_info = CVESearch('https://cve.circl.lu').id(cve_id) if not cve_info: logger.error(f'Could not fetch CVE info for cve {cve_id}. Skipping.') return None vuln_cve_id = cve_info['id'] vuln_name = vuln_cve_id vuln_description = cve_info.get('summary', 'none').replace(vuln_cve_id, '').strip() try: vuln_cvss = float(cve_info.get('cvss', -1)) except (ValueError, TypeError): vuln_cvss = -1 vuln_cwe_id = cve_info.get('cwe', '') exploit_ids = cve_info.get('refmap', {}).get('exploit-db', []) osvdb_ids = cve_info.get('refmap', {}).get('osvdb', []) references = cve_info.get('references', []) capec_objects = cve_info.get('capec', []) # Parse ovals for a better vuln name / type ovals = cve_info.get('oval', []) if ovals: vuln_name = ovals[0]['title'] vuln_type = ovals[0]['family'] # Set vulnerability severity based on CVSS score vuln_severity = 'info' if vuln_cvss < 4: vuln_severity = 'low' elif vuln_cvss < 7: vuln_severity = 'medium' elif vuln_cvss < 9: vuln_severity = 'high' else: vuln_severity = 'critical' # Build console warning message msg = f'{vuln_name} | {vuln_severity.upper()} | {vuln_cve_id} | {vuln_cwe_id} | {vuln_cvss}' for id in osvdb_ids: msg += f'\n\tOSVDB: {id}' for exploit_id in exploit_ids: msg += f'\n\tEXPLOITDB: {exploit_id}' logger.warning(msg) vuln = { 'name': vuln_name, 'type': vuln_type, 'severity': NUCLEI_SEVERITY_MAP[vuln_severity], 'description': vuln_description, 'cvss_score': vuln_cvss, 'references': references, 'cve_ids': [vuln_cve_id], 'cwe_ids': [vuln_cwe_id] } return vuln def parse_s3scanner_result(line): ''' Parses and returns s3Scanner Data ''' bucket = line['bucket'] return { 'name': bucket['name'], 'region': bucket['region'], 'provider': bucket['provider'], 'owner_display_name': bucket['owner_display_name'], 'owner_id': bucket['owner_id'], 'perm_auth_users_read': bucket['perm_auth_users_read'], 'perm_auth_users_write': bucket['perm_auth_users_write'], 'perm_auth_users_read_acl': bucket['perm_auth_users_read_acl'], 'perm_auth_users_write_acl': bucket['perm_auth_users_write_acl'], 'perm_auth_users_full_control': bucket['perm_auth_users_full_control'], 'perm_all_users_read': bucket['perm_all_users_read'], 'perm_all_users_write': bucket['perm_all_users_write'], 'perm_all_users_read_acl': bucket['perm_all_users_read_acl'], 'perm_all_users_write_acl': bucket['perm_all_users_write_acl'], 'perm_all_users_full_control': bucket['perm_all_users_full_control'], 'num_objects': bucket['num_objects'], 'size': bucket['bucket_size'] } def parse_nuclei_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ return { 'name': line['info'].get('name', ''), 'type': line['type'], 'severity': NUCLEI_SEVERITY_MAP[line['info'].get('severity', 'unknown')], 'template': line['template'], 'template_url': line['template-url'], 'template_id': line['template-id'], 'description': line['info'].get('description', ''), 'matcher_name': line.get('matcher-name', ''), 'curl_command': line.get('curl-command'), 'request': line.get('request'), 'response': line.get('response'), 'extracted_results': line.get('extracted-results', []), 'cvss_metrics': line['info'].get('classification', {}).get('cvss-metrics', ''), 'cvss_score': line['info'].get('classification', {}).get('cvss-score'), 'cve_ids': line['info'].get('classification', {}).get('cve_id', []) or [], 'cwe_ids': line['info'].get('classification', {}).get('cwe_id', []) or [], 'references': line['info'].get('reference', []) or [], 'tags': line['info'].get('tags', []), 'source': NUCLEI, } def parse_dalfox_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ description = '' description += f" Evidence: {line.get('evidence')} <br>" if line.get('evidence') else '' description += f" Message: {line.get('message')} <br>" if line.get('message') else '' description += f" Payload: {line.get('message_str')} <br>" if line.get('message_str') else '' description += f" Vulnerable Parameter: {line.get('param')} <br>" if line.get('param') else '' return { 'name': 'XSS (Cross Site Scripting)', 'type': 'XSS', 'severity': DALFOX_SEVERITY_MAP[line.get('severity', 'unknown')], 'description': description, 'source': DALFOX, 'cwe_ids': [line.get('cwe')] } def parse_crlfuzz_result(url): """Parse CRLF results Args: url (str): CRLF Vulnerable URL Returns: dict: Vulnerability data. """ return { 'name': 'CRLF (HTTP Response Splitting)', 'type': 'CRLF', 'severity': 2, 'description': 'A CRLF (HTTP Response Splitting) vulnerability has been discovered.', 'source': CRLFUZZ, } def record_exists(model, data, exclude_keys=[]): """ Check if a record already exists in the database based on the given data. Args: model (django.db.models.Model): The Django model to check against. data (dict): Data dictionary containing fields and values. exclude_keys (list): List of keys to exclude from the lookup. Returns: bool: True if the record exists, False otherwise. """ # Extract the keys that will be used for the lookup lookup_fields = {key: data[key] for key in data if key not in exclude_keys} # Return True if a record exists based on the lookup fields, False otherwise return model.objects.filter(**lookup_fields).exists() @app.task(name='geo_localize', bind=False, queue='geo_localize_queue') def geo_localize(host, ip_id=None): """Uses geoiplookup to find location associated with host. Args: host (str): Hostname. ip_id (int): IpAddress object id. Returns: startScan.models.CountryISO: CountryISO object from DB or None. """ if validators.ipv6(host): logger.info(f'Ipv6 "{host}" is not supported by geoiplookup. Skipping.') return None cmd = f'geoiplookup {host}' _, out = run_command(cmd) if 'IP Address not found' not in out and "can't resolve hostname" not in out: country_iso = out.split(':')[1].strip().split(',')[0] country_name = out.split(':')[1].strip().split(',')[1].strip() geo_object, _ = CountryISO.objects.get_or_create( iso=country_iso, name=country_name ) geo_json = { 'iso': country_iso, 'name': country_name } if ip_id: ip = IpAddress.objects.get(pk=ip_id) ip.geo_iso = geo_object ip.save() return geo_json logger.info(f'Geo IP lookup failed for host "{host}"') return None @app.task(name='query_whois', bind=False, queue='query_whois_queue') def query_whois(ip_domain, force_reload_whois=False): """Query WHOIS information for an IP or a domain name. Args: ip_domain (str): IP address or domain name. save_domain (bool): Whether to save domain or not, default False Returns: dict: WHOIS information. """ if not force_reload_whois and Domain.objects.filter(name=ip_domain).exists() and Domain.objects.get(name=ip_domain).domain_info: domain = Domain.objects.get(name=ip_domain) if not domain.insert_date: domain.insert_date = timezone.now() domain.save() domain_info_db = domain.domain_info domain_info = DottedDict( dnssec=domain_info_db.dnssec, created=domain_info_db.created, updated=domain_info_db.updated, expires=domain_info_db.expires, geolocation_iso=domain_info_db.geolocation_iso, status=[status['name'] for status in DomainWhoisStatusSerializer(domain_info_db.status, many=True).data], whois_server=domain_info_db.whois_server, ns_records=[ns['name'] for ns in NameServersSerializer(domain_info_db.name_servers, many=True).data], registrar_name=domain_info_db.registrar.name, registrar_phone=domain_info_db.registrar.phone, registrar_email=domain_info_db.registrar.email, registrar_url=domain_info_db.registrar.url, registrant_name=domain_info_db.registrant.name, registrant_id=domain_info_db.registrant.id_str, registrant_organization=domain_info_db.registrant.organization, registrant_city=domain_info_db.registrant.city, registrant_state=domain_info_db.registrant.state, registrant_zip_code=domain_info_db.registrant.zip_code, registrant_country=domain_info_db.registrant.country, registrant_phone=domain_info_db.registrant.phone, registrant_fax=domain_info_db.registrant.fax, registrant_email=domain_info_db.registrant.email, registrant_address=domain_info_db.registrant.address, admin_name=domain_info_db.admin.name, admin_id=domain_info_db.admin.id_str, admin_organization=domain_info_db.admin.organization, admin_city=domain_info_db.admin.city, admin_state=domain_info_db.admin.state, admin_zip_code=domain_info_db.admin.zip_code, admin_country=domain_info_db.admin.country, admin_phone=domain_info_db.admin.phone, admin_fax=domain_info_db.admin.fax, admin_email=domain_info_db.admin.email, admin_address=domain_info_db.admin.address, tech_name=domain_info_db.tech.name, tech_id=domain_info_db.tech.id_str, tech_organization=domain_info_db.tech.organization, tech_city=domain_info_db.tech.city, tech_state=domain_info_db.tech.state, tech_zip_code=domain_info_db.tech.zip_code, tech_country=domain_info_db.tech.country, tech_phone=domain_info_db.tech.phone, tech_fax=domain_info_db.tech.fax, tech_email=domain_info_db.tech.email, tech_address=domain_info_db.tech.address, related_tlds=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_tlds, many=True).data], related_domains=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_domains, many=True).data], historical_ips=[ip for ip in HistoricalIPSerializer(domain_info_db.historical_ips, many=True).data], ) if domain_info_db.dns_records: a_records = [] txt_records = [] mx_records = [] dns_records = [{'name': dns['name'], 'type': dns['type']} for dns in DomainDNSRecordSerializer(domain_info_db.dns_records, many=True).data] for dns in dns_records: if dns['type'] == 'a': a_records.append(dns['name']) elif dns['type'] == 'txt': txt_records.append(dns['name']) elif dns['type'] == 'mx': mx_records.append(dns['name']) domain_info.a_records = a_records domain_info.txt_records = txt_records domain_info.mx_records = mx_records else: logger.info(f'Domain info for "{ip_domain}" not found in DB, querying whois') domain_info = DottedDict() # find domain historical ip try: historical_ips = get_domain_historical_ip_address(ip_domain) domain_info.historical_ips = historical_ips except Exception as e: logger.error(f'HistoricalIP for {ip_domain} not found!\nError: {str(e)}') historical_ips = [] # find associated domains using ip_domain try: related_domains = reverse_whois(ip_domain.split('.')[0]) except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] # find related tlds using TLSx try: related_tlds = [] output_path = '/tmp/ip_domain_tlsx.txt' tlsx_command = f'tlsx -san -cn -silent -ro -host {ip_domain} -o {output_path}' run_command( tlsx_command, shell=True, ) tlsx_output = [] with open(output_path) as f: tlsx_output = f.readlines() tldextract_target = tldextract.extract(ip_domain) for doms in tlsx_output: doms = doms.strip() tldextract_res = tldextract.extract(doms) if ip_domain != doms and tldextract_res.domain == tldextract_target.domain and tldextract_res.subdomain == '': related_tlds.append(doms) related_tlds = list(set(related_tlds)) domain_info.related_tlds = related_tlds except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] related_domains_list = [] if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_related = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_related) related_domains_list.append(_domain['name']) for _domain in related_tlds: domain_related = RelatedDomain.objects.get_or_create( name=_domain, )[0] db_domain_info.related_tlds.add(domain_related) for _ip in historical_ips: historical_ip = HistoricalIP.objects.get_or_create( ip=_ip['ip'], owner=_ip['owner'], location=_ip['location'], last_seen=_ip['last_seen'], )[0] db_domain_info.historical_ips.add(historical_ip) domain.domain_info = db_domain_info domain.save() command = f'netlas host {ip_domain} -f json' # check if netlas key is provided netlas_key = get_netlas_key() command += f' -a {netlas_key}' if netlas_key else '' result = subprocess.check_output(command.split()).decode('utf-8') if 'Failed to parse response data' in result: # do fallback return { 'status': False, 'ip_domain': ip_domain, 'result': "Netlas limit exceeded.", 'message': 'Netlas limit exceeded.' } try: result = json.loads(result) logger.info(result) whois = result.get('whois') if result.get('whois') else {} domain_info.created = whois.get('created_date') domain_info.expires = whois.get('expiration_date') domain_info.updated = whois.get('updated_date') domain_info.whois_server = whois.get('whois_server') if 'registrant' in whois: registrant = whois.get('registrant') domain_info.registrant_name = registrant.get('name') domain_info.registrant_country = registrant.get('country') domain_info.registrant_id = registrant.get('id') domain_info.registrant_state = registrant.get('province') domain_info.registrant_city = registrant.get('city') domain_info.registrant_phone = registrant.get('phone') domain_info.registrant_address = registrant.get('street') domain_info.registrant_organization = registrant.get('organization') domain_info.registrant_fax = registrant.get('fax') domain_info.registrant_zip_code = registrant.get('postal_code') email_search = EMAIL_REGEX.search(str(registrant.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.registrant_email = field_content if 'administrative' in whois: administrative = whois.get('administrative') domain_info.admin_name = administrative.get('name') domain_info.admin_country = administrative.get('country') domain_info.admin_id = administrative.get('id') domain_info.admin_state = administrative.get('province') domain_info.admin_city = administrative.get('city') domain_info.admin_phone = administrative.get('phone') domain_info.admin_address = administrative.get('street') domain_info.admin_organization = administrative.get('organization') domain_info.admin_fax = administrative.get('fax') domain_info.admin_zip_code = administrative.get('postal_code') mail_search = EMAIL_REGEX.search(str(administrative.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.admin_email = field_content if 'technical' in whois: technical = whois.get('technical') domain_info.tech_name = technical.get('name') domain_info.tech_country = technical.get('country') domain_info.tech_state = technical.get('province') domain_info.tech_id = technical.get('id') domain_info.tech_city = technical.get('city') domain_info.tech_phone = technical.get('phone') domain_info.tech_address = technical.get('street') domain_info.tech_organization = technical.get('organization') domain_info.tech_fax = technical.get('fax') domain_info.tech_zip_code = technical.get('postal_code') mail_search = EMAIL_REGEX.search(str(technical.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.tech_email = field_content if 'dns' in result: dns = result.get('dns') domain_info.mx_records = dns.get('mx') domain_info.txt_records = dns.get('txt') domain_info.a_records = dns.get('a') domain_info.ns_records = whois.get('name_servers') domain_info.dnssec = True if whois.get('dnssec') else False domain_info.status = whois.get('status') if 'registrar' in whois: registrar = whois.get('registrar') domain_info.registrar_name = registrar.get('name') domain_info.registrar_email = registrar.get('email') domain_info.registrar_phone = registrar.get('phone') domain_info.registrar_url = registrar.get('url') # find associated domains if registrant email is found related_domains = reverse_whois(domain_info.get('registrant_email')) if domain_info.get('registrant_email') else [] for _domain in related_domains: related_domains_list.append(_domain['name']) # remove duplicate domains from related domains list related_domains_list = list(set(related_domains_list)) domain_info.related_domains = related_domains_list # save to db if domain exists if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_rel = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_rel) db_domain_info.dnssec = domain_info.get('dnssec') #dates db_domain_info.created = domain_info.get('created') db_domain_info.updated = domain_info.get('updated') db_domain_info.expires = domain_info.get('expires') #registrar db_domain_info.registrar = Registrar.objects.get_or_create( name=domain_info.get('registrar_name'), email=domain_info.get('registrar_email'), phone=domain_info.get('registrar_phone'), url=domain_info.get('registrar_url'), )[0] db_domain_info.registrant = DomainRegistration.objects.get_or_create( name=domain_info.get('registrant_name'), organization=domain_info.get('registrant_organization'), address=domain_info.get('registrant_address'), city=domain_info.get('registrant_city'), state=domain_info.get('registrant_state'), zip_code=domain_info.get('registrant_zip_code'), country=domain_info.get('registrant_country'), email=domain_info.get('registrant_email'), phone=domain_info.get('registrant_phone'), fax=domain_info.get('registrant_fax'), id_str=domain_info.get('registrant_id'), )[0] db_domain_info.admin = DomainRegistration.objects.get_or_create( name=domain_info.get('admin_name'), organization=domain_info.get('admin_organization'), address=domain_info.get('admin_address'), city=domain_info.get('admin_city'), state=domain_info.get('admin_state'), zip_code=domain_info.get('admin_zip_code'), country=domain_info.get('admin_country'), email=domain_info.get('admin_email'), phone=domain_info.get('admin_phone'), fax=domain_info.get('admin_fax'), id_str=domain_info.get('admin_id'), )[0] db_domain_info.tech = DomainRegistration.objects.get_or_create( name=domain_info.get('tech_name'), organization=domain_info.get('tech_organization'), address=domain_info.get('tech_address'), city=domain_info.get('tech_city'), state=domain_info.get('tech_state'), zip_code=domain_info.get('tech_zip_code'), country=domain_info.get('tech_country'), email=domain_info.get('tech_email'), phone=domain_info.get('tech_phone'), fax=domain_info.get('tech_fax'), id_str=domain_info.get('tech_id'), )[0] for status in domain_info.get('status') or []: _status = WhoisStatus.objects.get_or_create( name=status )[0] _status.save() db_domain_info.status.add(_status) for ns in domain_info.get('ns_records') or []: _ns = NameServer.objects.get_or_create( name=ns )[0] _ns.save() db_domain_info.name_servers.add(_ns) for a in domain_info.get('a_records') or []: _a = DNSRecord.objects.get_or_create( name=a, type='a' )[0] _a.save() db_domain_info.dns_records.add(_a) for mx in domain_info.get('mx_records') or []: _mx = DNSRecord.objects.get_or_create( name=mx, type='mx' )[0] _mx.save() db_domain_info.dns_records.add(_mx) for txt in domain_info.get('txt_records') or []: _txt = DNSRecord.objects.get_or_create( name=txt, type='txt' )[0] _txt.save() db_domain_info.dns_records.add(_txt) db_domain_info.geolocation_iso = domain_info.get('registrant_country') db_domain_info.whois_server = domain_info.get('whois_server') db_domain_info.save() domain.domain_info = db_domain_info domain.save() except Exception as e: return { 'status': False, 'ip_domain': ip_domain, 'result': "unable to fetch records from WHOIS database.", 'message': str(e) } return { 'status': True, 'ip_domain': ip_domain, 'dnssec': domain_info.get('dnssec'), 'created': domain_info.get('created'), 'updated': domain_info.get('updated'), 'expires': domain_info.get('expires'), 'geolocation_iso': domain_info.get('registrant_country'), 'domain_statuses': domain_info.get('status'), 'whois_server': domain_info.get('whois_server'), 'dns': { 'a': domain_info.get('a_records'), 'mx': domain_info.get('mx_records'), 'txt': domain_info.get('txt_records'), }, 'registrar': { 'name': domain_info.get('registrar_name'), 'phone': domain_info.get('registrar_phone'), 'email': domain_info.get('registrar_email'), 'url': domain_info.get('registrar_url'), }, 'registrant': { 'name': domain_info.get('registrant_name'), 'id': domain_info.get('registrant_id'), 'organization': domain_info.get('registrant_organization'), 'address': domain_info.get('registrant_address'), 'city': domain_info.get('registrant_city'), 'state': domain_info.get('registrant_state'), 'zipcode': domain_info.get('registrant_zip_code'), 'country': domain_info.get('registrant_country'), 'phone': domain_info.get('registrant_phone'), 'fax': domain_info.get('registrant_fax'), 'email': domain_info.get('registrant_email'), }, 'admin': { 'name': domain_info.get('admin_name'), 'id': domain_info.get('admin_id'), 'organization': domain_info.get('admin_organization'), 'address':domain_info.get('admin_address'), 'city': domain_info.get('admin_city'), 'state': domain_info.get('admin_state'), 'zipcode': domain_info.get('admin_zip_code'), 'country': domain_info.get('admin_country'), 'phone': domain_info.get('admin_phone'), 'fax': domain_info.get('admin_fax'), 'email': domain_info.get('admin_email'), }, 'technical_contact': { 'name': domain_info.get('tech_name'), 'id': domain_info.get('tech_id'), 'organization': domain_info.get('tech_organization'), 'address': domain_info.get('tech_address'), 'city': domain_info.get('tech_city'), 'state': domain_info.get('tech_state'), 'zipcode': domain_info.get('tech_zip_code'), 'country': domain_info.get('tech_country'), 'phone': domain_info.get('tech_phone'), 'fax': domain_info.get('tech_fax'), 'email': domain_info.get('tech_email'), }, 'nameservers': domain_info.get('ns_records'), # 'similar_domains': domain_info.get('similar_domains'), 'related_domains': domain_info.get('related_domains'), 'related_tlds': domain_info.get('related_tlds'), 'historical_ips': domain_info.get('historical_ips'), } @app.task(name='remove_duplicate_endpoints', bind=False, queue='remove_duplicate_endpoints_queue') def remove_duplicate_endpoints( scan_history_id, domain_id, subdomain_id=None, filter_ids=[], filter_status=[200, 301, 404], duplicate_removal_fields=ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS ): """Remove duplicate endpoints. Check for implicit redirections by comparing endpoints: - [x] `content_length` similarities indicating redirections - [x] `page_title` (check for same page title) - [ ] Sign-in / login page (check for endpoints with the same words) Args: scan_history_id: ScanHistory id. domain_id (int): Domain id. subdomain_id (int, optional): Subdomain id. filter_ids (list): List of endpoint ids to filter on. filter_status (list): List of HTTP status codes to filter on. duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates """ logger.info(f'Removing duplicate endpoints based on {duplicate_removal_fields}') endpoints = ( EndPoint.objects .filter(scan_history__id=scan_history_id) .filter(target_domain__id=domain_id) ) if filter_status: endpoints = endpoints.filter(http_status__in=filter_status) if subdomain_id: endpoints = endpoints.filter(subdomain__id=subdomain_id) if filter_ids: endpoints = endpoints.filter(id__in=filter_ids) for field_name in duplicate_removal_fields: cl_query = ( endpoints .values_list(field_name) .annotate(mc=Count(field_name)) .order_by('-mc') ) for (field_value, count) in cl_query: if count > DELETE_DUPLICATES_THRESHOLD: eps_to_delete = ( endpoints .filter(**{field_name: field_value}) .order_by('discovered_date') .all()[1:] ) msg = f'Deleting {len(eps_to_delete)} endpoints [reason: same {field_name} {field_value}]' for ep in eps_to_delete: url = urlparse(ep.http_url) if url.path in ['', '/', '/login']: # try do not delete the original page that other pages redirect to continue msg += f'\n\t {ep.http_url} [{ep.http_status}] [{field_name}={field_value}]' ep.delete() logger.warning(msg) @app.task(name='run_command', bind=False, queue='run_command_queue') def run_command(cmd, cwd=None, shell=False, history_file=None, scan_id=None, activity_id=None): """Run a given command using subprocess module. Args: cmd (str): Command to run. cwd (str): Current working directory. echo (bool): Log command. shell (bool): Run within separate shell if True. history_file (str): Write command + output to history file. Returns: tuple: Tuple with return_code, output. """ logger.info(cmd) logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Run the command using subprocess popen = subprocess.Popen( cmd if shell else cmd.split(), shell=shell, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, cwd=cwd, universal_newlines=True) output = '' for stdout_line in iter(popen.stdout.readline, ""): item = stdout_line.strip() output += '\n' + item logger.debug(item) popen.stdout.close() popen.wait() return_code = popen.returncode command_obj.output = output command_obj.return_code = return_code command_obj.save() if history_file: mode = 'a' if not os.path.exists(history_file): mode = 'w' with open(history_file, mode) as f: f.write(f'\n{cmd}\n{return_code}\n{output}\n------------------\n') return return_code, output #-------------# # Other utils # #-------------# def stream_command(cmd, cwd=None, shell=False, history_file=None, encoding='utf-8', scan_id=None, activity_id=None, trunc_char=None): # Log cmd logger.info(cmd) # logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Sanitize the cmd command = cmd if shell else cmd.split() # Run the command using subprocess process = subprocess.Popen( command, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, universal_newlines=True, shell=shell) # Log the output in real-time to the database output = "" # Process the output for line in iter(lambda: process.stdout.readline(), b''): if not line: break line = line.strip() ansi_escape = re.compile(r'\x1B(?:[@-Z\\-_]|\[[0-?]*[ -/]*[@-~])') line = ansi_escape.sub('', line) line = line.replace('\\x0d\\x0a', '\n') if trunc_char and line.endswith(trunc_char): line = line[:-1] item = line # Try to parse the line as JSON try: item = json.loads(line) except json.JSONDecodeError: pass # Yield the line #logger.debug(item) yield item # Add the log line to the output output += line + "\n" # Update the command record in the database command_obj.output = output command_obj.save() # Retrieve the return code and output process.wait() return_code = process.returncode # Update the return code and final output in the database command_obj.return_code = return_code command_obj.save() # Append the command, return code and output to the history file if history_file is not None: with open(history_file, "a") as f: f.write(f"{cmd}\n{return_code}\n{output}\n") def process_httpx_response(line): """TODO: implement this""" def extract_httpx_url(line): """Extract final URL from httpx results. Always follow redirects to find the last URL. Args: line (dict): URL data output by httpx. Returns: tuple: (final_url, redirect_bool) tuple. """ status_code = line.get('status_code', 0) final_url = line.get('final_url') location = line.get('location') chain_status_codes = line.get('chain_status_codes', []) # Final URL is already looking nice, if it exists return it if final_url: return final_url, False http_url = line['url'] # fallback to url field # Handle redirects manually REDIRECT_STATUS_CODES = [301, 302] is_redirect = ( status_code in REDIRECT_STATUS_CODES or any(x in REDIRECT_STATUS_CODES for x in chain_status_codes) ) if is_redirect and location: if location.startswith(('http', 'https')): http_url = location else: http_url = f'{http_url}/{location.lstrip("/")}' # Sanitize URL http_url = sanitize_url(http_url) return http_url, is_redirect #-------------# # OSInt utils # #-------------# def get_and_save_dork_results(lookup_target, results_dir, type, lookup_keywords=None, lookup_extensions=None, delay=3, page_count=2, scan_history=None): """ Uses gofuzz to dork and store information Args: lookup_target (str): target to look into such as stackoverflow or even the target itself results_dir (str): Results directory type (str): Dork Type Title lookup_keywords (str): comma separated keywords or paths to look for lookup_extensions (str): comma separated extensions to look for delay (int): delay between each requests page_count (int): pages in google to extract information scan_history (startScan.ScanHistory): Scan History Object """ results = [] gofuzz_command = f'{GOFUZZ_EXEC_PATH} -t {lookup_target} -d {delay} -p {page_count}' if lookup_extensions: gofuzz_command += f' -e {lookup_extensions}' elif lookup_keywords: gofuzz_command += f' -w {lookup_keywords}' output_file = f'{results_dir}/gofuzz.txt' gofuzz_command += f' -o {output_file}' history_file = f'{results_dir}/commands.txt' try: run_command( gofuzz_command, shell=False, history_file=history_file, scan_id=scan_history.id, ) if not os.path.isfile(output_file): return with open(output_file) as f: for line in f.readlines(): url = line.strip() if url: results.append(url) dork, created = Dork.objects.get_or_create( type=type, url=url ) if scan_history: scan_history.dorks.add(dork) # remove output file os.remove(output_file) except Exception as e: logger.exception(e) return results def get_and_save_emails(scan_history, activity_id, results_dir): """Get and save emails from Google, Bing and Baidu. Args: scan_history (startScan.ScanHistory): Scan history object. activity_id: ScanActivity Object results_dir (str): Results directory. Returns: list: List of emails found. """ emails = [] # Proxy settings # get_random_proxy() # Gather emails from Google, Bing and Baidu output_file = f'{results_dir}/emails_tmp.txt' history_file = f'{results_dir}/commands.txt' command = f'python3 /usr/src/github/Infoga/infoga.py --domain {scan_history.domain.name} --source all --report {output_file}' try: run_command( command, shell=False, history_file=history_file, scan_id=scan_history.id, activity_id=activity_id) if not os.path.isfile(output_file): logger.info('No Email results') return [] with open(output_file) as f: for line in f.readlines(): if 'Email' in line: split_email = line.split(' ')[2] emails.append(split_email) output_path = f'{results_dir}/emails.txt' with open(output_path, 'w') as output_file: for email_address in emails: save_email(email_address, scan_history) output_file.write(f'{email_address}\n') except Exception as e: logger.exception(e) return emails def save_metadata_info(meta_dict): """Extract metadata from Google Search. Args: meta_dict (dict): Info dict. Returns: list: List of startScan.MetaFinderDocument objects. """ logger.warning(f'Getting metadata for {meta_dict.osint_target}') scan_history = ScanHistory.objects.get(id=meta_dict.scan_id) # Proxy settings get_random_proxy() # Get metadata result = extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if not result: logger.error(f'No metadata result from Google Search for {meta_dict.osint_target}.') return [] # Add metadata info to DB results = [] for metadata_name, data in result.get_metadata().items(): subdomain = Subdomain.objects.get( scan_history=meta_dict.scan_id, name=meta_dict.osint_target) metadata = DottedDict({k: v for k, v in data.items()}) meta_finder_document = MetaFinderDocument( subdomain=subdomain, target_domain=meta_dict.domain, scan_history=scan_history, url=metadata.url, doc_name=metadata_name, http_status=metadata.status_code, producer=metadata.metadata.get('Producer'), creator=metadata.metadata.get('Creator'), creation_date=metadata.metadata.get('CreationDate'), modified_date=metadata.metadata.get('ModDate'), author=metadata.metadata.get('Author'), title=metadata.metadata.get('Title'), os=metadata.metadata.get('OSInfo')) meta_finder_document.save() results.append(data) return results #-----------------# # Utils functions # #-----------------# def create_scan_activity(scan_history_id, message, status): scan_activity = ScanActivity() scan_activity.scan_of = ScanHistory.objects.get(pk=scan_history_id) scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id #--------------------# # Database functions # #--------------------# def save_vulnerability(**vuln_data): references = vuln_data.pop('references', []) cve_ids = vuln_data.pop('cve_ids', []) cwe_ids = vuln_data.pop('cwe_ids', []) tags = vuln_data.pop('tags', []) subscan = vuln_data.pop('subscan', None) # remove nulls vuln_data = replace_nulls(vuln_data) # Create vulnerability vuln, created = Vulnerability.objects.get_or_create(**vuln_data) if created: vuln.discovered_date = timezone.now() vuln.open_status = True vuln.save() # Save vuln tags for tag_name in tags or []: tag, created = VulnerabilityTags.objects.get_or_create(name=tag_name) if tag: vuln.tags.add(tag) vuln.save() # Save CVEs for cve_id in cve_ids or []: cve, created = CveId.objects.get_or_create(name=cve_id) if cve: vuln.cve_ids.add(cve) vuln.save() # Save CWEs for cve_id in cwe_ids or []: cwe, created = CweId.objects.get_or_create(name=cve_id) if cwe: vuln.cwe_ids.add(cwe) vuln.save() # Save vuln reference for url in references or []: ref, created = VulnerabilityReference.objects.get_or_create(url=url) if created: vuln.references.add(ref) vuln.save() # Save subscan id in vuln object if subscan: vuln.vuln_subscan_ids.add(subscan) vuln.save() return vuln, created def save_endpoint( http_url, ctx={}, crawl=False, is_default=False, **endpoint_data): """Get or create EndPoint object. If crawl is True, also crawl the endpoint HTTP URL with httpx. Args: http_url (str): Input HTTP URL. is_default (bool): If the url is a default url for SubDomains. scan_history (startScan.models.ScanHistory): ScanHistory object. domain (startScan.models.Domain): Domain object. subdomain (starScan.models.Subdomain): Subdomain object. results_dir (str, optional): Results directory. crawl (bool, optional): Run httpx on endpoint if True. Default: False. force (bool, optional): Force crawl even if ENABLE_HTTP_CRAWL mode is on. subscan (startScan.models.SubScan, optional): SubScan object. Returns: tuple: (startScan.models.EndPoint, created) where `created` is a boolean indicating if the object is new or already existed. """ # remove nulls endpoint_data = replace_nulls(endpoint_data) scheme = urlparse(http_url).scheme endpoint = None created = False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in http_url: logger.error(f"{http_url} is not a URL of domain {domain.name}. Skipping.") return None, False if crawl: ctx['track'] = False results = http_crawl( urls=[http_url], method='HEAD', ctx=ctx) if results: endpoint_data = results[0] endpoint_id = endpoint_data['endpoint_id'] created = endpoint_data['endpoint_created'] endpoint = EndPoint.objects.get(pk=endpoint_id) elif not scheme: return None, False else: # add dumb endpoint without probing it scan = ScanHistory.objects.filter(pk=ctx.get('scan_history_id')).first() domain = Domain.objects.filter(pk=ctx.get('domain_id')).first() if not validators.url(http_url): return None, False http_url = sanitize_url(http_url) # Try to get the first matching record (prevent duplicate error) endpoints = EndPoint.objects.filter( scan_history=scan, target_domain=domain, http_url=http_url, **endpoint_data ) if endpoints.exists(): endpoint = endpoints.first() created = False else: # No existing record, create a new one endpoint = EndPoint.objects.create( scan_history=scan, target_domain=domain, http_url=http_url, **endpoint_data ) created = True if created: endpoint.is_default = is_default endpoint.discovered_date = timezone.now() endpoint.save() subscan_id = ctx.get('subscan_id') if subscan_id: endpoint.endpoint_subscan_ids.add(subscan_id) endpoint.save() return endpoint, created def save_subdomain(subdomain_name, ctx={}): """Get or create Subdomain object. Args: subdomain_name (str): Subdomain name. scan_history (startScan.models.ScanHistory): ScanHistory object. Returns: tuple: (startScan.models.Subdomain, created) where `created` is a boolean indicating if the object has been created in DB. """ scan_id = ctx.get('scan_history_id') subscan_id = ctx.get('subscan_id') out_of_scope_subdomains = ctx.get('out_of_scope_subdomains', []) valid_domain = ( validators.domain(subdomain_name) or validators.ipv4(subdomain_name) or validators.ipv6(subdomain_name) ) if not valid_domain: logger.error(f'{subdomain_name} is not an invalid domain. Skipping.') return None, False if subdomain_name in out_of_scope_subdomains: logger.error(f'{subdomain_name} is out-of-scope. Skipping.') return None, False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in subdomain_name: logger.error(f"{subdomain_name} is not a subdomain of domain {domain.name}. Skipping.") return None, False scan = ScanHistory.objects.filter(pk=scan_id).first() domain = scan.domain if scan else None subdomain, created = Subdomain.objects.get_or_create( scan_history=scan, target_domain=domain, name=subdomain_name) if created: # logger.warning(f'Found new subdomain {subdomain_name}') subdomain.discovered_date = timezone.now() if subscan_id: subdomain.subdomain_subscan_ids.add(subscan_id) subdomain.save() return subdomain, created def save_email(email_address, scan_history=None): if not validators.email(email_address): logger.info(f'Email {email_address} is invalid. Skipping.') return None, False email, created = Email.objects.get_or_create(address=email_address) # if created: # logger.warning(f'Found new email address {email_address}') # Add email to ScanHistory if scan_history: scan_history.emails.add(email) scan_history.save() return email, created def save_employee(name, designation, scan_history=None): employee, created = Employee.objects.get_or_create( name=name, designation=designation) # if created: # logger.warning(f'Found new employee {name}') # Add employee to ScanHistory if scan_history: scan_history.employees.add(employee) scan_history.save() return employee, created def save_ip_address(ip_address, subdomain=None, subscan=None, **kwargs): if not (validators.ipv4(ip_address) or validators.ipv6(ip_address)): logger.info(f'IP {ip_address} is not a valid IP. Skipping.') return None, False ip, created = IpAddress.objects.get_or_create(address=ip_address) # if created: # logger.warning(f'Found new IP {ip_address}') # Set extra attributes for key, value in kwargs.items(): setattr(ip, key, value) ip.save() # Add IP to subdomain if subdomain: subdomain.ip_addresses.add(ip) subdomain.save() # Add subscan to IP if subscan: ip.ip_subscan_ids.add(subscan) # Geo-localize IP asynchronously if created: geo_localize.delay(ip_address, ip.id) return ip, created def save_imported_subdomains(subdomains, ctx={}): """Take a list of subdomains imported and write them to from_imported.txt. Args: subdomains (list): List of subdomain names. scan_history (startScan.models.ScanHistory): ScanHistory instance. domain (startScan.models.Domain): Domain instance. results_dir (str): Results directory. """ domain_id = ctx['domain_id'] domain = Domain.objects.get(pk=domain_id) results_dir = ctx.get('results_dir', RENGINE_RESULTS) # Validate each subdomain and de-duplicate entries subdomains = list(set([ subdomain for subdomain in subdomains if validators.domain(subdomain) and domain.name == get_domain_from_subdomain(subdomain) ])) if not subdomains: return logger.warning(f'Found {len(subdomains)} imported subdomains.') with open(f'{results_dir}/from_imported.txt', 'w+') as output_file: for name in subdomains: subdomain_name = name.strip() subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain.is_imported_subdomain = True subdomain.save() output_file.write(f'{subdomain}\n') @app.task(name='query_reverse_whois', bind=False, queue='query_reverse_whois_queue') def query_reverse_whois(lookup_keyword): """Queries Reverse WHOIS information for an organization or email address. Args: lookup_keyword (str): Registrar Name or email Returns: dict: Reverse WHOIS information. """ return get_associated_domains(lookup_keyword) @app.task(name='query_ip_history', bind=False, queue='query_ip_history_queue') def query_ip_history(domain): """Queries the IP history for a domain Args: domain (str): domain_name Returns: list: list of historical ip addresses """ return get_domain_historical_ip_address(domain) @app.task(name='gpt_vulnerability_description', bind=False, queue='gpt_queue') def gpt_vulnerability_description(vulnerability_id): """Generate and store Vulnerability Description using GPT. Args: vulnerability_id (Vulnerability Model ID): Vulnerability ID to fetch Description. """ logger.info('Getting GPT Vulnerability Description') try: lookup_vulnerability = Vulnerability.objects.get(id=vulnerability_id) lookup_url = urlparse(lookup_vulnerability.http_url) path = lookup_url.path except Exception as e: return { 'status': False, 'error': str(e) } # check in db GPTVulnerabilityReport model if vulnerability description and path matches stored = GPTVulnerabilityReport.objects.filter(url_path=path).filter(title=lookup_vulnerability.name).first() if stored: response = { 'status': True, 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: vulnerability_description = get_gpt_vuln_input_description( lookup_vulnerability.name, path ) # one can add more description here later gpt_generator = GPTVulnerabilityReportGenerator() response = gpt_generator.get_vulnerability_description(vulnerability_description) add_gpt_description_db( lookup_vulnerability.name, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) # for all vulnerabilities with the same vulnerability name this description has to be stored. # also the consition is that the url must contain a part of this. for vuln in Vulnerability.objects.filter(name=lookup_vulnerability.name, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() return response

psyray

7c01a46cea370e74385682ba7c28eaf4e58f5d69

2e089dc62f1bd64aa481750da10fa750e3aa232d

@psyray Remove it please so I can merge it.

AnonymousWP

yogeshojha/rengine

fix: ffuf ANSI code processing preventing task to finish

Should - [ ] fix #1006 Needs to be tested for potential impact on other tasks (e.g: dalfox)

2023-11-21 11:54:34+00:00

2023-11-24 03:10:39+00:00

web/reNgine/tasks.py

import csv import json import os import pprint import subprocess import time import validators import whatportis import xmltodict import yaml import tldextract import concurrent.futures from datetime import datetime from urllib.parse import urlparse from api.serializers import SubdomainSerializer from celery import chain, chord, group from celery.result import allow_join_result from celery.utils.log import get_task_logger from django.db.models import Count from dotted_dict import DottedDict from django.utils import timezone from pycvesearch import CVESearch from metafinder.extractor import extract_metadata_from_google_search from reNgine.celery import app from reNgine.gpt import GPTVulnerabilityReportGenerator from reNgine.celery_custom_task import RengineTask from reNgine.common_func import * from reNgine.definitions import * from reNgine.settings import * from reNgine.gpt import * from reNgine.utilities import * from scanEngine.models import (EngineType, InstalledExternalTool, Notification, Proxy) from startScan.models import * from startScan.models import EndPoint, Subdomain, Vulnerability from targetApp.models import Domain """ Celery tasks. """ logger = get_task_logger(__name__) #----------------------# # Scan / Subscan tasks # #----------------------# @app.task(name='initiate_scan', bind=False, queue='initiate_scan_queue') def initiate_scan( scan_history_id, domain_id, engine_id=None, scan_type=LIVE_SCAN, results_dir=RENGINE_RESULTS, imported_subdomains=[], out_of_scope_subdomains=[], url_filter=''): """Initiate a new scan. Args: scan_history_id (int): ScanHistory id. domain_id (int): Domain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. imported_subdomains (list): Imported subdomains. out_of_scope_subdomains (list): Out-of-scope subdomains. url_filter (str): URL path. Default: '' """ # Get scan history scan = ScanHistory.objects.get(pk=scan_history_id) # Get scan engine engine_id = engine_id or scan.scan_type.id # scan history engine_id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, []) # Get domain and set last_scan_date domain = Domain.objects.get(pk=domain_id) domain.last_scan_date = timezone.now() domain.save() # Get path filter url_filter = url_filter.rstrip('/') # Get or create ScanHistory() object if scan_type == LIVE_SCAN: # immediate scan = ScanHistory.objects.get(pk=scan_history_id) scan.scan_status = RUNNING_TASK elif scan_type == SCHEDULED_SCAN: # scheduled scan = ScanHistory() scan.scan_status = INITIATED_TASK scan.scan_type = engine scan.celery_ids = [initiate_scan.request.id] scan.domain = domain scan.start_scan_date = timezone.now() scan.tasks = engine.tasks scan.results_dir = f'{results_dir}/{domain.name}_{scan.id}' add_gf_patterns = gf_patterns and 'fetch_url' in engine.tasks if add_gf_patterns: scan.used_gf_patterns = ','.join(gf_patterns) scan.save() # Create scan results dir os.makedirs(scan.results_dir) # Build task context ctx = { 'scan_history_id': scan_history_id, 'engine_id': engine_id, 'domain_id': domain.id, 'results_dir': scan.results_dir, 'url_filter': url_filter, 'yaml_configuration': config, 'out_of_scope_subdomains': out_of_scope_subdomains } ctx_str = json.dumps(ctx, indent=2) # Send start notif logger.warning(f'Starting scan {scan_history_id} with context:\n{ctx_str}') send_scan_notif.delay( scan_history_id, subscan_id=None, engine_id=engine_id, status=CELERY_TASK_STATUS_MAP[scan.scan_status]) # Save imported subdomains in DB save_imported_subdomains(imported_subdomains, ctx=ctx) # Create initial subdomain in DB: make a copy of domain as a subdomain so # that other tasks using subdomains can use it. subdomain_name = domain.name subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) # If enable_http_crawl is set, create an initial root HTTP endpoint so that # HTTP crawling can start somewhere http_url = f'{domain.name}{url_filter}' if url_filter else domain.name endpoint, _ = save_endpoint( http_url, ctx=ctx, crawl=enable_http_crawl, is_default=True, subdomain=subdomain ) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build Celery tasks, crafted according to the dependency graph below: # subdomain_discovery --> port_scan --> fetch_url --> dir_file_fuzz # osint vulnerability_scan # osint dalfox xss scan # screenshot # waf_detection workflow = chain( group( subdomain_discovery.si(ctx=ctx, description='Subdomain discovery'), osint.si(ctx=ctx, description='OS Intelligence') ), port_scan.si(ctx=ctx, description='Port scan'), fetch_url.si(ctx=ctx, description='Fetch URL'), group( dir_file_fuzz.si(ctx=ctx, description='Directories & files fuzz'), vulnerability_scan.si(ctx=ctx, description='Vulnerability scan'), screenshot.si(ctx=ctx, description='Screenshot'), waf_detection.si(ctx=ctx, description='WAF detection') ) ) # Build callback callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery chord logger.info(f'Running Celery workflow with {len(workflow.tasks) + 1} tasks') task = chain(workflow, callback).on_error(callback).delay() scan.celery_ids.append(task.id) scan.save() return { 'success': True, 'task_id': task.id } @app.task(name='initiate_subscan', bind=False, queue='subscan_queue') def initiate_subscan( scan_history_id, subdomain_id, engine_id=None, scan_type=None, results_dir=RENGINE_RESULTS, url_filter=''): """Initiate a new subscan. Args: scan_history_id (int): ScanHistory id. subdomain_id (int): Subdomain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. url_filter (str): URL path. Default: '' """ # Get Subdomain, Domain and ScanHistory subdomain = Subdomain.objects.get(pk=subdomain_id) scan = ScanHistory.objects.get(pk=subdomain.scan_history.id) domain = Domain.objects.get(pk=subdomain.target_domain.id) # Get EngineType engine_id = engine_id or scan.scan_type.id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Create scan activity of SubScan Model subscan = SubScan( start_scan_date=timezone.now(), celery_ids=[initiate_subscan.request.id], scan_history=scan, subdomain=subdomain, type=scan_type, status=RUNNING_TASK, engine=engine) subscan.save() # Get YAML configuration config = yaml.safe_load(engine.yaml_configuration) # Create results directory results_dir = f'{scan.results_dir}/subscans/{subscan.id}' os.makedirs(results_dir, exist_ok=True) # Run task method = globals().get(scan_type) if not method: logger.warning(f'Task {scan_type} is not supported by reNgine. Skipping') return scan.tasks.append(scan_type) scan.save() # Send start notif send_scan_notif.delay( scan.id, subscan_id=subscan.id, engine_id=engine_id, status='RUNNING') # Build context ctx = { 'scan_history_id': scan.id, 'subscan_id': subscan.id, 'engine_id': engine_id, 'domain_id': domain.id, 'subdomain_id': subdomain.id, 'yaml_configuration': config, 'results_dir': results_dir, 'url_filter': url_filter } # Create initial endpoints in DB: find domain HTTP endpoint so that HTTP # crawling can start somewhere base_url = f'{subdomain.name}{url_filter}' if url_filter else subdomain.name endpoint, _ = save_endpoint( base_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build header + callback workflow = method.si(ctx=ctx) callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery tasks task = chain(workflow, callback).on_error(callback).delay() subscan.celery_ids.append(task.id) subscan.save() return { 'success': True, 'task_id': task.id } @app.task(name='report', bind=False, queue='report_queue') def report(ctx={}, description=None): """Report task running after all other tasks. Mark ScanHistory or SubScan object as completed and update with final status, log run details and send notification. Args: description (str, optional): Task description shown in UI. """ # Get objects subscan_id = ctx.get('subscan_id') scan_id = ctx.get('scan_history_id') engine_id = ctx.get('engine_id') scan = ScanHistory.objects.filter(pk=scan_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() # Get failed tasks tasks = ScanActivity.objects.filter(scan_of=scan).all() if subscan: tasks = tasks.filter(celery_id__in=subscan.celery_ids) failed_tasks = tasks.filter(status=FAILED_TASK) # Get task status failed_count = failed_tasks.count() status = SUCCESS_TASK if failed_count == 0 else FAILED_TASK status_h = 'SUCCESS' if failed_count == 0 else 'FAILED' # Update scan / subscan status if subscan: subscan.stop_scan_date = timezone.now() subscan.status = status subscan.save() else: scan.scan_status = status scan.stop_scan_date = timezone.now() scan.save() # Send scan status notif send_scan_notif.delay( scan_history_id=scan_id, subscan_id=subscan_id, engine_id=engine_id, status=status_h) #------------------------- # # Tracked reNgine tasks # #--------------------------# @app.task(name='subdomain_discovery', queue='main_scan_queue', base=RengineTask, bind=True) def subdomain_discovery( self, host=None, ctx=None, description=None): """Uses a set of tools (see SUBDOMAIN_SCAN_DEFAULT_TOOLS) to scan all subdomains associated with a domain. Args: host (str): Hostname to scan. Returns: subdomains (list): List of subdomain names. """ if not host: host = self.subdomain.name if self.subdomain else self.domain.name if self.url_filter: logger.warning(f'Ignoring subdomains scan as an URL path filter was passed ({self.url_filter}).') return # Config config = self.yaml_configuration.get(SUBDOMAIN_DISCOVERY) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL) or self.yaml_configuration.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) tools = config.get(USES_TOOLS, SUBDOMAIN_SCAN_DEFAULT_TOOLS) default_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=True).filter(is_subdomain_gathering=True)] custom_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=False).filter(is_subdomain_gathering=True)] send_subdomain_changes, send_interesting = False, False notif = Notification.objects.first() if notif: send_subdomain_changes = notif.send_subdomain_changes_notif send_interesting = notif.send_interesting_notif # Gather tools to run for subdomain scan if ALL in tools: tools = SUBDOMAIN_SCAN_DEFAULT_TOOLS + custom_subdomain_tools tools = [t.lower() for t in tools] # Make exception for amass since tool name is amass, but command is amass-active/passive default_subdomain_tools.append('amass-passive') default_subdomain_tools.append('amass-active') # Run tools for tool in tools: cmd = None logger.info(f'Scanning subdomains for {host} with {tool}') proxy = get_random_proxy() if tool in default_subdomain_tools: if tool == 'amass-passive': cmd = f'amass enum -passive -d {host} -o {self.results_dir}/subdomains_amass.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' elif tool == 'amass-active': use_amass_config = config.get(USE_AMASS_CONFIG, False) amass_wordlist_name = config.get(AMASS_WORDLIST, 'deepmagic.com-prefixes-top50000') wordlist_path = f'/usr/src/wordlist/{amass_wordlist_name}.txt' cmd = f'amass enum -active -d {host} -o {self.results_dir}/subdomains_amass_active.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' cmd += f' -brute -w {wordlist_path}' elif tool == 'sublist3r': cmd = f'python3 /usr/src/github/Sublist3r/sublist3r.py -d {host} -t {threads} -o {self.results_dir}/subdomains_sublister.txt' elif tool == 'subfinder': cmd = f'subfinder -d {host} -o {self.results_dir}/subdomains_subfinder.txt' use_subfinder_config = config.get(USE_SUBFINDER_CONFIG, False) cmd += ' -config /root/.config/subfinder/config.yaml' if use_subfinder_config else '' cmd += f' -proxy {proxy}' if proxy else '' cmd += f' -timeout {timeout}' if timeout else '' cmd += f' -t {threads}' if threads else '' cmd += f' -silent' elif tool == 'oneforall': cmd = f'python3 /usr/src/github/OneForAll/oneforall.py --target {host} run' cmd_extract = f'cut -d\',\' -f6 /usr/src/github/OneForAll/results/{host}.csv > {self.results_dir}/subdomains_oneforall.txt' cmd_rm = f'rm -rf /usr/src/github/OneForAll/results/{host}.csv' cmd += f' && {cmd_extract} && {cmd_rm}' elif tool == 'ctfr': results_file = self.results_dir + '/subdomains_ctfr.txt' cmd = f'python3 /usr/src/github/ctfr/ctfr.py -d {host} -o {results_file}' cmd_extract = f"cat {results_file} | sed 's/\*.//g' | tail -n +12 | uniq | sort > {results_file}" cmd += f' && {cmd_extract}' elif tool == 'tlsx': results_file = self.results_dir + '/subdomains_tlsx.txt' cmd = f'tlsx -san -cn -silent -ro -host {host}' cmd += f" | sed -n '/^\([a-zA-Z0-9]\([-a-zA-Z0-9]*[a-zA-Z0-9]\)\?\.\)\+{host}$/p' | uniq | sort" cmd += f' > {results_file}' elif tool == 'netlas': results_file = self.results_dir + '/subdomains_netlas.txt' cmd = f'netlas search -d domain -i domain domain:"*.{host}" -f json' netlas_key = get_netlas_key() cmd += f' -a {netlas_key}' if netlas_key else '' cmd_extract = f"grep -oE '([a-zA-Z0-9]([-a-zA-Z0-9]*[a-zA-Z0-9])?\.)+{host}'" cmd += f' | {cmd_extract} > {results_file}' elif tool in custom_subdomain_tools: tool_query = InstalledExternalTool.objects.filter(name__icontains=tool.lower()) if not tool_query.exists(): logger.error(f'Missing {{TARGET}} and {{OUTPUT}} placeholders in {tool} configuration. Skipping.') continue custom_tool = tool_query.first() cmd = custom_tool.subdomain_gathering_command if '{TARGET}' in cmd and '{OUTPUT}' in cmd: cmd = cmd.replace('{TARGET}', host) cmd = cmd.replace('{OUTPUT}', f'{self.results_dir}/subdomains_{tool}.txt') cmd = cmd.replace('{PATH}', custom_tool.github_clone_path) if '{PATH}' in cmd else cmd else: logger.warning( f'Subdomain discovery tool "{tool}" is not supported by reNgine. Skipping.') continue # Run tool try: run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) except Exception as e: logger.error( f'Subdomain discovery tool "{tool}" raised an exception') logger.exception(e) # Gather all the tools' results in one single file. Write subdomains into # separate files, and sort all subdomains. run_command( f'cat {self.results_dir}/subdomains_*.txt > {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {self.output_path} -o {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) with open(self.output_path) as f: lines = f.readlines() # Parse the output_file file and store Subdomain and EndPoint objects found # in db. subdomain_count = 0 subdomains = [] urls = [] for line in lines: subdomain_name = line.strip() valid_url = bool(validators.url(subdomain_name)) valid_domain = ( bool(validators.domain(subdomain_name)) or bool(validators.ipv4(subdomain_name)) or bool(validators.ipv6(subdomain_name)) or valid_url ) if not valid_domain: logger.error(f'Subdomain {subdomain_name} is not a valid domain, IP or URL. Skipping.') continue if valid_url: subdomain_name = urlparse(subdomain_name).netloc if subdomain_name in self.out_of_scope_subdomains: logger.error(f'Subdomain {subdomain_name} is out of scope. Skipping.') continue # Add subdomain subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain_count += 1 subdomains.append(subdomain) urls.append(subdomain.name) # Bulk crawl subdomains if enable_http_crawl: ctx['track'] = True http_crawl(urls, ctx=ctx, is_ran_from_subdomain_scan=True) # Find root subdomain endpoints for subdomain in subdomains: pass # Send notifications subdomains_str = '\n'.join([f'• `{subdomain.name}`' for subdomain in subdomains]) self.notify(fields={ 'Subdomain count': len(subdomains), 'Subdomains': subdomains_str, }) if send_subdomain_changes and self.scan_id and self.domain_id: added = get_new_added_subdomain(self.scan_id, self.domain_id) removed = get_removed_subdomain(self.scan_id, self.domain_id) if added: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in added]) self.notify(fields={'Added subdomains': subdomains_str}) if removed: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in removed]) self.notify(fields={'Removed subdomains': subdomains_str}) if send_interesting and self.scan_id and self.domain_id: interesting_subdomains = get_interesting_subdomains(self.scan_id, self.domain_id) if interesting_subdomains: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in interesting_subdomains]) self.notify(fields={'Interesting subdomains': subdomains_str}) return SubdomainSerializer(subdomains, many=True).data @app.task(name='osint', queue='main_scan_queue', base=RengineTask, bind=True) def osint(self, host=None, ctx={}, description=None): """Run Open-Source Intelligence tools on selected domain. Args: host (str): Hostname to scan. Returns: dict: Results from osint discovery and dorking. """ config = self.yaml_configuration.get(OSINT) or OSINT_DEFAULT_CONFIG results = {} grouped_tasks = [] if 'discover' in config: ctx['track'] = False # results = osint_discovery(host=host, ctx=ctx) _task = osint_discovery.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, activity_id=self.activity_id, results_dir=self.results_dir, ctx=ctx ) grouped_tasks.append(_task) if OSINT_DORK in config or OSINT_CUSTOM_DORK in config: _task = dorking.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, results_dir=self.results_dir ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('OSINT Tasks finished...') # with open(self.output_path, 'w') as f: # json.dump(results, f, indent=4) # # return results @app.task(name='osint_discovery', queue='osint_discovery_queue', bind=False) def osint_discovery(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run OSINT discovery. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: dict: osint metadat and theHarvester and h8mail results. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) osint_lookup = config.get(OSINT_DISCOVER, []) osint_intensity = config.get(INTENSITY, 'normal') documents_limit = config.get(OSINT_DOCUMENTS_LIMIT, 50) results = {} meta_info = [] emails = [] creds = [] # Get and save meta info if 'metainfo' in osint_lookup: if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': host, 'domain': host, 'scan_id': scan_history_id, 'documents_limit': documents_limit }) meta_info.append(save_metadata_info(meta_dict)) # TODO: disabled for now # elif osint_intensity == 'deep': # subdomains = Subdomain.objects # if self.scan: # subdomains = subdomains.filter(scan_history=self.scan) # for subdomain in subdomains: # meta_dict = DottedDict({ # 'osint_target': subdomain.name, # 'domain': self.domain, # 'scan_id': self.scan_id, # 'documents_limit': documents_limit # }) # meta_info.append(save_metadata_info(meta_dict)) grouped_tasks = [] if 'emails' in osint_lookup: emails = get_and_save_emails(scan_history, activity_id, results_dir) emails_str = '\n'.join([f'• `{email}`' for email in emails]) # self.notify(fields={'Emails': emails_str}) # ctx['track'] = False _task = h8mail.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) if 'employees' in osint_lookup: ctx['track'] = False _task = theHarvester.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) # results['emails'] = results.get('emails', []) + emails # results['creds'] = creds # results['meta_info'] = meta_info return results @app.task(name='dorking', bind=False, queue='dorking_queue') def dorking(config, host, scan_history_id, results_dir): """Run Google dorks. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: list: Dorking results for each dork ran. """ # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh scan_history = ScanHistory.objects.get(pk=scan_history_id) dorks = config.get(OSINT_DORK, []) custom_dorks = config.get(OSINT_CUSTOM_DORK, []) results = [] # custom dorking has higher priority try: for custom_dork in custom_dorks: lookup_target = custom_dork.get('lookup_site') # replace with original host if _target_ lookup_target = host if lookup_target == '_target_' else lookup_target if 'lookup_extensions' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_extensions=custom_dork.get('lookup_extensions'), scan_history=scan_history ) elif 'lookup_keywords' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_keywords=custom_dork.get('lookup_keywords'), scan_history=scan_history ) except Exception as e: logger.exception(e) # default dorking try: for dork in dorks: logger.info(f'Getting dork information for {dork}') if dork == 'stackoverflow': results = get_and_save_dork_results( lookup_target='stackoverflow.com', results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'login_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/login/,login.html', page_count=5, scan_history=scan_history ) elif dork == 'admin_panels': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/admin/,admin.html', page_count=5, scan_history=scan_history ) elif dork == 'dashboard_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/dashboard/,dashboard.html', page_count=5, scan_history=scan_history ) elif dork == 'social_media' : social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'reddit.com' ] for site in social_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'project_management' : project_websites = [ 'trello.com', 'atlassian.net' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'code_sharing' : project_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'config_files' : config_file_exts = [ 'env', 'xml', 'conf', 'toml', 'yml', 'yaml', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(config_file_exts), page_count=4, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keyword = 'Jenkins' results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=lookup_keyword, page_count=1, scan_history=scan_history ) elif dork == 'wordpress_files' : lookup_keywords = [ '/wp-content/', '/wp-includes/' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'php_error' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'exposed_documents' : docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(docs_file_ext), page_count=7, scan_history=scan_history ) elif dork == 'db_files' : file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) elif dork == 'git_exposed' : file_ext = [ 'git', ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) except Exception as e: logger.exception(e) return results @app.task(name='theHarvester', queue='theHarvester_queue', bind=False) def theHarvester(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run theHarvester to get save emails, hosts, employees found in domain. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: dict: Dict of emails, employees, hosts and ips found during crawling. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) output_path_json = f'{results_dir}/theHarvester.json' theHarvester_dir = '/usr/src/github/theHarvester' history_file = f'{results_dir}/commands.txt' cmd = f'python3 {theHarvester_dir}/theHarvester.py -d {host} -b all -f {output_path_json}' # Update proxies.yaml proxy_query = Proxy.objects.all() if proxy_query.exists(): proxy = proxy_query.first() if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(f'{theHarvester_dir}/proxies.yaml', 'w') as file: yaml.dump(yaml_data, file) # Run cmd run_command( cmd, shell=False, cwd=theHarvester_dir, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) # Get file location if not os.path.isfile(output_path_json): logger.error(f'Could not open {output_path_json}') return {} # Load theHarvester results with open(output_path_json, 'r') as f: data = json.load(f) # Re-indent theHarvester JSON with open(output_path_json, 'w') as f: json.dump(data, f, indent=4) emails = data.get('emails', []) for email_address in emails: email, _ = save_email(email_address, scan_history=scan_history) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) linkedin_people = data.get('linkedin_people', []) for people in linkedin_people: employee, _ = save_employee( people, designation='linkedin', scan_history=scan_history) # if employee: # self.notify(fields={'LinkedIn people': f'• {employee.name}'}) twitter_people = data.get('twitter_people', []) for people in twitter_people: employee, _ = save_employee( people, designation='twitter', scan_history=scan_history) # if employee: # self.notify(fields={'Twitter people': f'• {employee.name}'}) hosts = data.get('hosts', []) urls = [] for host in hosts: split = tuple(host.split(':')) http_url = split[0] subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) endpoint, _ = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain) # if endpoint: # urls.append(endpoint.http_url) # self.notify(fields={'Hosts': f'• {endpoint.http_url}'}) # if enable_http_crawl: # ctx['track'] = False # http_crawl(urls, ctx=ctx) # TODO: Lots of ips unrelated with our domain are found, disabling # this for now. # ips = data.get('ips', []) # for ip_address in ips: # ip, created = save_ip_address( # ip_address, # subscan=subscan) # if ip: # send_task_notif.delay( # 'osint', # scan_history_id=scan_history_id, # subscan_id=subscan_id, # severity='success', # update_fields={'IPs': f'{ip.address}'}) return data @app.task(name='h8mail', queue='h8mail_queue', bind=False) def h8mail(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run h8mail. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: list[dict]: List of credentials info. """ logger.warning('Getting leaked credentials') scan_history = ScanHistory.objects.get(pk=scan_history_id) input_path = f'{results_dir}/emails.txt' output_file = f'{results_dir}/h8mail.json' cmd = f'h8mail -t {input_path} --json {output_file}' history_file = f'{results_dir}/commands.txt' run_command( cmd, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) with open(output_file) as f: data = json.load(f) creds = data.get('targets', []) # TODO: go through h8mail output and save emails to DB for cred in creds: logger.warning(cred) email_address = cred['target'] pwn_num = cred['pwn_num'] pwn_data = cred.get('data', []) email, created = save_email(email_address, scan_history=scan) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) return creds @app.task(name='screenshot', queue='main_scan_queue', base=RengineTask, bind=True) def screenshot(self, ctx={}, description=None): """Uses EyeWitness to gather screenshot of a domain and/or url. Args: description (str, optional): Task description shown in UI. """ # Config screenshots_path = f'{self.results_dir}/screenshots' output_path = f'{self.results_dir}/screenshots/{self.filename}' alive_endpoints_file = f'{self.results_dir}/endpoints_alive.txt' config = self.yaml_configuration.get(SCREENSHOT) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT + 5) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) # If intensity is normal, grab only the root endpoints of each subdomain strict = True if intensity == 'normal' else False # Get URLs to take screenshot of get_http_urls( is_alive=enable_http_crawl, strict=strict, write_filepath=alive_endpoints_file, get_only_default_urls=True, ctx=ctx ) # Send start notif notification = Notification.objects.first() send_output_file = notification.send_scan_output_file if notification else False # Run cmd cmd = f'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py -f {alive_endpoints_file} -d {screenshots_path} --no-prompt' cmd += f' --timeout {timeout}' if timeout > 0 else '' cmd += f' --threads {threads}' if threads > 0 else '' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(output_path): logger.error(f'Could not load EyeWitness results at {output_path} for {self.domain.name}.') return # Loop through results and save objects in DB screenshot_paths = [] with open(output_path, 'r') as file: reader = csv.reader(file) for row in reader: "Protocol,Port,Domain,Request Status,Screenshot Path, Source Path" protocol, port, subdomain_name, status, screenshot_path, source_path = tuple(row) logger.info(f'{protocol}:{port}:{subdomain_name}:{status}') subdomain_query = Subdomain.objects.filter(name=subdomain_name) if self.scan: subdomain_query = subdomain_query.filter(scan_history=self.scan) if status == 'Successful' and subdomain_query.exists(): subdomain = subdomain_query.first() screenshot_paths.append(screenshot_path) subdomain.screenshot_path = screenshot_path.replace('/usr/src/scan_results/', '') subdomain.save() logger.warning(f'Added screenshot for {subdomain.name} to DB') # Remove all db, html extra files in screenshot results run_command( 'rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format(screenshots_path), shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'rm -rf {screenshots_path}/source', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Send finish notifs screenshots_str = '• ' + '\n• '.join([f'`{path}`' for path in screenshot_paths]) self.notify(fields={'Screenshots': screenshots_str}) if send_output_file: for path in screenshot_paths: title = get_output_file_name( self.scan_id, self.subscan_id, self.filename) send_file_to_discord.delay(path, title) @app.task(name='port_scan', queue='main_scan_queue', base=RengineTask, bind=True) def port_scan(self, hosts=[], ctx={}, description=None): """Run port scan. Args: hosts (list, optional): Hosts to run port scan on. description (str, optional): Task description shown in UI. Returns: list: List of open ports (dict). """ input_file = f'{self.results_dir}/input_subdomains_port_scan.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(PORT_SCAN) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) exclude_ports = config.get(NAABU_EXCLUDE_PORTS, []) exclude_subdomains = config.get(NAABU_EXCLUDE_SUBDOMAINS, False) ports = config.get(PORTS, NAABU_DEFAULT_PORTS) ports = [str(port) for port in ports] rate_limit = config.get(NAABU_RATE) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) passive = config.get(NAABU_PASSIVE, False) use_naabu_config = config.get(USE_NAABU_CONFIG, False) exclude_ports_str = ','.join(return_iterable(exclude_ports)) # nmap args nmap_enabled = config.get(ENABLE_NMAP, False) nmap_cmd = config.get(NMAP_COMMAND, '') nmap_script = config.get(NMAP_SCRIPT, '') nmap_script = ','.join(return_iterable(nmap_script)) nmap_script_args = config.get(NMAP_SCRIPT_ARGS) if hosts: with open(input_file, 'w') as f: f.write('\n'.join(hosts)) else: hosts = get_subdomains( write_filepath=input_file, exclude_subdomains=exclude_subdomains, ctx=ctx) # Build cmd cmd = 'naabu -json -exclude-cdn' cmd += f' -list {input_file}' if len(hosts) > 0 else f' -host {hosts[0]}' if 'full' in ports or 'all' in ports: ports_str = ' -p "-"' elif 'top-100' in ports: ports_str = ' -top-ports 100' elif 'top-1000' in ports: ports_str = ' -top-ports 1000' else: ports_str = ','.join(ports) ports_str = f' -p {ports_str}' cmd += ports_str cmd += ' -config /root/.config/naabu/config.yaml' if use_naabu_config else '' cmd += f' -proxy "{proxy}"' if proxy else '' cmd += f' -c {threads}' if threads else '' cmd += f' -rate {rate_limit}' if rate_limit > 0 else '' cmd += f' -timeout {timeout*1000}' if timeout > 0 else '' cmd += f' -passive' if passive else '' cmd += f' -exclude-ports {exclude_ports_str}' if exclude_ports else '' cmd += f' -silent' # Execute cmd and gather results results = [] urls = [] ports_data = {} for line in stream_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) port_number = line['port'] ip_address = line['ip'] host = line.get('host') or ip_address if port_number == 0: continue # Grab subdomain subdomain = Subdomain.objects.filter( name=host, target_domain=self.domain, scan_history=self.scan ).first() # Add IP DB ip, _ = save_ip_address(ip_address, subdomain, subscan=self.subscan) if self.subscan: ip.ip_subscan_ids.add(self.subscan) ip.save() # Add endpoint to DB # port 80 and 443 not needed as http crawl already does that. if port_number not in [80, 443]: http_url = f'{host}:{port_number}' endpoint, _ = save_endpoint( http_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint: http_url = endpoint.http_url urls.append(http_url) # Add Port in DB port_details = whatportis.get_ports(str(port_number)) service_name = port_details[0].name if len(port_details) > 0 else 'unknown' description = port_details[0].description if len(port_details) > 0 else '' # get or create port port, created = Port.objects.get_or_create( number=port_number, service_name=service_name, description=description ) if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port.save() ip.ports.add(port) ip.save() if host in ports_data: ports_data[host].append(port_number) else: ports_data[host] = [port_number] # Send notification logger.warning(f'Found opened port {port_number} on {ip_address} ({host})') if len(ports_data) == 0: logger.info('Finished running naabu port scan - No open ports found.') if nmap_enabled: logger.info('Nmap scans skipped') return ports_data # Send notification fields_str = '' for host, ports in ports_data.items(): ports_str = ', '.join([f'`{port}`' for port in ports]) fields_str += f'• `{host}`: {ports_str}\n' self.notify(fields={'Ports discovered': fields_str}) # Save output to file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) logger.info('Finished running naabu port scan.') # Process nmap results: 1 process per host sigs = [] if nmap_enabled: logger.warning(f'Starting nmap scans ...') logger.warning(ports_data) for host, port_list in ports_data.items(): ports_str = '_'.join([str(p) for p in port_list]) ctx_nmap = ctx.copy() ctx_nmap['description'] = get_task_title(f'nmap_{host}', self.scan_id, self.subscan_id) ctx_nmap['track'] = False sig = nmap.si( cmd=nmap_cmd, ports=port_list, host=host, script=nmap_script, script_args=nmap_script_args, max_rate=rate_limit, ctx=ctx_nmap) sigs.append(sig) task = group(sigs).apply_async() with allow_join_result(): results = task.get() return ports_data @app.task(name='nmap', queue='main_scan_queue', base=RengineTask, bind=True) def nmap( self, cmd=None, ports=[], host=None, input_file=None, script=None, script_args=None, max_rate=None, ctx={}, description=None): """Run nmap on a host. Args: cmd (str, optional): Existing nmap command to complete. ports (list, optional): List of ports to scan. host (str, optional): Host to scan. input_file (str, optional): Input hosts file. script (str, optional): NSE script to run. script_args (str, optional): NSE script args. max_rate (int): Max rate. description (str, optional): Task description shown in UI. """ notif = Notification.objects.first() ports_str = ','.join(str(port) for port in ports) self.filename = self.filename.replace('.txt', '.xml') filename_vulns = self.filename.replace('.xml', '_vulns.json') output_file = self.output_path output_file_xml = f'{self.results_dir}/{host}_{self.filename}' vulns_file = f'{self.results_dir}/{host}_{filename_vulns}' logger.warning(f'Running nmap on {host}:{ports}') # Build cmd nmap_cmd = get_nmap_cmd( cmd=cmd, ports=ports_str, script=script, script_args=script_args, max_rate=max_rate, host=host, input_file=input_file, output_file=output_file_xml) # Run cmd run_command( nmap_cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Get nmap XML results and convert to JSON vulns = parse_nmap_results(output_file_xml, output_file) with open(vulns_file, 'w') as f: json.dump(vulns, f, indent=4) # Save vulnerabilities found by nmap vulns_str = '' for vuln_data in vulns: # URL is not necessarily an HTTP URL when running nmap (can be any # other vulnerable protocols). Look for existing endpoint and use its # URL as vulnerability.http_url if it exists. url = vuln_data['http_url'] endpoint = EndPoint.objects.filter(http_url__contains=url).first() if endpoint: vuln_data['http_url'] = endpoint.http_url vuln, created = save_vulnerability( target_domain=self.domain, subdomain=self.subdomain, scan_history=self.scan, subscan=self.subscan, endpoint=endpoint, **vuln_data) vulns_str += f'• {str(vuln)}\n' if created: logger.warning(str(vuln)) # Send only 1 notif for all vulns to reduce number of notifs if notif and notif.send_vuln_notif and vulns_str: logger.warning(vulns_str) self.notify(fields={'CVEs': vulns_str}) return vulns @app.task(name='waf_detection', queue='main_scan_queue', base=RengineTask, bind=True) def waf_detection(self, ctx={}, description=None): """ Uses wafw00f to check for the presence of a WAF. Args: description (str, optional): Task description shown in UI. Returns: list: List of startScan.models.Waf objects. """ input_path = f'{self.results_dir}/input_endpoints_waf_detection.txt' config = self.yaml_configuration.get(WAF_DETECTION) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Get alive endpoints from DB get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) cmd = f'wafw00f -i {input_path} -o {self.output_path}' run_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(self.output_path): logger.error(f'Could not find {self.output_path}') return with open(self.output_path) as file: wafs = file.readlines() for line in wafs: line = " ".join(line.split()) splitted = line.split(' ', 1) waf_info = splitted[1].strip() waf_name = waf_info[:waf_info.find('(')].strip() waf_manufacturer = waf_info[waf_info.find('(')+1:waf_info.find(')')].strip().replace('.', '') http_url = sanitize_url(splitted[0].strip()) if not waf_name or waf_name == 'None': continue # Add waf to db waf, _ = Waf.objects.get_or_create( name=waf_name, manufacturer=waf_manufacturer ) # Add waf info to Subdomain in DB subdomain = get_subdomain_from_url(http_url) logger.info(f'Wafw00f Subdomain : {subdomain}') subdomain_query, _ = Subdomain.objects.get_or_create(scan_history=self.scan, name=subdomain) subdomain_query.waf.add(waf) subdomain_query.save() return wafs @app.task(name='dir_file_fuzz', queue='main_scan_queue', base=RengineTask, bind=True) def dir_file_fuzz(self, ctx={}, description=None): """Perform directory scan, and currently uses `ffuf` as a default tool. Args: description (str, optional): Task description shown in UI. Returns: list: List of URLs discovered. """ # Config cmd = 'ffuf' config = self.yaml_configuration.get(DIR_FILE_FUZZ) or {} custom_header = self.yaml_configuration.get(CUSTOM_HEADER) auto_calibration = config.get(AUTO_CALIBRATION, True) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) extensions = config.get(EXTENSIONS, DEFAULT_DIR_FILE_FUZZ_EXTENSIONS) extensions_str = ','.join(map(str, extensions)) follow_redirect = config.get(FOLLOW_REDIRECT, FFUF_DEFAULT_FOLLOW_REDIRECT) max_time = config.get(MAX_TIME, 0) match_http_status = config.get(MATCH_HTTP_STATUS, FFUF_DEFAULT_MATCH_HTTP_STATUS) mc = ','.join([str(c) for c in match_http_status]) recursive_level = config.get(RECURSIVE_LEVEL, FFUF_DEFAULT_RECURSIVE_LEVEL) stop_on_error = config.get(STOP_ON_ERROR, False) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) wordlist_name = config.get(WORDLIST, 'dicc') delay = rate_limit / (threads * 100) # calculate request pause delay from rate_limit and number of threads input_path = f'{self.results_dir}/input_dir_file_fuzz.txt' # Get wordlist wordlist_name = 'dicc' if wordlist_name == 'default' else wordlist_name wordlist_path = f'/usr/src/wordlist/{wordlist_name}.txt' # Build command cmd += f' -w {wordlist_path}' cmd += f' -e {extensions_str}' if extensions else '' cmd += f' -maxtime {max_time}' if max_time > 0 else '' cmd += f' -p {delay}' if delay > 0 else '' cmd += f' -recursion -recursion-depth {recursive_level} ' if recursive_level > 0 else '' cmd += f' -t {threads}' if threads and threads > 0 else '' cmd += f' -timeout {timeout}' if timeout and timeout > 0 else '' cmd += ' -se' if stop_on_error else '' cmd += ' -fr' if follow_redirect else '' cmd += ' -ac' if auto_calibration else '' cmd += f' -mc {mc}' if mc else '' cmd += f' -H "{custom_header}"' if custom_header else '' # Grab URLs to fuzz urls = get_http_urls( is_alive=True, ignore_files=False, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) logger.warning(urls) # Loop through URLs and run command results = [] for url in urls: ''' Above while fetching urls, we are not ignoring files, because some default urls may redirect to https://example.com/login.php so, ignore_files is set to False but, during fuzzing, we will only need part of the path, in above example it is still a good idea to ffuf base url https://example.com so files from base url ''' url_parse = urlparse(url) url = url_parse.scheme + '://' + url_parse.netloc url += '/FUZZ' # TODO: fuzz not only URL but also POST / PUT / headers proxy = get_random_proxy() # Build final cmd fcmd = cmd fcmd += f' -x {proxy}' if proxy else '' fcmd += f' -u {url} -json' # Initialize DirectoryScan object dirscan = DirectoryScan() dirscan.scanned_date = timezone.now() dirscan.command_line = fcmd dirscan.save() # Loop through results and populate EndPoint and DirectoryFile in DB results = [] for line in stream_command( fcmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) name = line['input'].get('FUZZ') length = line['length'] status = line['status'] words = line['words'] url = line['url'] lines = line['lines'] content_type = line['content-type'] duration = line['duration'] if not name: logger.error(f'FUZZ not found for "{url}"') continue endpoint, created = save_endpoint(url, crawl=False, ctx=ctx) # endpoint.is_default = False endpoint.http_status = status endpoint.content_length = length endpoint.response_time = duration / 1000000000 endpoint.save() if created: urls.append(endpoint.http_url) endpoint.status = status endpoint.content_type = content_type endpoint.content_length = length dfile, created = DirectoryFile.objects.get_or_create( name=name, length=length, words=words, lines=lines, content_type=content_type, url=url) dfile.http_status = status dfile.save() # if created: # logger.warning(f'Found new directory or file {url}') dirscan.directory_files.add(dfile) dirscan.save() if self.subscan: dirscan.dir_subscan_ids.add(self.subscan) subdomain_name = get_subdomain_from_url(endpoint.http_url) subdomain = Subdomain.objects.get(name=subdomain_name, scan_history=self.scan) subdomain.directories.add(dirscan) subdomain.save() # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl(urls, ctx=ctx) return results @app.task(name='fetch_url', queue='main_scan_queue', base=RengineTask, bind=True) def fetch_url(self, urls=[], ctx={}, description=None): """Fetch URLs using different tools like gauplus, gau, gospider, waybackurls ... Args: urls (list): List of URLs to start from. description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/input_endpoints_fetch_url.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(FETCH_URL) or {} should_remove_duplicate_endpoints = config.get(REMOVE_DUPLICATE_ENDPOINTS, True) duplicate_removal_fields = config.get(DUPLICATE_REMOVAL_FIELDS, ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, DEFAULT_GF_PATTERNS) ignore_file_extension = config.get(IGNORE_FILE_EXTENSION, DEFAULT_IGNORE_FILE_EXTENSIONS) tools = config.get(USES_TOOLS, ENDPOINT_SCAN_DEFAULT_TOOLS) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) domain_request_headers = self.domain.request_headers if self.domain else None custom_header = domain_request_headers or self.yaml_configuration.get(CUSTOM_HEADER) exclude_subdomains = config.get(EXCLUDED_SUBDOMAINS, False) # Get URLs to scan and save to input file if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, exclude_subdomains=exclude_subdomains, get_only_default_urls=True, ctx=ctx ) # Domain regex host = self.domain.name if self.domain else urlparse(urls[0]).netloc host_regex = f"\'https?://([a-z0-9]+[.])*{host}.*\'" # Tools cmds cmd_map = { 'gau': f'gau', 'gauplus': f'gauplus -random-agent', 'hakrawler': 'hakrawler -subs -u', 'waybackurls': 'waybackurls', 'gospider': f'gospider -S {input_path} --js -d 2 --sitemap --robots -w -r', 'katana': f'katana -list {input_path} -silent -jc -kf all -d 3 -fs rdn', } if proxy: cmd_map['gau'] += f' --proxy "{proxy}"' cmd_map['gauplus'] += f' -p "{proxy}"' cmd_map['gospider'] += f' -p {proxy}' cmd_map['hakrawler'] += f' -proxy {proxy}' cmd_map['katana'] += f' -proxy {proxy}' if threads > 0: cmd_map['gau'] += f' --threads {threads}' cmd_map['gauplus'] += f' -t {threads}' cmd_map['gospider'] += f' -t {threads}' cmd_map['katana'] += f' -c {threads}' if custom_header: header_string = ';;'.join([ f'{key}: {value}' for key, value in custom_header.items() ]) cmd_map['hakrawler'] += f' -h {header_string}' cmd_map['katana'] += f' -H {header_string}' header_flags = [':'.join(h) for h in header_string.split(';;')] for flag in header_flags: cmd_map['gospider'] += f' -H {flag}' cat_input = f'cat {input_path}' grep_output = f'grep -Eo {host_regex}' cmd_map = { tool: f'{cat_input} | {cmd} | {grep_output} > {self.results_dir}/urls_{tool}.txt' for tool, cmd in cmd_map.items() } tasks = group( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for tool, cmd in cmd_map.items() if tool in tools ) # Cleanup task sort_output = [ f'cat {self.results_dir}/urls_* > {self.output_path}', f'cat {input_path} >> {self.output_path}', f'sort -u {self.output_path} -o {self.output_path}', ] if ignore_file_extension: ignore_exts = '|'.join(ignore_file_extension) grep_ext_filtered_output = [ f'cat {self.output_path} | grep -Eiv "\\.({ignore_exts}).*" > {self.results_dir}/urls_filtered.txt', f'mv {self.results_dir}/urls_filtered.txt {self.output_path}' ] sort_output.extend(grep_ext_filtered_output) cleanup = chain( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for cmd in sort_output ) # Run all commands task = chord(tasks)(cleanup) with allow_join_result(): task.get() # Store all the endpoints and run httpx with open(self.output_path) as f: discovered_urls = f.readlines() self.notify(fields={'Discovered URLs': len(discovered_urls)}) # Some tools can have an URL in the format <URL>] - <PATH> or <URL> - <PATH>, add them # to the final URL list all_urls = [] for url in discovered_urls: url = url.strip() urlpath = None base_url = None if '] ' in url: # found JS scraped endpoint e.g from gospider split = tuple(url.split('] ')) if not len(split) == 2: logger.warning(f'URL format not recognized for "{url}". Skipping.') continue base_url, urlpath = split urlpath = urlpath.lstrip('- ') elif ' - ' in url: # found JS scraped endpoint e.g from gospider base_url, urlpath = tuple(url.split(' - ')) if base_url and urlpath: subdomain = urlparse(base_url) url = f'{subdomain.scheme}://{subdomain.netloc}{self.url_filter}' if not validators.url(url): logger.warning(f'Invalid URL "{url}". Skipping.') if url not in all_urls: all_urls.append(url) # Filter out URLs if a path filter was passed if self.url_filter: all_urls = [url for url in all_urls if self.url_filter in url] # Write result to output path with open(self.output_path, 'w') as f: f.write('\n'.join(all_urls)) logger.warning(f'Found {len(all_urls)} usable URLs') # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl( all_urls, ctx=ctx, should_remove_duplicate_endpoints=should_remove_duplicate_endpoints, duplicate_removal_fields=duplicate_removal_fields ) #-------------------# # GF PATTERNS MATCH # #-------------------# # Combine old gf patterns with new ones if gf_patterns: self.scan.used_gf_patterns = ','.join(gf_patterns) self.scan.save() # Run gf patterns on saved endpoints # TODO: refactor to Celery task for gf_pattern in gf_patterns: # TODO: js var is causing issues, removing for now if gf_pattern == 'jsvar': logger.info('Ignoring jsvar as it is causing issues.') continue # Run gf on current pattern logger.warning(f'Running gf on pattern "{gf_pattern}"') gf_output_file = f'{self.results_dir}/gf_patterns_{gf_pattern}.txt' cmd = f'cat {self.output_path} | gf {gf_pattern} | grep -Eo {host_regex} >> {gf_output_file}' run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Check output file if not os.path.exists(gf_output_file): logger.error(f'Could not find GF output file {gf_output_file}. Skipping GF pattern "{gf_pattern}"') continue # Read output file line by line and with open(gf_output_file, 'r') as f: lines = f.readlines() # Add endpoints / subdomains to DB for url in lines: http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue endpoint, created = save_endpoint( http_url, crawl=False, subdomain=subdomain, ctx=ctx) if not endpoint: continue earlier_pattern = None if not created: earlier_pattern = endpoint.matched_gf_patterns pattern = f'{earlier_pattern},{gf_pattern}' if earlier_pattern else gf_pattern endpoint.matched_gf_patterns = pattern endpoint.save() return all_urls def parse_curl_output(response): # TODO: Enrich from other cURL fields. CURL_REGEX_HTTP_STATUS = f'HTTP\/(?:(?:\d\.?)+)\s(\d+)\s(?:\w+)' http_status = 0 if response: failed = False regex = re.compile(CURL_REGEX_HTTP_STATUS, re.MULTILINE) try: http_status = int(regex.findall(response)[0]) except (KeyError, TypeError, IndexError): pass return { 'http_status': http_status, } @app.task(name='vulnerability_scan', queue='main_scan_queue', bind=True, base=RengineTask) def vulnerability_scan(self, urls=[], ctx={}, description=None): """ This function will serve as an entrypoint to vulnerability scan. All other vulnerability scan will be run from here including nuclei, crlfuzz, etc """ logger.info('Running Vulnerability Scan Queue') config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_run_nuclei = config.get(RUN_NUCLEI, True) should_run_crlfuzz = config.get(RUN_CRLFUZZ, False) should_run_dalfox = config.get(RUN_DALFOX, False) should_run_s3scanner = config.get(RUN_S3SCANNER, True) grouped_tasks = [] if should_run_nuclei: _task = nuclei_scan.si( urls=urls, ctx=ctx, description=f'Nuclei Scan' ) grouped_tasks.append(_task) if should_run_crlfuzz: _task = crlfuzz_scan.si( urls=urls, ctx=ctx, description=f'CRLFuzz Scan' ) grouped_tasks.append(_task) if should_run_dalfox: _task = dalfox_xss_scan.si( urls=urls, ctx=ctx, description=f'Dalfox XSS Scan' ) grouped_tasks.append(_task) if should_run_s3scanner: _task = s3scanner.si( ctx=ctx, description=f'Misconfigured S3 Buckets Scanner' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan completed...') # return results return None @app.task(name='nuclei_individual_severity_module', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_individual_severity_module(self, cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx={}, description=None): ''' This celery task will run vulnerability scan in parallel. All severities supplied should run in parallel as grouped tasks. ''' results = [] logger.info(f'Running vulnerability scan with severity: {severity}') cmd += f' -severity {severity}' # Send start notification notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) # Gather nuclei results vuln_data = parse_nuclei_result(line) # Get corresponding subdomain http_url = sanitize_url(line.get('matched-at')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) # Look for duplicate vulnerabilities by excluding records that might change but are irrelevant. object_comparison_exclude = ['response', 'curl_command', 'tags', 'references', 'cve_ids', 'cwe_ids'] # Add subdomain and target domain to the duplicate check vuln_data_copy = vuln_data.copy() vuln_data_copy['subdomain'] = subdomain vuln_data_copy['target_domain'] = self.domain # Check if record exists, if exists do not save it if record_exists(Vulnerability, data=vuln_data_copy, exclude_keys=object_comparison_exclude): logger.warning(f'Nuclei vulnerability of severity {severity} : {vuln_data_copy["name"]} for {subdomain_name} already exists') continue # Get or create EndPoint object response = line.get('response') httpx_crawl = False if response else enable_http_crawl # avoid yet another httpx crawl endpoint, _ = save_endpoint( http_url, crawl=httpx_crawl, subdomain=subdomain, ctx=ctx) if endpoint: http_url = endpoint.http_url if not httpx_crawl: output = parse_curl_output(response) endpoint.http_status = output['http_status'] endpoint.save() # Get or create Vulnerability object vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, subdomain=subdomain, **vuln_data) if not vuln: continue # Print vuln severity = line['info'].get('severity', 'unknown') logger.warning(str(vuln)) # Send notification for all vulnerabilities except info url = vuln.http_url or vuln.subdomain send_vuln = ( notif and notif.send_vuln_notif and vuln and severity in ['low', 'medium', 'high', 'critical']) if send_vuln: fields = { 'Severity': f'**{severity.upper()}**', 'URL': http_url, 'Subdomain': subdomain_name, 'Name': vuln.name, 'Type': vuln.type, 'Description': vuln.description, 'Template': vuln.template_url, 'Tags': vuln.get_tags_str(), 'CVEs': vuln.get_cve_str(), 'CWEs': vuln.get_cwe_str(), 'References': vuln.get_refs_str() } severity_map = { 'low': 'info', 'medium': 'warning', 'high': 'error', 'critical': 'error' } self.notify( f'vulnerability_scan_#{vuln.id}', severity_map[severity], fields, add_meta_info=False) # Send report to hackerone hackerone_query = Hackerone.objects.all() send_report = ( hackerone_query.exists() and severity not in ('info', 'low') and vuln.target_domain.h1_team_handle ) if send_report: hackerone = hackerone_query.first() if hackerone.send_critical and severity == 'critical': send_hackerone_report.delay(vuln.id) elif hackerone.send_high and severity == 'high': send_hackerone_report.delay(vuln.id) elif hackerone.send_medium and severity == 'medium': send_hackerone_report.delay(vuln.id) # Write results to JSON file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) # Send finish notif if send_status: vulns = Vulnerability.objects.filter(scan_history__id=self.scan_id) info_count = vulns.filter(severity=0).count() low_count = vulns.filter(severity=1).count() medium_count = vulns.filter(severity=2).count() high_count = vulns.filter(severity=3).count() critical_count = vulns.filter(severity=4).count() unknown_count = vulns.filter(severity=-1).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count + unknown_count fields = { 'Total': vulnerability_count, 'Critical': critical_count, 'High': high_count, 'Medium': medium_count, 'Low': low_count, 'Info': info_count, 'Unknown': unknown_count } self.notify(fields=fields) # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=NUCLEI ).exclude( severity=0 ) # find all unique vulnerabilities based on path and title # all unique vulnerability will go thru gpt function and get report # once report is got, it will be matched with other vulnerabilities and saved unique_vulns = set() for vuln in vulns: unique_vulns.add((vuln.name, vuln.get_path())) unique_vulns = list(unique_vulns) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in unique_vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return None def get_vulnerability_gpt_report(vuln): title = vuln[0] path = vuln[1] logger.info(f'Getting GPT Report for {title}, PATH: {path}') # check if in db already exists stored = GPTVulnerabilityReport.objects.filter( url_path=path ).filter( title=title ).first() if stored: response = { 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: report = GPTVulnerabilityReportGenerator() vulnerability_description = get_gpt_vuln_input_description( title, path ) response = report.get_vulnerability_description(vulnerability_description) add_gpt_description_db( title, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) for vuln in Vulnerability.objects.filter(name=title, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() def add_gpt_description_db(title, path, description, impact, remediation, references): gpt_report = GPTVulnerabilityReport() gpt_report.url_path = path gpt_report.title = title gpt_report.description = description gpt_report.impact = impact gpt_report.remediation = remediation gpt_report.save() for url in references: ref, created = VulnerabilityReference.objects.get_or_create(url=url) gpt_report.references.add(ref) gpt_report.save() @app.task(name='nuclei_scan', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_scan(self, urls=[], ctx={}, description=None): """HTTP vulnerability scan using Nuclei Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. Notes: Unfurl the urls to keep only domain and path, will be sent to vuln scan and ignore certain file extensions. Thanks: https://github.com/six2dez/reconftw """ # Config config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} input_path = f'{self.results_dir}/input_endpoints_vulnerability_scan.txt' enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) concurrency = config.get(NUCLEI_CONCURRENCY) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) retries = config.get(RETRIES) or self.yaml_configuration.get(RETRIES, DEFAULT_RETRIES) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) custom_header = config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) should_fetch_gpt_report = config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) proxy = get_random_proxy() nuclei_specific_config = config.get('nuclei', {}) use_nuclei_conf = nuclei_specific_config.get(USE_CONFIG, False) severities = nuclei_specific_config.get(NUCLEI_SEVERITY, NUCLEI_DEFAULT_SEVERITIES) tags = nuclei_specific_config.get(NUCLEI_TAGS, []) tags = ','.join(tags) nuclei_templates = nuclei_specific_config.get(NUCLEI_TEMPLATE) custom_nuclei_templates = nuclei_specific_config.get(NUCLEI_CUSTOM_TEMPLATE) # severities_str = ','.join(severities) # Get alive endpoints if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=enable_http_crawl, ignore_files=True, write_filepath=input_path, ctx=ctx ) if intensity == 'normal': # reduce number of endpoints to scan unfurl_filter = f'{self.results_dir}/urls_unfurled.txt' run_command( f"cat {input_path} | unfurl -u format %s://%d%p |uro > {unfurl_filter}", shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {unfurl_filter} -o {unfurl_filter}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) input_path = unfurl_filter # Build templates # logger.info('Updating Nuclei templates ...') run_command( 'nuclei -update-templates', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) templates = [] if not (nuclei_templates or custom_nuclei_templates): templates.append(NUCLEI_DEFAULT_TEMPLATES_PATH) if nuclei_templates: if ALL in nuclei_templates: template = NUCLEI_DEFAULT_TEMPLATES_PATH templates.append(template) else: templates.extend(nuclei_templates) if custom_nuclei_templates: custom_nuclei_template_paths = [f'{str(elem)}.yaml' for elem in custom_nuclei_templates] template = templates.extend(custom_nuclei_template_paths) # Build CMD cmd = 'nuclei -j' cmd += ' -config /root/.config/nuclei/config.yaml' if use_nuclei_conf else '' cmd += f' -irr' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -l {input_path}' cmd += f' -c {str(concurrency)}' if concurrency > 0 else '' cmd += f' -proxy {proxy} ' if proxy else '' cmd += f' -retries {retries}' if retries > 0 else '' cmd += f' -rl {rate_limit}' if rate_limit > 0 else '' # cmd += f' -severity {severities_str}' cmd += f' -timeout {str(timeout)}' if timeout and timeout > 0 else '' cmd += f' -tags {tags}' if tags else '' cmd += f' -silent' for tpl in templates: cmd += f' -t {tpl}' grouped_tasks = [] custom_ctx = ctx for severity in severities: custom_ctx['track'] = True _task = nuclei_individual_severity_module.si( cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx=custom_ctx, description=f'Nuclei Scan with severity {severity}' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan with all severities completed...') return None @app.task(name='dalfox_xss_scan', queue='main_scan_queue', base=RengineTask, bind=True) def dalfox_xss_scan(self, urls=[], ctx={}, description=None): """XSS Scan using dalfox Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) dalfox_config = vuln_config.get(DALFOX) or {} custom_header = dalfox_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() is_waf_evasion = dalfox_config.get(WAF_EVASION, False) blind_xss_server = dalfox_config.get(BLIND_XSS_SERVER) user_agent = dalfox_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) timeout = dalfox_config.get(TIMEOUT) delay = dalfox_config.get(DELAY) threads = dalfox_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_dalfox_xss.txt' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=False, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'dalfox --silence --no-color --no-spinner' cmd += f' --only-poc r ' cmd += f' --ignore-return 302,404,403' cmd += f' --skip-bav' cmd += f' file {input_path}' cmd += f' --proxy {proxy}' if proxy else '' cmd += f' --waf-evasion' if is_waf_evasion else '' cmd += f' -b {blind_xss_server}' if blind_xss_server else '' cmd += f' --delay {delay}' if delay else '' cmd += f' --timeout {timeout}' if timeout else '' cmd += f' --user-agent {user_agent}' if user_agent else '' cmd += f' --header {custom_header}' if custom_header else '' cmd += f' --worker {threads}' if threads else '' cmd += f' --format json' results = [] for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id, trunc_char=',' ): if not isinstance(line, dict): continue results.append(line) vuln_data = parse_dalfox_result(line) http_url = sanitize_url(line.get('data')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Dalfox Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=DALFOX ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='crlfuzz_scan', queue='main_scan_queue', base=RengineTask, bind=True) def crlfuzz_scan(self, urls=[], ctx={}, description=None): """CRLF Fuzzing with CRLFuzz Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) custom_header = vuln_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() user_agent = vuln_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) threads = vuln_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_crlf.txt' output_path = f'{self.results_dir}/{self.filename}' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=True, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'crlfuzz -s' cmd += f' -l {input_path}' cmd += f' -x {proxy}' if proxy else '' cmd += f' --H {custom_header}' if custom_header else '' cmd += f' -o {output_path}' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id ) if not os.path.isfile(output_path): logger.info('No Results from CRLFuzz') return crlfs = [] results = [] with open(output_path, 'r') as file: crlfs = file.readlines() for crlf in crlfs: url = crlf.strip() vuln_data = parse_crlfuzz_result(url) http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting CRLFuzz Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=CRLFUZZ ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='s3scanner', queue='main_scan_queue', base=RengineTask, bind=True) def s3scanner(self, ctx={}, description=None): """Bucket Scanner Args: ctx (dict): Context description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/#{self.scan_id}_subdomain_discovery.txt' vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} s3_config = vuln_config.get(S3SCANNER) or {} threads = s3_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) providers = s3_config.get(PROVIDERS, S3SCANNER_DEFAULT_PROVIDERS) scan_history = ScanHistory.objects.filter(pk=self.scan_id).first() for provider in providers: cmd = f's3scanner -bucket-file {input_path} -enumerate -provider {provider} -threads {threads} -json' for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue if line.get('bucket', {}).get('exists', 0) == 1: result = parse_s3scanner_result(line) s3bucket, created = S3Bucket.objects.get_or_create(**result) scan_history.buckets.add(s3bucket) logger.info(f"s3 bucket added {result['provider']}-{result['name']}-{result['region']}") @app.task(name='http_crawl', queue='main_scan_queue', base=RengineTask, bind=True) def http_crawl( self, urls=[], method=None, recrawl=False, ctx={}, track=True, description=None, is_ran_from_subdomain_scan=False, should_remove_duplicate_endpoints=True, duplicate_removal_fields=[]): """Use httpx to query HTTP URLs for important info like page titles, http status, etc... Args: urls (list, optional): A set of URLs to check. Overrides default behavior which queries all endpoints related to this scan. method (str): HTTP method to use (GET, HEAD, POST, PUT, DELETE). recrawl (bool, optional): If False, filter out URLs that have already been crawled. should_remove_duplicate_endpoints (bool): Whether to remove duplicate endpoints duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates Returns: list: httpx results. """ logger.info('Initiating HTTP Crawl') if is_ran_from_subdomain_scan: logger.info('Running From Subdomain Scan...') cmd = '/go/bin/httpx' cfg = self.yaml_configuration.get(HTTP_CRAWL) or {} custom_header = cfg.get(CUSTOM_HEADER, '') threads = cfg.get(THREADS, DEFAULT_THREADS) follow_redirect = cfg.get(FOLLOW_REDIRECT, True) self.output_path = None input_path = f'{self.results_dir}/httpx_input.txt' history_file = f'{self.results_dir}/commands.txt' if urls: # direct passing URLs to check if self.url_filter: urls = [u for u in urls if self.url_filter in u] with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_uncrawled=not recrawl, write_filepath=input_path, ctx=ctx ) # logger.debug(urls) # If no URLs found, skip it if not urls: return # Re-adjust thread number if few URLs to avoid spinning up a monster to # kill a fly. if len(urls) < threads: threads = len(urls) # Get random proxy proxy = get_random_proxy() # Run command cmd += f' -cl -ct -rt -location -td -websocket -cname -asn -cdn -probe -random-agent' cmd += f' -t {threads}' if threads > 0 else '' cmd += f' --http-proxy {proxy}' if proxy else '' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -json' cmd += f' -u {urls[0]}' if len(urls) == 1 else f' -l {input_path}' cmd += f' -x {method}' if method else '' cmd += f' -silent' if follow_redirect: cmd += ' -fr' results = [] endpoint_ids = [] for line in stream_command( cmd, history_file=history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not line or not isinstance(line, dict): continue logger.debug(line) # No response from endpoint if line.get('failed', False): continue # Parse httpx output host = line.get('host', '') content_length = line.get('content_length', 0) http_status = line.get('status_code') http_url, is_redirect = extract_httpx_url(line) page_title = line.get('title') webserver = line.get('webserver') cdn = line.get('cdn', False) rt = line.get('time') techs = line.get('tech', []) cname = line.get('cname', '') content_type = line.get('content_type', '') response_time = -1 if rt: response_time = float(''.join(ch for ch in rt if not ch.isalpha())) if rt[-2:] == 'ms': response_time = response_time / 1000 # Create Subdomain object in DB subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue # Save default HTTP URL to endpoint object in DB endpoint, created = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain, is_default=is_ran_from_subdomain_scan ) if not endpoint: continue endpoint.http_status = http_status endpoint.page_title = page_title endpoint.content_length = content_length endpoint.webserver = webserver endpoint.response_time = response_time endpoint.content_type = content_type endpoint.save() endpoint_str = f'{http_url} [{http_status}] `{content_length}B` `{webserver}` `{rt}`' logger.warning(endpoint_str) if endpoint and endpoint.is_alive and endpoint.http_status != 403: self.notify( fields={'Alive endpoint': f'• {endpoint_str}'}, add_meta_info=False) # Add endpoint to results line['_cmd'] = cmd line['final_url'] = http_url line['endpoint_id'] = endpoint.id line['endpoint_created'] = created line['is_redirect'] = is_redirect results.append(line) # Add technology objects to DB for technology in techs: tech, _ = Technology.objects.get_or_create(name=technology) endpoint.techs.add(tech) if is_ran_from_subdomain_scan: subdomain.technologies.add(tech) subdomain.save() endpoint.save() techs_str = ', '.join([f'`{tech}`' for tech in techs]) self.notify( fields={'Technologies': techs_str}, add_meta_info=False) # Add IP objects for 'a' records to DB a_records = line.get('a', []) for ip_address in a_records: ip, created = save_ip_address( ip_address, subdomain, subscan=self.subscan, cdn=cdn) ips_str = '• ' + '\n• '.join([f'`{ip}`' for ip in a_records]) self.notify( fields={'IPs': ips_str}, add_meta_info=False) # Add IP object for host in DB if host: ip, created = save_ip_address( host, subdomain, subscan=self.subscan, cdn=cdn) self.notify( fields={'IPs': f'• `{ip.address}`'}, add_meta_info=False) # Save subdomain and endpoint if is_ran_from_subdomain_scan: # save subdomain stuffs subdomain.http_url = http_url subdomain.http_status = http_status subdomain.page_title = page_title subdomain.content_length = content_length subdomain.webserver = webserver subdomain.response_time = response_time subdomain.content_type = content_type subdomain.cname = ','.join(cname) subdomain.is_cdn = cdn if cdn: subdomain.cdn_name = line.get('cdn_name') subdomain.save() endpoint.save() endpoint_ids.append(endpoint.id) if should_remove_duplicate_endpoints: # Remove 'fake' alive endpoints that are just redirects to the same page remove_duplicate_endpoints( self.scan_id, self.domain_id, self.subdomain_id, filter_ids=endpoint_ids ) # Remove input file run_command( f'rm {input_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) return results #---------------------# # Notifications tasks # #---------------------# @app.task(name='send_notif', bind=False, queue='send_notif_queue') def send_notif( message, scan_history_id=None, subscan_id=None, **options): if not 'title' in options: message = enrich_notification(message, scan_history_id, subscan_id) send_discord_message(message, **options) send_slack_message(message) send_telegram_message(message) @app.task(name='send_scan_notif', bind=False, queue='send_scan_notif_queue') def send_scan_notif( scan_history_id, subscan_id=None, engine_id=None, status='RUNNING'): """Send scan status notification. Works for scan or a subscan if subscan_id is passed. Args: scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Get domain, engine, scan_history objects engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() tasks = ScanActivity.objects.filter(scan_of=scan) if scan else 0 # Build notif options url = get_scan_url(scan_history_id, subscan_id) title = get_scan_title(scan_history_id, subscan_id) fields = get_scan_fields(engine, scan, subscan, status, tasks) severity = None msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) if status: severity = STATUS_TO_SEVERITIES.get(status) opts = { 'title': title, 'url': url, 'fields': fields, 'severity': severity } logger.warning(f'Sending notification "{title}" [{severity}]') # Send notification send_notif( msg, scan_history_id, subscan_id, **opts) @app.task(name='send_task_notif', bind=False, queue='send_task_notif_queue') def send_task_notif( task_name, status=None, result=None, output_path=None, traceback=None, scan_history_id=None, engine_id=None, subscan_id=None, severity=None, add_meta_info=True, update_fields={}): """Send task status notification. Args: task_name (str): Task name. status (str, optional): Task status. result (str, optional): Task result. output_path (str, optional): Task output path. traceback (str, optional): Task traceback. scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. severity (str, optional): Severity (will be mapped to notif colors) add_meta_info (bool, optional): Wheter to add scan / subscan info to notif. update_fields (dict, optional): Fields key / value to update. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Build fields url = None fields = {} if add_meta_info: engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() url = get_scan_url(scan_history_id) if status: fields['Status'] = f'**{status}**' if engine: fields['Engine'] = engine.engine_name if scan: fields['Scan ID'] = f'[#{scan.id}]({url})' if subscan: url = get_scan_url(scan_history_id, subscan_id) fields['Subscan ID'] = f'[#{subscan.id}]({url})' title = get_task_title(task_name, scan_history_id, subscan_id) if status: severity = STATUS_TO_SEVERITIES.get(status) msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) # Add fields to update for k, v in update_fields.items(): fields[k] = v # Add traceback to notif if traceback and notif.send_scan_tracebacks: fields['Traceback'] = f'```\n{traceback}\n```' # Add files to notif files = [] attach_file = ( notif.send_scan_output_file and output_path and result and not traceback ) if attach_file: output_title = output_path.split('/')[-1] files = [(output_path, output_title)] # Send notif opts = { 'title': title, 'url': url, 'files': files, 'severity': severity, 'fields': fields, 'fields_append': update_fields.keys() } send_notif( msg, scan_history_id=scan_history_id, subscan_id=subscan_id, **opts) @app.task(name='send_file_to_discord', bind=False, queue='send_file_to_discord_queue') def send_file_to_discord(file_path, title=None): notif = Notification.objects.first() do_send = notif and notif.send_to_discord and notif.discord_hook_url if not do_send: return False webhook = DiscordWebhook( url=notif.discord_hook_url, rate_limit_retry=True, username=title or "reNgine Discord Plugin" ) with open(file_path, "rb") as f: head, tail = os.path.split(file_path) webhook.add_file(file=f.read(), filename=tail) webhook.execute() @app.task(name='send_hackerone_report', bind=False, queue='send_hackerone_report_queue') def send_hackerone_report(vulnerability_id): """Send HackerOne vulnerability report. Args: vulnerability_id (int): Vulnerability id. Returns: int: HTTP response status code. """ vulnerability = Vulnerability.objects.get(id=vulnerability_id) severities = {v: k for k,v in NUCLEI_SEVERITY_MAP.items()} headers = { 'Content-Type': 'application/json', 'Accept': 'application/json' } # can only send vulnerability report if team_handle exists if len(vulnerability.target_domain.h1_team_handle) !=0: hackerone_query = Hackerone.objects.all() if hackerone_query.exists(): hackerone = Hackerone.objects.first() severity_value = severities[vulnerability.severity] tpl = hackerone.report_template # Replace syntax of report template with actual content tpl = tpl.replace('{vulnerability_name}', vulnerability.name) tpl = tpl.replace('{vulnerable_url}', vulnerability.http_url) tpl = tpl.replace('{vulnerability_severity}', severity_value) tpl = tpl.replace('{vulnerability_description}', vulnerability.description if vulnerability.description else '') tpl = tpl.replace('{vulnerability_extracted_results}', vulnerability.extracted_results if vulnerability.extracted_results else '') tpl = tpl.replace('{vulnerability_reference}', vulnerability.reference if vulnerability.reference else '') data = { "data": { "type": "report", "attributes": { "team_handle": vulnerability.target_domain.h1_team_handle, "title": '{} found in {}'.format(vulnerability.name, vulnerability.http_url), "vulnerability_information": tpl, "severity_rating": severity_value, "impact": "More information about the impact and vulnerability can be found here: \n" + vulnerability.reference if vulnerability.reference else "NA", } } } r = requests.post( 'https://api.hackerone.com/v1/hackers/reports', auth=(hackerone.username, hackerone.api_key), json=data, headers=headers ) response = r.json() status_code = r.status_code if status_code == 201: vulnerability.hackerone_report_id = response['data']["id"] vulnerability.open_status = False vulnerability.save() return status_code else: logger.error('No team handle found.') status_code = 111 return status_code #-------------# # Utils tasks # #-------------# @app.task(name='parse_nmap_results', bind=False, queue='parse_nmap_results_queue') def parse_nmap_results(xml_file, output_file=None): """Parse results from nmap output file. Args: xml_file (str): nmap XML report file path. Returns: list: List of vulnerabilities found from nmap results. """ with open(xml_file, encoding='utf8') as f: content = f.read() try: nmap_results = xmltodict.parse(content) # parse XML to dict except Exception as e: logger.exception(e) logger.error(f'Cannot parse {xml_file} to valid JSON. Skipping.') return [] # Write JSON to output file if output_file: with open(output_file, 'w') as f: json.dump(nmap_results, f, indent=4) logger.warning(json.dumps(nmap_results, indent=4)) hosts = ( nmap_results .get('nmaprun', {}) .get('host', {}) ) all_vulns = [] if isinstance(hosts, dict): hosts = [hosts] for host in hosts: # Grab hostname / IP from output hostnames_dict = host.get('hostnames', {}) if hostnames_dict: # Ensure that hostnames['hostname'] is a list for consistency hostnames_list = hostnames_dict['hostname'] if isinstance(hostnames_dict['hostname'], list) else [hostnames_dict['hostname']] # Extract all the @name values from the list of dictionaries hostnames = [entry.get('@name') for entry in hostnames_list] else: hostnames = [host.get('address')['@addr']] # Iterate over each hostname for each port for hostname in hostnames: # Grab ports from output ports = host.get('ports', {}).get('port', []) if isinstance(ports, dict): ports = [ports] for port in ports: url_vulns = [] port_number = port['@portid'] url = sanitize_url(f'{hostname}:{port_number}') logger.info(f'Parsing nmap results for {hostname}:{port_number} ...') if not port_number or not port_number.isdigit(): continue port_protocol = port['@protocol'] scripts = port.get('script', []) if isinstance(scripts, dict): scripts = [scripts] for script in scripts: script_id = script['@id'] script_output = script['@output'] script_output_table = script.get('table', []) logger.debug(f'Ran nmap script "{script_id}" on {port_number}/{port_protocol}:\n{script_output}\n') if script_id == 'vulscan': vulns = parse_nmap_vulscan_output(script_output) url_vulns.extend(vulns) elif script_id == 'vulners': vulns = parse_nmap_vulners_output(script_output) url_vulns.extend(vulns) # elif script_id == 'http-server-header': # TODO: nmap can help find technologies as well using the http-server-header script # regex = r'(\w+)/([\d.]+)\s?(?:\((\w+)\))?' # tech_name, tech_version, tech_os = re.match(regex, test_string).groups() # Technology.objects.get_or_create(...) # elif script_id == 'http_csrf': # vulns = parse_nmap_http_csrf_output(script_output) # url_vulns.extend(vulns) else: logger.warning(f'Script output parsing for script "{script_id}" is not supported yet.') # Add URL to vuln for vuln in url_vulns: # TODO: This should extend to any URL, not just HTTP vuln['http_url'] = url if 'http_path' in vuln: vuln['http_url'] += vuln['http_path'] all_vulns.append(vuln) return all_vulns def parse_nmap_http_csrf_output(script_output): pass def parse_nmap_vulscan_output(script_output): """Parse nmap vulscan script output. Args: script_output (str): Vulscan script output. Returns: list: List of Vulnerability dicts. """ data = {} vulns = [] provider_name = '' # Sort all vulns found by provider so that we can match each provider with # a function that pulls from its API to get more info about the # vulnerability. for line in script_output.splitlines(): if not line: continue if not line.startswith('['): # provider line if "No findings" in line: logger.info(f"No findings: {line}") continue elif ' - ' in line: provider_name, provider_url = tuple(line.split(' - ')) data[provider_name] = {'url': provider_url.rstrip(':'), 'entries': []} continue else: # Log a warning logger.warning(f"Unexpected line format: {line}") continue reg = r'\[(.*)\] (.*)' matches = re.match(reg, line) id, title = matches.groups() entry = {'id': id, 'title': title} data[provider_name]['entries'].append(entry) logger.warning('Vulscan parsed output:') logger.warning(pprint.pformat(data)) for provider_name in data: if provider_name == 'Exploit-DB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'IBM X-Force': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'MITRE CVE': logger.error(f'Provider {provider_name} is not supported YET.') for entry in data[provider_name]['entries']: cve_id = entry['id'] vuln = cve_to_vuln(cve_id) vulns.append(vuln) elif provider_name == 'OSVDB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'OpenVAS (Nessus)': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'SecurityFocus': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'VulDB': logger.error(f'Provider {provider_name} is not supported YET.') pass else: logger.error(f'Provider {provider_name} is not supported.') return vulns def parse_nmap_vulners_output(script_output, url=''): """Parse nmap vulners script output. TODO: Rework this as it's currently matching all CVEs no matter the confidence. Args: script_output (str): Script output. Returns: list: List of found vulnerabilities. """ vulns = [] # Check for CVE in script output CVE_REGEX = re.compile(r'.*(CVE-\d\d\d\d-\d+).*') matches = CVE_REGEX.findall(script_output) matches = list(dict.fromkeys(matches)) for cve_id in matches: # get CVE info vuln = cve_to_vuln(cve_id, vuln_type='nmap-vulners-nse') if vuln: vulns.append(vuln) return vulns def cve_to_vuln(cve_id, vuln_type=''): """Search for a CVE using CVESearch and return Vulnerability data. Args: cve_id (str): CVE ID in the form CVE-* Returns: dict: Vulnerability dict. """ cve_info = CVESearch('https://cve.circl.lu').id(cve_id) if not cve_info: logger.error(f'Could not fetch CVE info for cve {cve_id}. Skipping.') return None vuln_cve_id = cve_info['id'] vuln_name = vuln_cve_id vuln_description = cve_info.get('summary', 'none').replace(vuln_cve_id, '').strip() try: vuln_cvss = float(cve_info.get('cvss', -1)) except (ValueError, TypeError): vuln_cvss = -1 vuln_cwe_id = cve_info.get('cwe', '') exploit_ids = cve_info.get('refmap', {}).get('exploit-db', []) osvdb_ids = cve_info.get('refmap', {}).get('osvdb', []) references = cve_info.get('references', []) capec_objects = cve_info.get('capec', []) # Parse ovals for a better vuln name / type ovals = cve_info.get('oval', []) if ovals: vuln_name = ovals[0]['title'] vuln_type = ovals[0]['family'] # Set vulnerability severity based on CVSS score vuln_severity = 'info' if vuln_cvss < 4: vuln_severity = 'low' elif vuln_cvss < 7: vuln_severity = 'medium' elif vuln_cvss < 9: vuln_severity = 'high' else: vuln_severity = 'critical' # Build console warning message msg = f'{vuln_name} | {vuln_severity.upper()} | {vuln_cve_id} | {vuln_cwe_id} | {vuln_cvss}' for id in osvdb_ids: msg += f'\n\tOSVDB: {id}' for exploit_id in exploit_ids: msg += f'\n\tEXPLOITDB: {exploit_id}' logger.warning(msg) vuln = { 'name': vuln_name, 'type': vuln_type, 'severity': NUCLEI_SEVERITY_MAP[vuln_severity], 'description': vuln_description, 'cvss_score': vuln_cvss, 'references': references, 'cve_ids': [vuln_cve_id], 'cwe_ids': [vuln_cwe_id] } return vuln def parse_s3scanner_result(line): ''' Parses and returns s3Scanner Data ''' bucket = line['bucket'] return { 'name': bucket['name'], 'region': bucket['region'], 'provider': bucket['provider'], 'owner_display_name': bucket['owner_display_name'], 'owner_id': bucket['owner_id'], 'perm_auth_users_read': bucket['perm_auth_users_read'], 'perm_auth_users_write': bucket['perm_auth_users_write'], 'perm_auth_users_read_acl': bucket['perm_auth_users_read_acl'], 'perm_auth_users_write_acl': bucket['perm_auth_users_write_acl'], 'perm_auth_users_full_control': bucket['perm_auth_users_full_control'], 'perm_all_users_read': bucket['perm_all_users_read'], 'perm_all_users_write': bucket['perm_all_users_write'], 'perm_all_users_read_acl': bucket['perm_all_users_read_acl'], 'perm_all_users_write_acl': bucket['perm_all_users_write_acl'], 'perm_all_users_full_control': bucket['perm_all_users_full_control'], 'num_objects': bucket['num_objects'], 'size': bucket['bucket_size'] } def parse_nuclei_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ return { 'name': line['info'].get('name', ''), 'type': line['type'], 'severity': NUCLEI_SEVERITY_MAP[line['info'].get('severity', 'unknown')], 'template': line['template'], 'template_url': line['template-url'], 'template_id': line['template-id'], 'description': line['info'].get('description', ''), 'matcher_name': line.get('matcher-name', ''), 'curl_command': line.get('curl-command'), 'request': line.get('request'), 'response': line.get('response'), 'extracted_results': line.get('extracted-results', []), 'cvss_metrics': line['info'].get('classification', {}).get('cvss-metrics', ''), 'cvss_score': line['info'].get('classification', {}).get('cvss-score'), 'cve_ids': line['info'].get('classification', {}).get('cve_id', []) or [], 'cwe_ids': line['info'].get('classification', {}).get('cwe_id', []) or [], 'references': line['info'].get('reference', []) or [], 'tags': line['info'].get('tags', []), 'source': NUCLEI, } def parse_dalfox_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ description = '' description += f" Evidence: {line.get('evidence')} <br>" if line.get('evidence') else '' description += f" Message: {line.get('message')} <br>" if line.get('message') else '' description += f" Payload: {line.get('message_str')} <br>" if line.get('message_str') else '' description += f" Vulnerable Parameter: {line.get('param')} <br>" if line.get('param') else '' return { 'name': 'XSS (Cross Site Scripting)', 'type': 'XSS', 'severity': DALFOX_SEVERITY_MAP[line.get('severity', 'unknown')], 'description': description, 'source': DALFOX, 'cwe_ids': [line.get('cwe')] } def parse_crlfuzz_result(url): """Parse CRLF results Args: url (str): CRLF Vulnerable URL Returns: dict: Vulnerability data. """ return { 'name': 'CRLF (HTTP Response Splitting)', 'type': 'CRLF', 'severity': 2, 'description': 'A CRLF (HTTP Response Splitting) vulnerability has been discovered.', 'source': CRLFUZZ, } def record_exists(model, data, exclude_keys=[]): """ Check if a record already exists in the database based on the given data. Args: model (django.db.models.Model): The Django model to check against. data (dict): Data dictionary containing fields and values. exclude_keys (list): List of keys to exclude from the lookup. Returns: bool: True if the record exists, False otherwise. """ # Extract the keys that will be used for the lookup lookup_fields = {key: data[key] for key in data if key not in exclude_keys} # Return True if a record exists based on the lookup fields, False otherwise return model.objects.filter(**lookup_fields).exists() @app.task(name='geo_localize', bind=False, queue='geo_localize_queue') def geo_localize(host, ip_id=None): """Uses geoiplookup to find location associated with host. Args: host (str): Hostname. ip_id (int): IpAddress object id. Returns: startScan.models.CountryISO: CountryISO object from DB or None. """ if validators.ipv6(host): logger.info(f'Ipv6 "{host}" is not supported by geoiplookup. Skipping.') return None cmd = f'geoiplookup {host}' _, out = run_command(cmd) if 'IP Address not found' not in out and "can't resolve hostname" not in out: country_iso = out.split(':')[1].strip().split(',')[0] country_name = out.split(':')[1].strip().split(',')[1].strip() geo_object, _ = CountryISO.objects.get_or_create( iso=country_iso, name=country_name ) geo_json = { 'iso': country_iso, 'name': country_name } if ip_id: ip = IpAddress.objects.get(pk=ip_id) ip.geo_iso = geo_object ip.save() return geo_json logger.info(f'Geo IP lookup failed for host "{host}"') return None @app.task(name='query_whois', bind=False, queue='query_whois_queue') def query_whois(ip_domain, force_reload_whois=False): """Query WHOIS information for an IP or a domain name. Args: ip_domain (str): IP address or domain name. save_domain (bool): Whether to save domain or not, default False Returns: dict: WHOIS information. """ if not force_reload_whois and Domain.objects.filter(name=ip_domain).exists() and Domain.objects.get(name=ip_domain).domain_info: domain = Domain.objects.get(name=ip_domain) if not domain.insert_date: domain.insert_date = timezone.now() domain.save() domain_info_db = domain.domain_info domain_info = DottedDict( dnssec=domain_info_db.dnssec, created=domain_info_db.created, updated=domain_info_db.updated, expires=domain_info_db.expires, geolocation_iso=domain_info_db.geolocation_iso, status=[status['name'] for status in DomainWhoisStatusSerializer(domain_info_db.status, many=True).data], whois_server=domain_info_db.whois_server, ns_records=[ns['name'] for ns in NameServersSerializer(domain_info_db.name_servers, many=True).data], registrar_name=domain_info_db.registrar.name, registrar_phone=domain_info_db.registrar.phone, registrar_email=domain_info_db.registrar.email, registrar_url=domain_info_db.registrar.url, registrant_name=domain_info_db.registrant.name, registrant_id=domain_info_db.registrant.id_str, registrant_organization=domain_info_db.registrant.organization, registrant_city=domain_info_db.registrant.city, registrant_state=domain_info_db.registrant.state, registrant_zip_code=domain_info_db.registrant.zip_code, registrant_country=domain_info_db.registrant.country, registrant_phone=domain_info_db.registrant.phone, registrant_fax=domain_info_db.registrant.fax, registrant_email=domain_info_db.registrant.email, registrant_address=domain_info_db.registrant.address, admin_name=domain_info_db.admin.name, admin_id=domain_info_db.admin.id_str, admin_organization=domain_info_db.admin.organization, admin_city=domain_info_db.admin.city, admin_state=domain_info_db.admin.state, admin_zip_code=domain_info_db.admin.zip_code, admin_country=domain_info_db.admin.country, admin_phone=domain_info_db.admin.phone, admin_fax=domain_info_db.admin.fax, admin_email=domain_info_db.admin.email, admin_address=domain_info_db.admin.address, tech_name=domain_info_db.tech.name, tech_id=domain_info_db.tech.id_str, tech_organization=domain_info_db.tech.organization, tech_city=domain_info_db.tech.city, tech_state=domain_info_db.tech.state, tech_zip_code=domain_info_db.tech.zip_code, tech_country=domain_info_db.tech.country, tech_phone=domain_info_db.tech.phone, tech_fax=domain_info_db.tech.fax, tech_email=domain_info_db.tech.email, tech_address=domain_info_db.tech.address, related_tlds=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_tlds, many=True).data], related_domains=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_domains, many=True).data], historical_ips=[ip for ip in HistoricalIPSerializer(domain_info_db.historical_ips, many=True).data], ) if domain_info_db.dns_records: a_records = [] txt_records = [] mx_records = [] dns_records = [{'name': dns['name'], 'type': dns['type']} for dns in DomainDNSRecordSerializer(domain_info_db.dns_records, many=True).data] for dns in dns_records: if dns['type'] == 'a': a_records.append(dns['name']) elif dns['type'] == 'txt': txt_records.append(dns['name']) elif dns['type'] == 'mx': mx_records.append(dns['name']) domain_info.a_records = a_records domain_info.txt_records = txt_records domain_info.mx_records = mx_records else: logger.info(f'Domain info for "{ip_domain}" not found in DB, querying whois') domain_info = DottedDict() # find domain historical ip try: historical_ips = get_domain_historical_ip_address(ip_domain) domain_info.historical_ips = historical_ips except Exception as e: logger.error(f'HistoricalIP for {ip_domain} not found!\nError: {str(e)}') historical_ips = [] # find associated domains using ip_domain try: related_domains = reverse_whois(ip_domain.split('.')[0]) except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] # find related tlds using TLSx try: related_tlds = [] output_path = '/tmp/ip_domain_tlsx.txt' tlsx_command = f'tlsx -san -cn -silent -ro -host {ip_domain} -o {output_path}' run_command( tlsx_command, shell=True, ) tlsx_output = [] with open(output_path) as f: tlsx_output = f.readlines() tldextract_target = tldextract.extract(ip_domain) for doms in tlsx_output: doms = doms.strip() tldextract_res = tldextract.extract(doms) if ip_domain != doms and tldextract_res.domain == tldextract_target.domain and tldextract_res.subdomain == '': related_tlds.append(doms) related_tlds = list(set(related_tlds)) domain_info.related_tlds = related_tlds except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] related_domains_list = [] if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_related = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_related) related_domains_list.append(_domain['name']) for _domain in related_tlds: domain_related = RelatedDomain.objects.get_or_create( name=_domain, )[0] db_domain_info.related_tlds.add(domain_related) for _ip in historical_ips: historical_ip = HistoricalIP.objects.get_or_create( ip=_ip['ip'], owner=_ip['owner'], location=_ip['location'], last_seen=_ip['last_seen'], )[0] db_domain_info.historical_ips.add(historical_ip) domain.domain_info = db_domain_info domain.save() command = f'netlas host {ip_domain} -f json' # check if netlas key is provided netlas_key = get_netlas_key() command += f' -a {netlas_key}' if netlas_key else '' result = subprocess.check_output(command.split()).decode('utf-8') if 'Failed to parse response data' in result: # do fallback return { 'status': False, 'ip_domain': ip_domain, 'result': "Netlas limit exceeded.", 'message': 'Netlas limit exceeded.' } try: result = json.loads(result) logger.info(result) whois = result.get('whois') if result.get('whois') else {} domain_info.created = whois.get('created_date') domain_info.expires = whois.get('expiration_date') domain_info.updated = whois.get('updated_date') domain_info.whois_server = whois.get('whois_server') if 'registrant' in whois: registrant = whois.get('registrant') domain_info.registrant_name = registrant.get('name') domain_info.registrant_country = registrant.get('country') domain_info.registrant_id = registrant.get('id') domain_info.registrant_state = registrant.get('province') domain_info.registrant_city = registrant.get('city') domain_info.registrant_phone = registrant.get('phone') domain_info.registrant_address = registrant.get('street') domain_info.registrant_organization = registrant.get('organization') domain_info.registrant_fax = registrant.get('fax') domain_info.registrant_zip_code = registrant.get('postal_code') email_search = EMAIL_REGEX.search(str(registrant.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.registrant_email = field_content if 'administrative' in whois: administrative = whois.get('administrative') domain_info.admin_name = administrative.get('name') domain_info.admin_country = administrative.get('country') domain_info.admin_id = administrative.get('id') domain_info.admin_state = administrative.get('province') domain_info.admin_city = administrative.get('city') domain_info.admin_phone = administrative.get('phone') domain_info.admin_address = administrative.get('street') domain_info.admin_organization = administrative.get('organization') domain_info.admin_fax = administrative.get('fax') domain_info.admin_zip_code = administrative.get('postal_code') mail_search = EMAIL_REGEX.search(str(administrative.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.admin_email = field_content if 'technical' in whois: technical = whois.get('technical') domain_info.tech_name = technical.get('name') domain_info.tech_country = technical.get('country') domain_info.tech_state = technical.get('province') domain_info.tech_id = technical.get('id') domain_info.tech_city = technical.get('city') domain_info.tech_phone = technical.get('phone') domain_info.tech_address = technical.get('street') domain_info.tech_organization = technical.get('organization') domain_info.tech_fax = technical.get('fax') domain_info.tech_zip_code = technical.get('postal_code') mail_search = EMAIL_REGEX.search(str(technical.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.tech_email = field_content if 'dns' in result: dns = result.get('dns') domain_info.mx_records = dns.get('mx') domain_info.txt_records = dns.get('txt') domain_info.a_records = dns.get('a') domain_info.ns_records = whois.get('name_servers') domain_info.dnssec = True if whois.get('dnssec') else False domain_info.status = whois.get('status') if 'registrar' in whois: registrar = whois.get('registrar') domain_info.registrar_name = registrar.get('name') domain_info.registrar_email = registrar.get('email') domain_info.registrar_phone = registrar.get('phone') domain_info.registrar_url = registrar.get('url') # find associated domains if registrant email is found related_domains = reverse_whois(domain_info.get('registrant_email')) if domain_info.get('registrant_email') else [] for _domain in related_domains: related_domains_list.append(_domain['name']) # remove duplicate domains from related domains list related_domains_list = list(set(related_domains_list)) domain_info.related_domains = related_domains_list # save to db if domain exists if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_rel = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_rel) db_domain_info.dnssec = domain_info.get('dnssec') #dates db_domain_info.created = domain_info.get('created') db_domain_info.updated = domain_info.get('updated') db_domain_info.expires = domain_info.get('expires') #registrar db_domain_info.registrar = Registrar.objects.get_or_create( name=domain_info.get('registrar_name'), email=domain_info.get('registrar_email'), phone=domain_info.get('registrar_phone'), url=domain_info.get('registrar_url'), )[0] db_domain_info.registrant = DomainRegistration.objects.get_or_create( name=domain_info.get('registrant_name'), organization=domain_info.get('registrant_organization'), address=domain_info.get('registrant_address'), city=domain_info.get('registrant_city'), state=domain_info.get('registrant_state'), zip_code=domain_info.get('registrant_zip_code'), country=domain_info.get('registrant_country'), email=domain_info.get('registrant_email'), phone=domain_info.get('registrant_phone'), fax=domain_info.get('registrant_fax'), id_str=domain_info.get('registrant_id'), )[0] db_domain_info.admin = DomainRegistration.objects.get_or_create( name=domain_info.get('admin_name'), organization=domain_info.get('admin_organization'), address=domain_info.get('admin_address'), city=domain_info.get('admin_city'), state=domain_info.get('admin_state'), zip_code=domain_info.get('admin_zip_code'), country=domain_info.get('admin_country'), email=domain_info.get('admin_email'), phone=domain_info.get('admin_phone'), fax=domain_info.get('admin_fax'), id_str=domain_info.get('admin_id'), )[0] db_domain_info.tech = DomainRegistration.objects.get_or_create( name=domain_info.get('tech_name'), organization=domain_info.get('tech_organization'), address=domain_info.get('tech_address'), city=domain_info.get('tech_city'), state=domain_info.get('tech_state'), zip_code=domain_info.get('tech_zip_code'), country=domain_info.get('tech_country'), email=domain_info.get('tech_email'), phone=domain_info.get('tech_phone'), fax=domain_info.get('tech_fax'), id_str=domain_info.get('tech_id'), )[0] for status in domain_info.get('status') or []: _status = WhoisStatus.objects.get_or_create( name=status )[0] _status.save() db_domain_info.status.add(_status) for ns in domain_info.get('ns_records') or []: _ns = NameServer.objects.get_or_create( name=ns )[0] _ns.save() db_domain_info.name_servers.add(_ns) for a in domain_info.get('a_records') or []: _a = DNSRecord.objects.get_or_create( name=a, type='a' )[0] _a.save() db_domain_info.dns_records.add(_a) for mx in domain_info.get('mx_records') or []: _mx = DNSRecord.objects.get_or_create( name=mx, type='mx' )[0] _mx.save() db_domain_info.dns_records.add(_mx) for txt in domain_info.get('txt_records') or []: _txt = DNSRecord.objects.get_or_create( name=txt, type='txt' )[0] _txt.save() db_domain_info.dns_records.add(_txt) db_domain_info.geolocation_iso = domain_info.get('registrant_country') db_domain_info.whois_server = domain_info.get('whois_server') db_domain_info.save() domain.domain_info = db_domain_info domain.save() except Exception as e: return { 'status': False, 'ip_domain': ip_domain, 'result': "unable to fetch records from WHOIS database.", 'message': str(e) } return { 'status': True, 'ip_domain': ip_domain, 'dnssec': domain_info.get('dnssec'), 'created': domain_info.get('created'), 'updated': domain_info.get('updated'), 'expires': domain_info.get('expires'), 'geolocation_iso': domain_info.get('registrant_country'), 'domain_statuses': domain_info.get('status'), 'whois_server': domain_info.get('whois_server'), 'dns': { 'a': domain_info.get('a_records'), 'mx': domain_info.get('mx_records'), 'txt': domain_info.get('txt_records'), }, 'registrar': { 'name': domain_info.get('registrar_name'), 'phone': domain_info.get('registrar_phone'), 'email': domain_info.get('registrar_email'), 'url': domain_info.get('registrar_url'), }, 'registrant': { 'name': domain_info.get('registrant_name'), 'id': domain_info.get('registrant_id'), 'organization': domain_info.get('registrant_organization'), 'address': domain_info.get('registrant_address'), 'city': domain_info.get('registrant_city'), 'state': domain_info.get('registrant_state'), 'zipcode': domain_info.get('registrant_zip_code'), 'country': domain_info.get('registrant_country'), 'phone': domain_info.get('registrant_phone'), 'fax': domain_info.get('registrant_fax'), 'email': domain_info.get('registrant_email'), }, 'admin': { 'name': domain_info.get('admin_name'), 'id': domain_info.get('admin_id'), 'organization': domain_info.get('admin_organization'), 'address':domain_info.get('admin_address'), 'city': domain_info.get('admin_city'), 'state': domain_info.get('admin_state'), 'zipcode': domain_info.get('admin_zip_code'), 'country': domain_info.get('admin_country'), 'phone': domain_info.get('admin_phone'), 'fax': domain_info.get('admin_fax'), 'email': domain_info.get('admin_email'), }, 'technical_contact': { 'name': domain_info.get('tech_name'), 'id': domain_info.get('tech_id'), 'organization': domain_info.get('tech_organization'), 'address': domain_info.get('tech_address'), 'city': domain_info.get('tech_city'), 'state': domain_info.get('tech_state'), 'zipcode': domain_info.get('tech_zip_code'), 'country': domain_info.get('tech_country'), 'phone': domain_info.get('tech_phone'), 'fax': domain_info.get('tech_fax'), 'email': domain_info.get('tech_email'), }, 'nameservers': domain_info.get('ns_records'), # 'similar_domains': domain_info.get('similar_domains'), 'related_domains': domain_info.get('related_domains'), 'related_tlds': domain_info.get('related_tlds'), 'historical_ips': domain_info.get('historical_ips'), } @app.task(name='remove_duplicate_endpoints', bind=False, queue='remove_duplicate_endpoints_queue') def remove_duplicate_endpoints( scan_history_id, domain_id, subdomain_id=None, filter_ids=[], filter_status=[200, 301, 404], duplicate_removal_fields=ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS ): """Remove duplicate endpoints. Check for implicit redirections by comparing endpoints: - [x] `content_length` similarities indicating redirections - [x] `page_title` (check for same page title) - [ ] Sign-in / login page (check for endpoints with the same words) Args: scan_history_id: ScanHistory id. domain_id (int): Domain id. subdomain_id (int, optional): Subdomain id. filter_ids (list): List of endpoint ids to filter on. filter_status (list): List of HTTP status codes to filter on. duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates """ logger.info(f'Removing duplicate endpoints based on {duplicate_removal_fields}') endpoints = ( EndPoint.objects .filter(scan_history__id=scan_history_id) .filter(target_domain__id=domain_id) ) if filter_status: endpoints = endpoints.filter(http_status__in=filter_status) if subdomain_id: endpoints = endpoints.filter(subdomain__id=subdomain_id) if filter_ids: endpoints = endpoints.filter(id__in=filter_ids) for field_name in duplicate_removal_fields: cl_query = ( endpoints .values_list(field_name) .annotate(mc=Count(field_name)) .order_by('-mc') ) for (field_value, count) in cl_query: if count > DELETE_DUPLICATES_THRESHOLD: eps_to_delete = ( endpoints .filter(**{field_name: field_value}) .order_by('discovered_date') .all()[1:] ) msg = f'Deleting {len(eps_to_delete)} endpoints [reason: same {field_name} {field_value}]' for ep in eps_to_delete: url = urlparse(ep.http_url) if url.path in ['', '/', '/login']: # try do not delete the original page that other pages redirect to continue msg += f'\n\t {ep.http_url} [{ep.http_status}] [{field_name}={field_value}]' ep.delete() logger.warning(msg) @app.task(name='run_command', bind=False, queue='run_command_queue') def run_command(cmd, cwd=None, shell=False, history_file=None, scan_id=None, activity_id=None): """Run a given command using subprocess module. Args: cmd (str): Command to run. cwd (str): Current working directory. echo (bool): Log command. shell (bool): Run within separate shell if True. history_file (str): Write command + output to history file. Returns: tuple: Tuple with return_code, output. """ logger.info(cmd) logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Run the command using subprocess popen = subprocess.Popen( cmd if shell else cmd.split(), shell=shell, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, cwd=cwd, universal_newlines=True) output = '' for stdout_line in iter(popen.stdout.readline, ""): item = stdout_line.strip() output += '\n' + item logger.debug(item) popen.stdout.close() popen.wait() return_code = popen.returncode command_obj.output = output command_obj.return_code = return_code command_obj.save() if history_file: mode = 'a' if not os.path.exists(history_file): mode = 'w' with open(history_file, mode) as f: f.write(f'\n{cmd}\n{return_code}\n{output}\n------------------\n') return return_code, output #-------------# # Other utils # #-------------# def stream_command(cmd, cwd=None, shell=False, history_file=None, encoding='utf-8', scan_id=None, activity_id=None, trunc_char=None): # Log cmd logger.info(cmd) # logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Sanitize the cmd command = cmd if shell else cmd.split() # Run the command using subprocess process = subprocess.Popen( command, stdout=subprocess.PIPE, stderr=subprocess.PIPE, shell=shell) # Log the output in real-time to the database output = "" # Process the output for line in iter(lambda: process.stdout.readline() or process.stderr.readline(), b''): line = re.sub(r'\x1b[^m]*m', '', line.decode('utf-8').strip()) if trunc_char and line.endswith(trunc_char): line = line[:-1] item = line # Try to parse the line as JSON try: item = json.loads(line) except json.JSONDecodeError: pass # Yield the line #logger.debug(item) yield item # Add the log line to the output output += line + "\n" # Update the command record in the database command_obj.output = output command_obj.save() # Retrieve the return code and output process.wait() return_code = process.returncode # Update the return code and final output in the database command_obj.return_code = return_code command_obj.save() # Append the command, return code and output to the history file if history_file is not None: with open(history_file, "a") as f: f.write(f"{cmd}\n{return_code}\n{output}\n") def process_httpx_response(line): """TODO: implement this""" def extract_httpx_url(line): """Extract final URL from httpx results. Always follow redirects to find the last URL. Args: line (dict): URL data output by httpx. Returns: tuple: (final_url, redirect_bool) tuple. """ status_code = line.get('status_code', 0) final_url = line.get('final_url') location = line.get('location') chain_status_codes = line.get('chain_status_codes', []) # Final URL is already looking nice, if it exists return it if final_url: return final_url, False http_url = line['url'] # fallback to url field # Handle redirects manually REDIRECT_STATUS_CODES = [301, 302] is_redirect = ( status_code in REDIRECT_STATUS_CODES or any(x in REDIRECT_STATUS_CODES for x in chain_status_codes) ) if is_redirect and location: if location.startswith(('http', 'https')): http_url = location else: http_url = f'{http_url}/{location.lstrip("/")}' # Sanitize URL http_url = sanitize_url(http_url) return http_url, is_redirect #-------------# # OSInt utils # #-------------# def get_and_save_dork_results(lookup_target, results_dir, type, lookup_keywords=None, lookup_extensions=None, delay=3, page_count=2, scan_history=None): """ Uses gofuzz to dork and store information Args: lookup_target (str): target to look into such as stackoverflow or even the target itself results_dir (str): Results directory type (str): Dork Type Title lookup_keywords (str): comma separated keywords or paths to look for lookup_extensions (str): comma separated extensions to look for delay (int): delay between each requests page_count (int): pages in google to extract information scan_history (startScan.ScanHistory): Scan History Object """ results = [] gofuzz_command = f'{GOFUZZ_EXEC_PATH} -t {lookup_target} -d {delay} -p {page_count}' if lookup_extensions: gofuzz_command += f' -e {lookup_extensions}' elif lookup_keywords: gofuzz_command += f' -w {lookup_keywords}' output_file = f'{results_dir}/gofuzz.txt' gofuzz_command += f' -o {output_file}' history_file = f'{results_dir}/commands.txt' try: run_command( gofuzz_command, shell=False, history_file=history_file, scan_id=scan_history.id, ) if not os.path.isfile(output_file): return with open(output_file) as f: for line in f.readlines(): url = line.strip() if url: results.append(url) dork, created = Dork.objects.get_or_create( type=type, url=url ) if scan_history: scan_history.dorks.add(dork) # remove output file os.remove(output_file) except Exception as e: logger.exception(e) return results def get_and_save_emails(scan_history, activity_id, results_dir): """Get and save emails from Google, Bing and Baidu. Args: scan_history (startScan.ScanHistory): Scan history object. activity_id: ScanActivity Object results_dir (str): Results directory. Returns: list: List of emails found. """ emails = [] # Proxy settings # get_random_proxy() # Gather emails from Google, Bing and Baidu output_file = f'{results_dir}/emails_tmp.txt' history_file = f'{results_dir}/commands.txt' command = f'python3 /usr/src/github/Infoga/infoga.py --domain {scan_history.domain.name} --source all --report {output_file}' try: run_command( command, shell=False, history_file=history_file, scan_id=scan_history.id, activity_id=activity_id) if not os.path.isfile(output_file): logger.info('No Email results') return [] with open(output_file) as f: for line in f.readlines(): if 'Email' in line: split_email = line.split(' ')[2] emails.append(split_email) output_path = f'{results_dir}/emails.txt' with open(output_path, 'w') as output_file: for email_address in emails: save_email(email_address, scan_history) output_file.write(f'{email_address}\n') except Exception as e: logger.exception(e) return emails def save_metadata_info(meta_dict): """Extract metadata from Google Search. Args: meta_dict (dict): Info dict. Returns: list: List of startScan.MetaFinderDocument objects. """ logger.warning(f'Getting metadata for {meta_dict.osint_target}') scan_history = ScanHistory.objects.get(id=meta_dict.scan_id) # Proxy settings get_random_proxy() # Get metadata result = extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if not result: logger.error(f'No metadata result from Google Search for {meta_dict.osint_target}.') return [] # Add metadata info to DB results = [] for metadata_name, data in result.get_metadata().items(): subdomain = Subdomain.objects.get( scan_history=meta_dict.scan_id, name=meta_dict.osint_target) metadata = DottedDict({k: v for k, v in data.items()}) meta_finder_document = MetaFinderDocument( subdomain=subdomain, target_domain=meta_dict.domain, scan_history=scan_history, url=metadata.url, doc_name=metadata_name, http_status=metadata.status_code, producer=metadata.metadata.get('Producer'), creator=metadata.metadata.get('Creator'), creation_date=metadata.metadata.get('CreationDate'), modified_date=metadata.metadata.get('ModDate'), author=metadata.metadata.get('Author'), title=metadata.metadata.get('Title'), os=metadata.metadata.get('OSInfo')) meta_finder_document.save() results.append(data) return results #-----------------# # Utils functions # #-----------------# def create_scan_activity(scan_history_id, message, status): scan_activity = ScanActivity() scan_activity.scan_of = ScanHistory.objects.get(pk=scan_history_id) scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id #--------------------# # Database functions # #--------------------# def save_vulnerability(**vuln_data): references = vuln_data.pop('references', []) cve_ids = vuln_data.pop('cve_ids', []) cwe_ids = vuln_data.pop('cwe_ids', []) tags = vuln_data.pop('tags', []) subscan = vuln_data.pop('subscan', None) # remove nulls vuln_data = replace_nulls(vuln_data) # Create vulnerability vuln, created = Vulnerability.objects.get_or_create(**vuln_data) if created: vuln.discovered_date = timezone.now() vuln.open_status = True vuln.save() # Save vuln tags for tag_name in tags or []: tag, created = VulnerabilityTags.objects.get_or_create(name=tag_name) if tag: vuln.tags.add(tag) vuln.save() # Save CVEs for cve_id in cve_ids or []: cve, created = CveId.objects.get_or_create(name=cve_id) if cve: vuln.cve_ids.add(cve) vuln.save() # Save CWEs for cve_id in cwe_ids or []: cwe, created = CweId.objects.get_or_create(name=cve_id) if cwe: vuln.cwe_ids.add(cwe) vuln.save() # Save vuln reference for url in references or []: ref, created = VulnerabilityReference.objects.get_or_create(url=url) if created: vuln.references.add(ref) vuln.save() # Save subscan id in vuln object if subscan: vuln.vuln_subscan_ids.add(subscan) vuln.save() return vuln, created def save_endpoint( http_url, ctx={}, crawl=False, is_default=False, **endpoint_data): """Get or create EndPoint object. If crawl is True, also crawl the endpoint HTTP URL with httpx. Args: http_url (str): Input HTTP URL. is_default (bool): If the url is a default url for SubDomains. scan_history (startScan.models.ScanHistory): ScanHistory object. domain (startScan.models.Domain): Domain object. subdomain (starScan.models.Subdomain): Subdomain object. results_dir (str, optional): Results directory. crawl (bool, optional): Run httpx on endpoint if True. Default: False. force (bool, optional): Force crawl even if ENABLE_HTTP_CRAWL mode is on. subscan (startScan.models.SubScan, optional): SubScan object. Returns: tuple: (startScan.models.EndPoint, created) where `created` is a boolean indicating if the object is new or already existed. """ # remove nulls endpoint_data = replace_nulls(endpoint_data) scheme = urlparse(http_url).scheme endpoint = None created = False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in http_url: logger.error(f"{http_url} is not a URL of domain {domain.name}. Skipping.") return None, False if crawl: ctx['track'] = False results = http_crawl( urls=[http_url], method='HEAD', ctx=ctx) if results: endpoint_data = results[0] endpoint_id = endpoint_data['endpoint_id'] created = endpoint_data['endpoint_created'] endpoint = EndPoint.objects.get(pk=endpoint_id) elif not scheme: return None, False else: # add dumb endpoint without probing it scan = ScanHistory.objects.filter(pk=ctx.get('scan_history_id')).first() domain = Domain.objects.filter(pk=ctx.get('domain_id')).first() if not validators.url(http_url): return None, False http_url = sanitize_url(http_url) endpoint, created = EndPoint.objects.get_or_create( scan_history=scan, target_domain=domain, http_url=http_url, **endpoint_data) if created: endpoint.is_default = is_default endpoint.discovered_date = timezone.now() endpoint.save() subscan_id = ctx.get('subscan_id') if subscan_id: endpoint.endpoint_subscan_ids.add(subscan_id) endpoint.save() return endpoint, created def save_subdomain(subdomain_name, ctx={}): """Get or create Subdomain object. Args: subdomain_name (str): Subdomain name. scan_history (startScan.models.ScanHistory): ScanHistory object. Returns: tuple: (startScan.models.Subdomain, created) where `created` is a boolean indicating if the object has been created in DB. """ scan_id = ctx.get('scan_history_id') subscan_id = ctx.get('subscan_id') out_of_scope_subdomains = ctx.get('out_of_scope_subdomains', []) valid_domain = ( validators.domain(subdomain_name) or validators.ipv4(subdomain_name) or validators.ipv6(subdomain_name) ) if not valid_domain: logger.error(f'{subdomain_name} is not an invalid domain. Skipping.') return None, False if subdomain_name in out_of_scope_subdomains: logger.error(f'{subdomain_name} is out-of-scope. Skipping.') return None, False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in subdomain_name: logger.error(f"{subdomain_name} is not a subdomain of domain {domain.name}. Skipping.") return None, False scan = ScanHistory.objects.filter(pk=scan_id).first() domain = scan.domain if scan else None subdomain, created = Subdomain.objects.get_or_create( scan_history=scan, target_domain=domain, name=subdomain_name) if created: # logger.warning(f'Found new subdomain {subdomain_name}') subdomain.discovered_date = timezone.now() if subscan_id: subdomain.subdomain_subscan_ids.add(subscan_id) subdomain.save() return subdomain, created def save_email(email_address, scan_history=None): if not validators.email(email_address): logger.info(f'Email {email_address} is invalid. Skipping.') return None, False email, created = Email.objects.get_or_create(address=email_address) # if created: # logger.warning(f'Found new email address {email_address}') # Add email to ScanHistory if scan_history: scan_history.emails.add(email) scan_history.save() return email, created def save_employee(name, designation, scan_history=None): employee, created = Employee.objects.get_or_create( name=name, designation=designation) # if created: # logger.warning(f'Found new employee {name}') # Add employee to ScanHistory if scan_history: scan_history.employees.add(employee) scan_history.save() return employee, created def save_ip_address(ip_address, subdomain=None, subscan=None, **kwargs): if not (validators.ipv4(ip_address) or validators.ipv6(ip_address)): logger.info(f'IP {ip_address} is not a valid IP. Skipping.') return None, False ip, created = IpAddress.objects.get_or_create(address=ip_address) # if created: # logger.warning(f'Found new IP {ip_address}') # Set extra attributes for key, value in kwargs.items(): setattr(ip, key, value) ip.save() # Add IP to subdomain if subdomain: subdomain.ip_addresses.add(ip) subdomain.save() # Add subscan to IP if subscan: ip.ip_subscan_ids.add(subscan) # Geo-localize IP asynchronously if created: geo_localize.delay(ip_address, ip.id) return ip, created def save_imported_subdomains(subdomains, ctx={}): """Take a list of subdomains imported and write them to from_imported.txt. Args: subdomains (list): List of subdomain names. scan_history (startScan.models.ScanHistory): ScanHistory instance. domain (startScan.models.Domain): Domain instance. results_dir (str): Results directory. """ domain_id = ctx['domain_id'] domain = Domain.objects.get(pk=domain_id) results_dir = ctx.get('results_dir', RENGINE_RESULTS) # Validate each subdomain and de-duplicate entries subdomains = list(set([ subdomain for subdomain in subdomains if validators.domain(subdomain) and domain.name == get_domain_from_subdomain(subdomain) ])) if not subdomains: return logger.warning(f'Found {len(subdomains)} imported subdomains.') with open(f'{results_dir}/from_imported.txt', 'w+') as output_file: for name in subdomains: subdomain_name = name.strip() subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain.is_imported_subdomain = True subdomain.save() output_file.write(f'{subdomain}\n') @app.task(name='query_reverse_whois', bind=False, queue='query_reverse_whois_queue') def query_reverse_whois(lookup_keyword): """Queries Reverse WHOIS information for an organization or email address. Args: lookup_keyword (str): Registrar Name or email Returns: dict: Reverse WHOIS information. """ return get_associated_domains(lookup_keyword) @app.task(name='query_ip_history', bind=False, queue='query_ip_history_queue') def query_ip_history(domain): """Queries the IP history for a domain Args: domain (str): domain_name Returns: list: list of historical ip addresses """ return get_domain_historical_ip_address(domain) @app.task(name='gpt_vulnerability_description', bind=False, queue='gpt_queue') def gpt_vulnerability_description(vulnerability_id): """Generate and store Vulnerability Description using GPT. Args: vulnerability_id (Vulnerability Model ID): Vulnerability ID to fetch Description. """ logger.info('Getting GPT Vulnerability Description') try: lookup_vulnerability = Vulnerability.objects.get(id=vulnerability_id) lookup_url = urlparse(lookup_vulnerability.http_url) path = lookup_url.path except Exception as e: return { 'status': False, 'error': str(e) } # check in db GPTVulnerabilityReport model if vulnerability description and path matches stored = GPTVulnerabilityReport.objects.filter(url_path=path).filter(title=lookup_vulnerability.name).first() if stored: response = { 'status': True, 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: vulnerability_description = get_gpt_vuln_input_description( lookup_vulnerability.name, path ) # one can add more description here later gpt_generator = GPTVulnerabilityReportGenerator() response = gpt_generator.get_vulnerability_description(vulnerability_description) add_gpt_description_db( lookup_vulnerability.name, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) # for all vulnerabilities with the same vulnerability name this description has to be stored. # also the consition is that the url must contain a part of this. for vuln in Vulnerability.objects.filter(name=lookup_vulnerability.name, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() return response

import csv import json import os import pprint import subprocess import time import validators import whatportis import xmltodict import yaml import tldextract import concurrent.futures from datetime import datetime from urllib.parse import urlparse from api.serializers import SubdomainSerializer from celery import chain, chord, group from celery.result import allow_join_result from celery.utils.log import get_task_logger from django.db.models import Count from dotted_dict import DottedDict from django.utils import timezone from pycvesearch import CVESearch from metafinder.extractor import extract_metadata_from_google_search from reNgine.celery import app from reNgine.gpt import GPTVulnerabilityReportGenerator from reNgine.celery_custom_task import RengineTask from reNgine.common_func import * from reNgine.definitions import * from reNgine.settings import * from reNgine.gpt import * from reNgine.utilities import * from scanEngine.models import (EngineType, InstalledExternalTool, Notification, Proxy) from startScan.models import * from startScan.models import EndPoint, Subdomain, Vulnerability from targetApp.models import Domain """ Celery tasks. """ logger = get_task_logger(__name__) #----------------------# # Scan / Subscan tasks # #----------------------# @app.task(name='initiate_scan', bind=False, queue='initiate_scan_queue') def initiate_scan( scan_history_id, domain_id, engine_id=None, scan_type=LIVE_SCAN, results_dir=RENGINE_RESULTS, imported_subdomains=[], out_of_scope_subdomains=[], url_filter=''): """Initiate a new scan. Args: scan_history_id (int): ScanHistory id. domain_id (int): Domain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. imported_subdomains (list): Imported subdomains. out_of_scope_subdomains (list): Out-of-scope subdomains. url_filter (str): URL path. Default: '' """ # Get scan history scan = ScanHistory.objects.get(pk=scan_history_id) # Get scan engine engine_id = engine_id or scan.scan_type.id # scan history engine_id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, []) # Get domain and set last_scan_date domain = Domain.objects.get(pk=domain_id) domain.last_scan_date = timezone.now() domain.save() # Get path filter url_filter = url_filter.rstrip('/') # Get or create ScanHistory() object if scan_type == LIVE_SCAN: # immediate scan = ScanHistory.objects.get(pk=scan_history_id) scan.scan_status = RUNNING_TASK elif scan_type == SCHEDULED_SCAN: # scheduled scan = ScanHistory() scan.scan_status = INITIATED_TASK scan.scan_type = engine scan.celery_ids = [initiate_scan.request.id] scan.domain = domain scan.start_scan_date = timezone.now() scan.tasks = engine.tasks scan.results_dir = f'{results_dir}/{domain.name}_{scan.id}' add_gf_patterns = gf_patterns and 'fetch_url' in engine.tasks if add_gf_patterns: scan.used_gf_patterns = ','.join(gf_patterns) scan.save() # Create scan results dir os.makedirs(scan.results_dir) # Build task context ctx = { 'scan_history_id': scan_history_id, 'engine_id': engine_id, 'domain_id': domain.id, 'results_dir': scan.results_dir, 'url_filter': url_filter, 'yaml_configuration': config, 'out_of_scope_subdomains': out_of_scope_subdomains } ctx_str = json.dumps(ctx, indent=2) # Send start notif logger.warning(f'Starting scan {scan_history_id} with context:\n{ctx_str}') send_scan_notif.delay( scan_history_id, subscan_id=None, engine_id=engine_id, status=CELERY_TASK_STATUS_MAP[scan.scan_status]) # Save imported subdomains in DB save_imported_subdomains(imported_subdomains, ctx=ctx) # Create initial subdomain in DB: make a copy of domain as a subdomain so # that other tasks using subdomains can use it. subdomain_name = domain.name subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) # If enable_http_crawl is set, create an initial root HTTP endpoint so that # HTTP crawling can start somewhere http_url = f'{domain.name}{url_filter}' if url_filter else domain.name endpoint, _ = save_endpoint( http_url, ctx=ctx, crawl=enable_http_crawl, is_default=True, subdomain=subdomain ) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build Celery tasks, crafted according to the dependency graph below: # subdomain_discovery --> port_scan --> fetch_url --> dir_file_fuzz # osint vulnerability_scan # osint dalfox xss scan # screenshot # waf_detection workflow = chain( group( subdomain_discovery.si(ctx=ctx, description='Subdomain discovery'), osint.si(ctx=ctx, description='OS Intelligence') ), port_scan.si(ctx=ctx, description='Port scan'), fetch_url.si(ctx=ctx, description='Fetch URL'), group( dir_file_fuzz.si(ctx=ctx, description='Directories & files fuzz'), vulnerability_scan.si(ctx=ctx, description='Vulnerability scan'), screenshot.si(ctx=ctx, description='Screenshot'), waf_detection.si(ctx=ctx, description='WAF detection') ) ) # Build callback callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery chord logger.info(f'Running Celery workflow with {len(workflow.tasks) + 1} tasks') task = chain(workflow, callback).on_error(callback).delay() scan.celery_ids.append(task.id) scan.save() return { 'success': True, 'task_id': task.id } @app.task(name='initiate_subscan', bind=False, queue='subscan_queue') def initiate_subscan( scan_history_id, subdomain_id, engine_id=None, scan_type=None, results_dir=RENGINE_RESULTS, url_filter=''): """Initiate a new subscan. Args: scan_history_id (int): ScanHistory id. subdomain_id (int): Subdomain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. url_filter (str): URL path. Default: '' """ # Get Subdomain, Domain and ScanHistory subdomain = Subdomain.objects.get(pk=subdomain_id) scan = ScanHistory.objects.get(pk=subdomain.scan_history.id) domain = Domain.objects.get(pk=subdomain.target_domain.id) # Get EngineType engine_id = engine_id or scan.scan_type.id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Create scan activity of SubScan Model subscan = SubScan( start_scan_date=timezone.now(), celery_ids=[initiate_subscan.request.id], scan_history=scan, subdomain=subdomain, type=scan_type, status=RUNNING_TASK, engine=engine) subscan.save() # Get YAML configuration config = yaml.safe_load(engine.yaml_configuration) # Create results directory results_dir = f'{scan.results_dir}/subscans/{subscan.id}' os.makedirs(results_dir, exist_ok=True) # Run task method = globals().get(scan_type) if not method: logger.warning(f'Task {scan_type} is not supported by reNgine. Skipping') return scan.tasks.append(scan_type) scan.save() # Send start notif send_scan_notif.delay( scan.id, subscan_id=subscan.id, engine_id=engine_id, status='RUNNING') # Build context ctx = { 'scan_history_id': scan.id, 'subscan_id': subscan.id, 'engine_id': engine_id, 'domain_id': domain.id, 'subdomain_id': subdomain.id, 'yaml_configuration': config, 'results_dir': results_dir, 'url_filter': url_filter } # Create initial endpoints in DB: find domain HTTP endpoint so that HTTP # crawling can start somewhere base_url = f'{subdomain.name}{url_filter}' if url_filter else subdomain.name endpoint, _ = save_endpoint( base_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build header + callback workflow = method.si(ctx=ctx) callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery tasks task = chain(workflow, callback).on_error(callback).delay() subscan.celery_ids.append(task.id) subscan.save() return { 'success': True, 'task_id': task.id } @app.task(name='report', bind=False, queue='report_queue') def report(ctx={}, description=None): """Report task running after all other tasks. Mark ScanHistory or SubScan object as completed and update with final status, log run details and send notification. Args: description (str, optional): Task description shown in UI. """ # Get objects subscan_id = ctx.get('subscan_id') scan_id = ctx.get('scan_history_id') engine_id = ctx.get('engine_id') scan = ScanHistory.objects.filter(pk=scan_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() # Get failed tasks tasks = ScanActivity.objects.filter(scan_of=scan).all() if subscan: tasks = tasks.filter(celery_id__in=subscan.celery_ids) failed_tasks = tasks.filter(status=FAILED_TASK) # Get task status failed_count = failed_tasks.count() status = SUCCESS_TASK if failed_count == 0 else FAILED_TASK status_h = 'SUCCESS' if failed_count == 0 else 'FAILED' # Update scan / subscan status if subscan: subscan.stop_scan_date = timezone.now() subscan.status = status subscan.save() else: scan.scan_status = status scan.stop_scan_date = timezone.now() scan.save() # Send scan status notif send_scan_notif.delay( scan_history_id=scan_id, subscan_id=subscan_id, engine_id=engine_id, status=status_h) #------------------------- # # Tracked reNgine tasks # #--------------------------# @app.task(name='subdomain_discovery', queue='main_scan_queue', base=RengineTask, bind=True) def subdomain_discovery( self, host=None, ctx=None, description=None): """Uses a set of tools (see SUBDOMAIN_SCAN_DEFAULT_TOOLS) to scan all subdomains associated with a domain. Args: host (str): Hostname to scan. Returns: subdomains (list): List of subdomain names. """ if not host: host = self.subdomain.name if self.subdomain else self.domain.name if self.url_filter: logger.warning(f'Ignoring subdomains scan as an URL path filter was passed ({self.url_filter}).') return # Config config = self.yaml_configuration.get(SUBDOMAIN_DISCOVERY) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL) or self.yaml_configuration.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) tools = config.get(USES_TOOLS, SUBDOMAIN_SCAN_DEFAULT_TOOLS) default_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=True).filter(is_subdomain_gathering=True)] custom_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=False).filter(is_subdomain_gathering=True)] send_subdomain_changes, send_interesting = False, False notif = Notification.objects.first() if notif: send_subdomain_changes = notif.send_subdomain_changes_notif send_interesting = notif.send_interesting_notif # Gather tools to run for subdomain scan if ALL in tools: tools = SUBDOMAIN_SCAN_DEFAULT_TOOLS + custom_subdomain_tools tools = [t.lower() for t in tools] # Make exception for amass since tool name is amass, but command is amass-active/passive default_subdomain_tools.append('amass-passive') default_subdomain_tools.append('amass-active') # Run tools for tool in tools: cmd = None logger.info(f'Scanning subdomains for {host} with {tool}') proxy = get_random_proxy() if tool in default_subdomain_tools: if tool == 'amass-passive': cmd = f'amass enum -passive -d {host} -o {self.results_dir}/subdomains_amass.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' elif tool == 'amass-active': use_amass_config = config.get(USE_AMASS_CONFIG, False) amass_wordlist_name = config.get(AMASS_WORDLIST, 'deepmagic.com-prefixes-top50000') wordlist_path = f'/usr/src/wordlist/{amass_wordlist_name}.txt' cmd = f'amass enum -active -d {host} -o {self.results_dir}/subdomains_amass_active.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' cmd += f' -brute -w {wordlist_path}' elif tool == 'sublist3r': cmd = f'python3 /usr/src/github/Sublist3r/sublist3r.py -d {host} -t {threads} -o {self.results_dir}/subdomains_sublister.txt' elif tool == 'subfinder': cmd = f'subfinder -d {host} -o {self.results_dir}/subdomains_subfinder.txt' use_subfinder_config = config.get(USE_SUBFINDER_CONFIG, False) cmd += ' -config /root/.config/subfinder/config.yaml' if use_subfinder_config else '' cmd += f' -proxy {proxy}' if proxy else '' cmd += f' -timeout {timeout}' if timeout else '' cmd += f' -t {threads}' if threads else '' cmd += f' -silent' elif tool == 'oneforall': cmd = f'python3 /usr/src/github/OneForAll/oneforall.py --target {host} run' cmd_extract = f'cut -d\',\' -f6 /usr/src/github/OneForAll/results/{host}.csv > {self.results_dir}/subdomains_oneforall.txt' cmd_rm = f'rm -rf /usr/src/github/OneForAll/results/{host}.csv' cmd += f' && {cmd_extract} && {cmd_rm}' elif tool == 'ctfr': results_file = self.results_dir + '/subdomains_ctfr.txt' cmd = f'python3 /usr/src/github/ctfr/ctfr.py -d {host} -o {results_file}' cmd_extract = f"cat {results_file} | sed 's/\*.//g' | tail -n +12 | uniq | sort > {results_file}" cmd += f' && {cmd_extract}' elif tool == 'tlsx': results_file = self.results_dir + '/subdomains_tlsx.txt' cmd = f'tlsx -san -cn -silent -ro -host {host}' cmd += f" | sed -n '/^\([a-zA-Z0-9]\([-a-zA-Z0-9]*[a-zA-Z0-9]\)\?\.\)\+{host}$/p' | uniq | sort" cmd += f' > {results_file}' elif tool == 'netlas': results_file = self.results_dir + '/subdomains_netlas.txt' cmd = f'netlas search -d domain -i domain domain:"*.{host}" -f json' netlas_key = get_netlas_key() cmd += f' -a {netlas_key}' if netlas_key else '' cmd_extract = f"grep -oE '([a-zA-Z0-9]([-a-zA-Z0-9]*[a-zA-Z0-9])?\.)+{host}'" cmd += f' | {cmd_extract} > {results_file}' elif tool in custom_subdomain_tools: tool_query = InstalledExternalTool.objects.filter(name__icontains=tool.lower()) if not tool_query.exists(): logger.error(f'Missing {{TARGET}} and {{OUTPUT}} placeholders in {tool} configuration. Skipping.') continue custom_tool = tool_query.first() cmd = custom_tool.subdomain_gathering_command if '{TARGET}' in cmd and '{OUTPUT}' in cmd: cmd = cmd.replace('{TARGET}', host) cmd = cmd.replace('{OUTPUT}', f'{self.results_dir}/subdomains_{tool}.txt') cmd = cmd.replace('{PATH}', custom_tool.github_clone_path) if '{PATH}' in cmd else cmd else: logger.warning( f'Subdomain discovery tool "{tool}" is not supported by reNgine. Skipping.') continue # Run tool try: run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) except Exception as e: logger.error( f'Subdomain discovery tool "{tool}" raised an exception') logger.exception(e) # Gather all the tools' results in one single file. Write subdomains into # separate files, and sort all subdomains. run_command( f'cat {self.results_dir}/subdomains_*.txt > {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {self.output_path} -o {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) with open(self.output_path) as f: lines = f.readlines() # Parse the output_file file and store Subdomain and EndPoint objects found # in db. subdomain_count = 0 subdomains = [] urls = [] for line in lines: subdomain_name = line.strip() valid_url = bool(validators.url(subdomain_name)) valid_domain = ( bool(validators.domain(subdomain_name)) or bool(validators.ipv4(subdomain_name)) or bool(validators.ipv6(subdomain_name)) or valid_url ) if not valid_domain: logger.error(f'Subdomain {subdomain_name} is not a valid domain, IP or URL. Skipping.') continue if valid_url: subdomain_name = urlparse(subdomain_name).netloc if subdomain_name in self.out_of_scope_subdomains: logger.error(f'Subdomain {subdomain_name} is out of scope. Skipping.') continue # Add subdomain subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain_count += 1 subdomains.append(subdomain) urls.append(subdomain.name) # Bulk crawl subdomains if enable_http_crawl: ctx['track'] = True http_crawl(urls, ctx=ctx, is_ran_from_subdomain_scan=True) # Find root subdomain endpoints for subdomain in subdomains: pass # Send notifications subdomains_str = '\n'.join([f'• `{subdomain.name}`' for subdomain in subdomains]) self.notify(fields={ 'Subdomain count': len(subdomains), 'Subdomains': subdomains_str, }) if send_subdomain_changes and self.scan_id and self.domain_id: added = get_new_added_subdomain(self.scan_id, self.domain_id) removed = get_removed_subdomain(self.scan_id, self.domain_id) if added: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in added]) self.notify(fields={'Added subdomains': subdomains_str}) if removed: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in removed]) self.notify(fields={'Removed subdomains': subdomains_str}) if send_interesting and self.scan_id and self.domain_id: interesting_subdomains = get_interesting_subdomains(self.scan_id, self.domain_id) if interesting_subdomains: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in interesting_subdomains]) self.notify(fields={'Interesting subdomains': subdomains_str}) return SubdomainSerializer(subdomains, many=True).data @app.task(name='osint', queue='main_scan_queue', base=RengineTask, bind=True) def osint(self, host=None, ctx={}, description=None): """Run Open-Source Intelligence tools on selected domain. Args: host (str): Hostname to scan. Returns: dict: Results from osint discovery and dorking. """ config = self.yaml_configuration.get(OSINT) or OSINT_DEFAULT_CONFIG results = {} grouped_tasks = [] if 'discover' in config: ctx['track'] = False # results = osint_discovery(host=host, ctx=ctx) _task = osint_discovery.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, activity_id=self.activity_id, results_dir=self.results_dir, ctx=ctx ) grouped_tasks.append(_task) if OSINT_DORK in config or OSINT_CUSTOM_DORK in config: _task = dorking.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, results_dir=self.results_dir ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('OSINT Tasks finished...') # with open(self.output_path, 'w') as f: # json.dump(results, f, indent=4) # # return results @app.task(name='osint_discovery', queue='osint_discovery_queue', bind=False) def osint_discovery(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run OSINT discovery. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: dict: osint metadat and theHarvester and h8mail results. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) osint_lookup = config.get(OSINT_DISCOVER, []) osint_intensity = config.get(INTENSITY, 'normal') documents_limit = config.get(OSINT_DOCUMENTS_LIMIT, 50) results = {} meta_info = [] emails = [] creds = [] # Get and save meta info if 'metainfo' in osint_lookup: if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': host, 'domain': host, 'scan_id': scan_history_id, 'documents_limit': documents_limit }) meta_info.append(save_metadata_info(meta_dict)) # TODO: disabled for now # elif osint_intensity == 'deep': # subdomains = Subdomain.objects # if self.scan: # subdomains = subdomains.filter(scan_history=self.scan) # for subdomain in subdomains: # meta_dict = DottedDict({ # 'osint_target': subdomain.name, # 'domain': self.domain, # 'scan_id': self.scan_id, # 'documents_limit': documents_limit # }) # meta_info.append(save_metadata_info(meta_dict)) grouped_tasks = [] if 'emails' in osint_lookup: emails = get_and_save_emails(scan_history, activity_id, results_dir) emails_str = '\n'.join([f'• `{email}`' for email in emails]) # self.notify(fields={'Emails': emails_str}) # ctx['track'] = False _task = h8mail.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) if 'employees' in osint_lookup: ctx['track'] = False _task = theHarvester.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) # results['emails'] = results.get('emails', []) + emails # results['creds'] = creds # results['meta_info'] = meta_info return results @app.task(name='dorking', bind=False, queue='dorking_queue') def dorking(config, host, scan_history_id, results_dir): """Run Google dorks. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: list: Dorking results for each dork ran. """ # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh scan_history = ScanHistory.objects.get(pk=scan_history_id) dorks = config.get(OSINT_DORK, []) custom_dorks = config.get(OSINT_CUSTOM_DORK, []) results = [] # custom dorking has higher priority try: for custom_dork in custom_dorks: lookup_target = custom_dork.get('lookup_site') # replace with original host if _target_ lookup_target = host if lookup_target == '_target_' else lookup_target if 'lookup_extensions' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_extensions=custom_dork.get('lookup_extensions'), scan_history=scan_history ) elif 'lookup_keywords' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_keywords=custom_dork.get('lookup_keywords'), scan_history=scan_history ) except Exception as e: logger.exception(e) # default dorking try: for dork in dorks: logger.info(f'Getting dork information for {dork}') if dork == 'stackoverflow': results = get_and_save_dork_results( lookup_target='stackoverflow.com', results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'login_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/login/,login.html', page_count=5, scan_history=scan_history ) elif dork == 'admin_panels': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/admin/,admin.html', page_count=5, scan_history=scan_history ) elif dork == 'dashboard_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/dashboard/,dashboard.html', page_count=5, scan_history=scan_history ) elif dork == 'social_media' : social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'reddit.com' ] for site in social_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'project_management' : project_websites = [ 'trello.com', 'atlassian.net' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'code_sharing' : project_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'config_files' : config_file_exts = [ 'env', 'xml', 'conf', 'toml', 'yml', 'yaml', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(config_file_exts), page_count=4, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keyword = 'Jenkins' results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=lookup_keyword, page_count=1, scan_history=scan_history ) elif dork == 'wordpress_files' : lookup_keywords = [ '/wp-content/', '/wp-includes/' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'php_error' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'exposed_documents' : docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(docs_file_ext), page_count=7, scan_history=scan_history ) elif dork == 'db_files' : file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) elif dork == 'git_exposed' : file_ext = [ 'git', ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) except Exception as e: logger.exception(e) return results @app.task(name='theHarvester', queue='theHarvester_queue', bind=False) def theHarvester(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run theHarvester to get save emails, hosts, employees found in domain. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: dict: Dict of emails, employees, hosts and ips found during crawling. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) output_path_json = f'{results_dir}/theHarvester.json' theHarvester_dir = '/usr/src/github/theHarvester' history_file = f'{results_dir}/commands.txt' cmd = f'python3 {theHarvester_dir}/theHarvester.py -d {host} -b all -f {output_path_json}' # Update proxies.yaml proxy_query = Proxy.objects.all() if proxy_query.exists(): proxy = proxy_query.first() if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(f'{theHarvester_dir}/proxies.yaml', 'w') as file: yaml.dump(yaml_data, file) # Run cmd run_command( cmd, shell=False, cwd=theHarvester_dir, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) # Get file location if not os.path.isfile(output_path_json): logger.error(f'Could not open {output_path_json}') return {} # Load theHarvester results with open(output_path_json, 'r') as f: data = json.load(f) # Re-indent theHarvester JSON with open(output_path_json, 'w') as f: json.dump(data, f, indent=4) emails = data.get('emails', []) for email_address in emails: email, _ = save_email(email_address, scan_history=scan_history) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) linkedin_people = data.get('linkedin_people', []) for people in linkedin_people: employee, _ = save_employee( people, designation='linkedin', scan_history=scan_history) # if employee: # self.notify(fields={'LinkedIn people': f'• {employee.name}'}) twitter_people = data.get('twitter_people', []) for people in twitter_people: employee, _ = save_employee( people, designation='twitter', scan_history=scan_history) # if employee: # self.notify(fields={'Twitter people': f'• {employee.name}'}) hosts = data.get('hosts', []) urls = [] for host in hosts: split = tuple(host.split(':')) http_url = split[0] subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) endpoint, _ = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain) # if endpoint: # urls.append(endpoint.http_url) # self.notify(fields={'Hosts': f'• {endpoint.http_url}'}) # if enable_http_crawl: # ctx['track'] = False # http_crawl(urls, ctx=ctx) # TODO: Lots of ips unrelated with our domain are found, disabling # this for now. # ips = data.get('ips', []) # for ip_address in ips: # ip, created = save_ip_address( # ip_address, # subscan=subscan) # if ip: # send_task_notif.delay( # 'osint', # scan_history_id=scan_history_id, # subscan_id=subscan_id, # severity='success', # update_fields={'IPs': f'{ip.address}'}) return data @app.task(name='h8mail', queue='h8mail_queue', bind=False) def h8mail(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run h8mail. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: list[dict]: List of credentials info. """ logger.warning('Getting leaked credentials') scan_history = ScanHistory.objects.get(pk=scan_history_id) input_path = f'{results_dir}/emails.txt' output_file = f'{results_dir}/h8mail.json' cmd = f'h8mail -t {input_path} --json {output_file}' history_file = f'{results_dir}/commands.txt' run_command( cmd, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) with open(output_file) as f: data = json.load(f) creds = data.get('targets', []) # TODO: go through h8mail output and save emails to DB for cred in creds: logger.warning(cred) email_address = cred['target'] pwn_num = cred['pwn_num'] pwn_data = cred.get('data', []) email, created = save_email(email_address, scan_history=scan) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) return creds @app.task(name='screenshot', queue='main_scan_queue', base=RengineTask, bind=True) def screenshot(self, ctx={}, description=None): """Uses EyeWitness to gather screenshot of a domain and/or url. Args: description (str, optional): Task description shown in UI. """ # Config screenshots_path = f'{self.results_dir}/screenshots' output_path = f'{self.results_dir}/screenshots/{self.filename}' alive_endpoints_file = f'{self.results_dir}/endpoints_alive.txt' config = self.yaml_configuration.get(SCREENSHOT) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT + 5) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) # If intensity is normal, grab only the root endpoints of each subdomain strict = True if intensity == 'normal' else False # Get URLs to take screenshot of get_http_urls( is_alive=enable_http_crawl, strict=strict, write_filepath=alive_endpoints_file, get_only_default_urls=True, ctx=ctx ) # Send start notif notification = Notification.objects.first() send_output_file = notification.send_scan_output_file if notification else False # Run cmd cmd = f'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py -f {alive_endpoints_file} -d {screenshots_path} --no-prompt' cmd += f' --timeout {timeout}' if timeout > 0 else '' cmd += f' --threads {threads}' if threads > 0 else '' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(output_path): logger.error(f'Could not load EyeWitness results at {output_path} for {self.domain.name}.') return # Loop through results and save objects in DB screenshot_paths = [] with open(output_path, 'r') as file: reader = csv.reader(file) for row in reader: "Protocol,Port,Domain,Request Status,Screenshot Path, Source Path" protocol, port, subdomain_name, status, screenshot_path, source_path = tuple(row) logger.info(f'{protocol}:{port}:{subdomain_name}:{status}') subdomain_query = Subdomain.objects.filter(name=subdomain_name) if self.scan: subdomain_query = subdomain_query.filter(scan_history=self.scan) if status == 'Successful' and subdomain_query.exists(): subdomain = subdomain_query.first() screenshot_paths.append(screenshot_path) subdomain.screenshot_path = screenshot_path.replace('/usr/src/scan_results/', '') subdomain.save() logger.warning(f'Added screenshot for {subdomain.name} to DB') # Remove all db, html extra files in screenshot results run_command( 'rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format(screenshots_path), shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'rm -rf {screenshots_path}/source', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Send finish notifs screenshots_str = '• ' + '\n• '.join([f'`{path}`' for path in screenshot_paths]) self.notify(fields={'Screenshots': screenshots_str}) if send_output_file: for path in screenshot_paths: title = get_output_file_name( self.scan_id, self.subscan_id, self.filename) send_file_to_discord.delay(path, title) @app.task(name='port_scan', queue='main_scan_queue', base=RengineTask, bind=True) def port_scan(self, hosts=[], ctx={}, description=None): """Run port scan. Args: hosts (list, optional): Hosts to run port scan on. description (str, optional): Task description shown in UI. Returns: list: List of open ports (dict). """ input_file = f'{self.results_dir}/input_subdomains_port_scan.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(PORT_SCAN) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) exclude_ports = config.get(NAABU_EXCLUDE_PORTS, []) exclude_subdomains = config.get(NAABU_EXCLUDE_SUBDOMAINS, False) ports = config.get(PORTS, NAABU_DEFAULT_PORTS) ports = [str(port) for port in ports] rate_limit = config.get(NAABU_RATE) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) passive = config.get(NAABU_PASSIVE, False) use_naabu_config = config.get(USE_NAABU_CONFIG, False) exclude_ports_str = ','.join(return_iterable(exclude_ports)) # nmap args nmap_enabled = config.get(ENABLE_NMAP, False) nmap_cmd = config.get(NMAP_COMMAND, '') nmap_script = config.get(NMAP_SCRIPT, '') nmap_script = ','.join(return_iterable(nmap_script)) nmap_script_args = config.get(NMAP_SCRIPT_ARGS) if hosts: with open(input_file, 'w') as f: f.write('\n'.join(hosts)) else: hosts = get_subdomains( write_filepath=input_file, exclude_subdomains=exclude_subdomains, ctx=ctx) # Build cmd cmd = 'naabu -json -exclude-cdn' cmd += f' -list {input_file}' if len(hosts) > 0 else f' -host {hosts[0]}' if 'full' in ports or 'all' in ports: ports_str = ' -p "-"' elif 'top-100' in ports: ports_str = ' -top-ports 100' elif 'top-1000' in ports: ports_str = ' -top-ports 1000' else: ports_str = ','.join(ports) ports_str = f' -p {ports_str}' cmd += ports_str cmd += ' -config /root/.config/naabu/config.yaml' if use_naabu_config else '' cmd += f' -proxy "{proxy}"' if proxy else '' cmd += f' -c {threads}' if threads else '' cmd += f' -rate {rate_limit}' if rate_limit > 0 else '' cmd += f' -timeout {timeout*1000}' if timeout > 0 else '' cmd += f' -passive' if passive else '' cmd += f' -exclude-ports {exclude_ports_str}' if exclude_ports else '' cmd += f' -silent' # Execute cmd and gather results results = [] urls = [] ports_data = {} for line in stream_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) port_number = line['port'] ip_address = line['ip'] host = line.get('host') or ip_address if port_number == 0: continue # Grab subdomain subdomain = Subdomain.objects.filter( name=host, target_domain=self.domain, scan_history=self.scan ).first() # Add IP DB ip, _ = save_ip_address(ip_address, subdomain, subscan=self.subscan) if self.subscan: ip.ip_subscan_ids.add(self.subscan) ip.save() # Add endpoint to DB # port 80 and 443 not needed as http crawl already does that. if port_number not in [80, 443]: http_url = f'{host}:{port_number}' endpoint, _ = save_endpoint( http_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint: http_url = endpoint.http_url urls.append(http_url) # Add Port in DB port_details = whatportis.get_ports(str(port_number)) service_name = port_details[0].name if len(port_details) > 0 else 'unknown' description = port_details[0].description if len(port_details) > 0 else '' # get or create port port, created = Port.objects.get_or_create( number=port_number, service_name=service_name, description=description ) if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port.save() ip.ports.add(port) ip.save() if host in ports_data: ports_data[host].append(port_number) else: ports_data[host] = [port_number] # Send notification logger.warning(f'Found opened port {port_number} on {ip_address} ({host})') if len(ports_data) == 0: logger.info('Finished running naabu port scan - No open ports found.') if nmap_enabled: logger.info('Nmap scans skipped') return ports_data # Send notification fields_str = '' for host, ports in ports_data.items(): ports_str = ', '.join([f'`{port}`' for port in ports]) fields_str += f'• `{host}`: {ports_str}\n' self.notify(fields={'Ports discovered': fields_str}) # Save output to file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) logger.info('Finished running naabu port scan.') # Process nmap results: 1 process per host sigs = [] if nmap_enabled: logger.warning(f'Starting nmap scans ...') logger.warning(ports_data) for host, port_list in ports_data.items(): ports_str = '_'.join([str(p) for p in port_list]) ctx_nmap = ctx.copy() ctx_nmap['description'] = get_task_title(f'nmap_{host}', self.scan_id, self.subscan_id) ctx_nmap['track'] = False sig = nmap.si( cmd=nmap_cmd, ports=port_list, host=host, script=nmap_script, script_args=nmap_script_args, max_rate=rate_limit, ctx=ctx_nmap) sigs.append(sig) task = group(sigs).apply_async() with allow_join_result(): results = task.get() return ports_data @app.task(name='nmap', queue='main_scan_queue', base=RengineTask, bind=True) def nmap( self, cmd=None, ports=[], host=None, input_file=None, script=None, script_args=None, max_rate=None, ctx={}, description=None): """Run nmap on a host. Args: cmd (str, optional): Existing nmap command to complete. ports (list, optional): List of ports to scan. host (str, optional): Host to scan. input_file (str, optional): Input hosts file. script (str, optional): NSE script to run. script_args (str, optional): NSE script args. max_rate (int): Max rate. description (str, optional): Task description shown in UI. """ notif = Notification.objects.first() ports_str = ','.join(str(port) for port in ports) self.filename = self.filename.replace('.txt', '.xml') filename_vulns = self.filename.replace('.xml', '_vulns.json') output_file = self.output_path output_file_xml = f'{self.results_dir}/{host}_{self.filename}' vulns_file = f'{self.results_dir}/{host}_{filename_vulns}' logger.warning(f'Running nmap on {host}:{ports}') # Build cmd nmap_cmd = get_nmap_cmd( cmd=cmd, ports=ports_str, script=script, script_args=script_args, max_rate=max_rate, host=host, input_file=input_file, output_file=output_file_xml) # Run cmd run_command( nmap_cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Get nmap XML results and convert to JSON vulns = parse_nmap_results(output_file_xml, output_file) with open(vulns_file, 'w') as f: json.dump(vulns, f, indent=4) # Save vulnerabilities found by nmap vulns_str = '' for vuln_data in vulns: # URL is not necessarily an HTTP URL when running nmap (can be any # other vulnerable protocols). Look for existing endpoint and use its # URL as vulnerability.http_url if it exists. url = vuln_data['http_url'] endpoint = EndPoint.objects.filter(http_url__contains=url).first() if endpoint: vuln_data['http_url'] = endpoint.http_url vuln, created = save_vulnerability( target_domain=self.domain, subdomain=self.subdomain, scan_history=self.scan, subscan=self.subscan, endpoint=endpoint, **vuln_data) vulns_str += f'• {str(vuln)}\n' if created: logger.warning(str(vuln)) # Send only 1 notif for all vulns to reduce number of notifs if notif and notif.send_vuln_notif and vulns_str: logger.warning(vulns_str) self.notify(fields={'CVEs': vulns_str}) return vulns @app.task(name='waf_detection', queue='main_scan_queue', base=RengineTask, bind=True) def waf_detection(self, ctx={}, description=None): """ Uses wafw00f to check for the presence of a WAF. Args: description (str, optional): Task description shown in UI. Returns: list: List of startScan.models.Waf objects. """ input_path = f'{self.results_dir}/input_endpoints_waf_detection.txt' config = self.yaml_configuration.get(WAF_DETECTION) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Get alive endpoints from DB get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) cmd = f'wafw00f -i {input_path} -o {self.output_path}' run_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(self.output_path): logger.error(f'Could not find {self.output_path}') return with open(self.output_path) as file: wafs = file.readlines() for line in wafs: line = " ".join(line.split()) splitted = line.split(' ', 1) waf_info = splitted[1].strip() waf_name = waf_info[:waf_info.find('(')].strip() waf_manufacturer = waf_info[waf_info.find('(')+1:waf_info.find(')')].strip().replace('.', '') http_url = sanitize_url(splitted[0].strip()) if not waf_name or waf_name == 'None': continue # Add waf to db waf, _ = Waf.objects.get_or_create( name=waf_name, manufacturer=waf_manufacturer ) # Add waf info to Subdomain in DB subdomain = get_subdomain_from_url(http_url) logger.info(f'Wafw00f Subdomain : {subdomain}') subdomain_query, _ = Subdomain.objects.get_or_create(scan_history=self.scan, name=subdomain) subdomain_query.waf.add(waf) subdomain_query.save() return wafs @app.task(name='dir_file_fuzz', queue='main_scan_queue', base=RengineTask, bind=True) def dir_file_fuzz(self, ctx={}, description=None): """Perform directory scan, and currently uses `ffuf` as a default tool. Args: description (str, optional): Task description shown in UI. Returns: list: List of URLs discovered. """ # Config cmd = 'ffuf' config = self.yaml_configuration.get(DIR_FILE_FUZZ) or {} custom_header = self.yaml_configuration.get(CUSTOM_HEADER) auto_calibration = config.get(AUTO_CALIBRATION, True) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) extensions = config.get(EXTENSIONS, DEFAULT_DIR_FILE_FUZZ_EXTENSIONS) # prepend . on extensions extensions = [ext if ext.startswith('.') else '.' + ext for ext in extensions] extensions_str = ','.join(map(str, extensions)) follow_redirect = config.get(FOLLOW_REDIRECT, FFUF_DEFAULT_FOLLOW_REDIRECT) max_time = config.get(MAX_TIME, 0) match_http_status = config.get(MATCH_HTTP_STATUS, FFUF_DEFAULT_MATCH_HTTP_STATUS) mc = ','.join([str(c) for c in match_http_status]) recursive_level = config.get(RECURSIVE_LEVEL, FFUF_DEFAULT_RECURSIVE_LEVEL) stop_on_error = config.get(STOP_ON_ERROR, False) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) wordlist_name = config.get(WORDLIST, 'dicc') delay = rate_limit / (threads * 100) # calculate request pause delay from rate_limit and number of threads input_path = f'{self.results_dir}/input_dir_file_fuzz.txt' # Get wordlist wordlist_name = 'dicc' if wordlist_name == 'default' else wordlist_name wordlist_path = f'/usr/src/wordlist/{wordlist_name}.txt' # Build command cmd += f' -w {wordlist_path}' cmd += f' -e {extensions_str}' if extensions else '' cmd += f' -maxtime {max_time}' if max_time > 0 else '' cmd += f' -p {delay}' if delay > 0 else '' cmd += f' -recursion -recursion-depth {recursive_level} ' if recursive_level > 0 else '' cmd += f' -t {threads}' if threads and threads > 0 else '' cmd += f' -timeout {timeout}' if timeout and timeout > 0 else '' cmd += ' -se' if stop_on_error else '' cmd += ' -fr' if follow_redirect else '' cmd += ' -ac' if auto_calibration else '' cmd += f' -mc {mc}' if mc else '' cmd += f' -H "{custom_header}"' if custom_header else '' # Grab URLs to fuzz urls = get_http_urls( is_alive=True, ignore_files=False, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) logger.warning(urls) # Loop through URLs and run command results = [] for url in urls: ''' Above while fetching urls, we are not ignoring files, because some default urls may redirect to https://example.com/login.php so, ignore_files is set to False but, during fuzzing, we will only need part of the path, in above example it is still a good idea to ffuf base url https://example.com so files from base url ''' url_parse = urlparse(url) url = url_parse.scheme + '://' + url_parse.netloc url += '/FUZZ' # TODO: fuzz not only URL but also POST / PUT / headers proxy = get_random_proxy() # Build final cmd fcmd = cmd fcmd += f' -x {proxy}' if proxy else '' fcmd += f' -u {url} -json' # Initialize DirectoryScan object dirscan = DirectoryScan() dirscan.scanned_date = timezone.now() dirscan.command_line = fcmd dirscan.save() # Loop through results and populate EndPoint and DirectoryFile in DB results = [] for line in stream_command( fcmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) name = line['input'].get('FUZZ') length = line['length'] status = line['status'] words = line['words'] url = line['url'] lines = line['lines'] content_type = line['content-type'] duration = line['duration'] if not name: logger.error(f'FUZZ not found for "{url}"') continue endpoint, created = save_endpoint(url, crawl=False, ctx=ctx) # endpoint.is_default = False endpoint.http_status = status endpoint.content_length = length endpoint.response_time = duration / 1000000000 endpoint.save() if created: urls.append(endpoint.http_url) endpoint.status = status endpoint.content_type = content_type endpoint.content_length = length dfile, created = DirectoryFile.objects.get_or_create( name=name, length=length, words=words, lines=lines, content_type=content_type, url=url) dfile.http_status = status dfile.save() # if created: # logger.warning(f'Found new directory or file {url}') dirscan.directory_files.add(dfile) dirscan.save() if self.subscan: dirscan.dir_subscan_ids.add(self.subscan) subdomain_name = get_subdomain_from_url(endpoint.http_url) subdomain = Subdomain.objects.get(name=subdomain_name, scan_history=self.scan) subdomain.directories.add(dirscan) subdomain.save() # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl(urls, ctx=ctx) return results @app.task(name='fetch_url', queue='main_scan_queue', base=RengineTask, bind=True) def fetch_url(self, urls=[], ctx={}, description=None): """Fetch URLs using different tools like gauplus, gau, gospider, waybackurls ... Args: urls (list): List of URLs to start from. description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/input_endpoints_fetch_url.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(FETCH_URL) or {} should_remove_duplicate_endpoints = config.get(REMOVE_DUPLICATE_ENDPOINTS, True) duplicate_removal_fields = config.get(DUPLICATE_REMOVAL_FIELDS, ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, DEFAULT_GF_PATTERNS) ignore_file_extension = config.get(IGNORE_FILE_EXTENSION, DEFAULT_IGNORE_FILE_EXTENSIONS) tools = config.get(USES_TOOLS, ENDPOINT_SCAN_DEFAULT_TOOLS) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) domain_request_headers = self.domain.request_headers if self.domain else None custom_header = domain_request_headers or self.yaml_configuration.get(CUSTOM_HEADER) exclude_subdomains = config.get(EXCLUDED_SUBDOMAINS, False) # Get URLs to scan and save to input file if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, exclude_subdomains=exclude_subdomains, get_only_default_urls=True, ctx=ctx ) # Domain regex host = self.domain.name if self.domain else urlparse(urls[0]).netloc host_regex = f"\'https?://([a-z0-9]+[.])*{host}.*\'" # Tools cmds cmd_map = { 'gau': f'gau', 'gauplus': f'gauplus -random-agent', 'hakrawler': 'hakrawler -subs -u', 'waybackurls': 'waybackurls', 'gospider': f'gospider -S {input_path} --js -d 2 --sitemap --robots -w -r', 'katana': f'katana -list {input_path} -silent -jc -kf all -d 3 -fs rdn', } if proxy: cmd_map['gau'] += f' --proxy "{proxy}"' cmd_map['gauplus'] += f' -p "{proxy}"' cmd_map['gospider'] += f' -p {proxy}' cmd_map['hakrawler'] += f' -proxy {proxy}' cmd_map['katana'] += f' -proxy {proxy}' if threads > 0: cmd_map['gau'] += f' --threads {threads}' cmd_map['gauplus'] += f' -t {threads}' cmd_map['gospider'] += f' -t {threads}' cmd_map['katana'] += f' -c {threads}' if custom_header: header_string = ';;'.join([ f'{key}: {value}' for key, value in custom_header.items() ]) cmd_map['hakrawler'] += f' -h {header_string}' cmd_map['katana'] += f' -H {header_string}' header_flags = [':'.join(h) for h in header_string.split(';;')] for flag in header_flags: cmd_map['gospider'] += f' -H {flag}' cat_input = f'cat {input_path}' grep_output = f'grep -Eo {host_regex}' cmd_map = { tool: f'{cat_input} | {cmd} | {grep_output} > {self.results_dir}/urls_{tool}.txt' for tool, cmd in cmd_map.items() } tasks = group( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for tool, cmd in cmd_map.items() if tool in tools ) # Cleanup task sort_output = [ f'cat {self.results_dir}/urls_* > {self.output_path}', f'cat {input_path} >> {self.output_path}', f'sort -u {self.output_path} -o {self.output_path}', ] if ignore_file_extension: ignore_exts = '|'.join(ignore_file_extension) grep_ext_filtered_output = [ f'cat {self.output_path} | grep -Eiv "\\.({ignore_exts}).*" > {self.results_dir}/urls_filtered.txt', f'mv {self.results_dir}/urls_filtered.txt {self.output_path}' ] sort_output.extend(grep_ext_filtered_output) cleanup = chain( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for cmd in sort_output ) # Run all commands task = chord(tasks)(cleanup) with allow_join_result(): task.get() # Store all the endpoints and run httpx with open(self.output_path) as f: discovered_urls = f.readlines() self.notify(fields={'Discovered URLs': len(discovered_urls)}) # Some tools can have an URL in the format <URL>] - <PATH> or <URL> - <PATH>, add them # to the final URL list all_urls = [] for url in discovered_urls: url = url.strip() urlpath = None base_url = None if '] ' in url: # found JS scraped endpoint e.g from gospider split = tuple(url.split('] ')) if not len(split) == 2: logger.warning(f'URL format not recognized for "{url}". Skipping.') continue base_url, urlpath = split urlpath = urlpath.lstrip('- ') elif ' - ' in url: # found JS scraped endpoint e.g from gospider base_url, urlpath = tuple(url.split(' - ')) if base_url and urlpath: subdomain = urlparse(base_url) url = f'{subdomain.scheme}://{subdomain.netloc}{self.url_filter}' if not validators.url(url): logger.warning(f'Invalid URL "{url}". Skipping.') if url not in all_urls: all_urls.append(url) # Filter out URLs if a path filter was passed if self.url_filter: all_urls = [url for url in all_urls if self.url_filter in url] # Write result to output path with open(self.output_path, 'w') as f: f.write('\n'.join(all_urls)) logger.warning(f'Found {len(all_urls)} usable URLs') # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl( all_urls, ctx=ctx, should_remove_duplicate_endpoints=should_remove_duplicate_endpoints, duplicate_removal_fields=duplicate_removal_fields ) #-------------------# # GF PATTERNS MATCH # #-------------------# # Combine old gf patterns with new ones if gf_patterns: self.scan.used_gf_patterns = ','.join(gf_patterns) self.scan.save() # Run gf patterns on saved endpoints # TODO: refactor to Celery task for gf_pattern in gf_patterns: # TODO: js var is causing issues, removing for now if gf_pattern == 'jsvar': logger.info('Ignoring jsvar as it is causing issues.') continue # Run gf on current pattern logger.warning(f'Running gf on pattern "{gf_pattern}"') gf_output_file = f'{self.results_dir}/gf_patterns_{gf_pattern}.txt' cmd = f'cat {self.output_path} | gf {gf_pattern} | grep -Eo {host_regex} >> {gf_output_file}' run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Check output file if not os.path.exists(gf_output_file): logger.error(f'Could not find GF output file {gf_output_file}. Skipping GF pattern "{gf_pattern}"') continue # Read output file line by line and with open(gf_output_file, 'r') as f: lines = f.readlines() # Add endpoints / subdomains to DB for url in lines: http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue endpoint, created = save_endpoint( http_url, crawl=False, subdomain=subdomain, ctx=ctx) if not endpoint: continue earlier_pattern = None if not created: earlier_pattern = endpoint.matched_gf_patterns pattern = f'{earlier_pattern},{gf_pattern}' if earlier_pattern else gf_pattern endpoint.matched_gf_patterns = pattern endpoint.save() return all_urls def parse_curl_output(response): # TODO: Enrich from other cURL fields. CURL_REGEX_HTTP_STATUS = f'HTTP\/(?:(?:\d\.?)+)\s(\d+)\s(?:\w+)' http_status = 0 if response: failed = False regex = re.compile(CURL_REGEX_HTTP_STATUS, re.MULTILINE) try: http_status = int(regex.findall(response)[0]) except (KeyError, TypeError, IndexError): pass return { 'http_status': http_status, } @app.task(name='vulnerability_scan', queue='main_scan_queue', bind=True, base=RengineTask) def vulnerability_scan(self, urls=[], ctx={}, description=None): """ This function will serve as an entrypoint to vulnerability scan. All other vulnerability scan will be run from here including nuclei, crlfuzz, etc """ logger.info('Running Vulnerability Scan Queue') config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_run_nuclei = config.get(RUN_NUCLEI, True) should_run_crlfuzz = config.get(RUN_CRLFUZZ, False) should_run_dalfox = config.get(RUN_DALFOX, False) should_run_s3scanner = config.get(RUN_S3SCANNER, True) grouped_tasks = [] if should_run_nuclei: _task = nuclei_scan.si( urls=urls, ctx=ctx, description=f'Nuclei Scan' ) grouped_tasks.append(_task) if should_run_crlfuzz: _task = crlfuzz_scan.si( urls=urls, ctx=ctx, description=f'CRLFuzz Scan' ) grouped_tasks.append(_task) if should_run_dalfox: _task = dalfox_xss_scan.si( urls=urls, ctx=ctx, description=f'Dalfox XSS Scan' ) grouped_tasks.append(_task) if should_run_s3scanner: _task = s3scanner.si( ctx=ctx, description=f'Misconfigured S3 Buckets Scanner' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan completed...') # return results return None @app.task(name='nuclei_individual_severity_module', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_individual_severity_module(self, cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx={}, description=None): ''' This celery task will run vulnerability scan in parallel. All severities supplied should run in parallel as grouped tasks. ''' results = [] logger.info(f'Running vulnerability scan with severity: {severity}') cmd += f' -severity {severity}' # Send start notification notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) # Gather nuclei results vuln_data = parse_nuclei_result(line) # Get corresponding subdomain http_url = sanitize_url(line.get('matched-at')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) # Look for duplicate vulnerabilities by excluding records that might change but are irrelevant. object_comparison_exclude = ['response', 'curl_command', 'tags', 'references', 'cve_ids', 'cwe_ids'] # Add subdomain and target domain to the duplicate check vuln_data_copy = vuln_data.copy() vuln_data_copy['subdomain'] = subdomain vuln_data_copy['target_domain'] = self.domain # Check if record exists, if exists do not save it if record_exists(Vulnerability, data=vuln_data_copy, exclude_keys=object_comparison_exclude): logger.warning(f'Nuclei vulnerability of severity {severity} : {vuln_data_copy["name"]} for {subdomain_name} already exists') continue # Get or create EndPoint object response = line.get('response') httpx_crawl = False if response else enable_http_crawl # avoid yet another httpx crawl endpoint, _ = save_endpoint( http_url, crawl=httpx_crawl, subdomain=subdomain, ctx=ctx) if endpoint: http_url = endpoint.http_url if not httpx_crawl: output = parse_curl_output(response) endpoint.http_status = output['http_status'] endpoint.save() # Get or create Vulnerability object vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, subdomain=subdomain, **vuln_data) if not vuln: continue # Print vuln severity = line['info'].get('severity', 'unknown') logger.warning(str(vuln)) # Send notification for all vulnerabilities except info url = vuln.http_url or vuln.subdomain send_vuln = ( notif and notif.send_vuln_notif and vuln and severity in ['low', 'medium', 'high', 'critical']) if send_vuln: fields = { 'Severity': f'**{severity.upper()}**', 'URL': http_url, 'Subdomain': subdomain_name, 'Name': vuln.name, 'Type': vuln.type, 'Description': vuln.description, 'Template': vuln.template_url, 'Tags': vuln.get_tags_str(), 'CVEs': vuln.get_cve_str(), 'CWEs': vuln.get_cwe_str(), 'References': vuln.get_refs_str() } severity_map = { 'low': 'info', 'medium': 'warning', 'high': 'error', 'critical': 'error' } self.notify( f'vulnerability_scan_#{vuln.id}', severity_map[severity], fields, add_meta_info=False) # Send report to hackerone hackerone_query = Hackerone.objects.all() send_report = ( hackerone_query.exists() and severity not in ('info', 'low') and vuln.target_domain.h1_team_handle ) if send_report: hackerone = hackerone_query.first() if hackerone.send_critical and severity == 'critical': send_hackerone_report.delay(vuln.id) elif hackerone.send_high and severity == 'high': send_hackerone_report.delay(vuln.id) elif hackerone.send_medium and severity == 'medium': send_hackerone_report.delay(vuln.id) # Write results to JSON file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) # Send finish notif if send_status: vulns = Vulnerability.objects.filter(scan_history__id=self.scan_id) info_count = vulns.filter(severity=0).count() low_count = vulns.filter(severity=1).count() medium_count = vulns.filter(severity=2).count() high_count = vulns.filter(severity=3).count() critical_count = vulns.filter(severity=4).count() unknown_count = vulns.filter(severity=-1).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count + unknown_count fields = { 'Total': vulnerability_count, 'Critical': critical_count, 'High': high_count, 'Medium': medium_count, 'Low': low_count, 'Info': info_count, 'Unknown': unknown_count } self.notify(fields=fields) # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=NUCLEI ).exclude( severity=0 ) # find all unique vulnerabilities based on path and title # all unique vulnerability will go thru gpt function and get report # once report is got, it will be matched with other vulnerabilities and saved unique_vulns = set() for vuln in vulns: unique_vulns.add((vuln.name, vuln.get_path())) unique_vulns = list(unique_vulns) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in unique_vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return None def get_vulnerability_gpt_report(vuln): title = vuln[0] path = vuln[1] logger.info(f'Getting GPT Report for {title}, PATH: {path}') # check if in db already exists stored = GPTVulnerabilityReport.objects.filter( url_path=path ).filter( title=title ).first() if stored: response = { 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: report = GPTVulnerabilityReportGenerator() vulnerability_description = get_gpt_vuln_input_description( title, path ) response = report.get_vulnerability_description(vulnerability_description) add_gpt_description_db( title, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) for vuln in Vulnerability.objects.filter(name=title, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() def add_gpt_description_db(title, path, description, impact, remediation, references): gpt_report = GPTVulnerabilityReport() gpt_report.url_path = path gpt_report.title = title gpt_report.description = description gpt_report.impact = impact gpt_report.remediation = remediation gpt_report.save() for url in references: ref, created = VulnerabilityReference.objects.get_or_create(url=url) gpt_report.references.add(ref) gpt_report.save() @app.task(name='nuclei_scan', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_scan(self, urls=[], ctx={}, description=None): """HTTP vulnerability scan using Nuclei Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. Notes: Unfurl the urls to keep only domain and path, will be sent to vuln scan and ignore certain file extensions. Thanks: https://github.com/six2dez/reconftw """ # Config config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} input_path = f'{self.results_dir}/input_endpoints_vulnerability_scan.txt' enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) concurrency = config.get(NUCLEI_CONCURRENCY) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) retries = config.get(RETRIES) or self.yaml_configuration.get(RETRIES, DEFAULT_RETRIES) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) custom_header = config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) should_fetch_gpt_report = config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) proxy = get_random_proxy() nuclei_specific_config = config.get('nuclei', {}) use_nuclei_conf = nuclei_specific_config.get(USE_CONFIG, False) severities = nuclei_specific_config.get(NUCLEI_SEVERITY, NUCLEI_DEFAULT_SEVERITIES) tags = nuclei_specific_config.get(NUCLEI_TAGS, []) tags = ','.join(tags) nuclei_templates = nuclei_specific_config.get(NUCLEI_TEMPLATE) custom_nuclei_templates = nuclei_specific_config.get(NUCLEI_CUSTOM_TEMPLATE) # severities_str = ','.join(severities) # Get alive endpoints if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=enable_http_crawl, ignore_files=True, write_filepath=input_path, ctx=ctx ) if intensity == 'normal': # reduce number of endpoints to scan unfurl_filter = f'{self.results_dir}/urls_unfurled.txt' run_command( f"cat {input_path} | unfurl -u format %s://%d%p |uro > {unfurl_filter}", shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {unfurl_filter} -o {unfurl_filter}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) input_path = unfurl_filter # Build templates # logger.info('Updating Nuclei templates ...') run_command( 'nuclei -update-templates', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) templates = [] if not (nuclei_templates or custom_nuclei_templates): templates.append(NUCLEI_DEFAULT_TEMPLATES_PATH) if nuclei_templates: if ALL in nuclei_templates: template = NUCLEI_DEFAULT_TEMPLATES_PATH templates.append(template) else: templates.extend(nuclei_templates) if custom_nuclei_templates: custom_nuclei_template_paths = [f'{str(elem)}.yaml' for elem in custom_nuclei_templates] template = templates.extend(custom_nuclei_template_paths) # Build CMD cmd = 'nuclei -j' cmd += ' -config /root/.config/nuclei/config.yaml' if use_nuclei_conf else '' cmd += f' -irr' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -l {input_path}' cmd += f' -c {str(concurrency)}' if concurrency > 0 else '' cmd += f' -proxy {proxy} ' if proxy else '' cmd += f' -retries {retries}' if retries > 0 else '' cmd += f' -rl {rate_limit}' if rate_limit > 0 else '' # cmd += f' -severity {severities_str}' cmd += f' -timeout {str(timeout)}' if timeout and timeout > 0 else '' cmd += f' -tags {tags}' if tags else '' cmd += f' -silent' for tpl in templates: cmd += f' -t {tpl}' grouped_tasks = [] custom_ctx = ctx for severity in severities: custom_ctx['track'] = True _task = nuclei_individual_severity_module.si( cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx=custom_ctx, description=f'Nuclei Scan with severity {severity}' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan with all severities completed...') return None @app.task(name='dalfox_xss_scan', queue='main_scan_queue', base=RengineTask, bind=True) def dalfox_xss_scan(self, urls=[], ctx={}, description=None): """XSS Scan using dalfox Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) dalfox_config = vuln_config.get(DALFOX) or {} custom_header = dalfox_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() is_waf_evasion = dalfox_config.get(WAF_EVASION, False) blind_xss_server = dalfox_config.get(BLIND_XSS_SERVER) user_agent = dalfox_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) timeout = dalfox_config.get(TIMEOUT) delay = dalfox_config.get(DELAY) threads = dalfox_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_dalfox_xss.txt' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=False, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'dalfox --silence --no-color --no-spinner' cmd += f' --only-poc r ' cmd += f' --ignore-return 302,404,403' cmd += f' --skip-bav' cmd += f' file {input_path}' cmd += f' --proxy {proxy}' if proxy else '' cmd += f' --waf-evasion' if is_waf_evasion else '' cmd += f' -b {blind_xss_server}' if blind_xss_server else '' cmd += f' --delay {delay}' if delay else '' cmd += f' --timeout {timeout}' if timeout else '' cmd += f' --user-agent {user_agent}' if user_agent else '' cmd += f' --header {custom_header}' if custom_header else '' cmd += f' --worker {threads}' if threads else '' cmd += f' --format json' results = [] for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id, trunc_char=',' ): if not isinstance(line, dict): continue results.append(line) vuln_data = parse_dalfox_result(line) http_url = sanitize_url(line.get('data')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Dalfox Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=DALFOX ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='crlfuzz_scan', queue='main_scan_queue', base=RengineTask, bind=True) def crlfuzz_scan(self, urls=[], ctx={}, description=None): """CRLF Fuzzing with CRLFuzz Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) custom_header = vuln_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() user_agent = vuln_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) threads = vuln_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_crlf.txt' output_path = f'{self.results_dir}/{self.filename}' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=True, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'crlfuzz -s' cmd += f' -l {input_path}' cmd += f' -x {proxy}' if proxy else '' cmd += f' --H {custom_header}' if custom_header else '' cmd += f' -o {output_path}' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id ) if not os.path.isfile(output_path): logger.info('No Results from CRLFuzz') return crlfs = [] results = [] with open(output_path, 'r') as file: crlfs = file.readlines() for crlf in crlfs: url = crlf.strip() vuln_data = parse_crlfuzz_result(url) http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting CRLFuzz Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=CRLFUZZ ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='s3scanner', queue='main_scan_queue', base=RengineTask, bind=True) def s3scanner(self, ctx={}, description=None): """Bucket Scanner Args: ctx (dict): Context description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/#{self.scan_id}_subdomain_discovery.txt' vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} s3_config = vuln_config.get(S3SCANNER) or {} threads = s3_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) providers = s3_config.get(PROVIDERS, S3SCANNER_DEFAULT_PROVIDERS) scan_history = ScanHistory.objects.filter(pk=self.scan_id).first() for provider in providers: cmd = f's3scanner -bucket-file {input_path} -enumerate -provider {provider} -threads {threads} -json' for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue if line.get('bucket', {}).get('exists', 0) == 1: result = parse_s3scanner_result(line) s3bucket, created = S3Bucket.objects.get_or_create(**result) scan_history.buckets.add(s3bucket) logger.info(f"s3 bucket added {result['provider']}-{result['name']}-{result['region']}") @app.task(name='http_crawl', queue='main_scan_queue', base=RengineTask, bind=True) def http_crawl( self, urls=[], method=None, recrawl=False, ctx={}, track=True, description=None, is_ran_from_subdomain_scan=False, should_remove_duplicate_endpoints=True, duplicate_removal_fields=[]): """Use httpx to query HTTP URLs for important info like page titles, http status, etc... Args: urls (list, optional): A set of URLs to check. Overrides default behavior which queries all endpoints related to this scan. method (str): HTTP method to use (GET, HEAD, POST, PUT, DELETE). recrawl (bool, optional): If False, filter out URLs that have already been crawled. should_remove_duplicate_endpoints (bool): Whether to remove duplicate endpoints duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates Returns: list: httpx results. """ logger.info('Initiating HTTP Crawl') if is_ran_from_subdomain_scan: logger.info('Running From Subdomain Scan...') cmd = '/go/bin/httpx' cfg = self.yaml_configuration.get(HTTP_CRAWL) or {} custom_header = cfg.get(CUSTOM_HEADER, '') threads = cfg.get(THREADS, DEFAULT_THREADS) follow_redirect = cfg.get(FOLLOW_REDIRECT, True) self.output_path = None input_path = f'{self.results_dir}/httpx_input.txt' history_file = f'{self.results_dir}/commands.txt' if urls: # direct passing URLs to check if self.url_filter: urls = [u for u in urls if self.url_filter in u] with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_uncrawled=not recrawl, write_filepath=input_path, ctx=ctx ) # logger.debug(urls) # If no URLs found, skip it if not urls: return # Re-adjust thread number if few URLs to avoid spinning up a monster to # kill a fly. if len(urls) < threads: threads = len(urls) # Get random proxy proxy = get_random_proxy() # Run command cmd += f' -cl -ct -rt -location -td -websocket -cname -asn -cdn -probe -random-agent' cmd += f' -t {threads}' if threads > 0 else '' cmd += f' --http-proxy {proxy}' if proxy else '' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -json' cmd += f' -u {urls[0]}' if len(urls) == 1 else f' -l {input_path}' cmd += f' -x {method}' if method else '' cmd += f' -silent' if follow_redirect: cmd += ' -fr' results = [] endpoint_ids = [] for line in stream_command( cmd, history_file=history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not line or not isinstance(line, dict): continue logger.debug(line) # No response from endpoint if line.get('failed', False): continue # Parse httpx output host = line.get('host', '') content_length = line.get('content_length', 0) http_status = line.get('status_code') http_url, is_redirect = extract_httpx_url(line) page_title = line.get('title') webserver = line.get('webserver') cdn = line.get('cdn', False) rt = line.get('time') techs = line.get('tech', []) cname = line.get('cname', '') content_type = line.get('content_type', '') response_time = -1 if rt: response_time = float(''.join(ch for ch in rt if not ch.isalpha())) if rt[-2:] == 'ms': response_time = response_time / 1000 # Create Subdomain object in DB subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue # Save default HTTP URL to endpoint object in DB endpoint, created = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain, is_default=is_ran_from_subdomain_scan ) if not endpoint: continue endpoint.http_status = http_status endpoint.page_title = page_title endpoint.content_length = content_length endpoint.webserver = webserver endpoint.response_time = response_time endpoint.content_type = content_type endpoint.save() endpoint_str = f'{http_url} [{http_status}] `{content_length}B` `{webserver}` `{rt}`' logger.warning(endpoint_str) if endpoint and endpoint.is_alive and endpoint.http_status != 403: self.notify( fields={'Alive endpoint': f'• {endpoint_str}'}, add_meta_info=False) # Add endpoint to results line['_cmd'] = cmd line['final_url'] = http_url line['endpoint_id'] = endpoint.id line['endpoint_created'] = created line['is_redirect'] = is_redirect results.append(line) # Add technology objects to DB for technology in techs: tech, _ = Technology.objects.get_or_create(name=technology) endpoint.techs.add(tech) if is_ran_from_subdomain_scan: subdomain.technologies.add(tech) subdomain.save() endpoint.save() techs_str = ', '.join([f'`{tech}`' for tech in techs]) self.notify( fields={'Technologies': techs_str}, add_meta_info=False) # Add IP objects for 'a' records to DB a_records = line.get('a', []) for ip_address in a_records: ip, created = save_ip_address( ip_address, subdomain, subscan=self.subscan, cdn=cdn) ips_str = '• ' + '\n• '.join([f'`{ip}`' for ip in a_records]) self.notify( fields={'IPs': ips_str}, add_meta_info=False) # Add IP object for host in DB if host: ip, created = save_ip_address( host, subdomain, subscan=self.subscan, cdn=cdn) self.notify( fields={'IPs': f'• `{ip.address}`'}, add_meta_info=False) # Save subdomain and endpoint if is_ran_from_subdomain_scan: # save subdomain stuffs subdomain.http_url = http_url subdomain.http_status = http_status subdomain.page_title = page_title subdomain.content_length = content_length subdomain.webserver = webserver subdomain.response_time = response_time subdomain.content_type = content_type subdomain.cname = ','.join(cname) subdomain.is_cdn = cdn if cdn: subdomain.cdn_name = line.get('cdn_name') subdomain.save() endpoint.save() endpoint_ids.append(endpoint.id) if should_remove_duplicate_endpoints: # Remove 'fake' alive endpoints that are just redirects to the same page remove_duplicate_endpoints( self.scan_id, self.domain_id, self.subdomain_id, filter_ids=endpoint_ids ) # Remove input file run_command( f'rm {input_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) return results #---------------------# # Notifications tasks # #---------------------# @app.task(name='send_notif', bind=False, queue='send_notif_queue') def send_notif( message, scan_history_id=None, subscan_id=None, **options): if not 'title' in options: message = enrich_notification(message, scan_history_id, subscan_id) send_discord_message(message, **options) send_slack_message(message) send_telegram_message(message) @app.task(name='send_scan_notif', bind=False, queue='send_scan_notif_queue') def send_scan_notif( scan_history_id, subscan_id=None, engine_id=None, status='RUNNING'): """Send scan status notification. Works for scan or a subscan if subscan_id is passed. Args: scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Get domain, engine, scan_history objects engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() tasks = ScanActivity.objects.filter(scan_of=scan) if scan else 0 # Build notif options url = get_scan_url(scan_history_id, subscan_id) title = get_scan_title(scan_history_id, subscan_id) fields = get_scan_fields(engine, scan, subscan, status, tasks) severity = None msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) if status: severity = STATUS_TO_SEVERITIES.get(status) opts = { 'title': title, 'url': url, 'fields': fields, 'severity': severity } logger.warning(f'Sending notification "{title}" [{severity}]') # Send notification send_notif( msg, scan_history_id, subscan_id, **opts) @app.task(name='send_task_notif', bind=False, queue='send_task_notif_queue') def send_task_notif( task_name, status=None, result=None, output_path=None, traceback=None, scan_history_id=None, engine_id=None, subscan_id=None, severity=None, add_meta_info=True, update_fields={}): """Send task status notification. Args: task_name (str): Task name. status (str, optional): Task status. result (str, optional): Task result. output_path (str, optional): Task output path. traceback (str, optional): Task traceback. scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. severity (str, optional): Severity (will be mapped to notif colors) add_meta_info (bool, optional): Wheter to add scan / subscan info to notif. update_fields (dict, optional): Fields key / value to update. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Build fields url = None fields = {} if add_meta_info: engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() url = get_scan_url(scan_history_id) if status: fields['Status'] = f'**{status}**' if engine: fields['Engine'] = engine.engine_name if scan: fields['Scan ID'] = f'[#{scan.id}]({url})' if subscan: url = get_scan_url(scan_history_id, subscan_id) fields['Subscan ID'] = f'[#{subscan.id}]({url})' title = get_task_title(task_name, scan_history_id, subscan_id) if status: severity = STATUS_TO_SEVERITIES.get(status) msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) # Add fields to update for k, v in update_fields.items(): fields[k] = v # Add traceback to notif if traceback and notif.send_scan_tracebacks: fields['Traceback'] = f'```\n{traceback}\n```' # Add files to notif files = [] attach_file = ( notif.send_scan_output_file and output_path and result and not traceback ) if attach_file: output_title = output_path.split('/')[-1] files = [(output_path, output_title)] # Send notif opts = { 'title': title, 'url': url, 'files': files, 'severity': severity, 'fields': fields, 'fields_append': update_fields.keys() } send_notif( msg, scan_history_id=scan_history_id, subscan_id=subscan_id, **opts) @app.task(name='send_file_to_discord', bind=False, queue='send_file_to_discord_queue') def send_file_to_discord(file_path, title=None): notif = Notification.objects.first() do_send = notif and notif.send_to_discord and notif.discord_hook_url if not do_send: return False webhook = DiscordWebhook( url=notif.discord_hook_url, rate_limit_retry=True, username=title or "reNgine Discord Plugin" ) with open(file_path, "rb") as f: head, tail = os.path.split(file_path) webhook.add_file(file=f.read(), filename=tail) webhook.execute() @app.task(name='send_hackerone_report', bind=False, queue='send_hackerone_report_queue') def send_hackerone_report(vulnerability_id): """Send HackerOne vulnerability report. Args: vulnerability_id (int): Vulnerability id. Returns: int: HTTP response status code. """ vulnerability = Vulnerability.objects.get(id=vulnerability_id) severities = {v: k for k,v in NUCLEI_SEVERITY_MAP.items()} headers = { 'Content-Type': 'application/json', 'Accept': 'application/json' } # can only send vulnerability report if team_handle exists if len(vulnerability.target_domain.h1_team_handle) !=0: hackerone_query = Hackerone.objects.all() if hackerone_query.exists(): hackerone = Hackerone.objects.first() severity_value = severities[vulnerability.severity] tpl = hackerone.report_template # Replace syntax of report template with actual content tpl = tpl.replace('{vulnerability_name}', vulnerability.name) tpl = tpl.replace('{vulnerable_url}', vulnerability.http_url) tpl = tpl.replace('{vulnerability_severity}', severity_value) tpl = tpl.replace('{vulnerability_description}', vulnerability.description if vulnerability.description else '') tpl = tpl.replace('{vulnerability_extracted_results}', vulnerability.extracted_results if vulnerability.extracted_results else '') tpl = tpl.replace('{vulnerability_reference}', vulnerability.reference if vulnerability.reference else '') data = { "data": { "type": "report", "attributes": { "team_handle": vulnerability.target_domain.h1_team_handle, "title": '{} found in {}'.format(vulnerability.name, vulnerability.http_url), "vulnerability_information": tpl, "severity_rating": severity_value, "impact": "More information about the impact and vulnerability can be found here: \n" + vulnerability.reference if vulnerability.reference else "NA", } } } r = requests.post( 'https://api.hackerone.com/v1/hackers/reports', auth=(hackerone.username, hackerone.api_key), json=data, headers=headers ) response = r.json() status_code = r.status_code if status_code == 201: vulnerability.hackerone_report_id = response['data']["id"] vulnerability.open_status = False vulnerability.save() return status_code else: logger.error('No team handle found.') status_code = 111 return status_code #-------------# # Utils tasks # #-------------# @app.task(name='parse_nmap_results', bind=False, queue='parse_nmap_results_queue') def parse_nmap_results(xml_file, output_file=None): """Parse results from nmap output file. Args: xml_file (str): nmap XML report file path. Returns: list: List of vulnerabilities found from nmap results. """ with open(xml_file, encoding='utf8') as f: content = f.read() try: nmap_results = xmltodict.parse(content) # parse XML to dict except Exception as e: logger.exception(e) logger.error(f'Cannot parse {xml_file} to valid JSON. Skipping.') return [] # Write JSON to output file if output_file: with open(output_file, 'w') as f: json.dump(nmap_results, f, indent=4) logger.warning(json.dumps(nmap_results, indent=4)) hosts = ( nmap_results .get('nmaprun', {}) .get('host', {}) ) all_vulns = [] if isinstance(hosts, dict): hosts = [hosts] for host in hosts: # Grab hostname / IP from output hostnames_dict = host.get('hostnames', {}) if hostnames_dict: # Ensure that hostnames['hostname'] is a list for consistency hostnames_list = hostnames_dict['hostname'] if isinstance(hostnames_dict['hostname'], list) else [hostnames_dict['hostname']] # Extract all the @name values from the list of dictionaries hostnames = [entry.get('@name') for entry in hostnames_list] else: hostnames = [host.get('address')['@addr']] # Iterate over each hostname for each port for hostname in hostnames: # Grab ports from output ports = host.get('ports', {}).get('port', []) if isinstance(ports, dict): ports = [ports] for port in ports: url_vulns = [] port_number = port['@portid'] url = sanitize_url(f'{hostname}:{port_number}') logger.info(f'Parsing nmap results for {hostname}:{port_number} ...') if not port_number or not port_number.isdigit(): continue port_protocol = port['@protocol'] scripts = port.get('script', []) if isinstance(scripts, dict): scripts = [scripts] for script in scripts: script_id = script['@id'] script_output = script['@output'] script_output_table = script.get('table', []) logger.debug(f'Ran nmap script "{script_id}" on {port_number}/{port_protocol}:\n{script_output}\n') if script_id == 'vulscan': vulns = parse_nmap_vulscan_output(script_output) url_vulns.extend(vulns) elif script_id == 'vulners': vulns = parse_nmap_vulners_output(script_output) url_vulns.extend(vulns) # elif script_id == 'http-server-header': # TODO: nmap can help find technologies as well using the http-server-header script # regex = r'(\w+)/([\d.]+)\s?(?:\((\w+)\))?' # tech_name, tech_version, tech_os = re.match(regex, test_string).groups() # Technology.objects.get_or_create(...) # elif script_id == 'http_csrf': # vulns = parse_nmap_http_csrf_output(script_output) # url_vulns.extend(vulns) else: logger.warning(f'Script output parsing for script "{script_id}" is not supported yet.') # Add URL to vuln for vuln in url_vulns: # TODO: This should extend to any URL, not just HTTP vuln['http_url'] = url if 'http_path' in vuln: vuln['http_url'] += vuln['http_path'] all_vulns.append(vuln) return all_vulns def parse_nmap_http_csrf_output(script_output): pass def parse_nmap_vulscan_output(script_output): """Parse nmap vulscan script output. Args: script_output (str): Vulscan script output. Returns: list: List of Vulnerability dicts. """ data = {} vulns = [] provider_name = '' # Sort all vulns found by provider so that we can match each provider with # a function that pulls from its API to get more info about the # vulnerability. for line in script_output.splitlines(): if not line: continue if not line.startswith('['): # provider line if "No findings" in line: logger.info(f"No findings: {line}") continue elif ' - ' in line: provider_name, provider_url = tuple(line.split(' - ')) data[provider_name] = {'url': provider_url.rstrip(':'), 'entries': []} continue else: # Log a warning logger.warning(f"Unexpected line format: {line}") continue reg = r'\[(.*)\] (.*)' matches = re.match(reg, line) id, title = matches.groups() entry = {'id': id, 'title': title} data[provider_name]['entries'].append(entry) logger.warning('Vulscan parsed output:') logger.warning(pprint.pformat(data)) for provider_name in data: if provider_name == 'Exploit-DB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'IBM X-Force': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'MITRE CVE': logger.error(f'Provider {provider_name} is not supported YET.') for entry in data[provider_name]['entries']: cve_id = entry['id'] vuln = cve_to_vuln(cve_id) vulns.append(vuln) elif provider_name == 'OSVDB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'OpenVAS (Nessus)': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'SecurityFocus': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'VulDB': logger.error(f'Provider {provider_name} is not supported YET.') pass else: logger.error(f'Provider {provider_name} is not supported.') return vulns def parse_nmap_vulners_output(script_output, url=''): """Parse nmap vulners script output. TODO: Rework this as it's currently matching all CVEs no matter the confidence. Args: script_output (str): Script output. Returns: list: List of found vulnerabilities. """ vulns = [] # Check for CVE in script output CVE_REGEX = re.compile(r'.*(CVE-\d\d\d\d-\d+).*') matches = CVE_REGEX.findall(script_output) matches = list(dict.fromkeys(matches)) for cve_id in matches: # get CVE info vuln = cve_to_vuln(cve_id, vuln_type='nmap-vulners-nse') if vuln: vulns.append(vuln) return vulns def cve_to_vuln(cve_id, vuln_type=''): """Search for a CVE using CVESearch and return Vulnerability data. Args: cve_id (str): CVE ID in the form CVE-* Returns: dict: Vulnerability dict. """ cve_info = CVESearch('https://cve.circl.lu').id(cve_id) if not cve_info: logger.error(f'Could not fetch CVE info for cve {cve_id}. Skipping.') return None vuln_cve_id = cve_info['id'] vuln_name = vuln_cve_id vuln_description = cve_info.get('summary', 'none').replace(vuln_cve_id, '').strip() try: vuln_cvss = float(cve_info.get('cvss', -1)) except (ValueError, TypeError): vuln_cvss = -1 vuln_cwe_id = cve_info.get('cwe', '') exploit_ids = cve_info.get('refmap', {}).get('exploit-db', []) osvdb_ids = cve_info.get('refmap', {}).get('osvdb', []) references = cve_info.get('references', []) capec_objects = cve_info.get('capec', []) # Parse ovals for a better vuln name / type ovals = cve_info.get('oval', []) if ovals: vuln_name = ovals[0]['title'] vuln_type = ovals[0]['family'] # Set vulnerability severity based on CVSS score vuln_severity = 'info' if vuln_cvss < 4: vuln_severity = 'low' elif vuln_cvss < 7: vuln_severity = 'medium' elif vuln_cvss < 9: vuln_severity = 'high' else: vuln_severity = 'critical' # Build console warning message msg = f'{vuln_name} | {vuln_severity.upper()} | {vuln_cve_id} | {vuln_cwe_id} | {vuln_cvss}' for id in osvdb_ids: msg += f'\n\tOSVDB: {id}' for exploit_id in exploit_ids: msg += f'\n\tEXPLOITDB: {exploit_id}' logger.warning(msg) vuln = { 'name': vuln_name, 'type': vuln_type, 'severity': NUCLEI_SEVERITY_MAP[vuln_severity], 'description': vuln_description, 'cvss_score': vuln_cvss, 'references': references, 'cve_ids': [vuln_cve_id], 'cwe_ids': [vuln_cwe_id] } return vuln def parse_s3scanner_result(line): ''' Parses and returns s3Scanner Data ''' bucket = line['bucket'] return { 'name': bucket['name'], 'region': bucket['region'], 'provider': bucket['provider'], 'owner_display_name': bucket['owner_display_name'], 'owner_id': bucket['owner_id'], 'perm_auth_users_read': bucket['perm_auth_users_read'], 'perm_auth_users_write': bucket['perm_auth_users_write'], 'perm_auth_users_read_acl': bucket['perm_auth_users_read_acl'], 'perm_auth_users_write_acl': bucket['perm_auth_users_write_acl'], 'perm_auth_users_full_control': bucket['perm_auth_users_full_control'], 'perm_all_users_read': bucket['perm_all_users_read'], 'perm_all_users_write': bucket['perm_all_users_write'], 'perm_all_users_read_acl': bucket['perm_all_users_read_acl'], 'perm_all_users_write_acl': bucket['perm_all_users_write_acl'], 'perm_all_users_full_control': bucket['perm_all_users_full_control'], 'num_objects': bucket['num_objects'], 'size': bucket['bucket_size'] } def parse_nuclei_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ return { 'name': line['info'].get('name', ''), 'type': line['type'], 'severity': NUCLEI_SEVERITY_MAP[line['info'].get('severity', 'unknown')], 'template': line['template'], 'template_url': line['template-url'], 'template_id': line['template-id'], 'description': line['info'].get('description', ''), 'matcher_name': line.get('matcher-name', ''), 'curl_command': line.get('curl-command'), 'request': line.get('request'), 'response': line.get('response'), 'extracted_results': line.get('extracted-results', []), 'cvss_metrics': line['info'].get('classification', {}).get('cvss-metrics', ''), 'cvss_score': line['info'].get('classification', {}).get('cvss-score'), 'cve_ids': line['info'].get('classification', {}).get('cve_id', []) or [], 'cwe_ids': line['info'].get('classification', {}).get('cwe_id', []) or [], 'references': line['info'].get('reference', []) or [], 'tags': line['info'].get('tags', []), 'source': NUCLEI, } def parse_dalfox_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ description = '' description += f" Evidence: {line.get('evidence')} <br>" if line.get('evidence') else '' description += f" Message: {line.get('message')} <br>" if line.get('message') else '' description += f" Payload: {line.get('message_str')} <br>" if line.get('message_str') else '' description += f" Vulnerable Parameter: {line.get('param')} <br>" if line.get('param') else '' return { 'name': 'XSS (Cross Site Scripting)', 'type': 'XSS', 'severity': DALFOX_SEVERITY_MAP[line.get('severity', 'unknown')], 'description': description, 'source': DALFOX, 'cwe_ids': [line.get('cwe')] } def parse_crlfuzz_result(url): """Parse CRLF results Args: url (str): CRLF Vulnerable URL Returns: dict: Vulnerability data. """ return { 'name': 'CRLF (HTTP Response Splitting)', 'type': 'CRLF', 'severity': 2, 'description': 'A CRLF (HTTP Response Splitting) vulnerability has been discovered.', 'source': CRLFUZZ, } def record_exists(model, data, exclude_keys=[]): """ Check if a record already exists in the database based on the given data. Args: model (django.db.models.Model): The Django model to check against. data (dict): Data dictionary containing fields and values. exclude_keys (list): List of keys to exclude from the lookup. Returns: bool: True if the record exists, False otherwise. """ # Extract the keys that will be used for the lookup lookup_fields = {key: data[key] for key in data if key not in exclude_keys} # Return True if a record exists based on the lookup fields, False otherwise return model.objects.filter(**lookup_fields).exists() @app.task(name='geo_localize', bind=False, queue='geo_localize_queue') def geo_localize(host, ip_id=None): """Uses geoiplookup to find location associated with host. Args: host (str): Hostname. ip_id (int): IpAddress object id. Returns: startScan.models.CountryISO: CountryISO object from DB or None. """ if validators.ipv6(host): logger.info(f'Ipv6 "{host}" is not supported by geoiplookup. Skipping.') return None cmd = f'geoiplookup {host}' _, out = run_command(cmd) if 'IP Address not found' not in out and "can't resolve hostname" not in out: country_iso = out.split(':')[1].strip().split(',')[0] country_name = out.split(':')[1].strip().split(',')[1].strip() geo_object, _ = CountryISO.objects.get_or_create( iso=country_iso, name=country_name ) geo_json = { 'iso': country_iso, 'name': country_name } if ip_id: ip = IpAddress.objects.get(pk=ip_id) ip.geo_iso = geo_object ip.save() return geo_json logger.info(f'Geo IP lookup failed for host "{host}"') return None @app.task(name='query_whois', bind=False, queue='query_whois_queue') def query_whois(ip_domain, force_reload_whois=False): """Query WHOIS information for an IP or a domain name. Args: ip_domain (str): IP address or domain name. save_domain (bool): Whether to save domain or not, default False Returns: dict: WHOIS information. """ if not force_reload_whois and Domain.objects.filter(name=ip_domain).exists() and Domain.objects.get(name=ip_domain).domain_info: domain = Domain.objects.get(name=ip_domain) if not domain.insert_date: domain.insert_date = timezone.now() domain.save() domain_info_db = domain.domain_info domain_info = DottedDict( dnssec=domain_info_db.dnssec, created=domain_info_db.created, updated=domain_info_db.updated, expires=domain_info_db.expires, geolocation_iso=domain_info_db.geolocation_iso, status=[status['name'] for status in DomainWhoisStatusSerializer(domain_info_db.status, many=True).data], whois_server=domain_info_db.whois_server, ns_records=[ns['name'] for ns in NameServersSerializer(domain_info_db.name_servers, many=True).data], registrar_name=domain_info_db.registrar.name, registrar_phone=domain_info_db.registrar.phone, registrar_email=domain_info_db.registrar.email, registrar_url=domain_info_db.registrar.url, registrant_name=domain_info_db.registrant.name, registrant_id=domain_info_db.registrant.id_str, registrant_organization=domain_info_db.registrant.organization, registrant_city=domain_info_db.registrant.city, registrant_state=domain_info_db.registrant.state, registrant_zip_code=domain_info_db.registrant.zip_code, registrant_country=domain_info_db.registrant.country, registrant_phone=domain_info_db.registrant.phone, registrant_fax=domain_info_db.registrant.fax, registrant_email=domain_info_db.registrant.email, registrant_address=domain_info_db.registrant.address, admin_name=domain_info_db.admin.name, admin_id=domain_info_db.admin.id_str, admin_organization=domain_info_db.admin.organization, admin_city=domain_info_db.admin.city, admin_state=domain_info_db.admin.state, admin_zip_code=domain_info_db.admin.zip_code, admin_country=domain_info_db.admin.country, admin_phone=domain_info_db.admin.phone, admin_fax=domain_info_db.admin.fax, admin_email=domain_info_db.admin.email, admin_address=domain_info_db.admin.address, tech_name=domain_info_db.tech.name, tech_id=domain_info_db.tech.id_str, tech_organization=domain_info_db.tech.organization, tech_city=domain_info_db.tech.city, tech_state=domain_info_db.tech.state, tech_zip_code=domain_info_db.tech.zip_code, tech_country=domain_info_db.tech.country, tech_phone=domain_info_db.tech.phone, tech_fax=domain_info_db.tech.fax, tech_email=domain_info_db.tech.email, tech_address=domain_info_db.tech.address, related_tlds=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_tlds, many=True).data], related_domains=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_domains, many=True).data], historical_ips=[ip for ip in HistoricalIPSerializer(domain_info_db.historical_ips, many=True).data], ) if domain_info_db.dns_records: a_records = [] txt_records = [] mx_records = [] dns_records = [{'name': dns['name'], 'type': dns['type']} for dns in DomainDNSRecordSerializer(domain_info_db.dns_records, many=True).data] for dns in dns_records: if dns['type'] == 'a': a_records.append(dns['name']) elif dns['type'] == 'txt': txt_records.append(dns['name']) elif dns['type'] == 'mx': mx_records.append(dns['name']) domain_info.a_records = a_records domain_info.txt_records = txt_records domain_info.mx_records = mx_records else: logger.info(f'Domain info for "{ip_domain}" not found in DB, querying whois') domain_info = DottedDict() # find domain historical ip try: historical_ips = get_domain_historical_ip_address(ip_domain) domain_info.historical_ips = historical_ips except Exception as e: logger.error(f'HistoricalIP for {ip_domain} not found!\nError: {str(e)}') historical_ips = [] # find associated domains using ip_domain try: related_domains = reverse_whois(ip_domain.split('.')[0]) except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] # find related tlds using TLSx try: related_tlds = [] output_path = '/tmp/ip_domain_tlsx.txt' tlsx_command = f'tlsx -san -cn -silent -ro -host {ip_domain} -o {output_path}' run_command( tlsx_command, shell=True, ) tlsx_output = [] with open(output_path) as f: tlsx_output = f.readlines() tldextract_target = tldextract.extract(ip_domain) for doms in tlsx_output: doms = doms.strip() tldextract_res = tldextract.extract(doms) if ip_domain != doms and tldextract_res.domain == tldextract_target.domain and tldextract_res.subdomain == '': related_tlds.append(doms) related_tlds = list(set(related_tlds)) domain_info.related_tlds = related_tlds except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] related_domains_list = [] if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_related = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_related) related_domains_list.append(_domain['name']) for _domain in related_tlds: domain_related = RelatedDomain.objects.get_or_create( name=_domain, )[0] db_domain_info.related_tlds.add(domain_related) for _ip in historical_ips: historical_ip = HistoricalIP.objects.get_or_create( ip=_ip['ip'], owner=_ip['owner'], location=_ip['location'], last_seen=_ip['last_seen'], )[0] db_domain_info.historical_ips.add(historical_ip) domain.domain_info = db_domain_info domain.save() command = f'netlas host {ip_domain} -f json' # check if netlas key is provided netlas_key = get_netlas_key() command += f' -a {netlas_key}' if netlas_key else '' result = subprocess.check_output(command.split()).decode('utf-8') if 'Failed to parse response data' in result: # do fallback return { 'status': False, 'ip_domain': ip_domain, 'result': "Netlas limit exceeded.", 'message': 'Netlas limit exceeded.' } try: result = json.loads(result) logger.info(result) whois = result.get('whois') if result.get('whois') else {} domain_info.created = whois.get('created_date') domain_info.expires = whois.get('expiration_date') domain_info.updated = whois.get('updated_date') domain_info.whois_server = whois.get('whois_server') if 'registrant' in whois: registrant = whois.get('registrant') domain_info.registrant_name = registrant.get('name') domain_info.registrant_country = registrant.get('country') domain_info.registrant_id = registrant.get('id') domain_info.registrant_state = registrant.get('province') domain_info.registrant_city = registrant.get('city') domain_info.registrant_phone = registrant.get('phone') domain_info.registrant_address = registrant.get('street') domain_info.registrant_organization = registrant.get('organization') domain_info.registrant_fax = registrant.get('fax') domain_info.registrant_zip_code = registrant.get('postal_code') email_search = EMAIL_REGEX.search(str(registrant.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.registrant_email = field_content if 'administrative' in whois: administrative = whois.get('administrative') domain_info.admin_name = administrative.get('name') domain_info.admin_country = administrative.get('country') domain_info.admin_id = administrative.get('id') domain_info.admin_state = administrative.get('province') domain_info.admin_city = administrative.get('city') domain_info.admin_phone = administrative.get('phone') domain_info.admin_address = administrative.get('street') domain_info.admin_organization = administrative.get('organization') domain_info.admin_fax = administrative.get('fax') domain_info.admin_zip_code = administrative.get('postal_code') mail_search = EMAIL_REGEX.search(str(administrative.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.admin_email = field_content if 'technical' in whois: technical = whois.get('technical') domain_info.tech_name = technical.get('name') domain_info.tech_country = technical.get('country') domain_info.tech_state = technical.get('province') domain_info.tech_id = technical.get('id') domain_info.tech_city = technical.get('city') domain_info.tech_phone = technical.get('phone') domain_info.tech_address = technical.get('street') domain_info.tech_organization = technical.get('organization') domain_info.tech_fax = technical.get('fax') domain_info.tech_zip_code = technical.get('postal_code') mail_search = EMAIL_REGEX.search(str(technical.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.tech_email = field_content if 'dns' in result: dns = result.get('dns') domain_info.mx_records = dns.get('mx') domain_info.txt_records = dns.get('txt') domain_info.a_records = dns.get('a') domain_info.ns_records = whois.get('name_servers') domain_info.dnssec = True if whois.get('dnssec') else False domain_info.status = whois.get('status') if 'registrar' in whois: registrar = whois.get('registrar') domain_info.registrar_name = registrar.get('name') domain_info.registrar_email = registrar.get('email') domain_info.registrar_phone = registrar.get('phone') domain_info.registrar_url = registrar.get('url') # find associated domains if registrant email is found related_domains = reverse_whois(domain_info.get('registrant_email')) if domain_info.get('registrant_email') else [] for _domain in related_domains: related_domains_list.append(_domain['name']) # remove duplicate domains from related domains list related_domains_list = list(set(related_domains_list)) domain_info.related_domains = related_domains_list # save to db if domain exists if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_rel = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_rel) db_domain_info.dnssec = domain_info.get('dnssec') #dates db_domain_info.created = domain_info.get('created') db_domain_info.updated = domain_info.get('updated') db_domain_info.expires = domain_info.get('expires') #registrar db_domain_info.registrar = Registrar.objects.get_or_create( name=domain_info.get('registrar_name'), email=domain_info.get('registrar_email'), phone=domain_info.get('registrar_phone'), url=domain_info.get('registrar_url'), )[0] db_domain_info.registrant = DomainRegistration.objects.get_or_create( name=domain_info.get('registrant_name'), organization=domain_info.get('registrant_organization'), address=domain_info.get('registrant_address'), city=domain_info.get('registrant_city'), state=domain_info.get('registrant_state'), zip_code=domain_info.get('registrant_zip_code'), country=domain_info.get('registrant_country'), email=domain_info.get('registrant_email'), phone=domain_info.get('registrant_phone'), fax=domain_info.get('registrant_fax'), id_str=domain_info.get('registrant_id'), )[0] db_domain_info.admin = DomainRegistration.objects.get_or_create( name=domain_info.get('admin_name'), organization=domain_info.get('admin_organization'), address=domain_info.get('admin_address'), city=domain_info.get('admin_city'), state=domain_info.get('admin_state'), zip_code=domain_info.get('admin_zip_code'), country=domain_info.get('admin_country'), email=domain_info.get('admin_email'), phone=domain_info.get('admin_phone'), fax=domain_info.get('admin_fax'), id_str=domain_info.get('admin_id'), )[0] db_domain_info.tech = DomainRegistration.objects.get_or_create( name=domain_info.get('tech_name'), organization=domain_info.get('tech_organization'), address=domain_info.get('tech_address'), city=domain_info.get('tech_city'), state=domain_info.get('tech_state'), zip_code=domain_info.get('tech_zip_code'), country=domain_info.get('tech_country'), email=domain_info.get('tech_email'), phone=domain_info.get('tech_phone'), fax=domain_info.get('tech_fax'), id_str=domain_info.get('tech_id'), )[0] for status in domain_info.get('status') or []: _status = WhoisStatus.objects.get_or_create( name=status )[0] _status.save() db_domain_info.status.add(_status) for ns in domain_info.get('ns_records') or []: _ns = NameServer.objects.get_or_create( name=ns )[0] _ns.save() db_domain_info.name_servers.add(_ns) for a in domain_info.get('a_records') or []: _a = DNSRecord.objects.get_or_create( name=a, type='a' )[0] _a.save() db_domain_info.dns_records.add(_a) for mx in domain_info.get('mx_records') or []: _mx = DNSRecord.objects.get_or_create( name=mx, type='mx' )[0] _mx.save() db_domain_info.dns_records.add(_mx) for txt in domain_info.get('txt_records') or []: _txt = DNSRecord.objects.get_or_create( name=txt, type='txt' )[0] _txt.save() db_domain_info.dns_records.add(_txt) db_domain_info.geolocation_iso = domain_info.get('registrant_country') db_domain_info.whois_server = domain_info.get('whois_server') db_domain_info.save() domain.domain_info = db_domain_info domain.save() except Exception as e: return { 'status': False, 'ip_domain': ip_domain, 'result': "unable to fetch records from WHOIS database.", 'message': str(e) } return { 'status': True, 'ip_domain': ip_domain, 'dnssec': domain_info.get('dnssec'), 'created': domain_info.get('created'), 'updated': domain_info.get('updated'), 'expires': domain_info.get('expires'), 'geolocation_iso': domain_info.get('registrant_country'), 'domain_statuses': domain_info.get('status'), 'whois_server': domain_info.get('whois_server'), 'dns': { 'a': domain_info.get('a_records'), 'mx': domain_info.get('mx_records'), 'txt': domain_info.get('txt_records'), }, 'registrar': { 'name': domain_info.get('registrar_name'), 'phone': domain_info.get('registrar_phone'), 'email': domain_info.get('registrar_email'), 'url': domain_info.get('registrar_url'), }, 'registrant': { 'name': domain_info.get('registrant_name'), 'id': domain_info.get('registrant_id'), 'organization': domain_info.get('registrant_organization'), 'address': domain_info.get('registrant_address'), 'city': domain_info.get('registrant_city'), 'state': domain_info.get('registrant_state'), 'zipcode': domain_info.get('registrant_zip_code'), 'country': domain_info.get('registrant_country'), 'phone': domain_info.get('registrant_phone'), 'fax': domain_info.get('registrant_fax'), 'email': domain_info.get('registrant_email'), }, 'admin': { 'name': domain_info.get('admin_name'), 'id': domain_info.get('admin_id'), 'organization': domain_info.get('admin_organization'), 'address':domain_info.get('admin_address'), 'city': domain_info.get('admin_city'), 'state': domain_info.get('admin_state'), 'zipcode': domain_info.get('admin_zip_code'), 'country': domain_info.get('admin_country'), 'phone': domain_info.get('admin_phone'), 'fax': domain_info.get('admin_fax'), 'email': domain_info.get('admin_email'), }, 'technical_contact': { 'name': domain_info.get('tech_name'), 'id': domain_info.get('tech_id'), 'organization': domain_info.get('tech_organization'), 'address': domain_info.get('tech_address'), 'city': domain_info.get('tech_city'), 'state': domain_info.get('tech_state'), 'zipcode': domain_info.get('tech_zip_code'), 'country': domain_info.get('tech_country'), 'phone': domain_info.get('tech_phone'), 'fax': domain_info.get('tech_fax'), 'email': domain_info.get('tech_email'), }, 'nameservers': domain_info.get('ns_records'), # 'similar_domains': domain_info.get('similar_domains'), 'related_domains': domain_info.get('related_domains'), 'related_tlds': domain_info.get('related_tlds'), 'historical_ips': domain_info.get('historical_ips'), } @app.task(name='remove_duplicate_endpoints', bind=False, queue='remove_duplicate_endpoints_queue') def remove_duplicate_endpoints( scan_history_id, domain_id, subdomain_id=None, filter_ids=[], filter_status=[200, 301, 404], duplicate_removal_fields=ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS ): """Remove duplicate endpoints. Check for implicit redirections by comparing endpoints: - [x] `content_length` similarities indicating redirections - [x] `page_title` (check for same page title) - [ ] Sign-in / login page (check for endpoints with the same words) Args: scan_history_id: ScanHistory id. domain_id (int): Domain id. subdomain_id (int, optional): Subdomain id. filter_ids (list): List of endpoint ids to filter on. filter_status (list): List of HTTP status codes to filter on. duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates """ logger.info(f'Removing duplicate endpoints based on {duplicate_removal_fields}') endpoints = ( EndPoint.objects .filter(scan_history__id=scan_history_id) .filter(target_domain__id=domain_id) ) if filter_status: endpoints = endpoints.filter(http_status__in=filter_status) if subdomain_id: endpoints = endpoints.filter(subdomain__id=subdomain_id) if filter_ids: endpoints = endpoints.filter(id__in=filter_ids) for field_name in duplicate_removal_fields: cl_query = ( endpoints .values_list(field_name) .annotate(mc=Count(field_name)) .order_by('-mc') ) for (field_value, count) in cl_query: if count > DELETE_DUPLICATES_THRESHOLD: eps_to_delete = ( endpoints .filter(**{field_name: field_value}) .order_by('discovered_date') .all()[1:] ) msg = f'Deleting {len(eps_to_delete)} endpoints [reason: same {field_name} {field_value}]' for ep in eps_to_delete: url = urlparse(ep.http_url) if url.path in ['', '/', '/login']: # try do not delete the original page that other pages redirect to continue msg += f'\n\t {ep.http_url} [{ep.http_status}] [{field_name}={field_value}]' ep.delete() logger.warning(msg) @app.task(name='run_command', bind=False, queue='run_command_queue') def run_command(cmd, cwd=None, shell=False, history_file=None, scan_id=None, activity_id=None): """Run a given command using subprocess module. Args: cmd (str): Command to run. cwd (str): Current working directory. echo (bool): Log command. shell (bool): Run within separate shell if True. history_file (str): Write command + output to history file. Returns: tuple: Tuple with return_code, output. """ logger.info(cmd) logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Run the command using subprocess popen = subprocess.Popen( cmd if shell else cmd.split(), shell=shell, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, cwd=cwd, universal_newlines=True) output = '' for stdout_line in iter(popen.stdout.readline, ""): item = stdout_line.strip() output += '\n' + item logger.debug(item) popen.stdout.close() popen.wait() return_code = popen.returncode command_obj.output = output command_obj.return_code = return_code command_obj.save() if history_file: mode = 'a' if not os.path.exists(history_file): mode = 'w' with open(history_file, mode) as f: f.write(f'\n{cmd}\n{return_code}\n{output}\n------------------\n') return return_code, output #-------------# # Other utils # #-------------# def stream_command(cmd, cwd=None, shell=False, history_file=None, encoding='utf-8', scan_id=None, activity_id=None, trunc_char=None): # Log cmd logger.info(cmd) # logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Sanitize the cmd command = cmd if shell else cmd.split() # Run the command using subprocess process = subprocess.Popen( command, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, universal_newlines=True, shell=shell) # Log the output in real-time to the database output = "" # Process the output for line in iter(lambda: process.stdout.readline(), b''): if not line: break line = line.strip() ansi_escape = re.compile(r'\x1B(?:[@-Z\\-_]|\[[0-?]*[ -/]*[@-~])') line = ansi_escape.sub('', line) line = line.replace('\\x0d\\x0a', '\n') if trunc_char and line.endswith(trunc_char): line = line[:-1] item = line # Try to parse the line as JSON try: item = json.loads(line) except json.JSONDecodeError: pass # Yield the line #logger.debug(item) yield item # Add the log line to the output output += line + "\n" # Update the command record in the database command_obj.output = output command_obj.save() # Retrieve the return code and output process.wait() return_code = process.returncode # Update the return code and final output in the database command_obj.return_code = return_code command_obj.save() # Append the command, return code and output to the history file if history_file is not None: with open(history_file, "a") as f: f.write(f"{cmd}\n{return_code}\n{output}\n") def process_httpx_response(line): """TODO: implement this""" def extract_httpx_url(line): """Extract final URL from httpx results. Always follow redirects to find the last URL. Args: line (dict): URL data output by httpx. Returns: tuple: (final_url, redirect_bool) tuple. """ status_code = line.get('status_code', 0) final_url = line.get('final_url') location = line.get('location') chain_status_codes = line.get('chain_status_codes', []) # Final URL is already looking nice, if it exists return it if final_url: return final_url, False http_url = line['url'] # fallback to url field # Handle redirects manually REDIRECT_STATUS_CODES = [301, 302] is_redirect = ( status_code in REDIRECT_STATUS_CODES or any(x in REDIRECT_STATUS_CODES for x in chain_status_codes) ) if is_redirect and location: if location.startswith(('http', 'https')): http_url = location else: http_url = f'{http_url}/{location.lstrip("/")}' # Sanitize URL http_url = sanitize_url(http_url) return http_url, is_redirect #-------------# # OSInt utils # #-------------# def get_and_save_dork_results(lookup_target, results_dir, type, lookup_keywords=None, lookup_extensions=None, delay=3, page_count=2, scan_history=None): """ Uses gofuzz to dork and store information Args: lookup_target (str): target to look into such as stackoverflow or even the target itself results_dir (str): Results directory type (str): Dork Type Title lookup_keywords (str): comma separated keywords or paths to look for lookup_extensions (str): comma separated extensions to look for delay (int): delay between each requests page_count (int): pages in google to extract information scan_history (startScan.ScanHistory): Scan History Object """ results = [] gofuzz_command = f'{GOFUZZ_EXEC_PATH} -t {lookup_target} -d {delay} -p {page_count}' if lookup_extensions: gofuzz_command += f' -e {lookup_extensions}' elif lookup_keywords: gofuzz_command += f' -w {lookup_keywords}' output_file = f'{results_dir}/gofuzz.txt' gofuzz_command += f' -o {output_file}' history_file = f'{results_dir}/commands.txt' try: run_command( gofuzz_command, shell=False, history_file=history_file, scan_id=scan_history.id, ) if not os.path.isfile(output_file): return with open(output_file) as f: for line in f.readlines(): url = line.strip() if url: results.append(url) dork, created = Dork.objects.get_or_create( type=type, url=url ) if scan_history: scan_history.dorks.add(dork) # remove output file os.remove(output_file) except Exception as e: logger.exception(e) return results def get_and_save_emails(scan_history, activity_id, results_dir): """Get and save emails from Google, Bing and Baidu. Args: scan_history (startScan.ScanHistory): Scan history object. activity_id: ScanActivity Object results_dir (str): Results directory. Returns: list: List of emails found. """ emails = [] # Proxy settings # get_random_proxy() # Gather emails from Google, Bing and Baidu output_file = f'{results_dir}/emails_tmp.txt' history_file = f'{results_dir}/commands.txt' command = f'python3 /usr/src/github/Infoga/infoga.py --domain {scan_history.domain.name} --source all --report {output_file}' try: run_command( command, shell=False, history_file=history_file, scan_id=scan_history.id, activity_id=activity_id) if not os.path.isfile(output_file): logger.info('No Email results') return [] with open(output_file) as f: for line in f.readlines(): if 'Email' in line: split_email = line.split(' ')[2] emails.append(split_email) output_path = f'{results_dir}/emails.txt' with open(output_path, 'w') as output_file: for email_address in emails: save_email(email_address, scan_history) output_file.write(f'{email_address}\n') except Exception as e: logger.exception(e) return emails def save_metadata_info(meta_dict): """Extract metadata from Google Search. Args: meta_dict (dict): Info dict. Returns: list: List of startScan.MetaFinderDocument objects. """ logger.warning(f'Getting metadata for {meta_dict.osint_target}') scan_history = ScanHistory.objects.get(id=meta_dict.scan_id) # Proxy settings get_random_proxy() # Get metadata result = extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if not result: logger.error(f'No metadata result from Google Search for {meta_dict.osint_target}.') return [] # Add metadata info to DB results = [] for metadata_name, data in result.get_metadata().items(): subdomain = Subdomain.objects.get( scan_history=meta_dict.scan_id, name=meta_dict.osint_target) metadata = DottedDict({k: v for k, v in data.items()}) meta_finder_document = MetaFinderDocument( subdomain=subdomain, target_domain=meta_dict.domain, scan_history=scan_history, url=metadata.url, doc_name=metadata_name, http_status=metadata.status_code, producer=metadata.metadata.get('Producer'), creator=metadata.metadata.get('Creator'), creation_date=metadata.metadata.get('CreationDate'), modified_date=metadata.metadata.get('ModDate'), author=metadata.metadata.get('Author'), title=metadata.metadata.get('Title'), os=metadata.metadata.get('OSInfo')) meta_finder_document.save() results.append(data) return results #-----------------# # Utils functions # #-----------------# def create_scan_activity(scan_history_id, message, status): scan_activity = ScanActivity() scan_activity.scan_of = ScanHistory.objects.get(pk=scan_history_id) scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id #--------------------# # Database functions # #--------------------# def save_vulnerability(**vuln_data): references = vuln_data.pop('references', []) cve_ids = vuln_data.pop('cve_ids', []) cwe_ids = vuln_data.pop('cwe_ids', []) tags = vuln_data.pop('tags', []) subscan = vuln_data.pop('subscan', None) # remove nulls vuln_data = replace_nulls(vuln_data) # Create vulnerability vuln, created = Vulnerability.objects.get_or_create(**vuln_data) if created: vuln.discovered_date = timezone.now() vuln.open_status = True vuln.save() # Save vuln tags for tag_name in tags or []: tag, created = VulnerabilityTags.objects.get_or_create(name=tag_name) if tag: vuln.tags.add(tag) vuln.save() # Save CVEs for cve_id in cve_ids or []: cve, created = CveId.objects.get_or_create(name=cve_id) if cve: vuln.cve_ids.add(cve) vuln.save() # Save CWEs for cve_id in cwe_ids or []: cwe, created = CweId.objects.get_or_create(name=cve_id) if cwe: vuln.cwe_ids.add(cwe) vuln.save() # Save vuln reference for url in references or []: ref, created = VulnerabilityReference.objects.get_or_create(url=url) if created: vuln.references.add(ref) vuln.save() # Save subscan id in vuln object if subscan: vuln.vuln_subscan_ids.add(subscan) vuln.save() return vuln, created def save_endpoint( http_url, ctx={}, crawl=False, is_default=False, **endpoint_data): """Get or create EndPoint object. If crawl is True, also crawl the endpoint HTTP URL with httpx. Args: http_url (str): Input HTTP URL. is_default (bool): If the url is a default url for SubDomains. scan_history (startScan.models.ScanHistory): ScanHistory object. domain (startScan.models.Domain): Domain object. subdomain (starScan.models.Subdomain): Subdomain object. results_dir (str, optional): Results directory. crawl (bool, optional): Run httpx on endpoint if True. Default: False. force (bool, optional): Force crawl even if ENABLE_HTTP_CRAWL mode is on. subscan (startScan.models.SubScan, optional): SubScan object. Returns: tuple: (startScan.models.EndPoint, created) where `created` is a boolean indicating if the object is new or already existed. """ # remove nulls endpoint_data = replace_nulls(endpoint_data) scheme = urlparse(http_url).scheme endpoint = None created = False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in http_url: logger.error(f"{http_url} is not a URL of domain {domain.name}. Skipping.") return None, False if crawl: ctx['track'] = False results = http_crawl( urls=[http_url], method='HEAD', ctx=ctx) if results: endpoint_data = results[0] endpoint_id = endpoint_data['endpoint_id'] created = endpoint_data['endpoint_created'] endpoint = EndPoint.objects.get(pk=endpoint_id) elif not scheme: return None, False else: # add dumb endpoint without probing it scan = ScanHistory.objects.filter(pk=ctx.get('scan_history_id')).first() domain = Domain.objects.filter(pk=ctx.get('domain_id')).first() if not validators.url(http_url): return None, False http_url = sanitize_url(http_url) endpoint, created = EndPoint.objects.get_or_create( scan_history=scan, target_domain=domain, http_url=http_url, **endpoint_data) if created: endpoint.is_default = is_default endpoint.discovered_date = timezone.now() endpoint.save() subscan_id = ctx.get('subscan_id') if subscan_id: endpoint.endpoint_subscan_ids.add(subscan_id) endpoint.save() return endpoint, created def save_subdomain(subdomain_name, ctx={}): """Get or create Subdomain object. Args: subdomain_name (str): Subdomain name. scan_history (startScan.models.ScanHistory): ScanHistory object. Returns: tuple: (startScan.models.Subdomain, created) where `created` is a boolean indicating if the object has been created in DB. """ scan_id = ctx.get('scan_history_id') subscan_id = ctx.get('subscan_id') out_of_scope_subdomains = ctx.get('out_of_scope_subdomains', []) valid_domain = ( validators.domain(subdomain_name) or validators.ipv4(subdomain_name) or validators.ipv6(subdomain_name) ) if not valid_domain: logger.error(f'{subdomain_name} is not an invalid domain. Skipping.') return None, False if subdomain_name in out_of_scope_subdomains: logger.error(f'{subdomain_name} is out-of-scope. Skipping.') return None, False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in subdomain_name: logger.error(f"{subdomain_name} is not a subdomain of domain {domain.name}. Skipping.") return None, False scan = ScanHistory.objects.filter(pk=scan_id).first() domain = scan.domain if scan else None subdomain, created = Subdomain.objects.get_or_create( scan_history=scan, target_domain=domain, name=subdomain_name) if created: # logger.warning(f'Found new subdomain {subdomain_name}') subdomain.discovered_date = timezone.now() if subscan_id: subdomain.subdomain_subscan_ids.add(subscan_id) subdomain.save() return subdomain, created def save_email(email_address, scan_history=None): if not validators.email(email_address): logger.info(f'Email {email_address} is invalid. Skipping.') return None, False email, created = Email.objects.get_or_create(address=email_address) # if created: # logger.warning(f'Found new email address {email_address}') # Add email to ScanHistory if scan_history: scan_history.emails.add(email) scan_history.save() return email, created def save_employee(name, designation, scan_history=None): employee, created = Employee.objects.get_or_create( name=name, designation=designation) # if created: # logger.warning(f'Found new employee {name}') # Add employee to ScanHistory if scan_history: scan_history.employees.add(employee) scan_history.save() return employee, created def save_ip_address(ip_address, subdomain=None, subscan=None, **kwargs): if not (validators.ipv4(ip_address) or validators.ipv6(ip_address)): logger.info(f'IP {ip_address} is not a valid IP. Skipping.') return None, False ip, created = IpAddress.objects.get_or_create(address=ip_address) # if created: # logger.warning(f'Found new IP {ip_address}') # Set extra attributes for key, value in kwargs.items(): setattr(ip, key, value) ip.save() # Add IP to subdomain if subdomain: subdomain.ip_addresses.add(ip) subdomain.save() # Add subscan to IP if subscan: ip.ip_subscan_ids.add(subscan) # Geo-localize IP asynchronously if created: geo_localize.delay(ip_address, ip.id) return ip, created def save_imported_subdomains(subdomains, ctx={}): """Take a list of subdomains imported and write them to from_imported.txt. Args: subdomains (list): List of subdomain names. scan_history (startScan.models.ScanHistory): ScanHistory instance. domain (startScan.models.Domain): Domain instance. results_dir (str): Results directory. """ domain_id = ctx['domain_id'] domain = Domain.objects.get(pk=domain_id) results_dir = ctx.get('results_dir', RENGINE_RESULTS) # Validate each subdomain and de-duplicate entries subdomains = list(set([ subdomain for subdomain in subdomains if validators.domain(subdomain) and domain.name == get_domain_from_subdomain(subdomain) ])) if not subdomains: return logger.warning(f'Found {len(subdomains)} imported subdomains.') with open(f'{results_dir}/from_imported.txt', 'w+') as output_file: for name in subdomains: subdomain_name = name.strip() subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain.is_imported_subdomain = True subdomain.save() output_file.write(f'{subdomain}\n') @app.task(name='query_reverse_whois', bind=False, queue='query_reverse_whois_queue') def query_reverse_whois(lookup_keyword): """Queries Reverse WHOIS information for an organization or email address. Args: lookup_keyword (str): Registrar Name or email Returns: dict: Reverse WHOIS information. """ return get_associated_domains(lookup_keyword) @app.task(name='query_ip_history', bind=False, queue='query_ip_history_queue') def query_ip_history(domain): """Queries the IP history for a domain Args: domain (str): domain_name Returns: list: list of historical ip addresses """ return get_domain_historical_ip_address(domain) @app.task(name='gpt_vulnerability_description', bind=False, queue='gpt_queue') def gpt_vulnerability_description(vulnerability_id): """Generate and store Vulnerability Description using GPT. Args: vulnerability_id (Vulnerability Model ID): Vulnerability ID to fetch Description. """ logger.info('Getting GPT Vulnerability Description') try: lookup_vulnerability = Vulnerability.objects.get(id=vulnerability_id) lookup_url = urlparse(lookup_vulnerability.http_url) path = lookup_url.path except Exception as e: return { 'status': False, 'error': str(e) } # check in db GPTVulnerabilityReport model if vulnerability description and path matches stored = GPTVulnerabilityReport.objects.filter(url_path=path).filter(title=lookup_vulnerability.name).first() if stored: response = { 'status': True, 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: vulnerability_description = get_gpt_vuln_input_description( lookup_vulnerability.name, path ) # one can add more description here later gpt_generator = GPTVulnerabilityReportGenerator() response = gpt_generator.get_vulnerability_description(vulnerability_description) add_gpt_description_db( lookup_vulnerability.name, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) # for all vulnerabilities with the same vulnerability name this description has to be stored. # also the consition is that the url must contain a part of this. for vuln in Vulnerability.objects.filter(name=lookup_vulnerability.name, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() return response

ocervell

b557c6b8b70ea554c232095bf2fbb213e6d3648f

0ded32c1bee7852e7fc5daea0fb6de999097400b

## Overly permissive regular expression range Suspicious character range that is equivalent to \[@A-Z\]. [Show more details](https://github.com/yogeshojha/rengine/security/code-scanning/167)

github-advanced-security[bot]

yogeshojha/rengine

fix: ffuf ANSI code processing preventing task to finish

Should - [ ] fix #1006 Needs to be tested for potential impact on other tasks (e.g: dalfox)

2023-11-21 11:54:34+00:00

2023-11-24 03:10:39+00:00

web/reNgine/tasks.py

import csv import json import os import pprint import subprocess import time import validators import whatportis import xmltodict import yaml import tldextract import concurrent.futures from datetime import datetime from urllib.parse import urlparse from api.serializers import SubdomainSerializer from celery import chain, chord, group from celery.result import allow_join_result from celery.utils.log import get_task_logger from django.db.models import Count from dotted_dict import DottedDict from django.utils import timezone from pycvesearch import CVESearch from metafinder.extractor import extract_metadata_from_google_search from reNgine.celery import app from reNgine.gpt import GPTVulnerabilityReportGenerator from reNgine.celery_custom_task import RengineTask from reNgine.common_func import * from reNgine.definitions import * from reNgine.settings import * from reNgine.gpt import * from reNgine.utilities import * from scanEngine.models import (EngineType, InstalledExternalTool, Notification, Proxy) from startScan.models import * from startScan.models import EndPoint, Subdomain, Vulnerability from targetApp.models import Domain """ Celery tasks. """ logger = get_task_logger(__name__) #----------------------# # Scan / Subscan tasks # #----------------------# @app.task(name='initiate_scan', bind=False, queue='initiate_scan_queue') def initiate_scan( scan_history_id, domain_id, engine_id=None, scan_type=LIVE_SCAN, results_dir=RENGINE_RESULTS, imported_subdomains=[], out_of_scope_subdomains=[], url_filter=''): """Initiate a new scan. Args: scan_history_id (int): ScanHistory id. domain_id (int): Domain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. imported_subdomains (list): Imported subdomains. out_of_scope_subdomains (list): Out-of-scope subdomains. url_filter (str): URL path. Default: '' """ # Get scan history scan = ScanHistory.objects.get(pk=scan_history_id) # Get scan engine engine_id = engine_id or scan.scan_type.id # scan history engine_id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, []) # Get domain and set last_scan_date domain = Domain.objects.get(pk=domain_id) domain.last_scan_date = timezone.now() domain.save() # Get path filter url_filter = url_filter.rstrip('/') # Get or create ScanHistory() object if scan_type == LIVE_SCAN: # immediate scan = ScanHistory.objects.get(pk=scan_history_id) scan.scan_status = RUNNING_TASK elif scan_type == SCHEDULED_SCAN: # scheduled scan = ScanHistory() scan.scan_status = INITIATED_TASK scan.scan_type = engine scan.celery_ids = [initiate_scan.request.id] scan.domain = domain scan.start_scan_date = timezone.now() scan.tasks = engine.tasks scan.results_dir = f'{results_dir}/{domain.name}_{scan.id}' add_gf_patterns = gf_patterns and 'fetch_url' in engine.tasks if add_gf_patterns: scan.used_gf_patterns = ','.join(gf_patterns) scan.save() # Create scan results dir os.makedirs(scan.results_dir) # Build task context ctx = { 'scan_history_id': scan_history_id, 'engine_id': engine_id, 'domain_id': domain.id, 'results_dir': scan.results_dir, 'url_filter': url_filter, 'yaml_configuration': config, 'out_of_scope_subdomains': out_of_scope_subdomains } ctx_str = json.dumps(ctx, indent=2) # Send start notif logger.warning(f'Starting scan {scan_history_id} with context:\n{ctx_str}') send_scan_notif.delay( scan_history_id, subscan_id=None, engine_id=engine_id, status=CELERY_TASK_STATUS_MAP[scan.scan_status]) # Save imported subdomains in DB save_imported_subdomains(imported_subdomains, ctx=ctx) # Create initial subdomain in DB: make a copy of domain as a subdomain so # that other tasks using subdomains can use it. subdomain_name = domain.name subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) # If enable_http_crawl is set, create an initial root HTTP endpoint so that # HTTP crawling can start somewhere http_url = f'{domain.name}{url_filter}' if url_filter else domain.name endpoint, _ = save_endpoint( http_url, ctx=ctx, crawl=enable_http_crawl, is_default=True, subdomain=subdomain ) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build Celery tasks, crafted according to the dependency graph below: # subdomain_discovery --> port_scan --> fetch_url --> dir_file_fuzz # osint vulnerability_scan # osint dalfox xss scan # screenshot # waf_detection workflow = chain( group( subdomain_discovery.si(ctx=ctx, description='Subdomain discovery'), osint.si(ctx=ctx, description='OS Intelligence') ), port_scan.si(ctx=ctx, description='Port scan'), fetch_url.si(ctx=ctx, description='Fetch URL'), group( dir_file_fuzz.si(ctx=ctx, description='Directories & files fuzz'), vulnerability_scan.si(ctx=ctx, description='Vulnerability scan'), screenshot.si(ctx=ctx, description='Screenshot'), waf_detection.si(ctx=ctx, description='WAF detection') ) ) # Build callback callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery chord logger.info(f'Running Celery workflow with {len(workflow.tasks) + 1} tasks') task = chain(workflow, callback).on_error(callback).delay() scan.celery_ids.append(task.id) scan.save() return { 'success': True, 'task_id': task.id } @app.task(name='initiate_subscan', bind=False, queue='subscan_queue') def initiate_subscan( scan_history_id, subdomain_id, engine_id=None, scan_type=None, results_dir=RENGINE_RESULTS, url_filter=''): """Initiate a new subscan. Args: scan_history_id (int): ScanHistory id. subdomain_id (int): Subdomain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. url_filter (str): URL path. Default: '' """ # Get Subdomain, Domain and ScanHistory subdomain = Subdomain.objects.get(pk=subdomain_id) scan = ScanHistory.objects.get(pk=subdomain.scan_history.id) domain = Domain.objects.get(pk=subdomain.target_domain.id) # Get EngineType engine_id = engine_id or scan.scan_type.id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Create scan activity of SubScan Model subscan = SubScan( start_scan_date=timezone.now(), celery_ids=[initiate_subscan.request.id], scan_history=scan, subdomain=subdomain, type=scan_type, status=RUNNING_TASK, engine=engine) subscan.save() # Get YAML configuration config = yaml.safe_load(engine.yaml_configuration) # Create results directory results_dir = f'{scan.results_dir}/subscans/{subscan.id}' os.makedirs(results_dir, exist_ok=True) # Run task method = globals().get(scan_type) if not method: logger.warning(f'Task {scan_type} is not supported by reNgine. Skipping') return scan.tasks.append(scan_type) scan.save() # Send start notif send_scan_notif.delay( scan.id, subscan_id=subscan.id, engine_id=engine_id, status='RUNNING') # Build context ctx = { 'scan_history_id': scan.id, 'subscan_id': subscan.id, 'engine_id': engine_id, 'domain_id': domain.id, 'subdomain_id': subdomain.id, 'yaml_configuration': config, 'results_dir': results_dir, 'url_filter': url_filter } # Create initial endpoints in DB: find domain HTTP endpoint so that HTTP # crawling can start somewhere base_url = f'{subdomain.name}{url_filter}' if url_filter else subdomain.name endpoint, _ = save_endpoint( base_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build header + callback workflow = method.si(ctx=ctx) callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery tasks task = chain(workflow, callback).on_error(callback).delay() subscan.celery_ids.append(task.id) subscan.save() return { 'success': True, 'task_id': task.id } @app.task(name='report', bind=False, queue='report_queue') def report(ctx={}, description=None): """Report task running after all other tasks. Mark ScanHistory or SubScan object as completed and update with final status, log run details and send notification. Args: description (str, optional): Task description shown in UI. """ # Get objects subscan_id = ctx.get('subscan_id') scan_id = ctx.get('scan_history_id') engine_id = ctx.get('engine_id') scan = ScanHistory.objects.filter(pk=scan_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() # Get failed tasks tasks = ScanActivity.objects.filter(scan_of=scan).all() if subscan: tasks = tasks.filter(celery_id__in=subscan.celery_ids) failed_tasks = tasks.filter(status=FAILED_TASK) # Get task status failed_count = failed_tasks.count() status = SUCCESS_TASK if failed_count == 0 else FAILED_TASK status_h = 'SUCCESS' if failed_count == 0 else 'FAILED' # Update scan / subscan status if subscan: subscan.stop_scan_date = timezone.now() subscan.status = status subscan.save() else: scan.scan_status = status scan.stop_scan_date = timezone.now() scan.save() # Send scan status notif send_scan_notif.delay( scan_history_id=scan_id, subscan_id=subscan_id, engine_id=engine_id, status=status_h) #------------------------- # # Tracked reNgine tasks # #--------------------------# @app.task(name='subdomain_discovery', queue='main_scan_queue', base=RengineTask, bind=True) def subdomain_discovery( self, host=None, ctx=None, description=None): """Uses a set of tools (see SUBDOMAIN_SCAN_DEFAULT_TOOLS) to scan all subdomains associated with a domain. Args: host (str): Hostname to scan. Returns: subdomains (list): List of subdomain names. """ if not host: host = self.subdomain.name if self.subdomain else self.domain.name if self.url_filter: logger.warning(f'Ignoring subdomains scan as an URL path filter was passed ({self.url_filter}).') return # Config config = self.yaml_configuration.get(SUBDOMAIN_DISCOVERY) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL) or self.yaml_configuration.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) tools = config.get(USES_TOOLS, SUBDOMAIN_SCAN_DEFAULT_TOOLS) default_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=True).filter(is_subdomain_gathering=True)] custom_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=False).filter(is_subdomain_gathering=True)] send_subdomain_changes, send_interesting = False, False notif = Notification.objects.first() if notif: send_subdomain_changes = notif.send_subdomain_changes_notif send_interesting = notif.send_interesting_notif # Gather tools to run for subdomain scan if ALL in tools: tools = SUBDOMAIN_SCAN_DEFAULT_TOOLS + custom_subdomain_tools tools = [t.lower() for t in tools] # Make exception for amass since tool name is amass, but command is amass-active/passive default_subdomain_tools.append('amass-passive') default_subdomain_tools.append('amass-active') # Run tools for tool in tools: cmd = None logger.info(f'Scanning subdomains for {host} with {tool}') proxy = get_random_proxy() if tool in default_subdomain_tools: if tool == 'amass-passive': cmd = f'amass enum -passive -d {host} -o {self.results_dir}/subdomains_amass.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' elif tool == 'amass-active': use_amass_config = config.get(USE_AMASS_CONFIG, False) amass_wordlist_name = config.get(AMASS_WORDLIST, 'deepmagic.com-prefixes-top50000') wordlist_path = f'/usr/src/wordlist/{amass_wordlist_name}.txt' cmd = f'amass enum -active -d {host} -o {self.results_dir}/subdomains_amass_active.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' cmd += f' -brute -w {wordlist_path}' elif tool == 'sublist3r': cmd = f'python3 /usr/src/github/Sublist3r/sublist3r.py -d {host} -t {threads} -o {self.results_dir}/subdomains_sublister.txt' elif tool == 'subfinder': cmd = f'subfinder -d {host} -o {self.results_dir}/subdomains_subfinder.txt' use_subfinder_config = config.get(USE_SUBFINDER_CONFIG, False) cmd += ' -config /root/.config/subfinder/config.yaml' if use_subfinder_config else '' cmd += f' -proxy {proxy}' if proxy else '' cmd += f' -timeout {timeout}' if timeout else '' cmd += f' -t {threads}' if threads else '' cmd += f' -silent' elif tool == 'oneforall': cmd = f'python3 /usr/src/github/OneForAll/oneforall.py --target {host} run' cmd_extract = f'cut -d\',\' -f6 /usr/src/github/OneForAll/results/{host}.csv > {self.results_dir}/subdomains_oneforall.txt' cmd_rm = f'rm -rf /usr/src/github/OneForAll/results/{host}.csv' cmd += f' && {cmd_extract} && {cmd_rm}' elif tool == 'ctfr': results_file = self.results_dir + '/subdomains_ctfr.txt' cmd = f'python3 /usr/src/github/ctfr/ctfr.py -d {host} -o {results_file}' cmd_extract = f"cat {results_file} | sed 's/\*.//g' | tail -n +12 | uniq | sort > {results_file}" cmd += f' && {cmd_extract}' elif tool == 'tlsx': results_file = self.results_dir + '/subdomains_tlsx.txt' cmd = f'tlsx -san -cn -silent -ro -host {host}' cmd += f" | sed -n '/^\([a-zA-Z0-9]\([-a-zA-Z0-9]*[a-zA-Z0-9]\)\?\.\)\+{host}$/p' | uniq | sort" cmd += f' > {results_file}' elif tool == 'netlas': results_file = self.results_dir + '/subdomains_netlas.txt' cmd = f'netlas search -d domain -i domain domain:"*.{host}" -f json' netlas_key = get_netlas_key() cmd += f' -a {netlas_key}' if netlas_key else '' cmd_extract = f"grep -oE '([a-zA-Z0-9]([-a-zA-Z0-9]*[a-zA-Z0-9])?\.)+{host}'" cmd += f' | {cmd_extract} > {results_file}' elif tool in custom_subdomain_tools: tool_query = InstalledExternalTool.objects.filter(name__icontains=tool.lower()) if not tool_query.exists(): logger.error(f'Missing {{TARGET}} and {{OUTPUT}} placeholders in {tool} configuration. Skipping.') continue custom_tool = tool_query.first() cmd = custom_tool.subdomain_gathering_command if '{TARGET}' in cmd and '{OUTPUT}' in cmd: cmd = cmd.replace('{TARGET}', host) cmd = cmd.replace('{OUTPUT}', f'{self.results_dir}/subdomains_{tool}.txt') cmd = cmd.replace('{PATH}', custom_tool.github_clone_path) if '{PATH}' in cmd else cmd else: logger.warning( f'Subdomain discovery tool "{tool}" is not supported by reNgine. Skipping.') continue # Run tool try: run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) except Exception as e: logger.error( f'Subdomain discovery tool "{tool}" raised an exception') logger.exception(e) # Gather all the tools' results in one single file. Write subdomains into # separate files, and sort all subdomains. run_command( f'cat {self.results_dir}/subdomains_*.txt > {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {self.output_path} -o {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) with open(self.output_path) as f: lines = f.readlines() # Parse the output_file file and store Subdomain and EndPoint objects found # in db. subdomain_count = 0 subdomains = [] urls = [] for line in lines: subdomain_name = line.strip() valid_url = bool(validators.url(subdomain_name)) valid_domain = ( bool(validators.domain(subdomain_name)) or bool(validators.ipv4(subdomain_name)) or bool(validators.ipv6(subdomain_name)) or valid_url ) if not valid_domain: logger.error(f'Subdomain {subdomain_name} is not a valid domain, IP or URL. Skipping.') continue if valid_url: subdomain_name = urlparse(subdomain_name).netloc if subdomain_name in self.out_of_scope_subdomains: logger.error(f'Subdomain {subdomain_name} is out of scope. Skipping.') continue # Add subdomain subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain_count += 1 subdomains.append(subdomain) urls.append(subdomain.name) # Bulk crawl subdomains if enable_http_crawl: ctx['track'] = True http_crawl(urls, ctx=ctx, is_ran_from_subdomain_scan=True) # Find root subdomain endpoints for subdomain in subdomains: pass # Send notifications subdomains_str = '\n'.join([f'• `{subdomain.name}`' for subdomain in subdomains]) self.notify(fields={ 'Subdomain count': len(subdomains), 'Subdomains': subdomains_str, }) if send_subdomain_changes and self.scan_id and self.domain_id: added = get_new_added_subdomain(self.scan_id, self.domain_id) removed = get_removed_subdomain(self.scan_id, self.domain_id) if added: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in added]) self.notify(fields={'Added subdomains': subdomains_str}) if removed: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in removed]) self.notify(fields={'Removed subdomains': subdomains_str}) if send_interesting and self.scan_id and self.domain_id: interesting_subdomains = get_interesting_subdomains(self.scan_id, self.domain_id) if interesting_subdomains: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in interesting_subdomains]) self.notify(fields={'Interesting subdomains': subdomains_str}) return SubdomainSerializer(subdomains, many=True).data @app.task(name='osint', queue='main_scan_queue', base=RengineTask, bind=True) def osint(self, host=None, ctx={}, description=None): """Run Open-Source Intelligence tools on selected domain. Args: host (str): Hostname to scan. Returns: dict: Results from osint discovery and dorking. """ config = self.yaml_configuration.get(OSINT) or OSINT_DEFAULT_CONFIG results = {} grouped_tasks = [] if 'discover' in config: ctx['track'] = False # results = osint_discovery(host=host, ctx=ctx) _task = osint_discovery.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, activity_id=self.activity_id, results_dir=self.results_dir, ctx=ctx ) grouped_tasks.append(_task) if OSINT_DORK in config or OSINT_CUSTOM_DORK in config: _task = dorking.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, results_dir=self.results_dir ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('OSINT Tasks finished...') # with open(self.output_path, 'w') as f: # json.dump(results, f, indent=4) # # return results @app.task(name='osint_discovery', queue='osint_discovery_queue', bind=False) def osint_discovery(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run OSINT discovery. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: dict: osint metadat and theHarvester and h8mail results. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) osint_lookup = config.get(OSINT_DISCOVER, []) osint_intensity = config.get(INTENSITY, 'normal') documents_limit = config.get(OSINT_DOCUMENTS_LIMIT, 50) results = {} meta_info = [] emails = [] creds = [] # Get and save meta info if 'metainfo' in osint_lookup: if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': host, 'domain': host, 'scan_id': scan_history_id, 'documents_limit': documents_limit }) meta_info.append(save_metadata_info(meta_dict)) # TODO: disabled for now # elif osint_intensity == 'deep': # subdomains = Subdomain.objects # if self.scan: # subdomains = subdomains.filter(scan_history=self.scan) # for subdomain in subdomains: # meta_dict = DottedDict({ # 'osint_target': subdomain.name, # 'domain': self.domain, # 'scan_id': self.scan_id, # 'documents_limit': documents_limit # }) # meta_info.append(save_metadata_info(meta_dict)) grouped_tasks = [] if 'emails' in osint_lookup: emails = get_and_save_emails(scan_history, activity_id, results_dir) emails_str = '\n'.join([f'• `{email}`' for email in emails]) # self.notify(fields={'Emails': emails_str}) # ctx['track'] = False _task = h8mail.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) if 'employees' in osint_lookup: ctx['track'] = False _task = theHarvester.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) # results['emails'] = results.get('emails', []) + emails # results['creds'] = creds # results['meta_info'] = meta_info return results @app.task(name='dorking', bind=False, queue='dorking_queue') def dorking(config, host, scan_history_id, results_dir): """Run Google dorks. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: list: Dorking results for each dork ran. """ # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh scan_history = ScanHistory.objects.get(pk=scan_history_id) dorks = config.get(OSINT_DORK, []) custom_dorks = config.get(OSINT_CUSTOM_DORK, []) results = [] # custom dorking has higher priority try: for custom_dork in custom_dorks: lookup_target = custom_dork.get('lookup_site') # replace with original host if _target_ lookup_target = host if lookup_target == '_target_' else lookup_target if 'lookup_extensions' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_extensions=custom_dork.get('lookup_extensions'), scan_history=scan_history ) elif 'lookup_keywords' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_keywords=custom_dork.get('lookup_keywords'), scan_history=scan_history ) except Exception as e: logger.exception(e) # default dorking try: for dork in dorks: logger.info(f'Getting dork information for {dork}') if dork == 'stackoverflow': results = get_and_save_dork_results( lookup_target='stackoverflow.com', results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'login_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/login/,login.html', page_count=5, scan_history=scan_history ) elif dork == 'admin_panels': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/admin/,admin.html', page_count=5, scan_history=scan_history ) elif dork == 'dashboard_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/dashboard/,dashboard.html', page_count=5, scan_history=scan_history ) elif dork == 'social_media' : social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'reddit.com' ] for site in social_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'project_management' : project_websites = [ 'trello.com', 'atlassian.net' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'code_sharing' : project_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'config_files' : config_file_exts = [ 'env', 'xml', 'conf', 'toml', 'yml', 'yaml', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(config_file_exts), page_count=4, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keyword = 'Jenkins' results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=lookup_keyword, page_count=1, scan_history=scan_history ) elif dork == 'wordpress_files' : lookup_keywords = [ '/wp-content/', '/wp-includes/' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'php_error' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'exposed_documents' : docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(docs_file_ext), page_count=7, scan_history=scan_history ) elif dork == 'db_files' : file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) elif dork == 'git_exposed' : file_ext = [ 'git', ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) except Exception as e: logger.exception(e) return results @app.task(name='theHarvester', queue='theHarvester_queue', bind=False) def theHarvester(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run theHarvester to get save emails, hosts, employees found in domain. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: dict: Dict of emails, employees, hosts and ips found during crawling. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) output_path_json = f'{results_dir}/theHarvester.json' theHarvester_dir = '/usr/src/github/theHarvester' history_file = f'{results_dir}/commands.txt' cmd = f'python3 {theHarvester_dir}/theHarvester.py -d {host} -b all -f {output_path_json}' # Update proxies.yaml proxy_query = Proxy.objects.all() if proxy_query.exists(): proxy = proxy_query.first() if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(f'{theHarvester_dir}/proxies.yaml', 'w') as file: yaml.dump(yaml_data, file) # Run cmd run_command( cmd, shell=False, cwd=theHarvester_dir, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) # Get file location if not os.path.isfile(output_path_json): logger.error(f'Could not open {output_path_json}') return {} # Load theHarvester results with open(output_path_json, 'r') as f: data = json.load(f) # Re-indent theHarvester JSON with open(output_path_json, 'w') as f: json.dump(data, f, indent=4) emails = data.get('emails', []) for email_address in emails: email, _ = save_email(email_address, scan_history=scan_history) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) linkedin_people = data.get('linkedin_people', []) for people in linkedin_people: employee, _ = save_employee( people, designation='linkedin', scan_history=scan_history) # if employee: # self.notify(fields={'LinkedIn people': f'• {employee.name}'}) twitter_people = data.get('twitter_people', []) for people in twitter_people: employee, _ = save_employee( people, designation='twitter', scan_history=scan_history) # if employee: # self.notify(fields={'Twitter people': f'• {employee.name}'}) hosts = data.get('hosts', []) urls = [] for host in hosts: split = tuple(host.split(':')) http_url = split[0] subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) endpoint, _ = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain) # if endpoint: # urls.append(endpoint.http_url) # self.notify(fields={'Hosts': f'• {endpoint.http_url}'}) # if enable_http_crawl: # ctx['track'] = False # http_crawl(urls, ctx=ctx) # TODO: Lots of ips unrelated with our domain are found, disabling # this for now. # ips = data.get('ips', []) # for ip_address in ips: # ip, created = save_ip_address( # ip_address, # subscan=subscan) # if ip: # send_task_notif.delay( # 'osint', # scan_history_id=scan_history_id, # subscan_id=subscan_id, # severity='success', # update_fields={'IPs': f'{ip.address}'}) return data @app.task(name='h8mail', queue='h8mail_queue', bind=False) def h8mail(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run h8mail. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: list[dict]: List of credentials info. """ logger.warning('Getting leaked credentials') scan_history = ScanHistory.objects.get(pk=scan_history_id) input_path = f'{results_dir}/emails.txt' output_file = f'{results_dir}/h8mail.json' cmd = f'h8mail -t {input_path} --json {output_file}' history_file = f'{results_dir}/commands.txt' run_command( cmd, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) with open(output_file) as f: data = json.load(f) creds = data.get('targets', []) # TODO: go through h8mail output and save emails to DB for cred in creds: logger.warning(cred) email_address = cred['target'] pwn_num = cred['pwn_num'] pwn_data = cred.get('data', []) email, created = save_email(email_address, scan_history=scan) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) return creds @app.task(name='screenshot', queue='main_scan_queue', base=RengineTask, bind=True) def screenshot(self, ctx={}, description=None): """Uses EyeWitness to gather screenshot of a domain and/or url. Args: description (str, optional): Task description shown in UI. """ # Config screenshots_path = f'{self.results_dir}/screenshots' output_path = f'{self.results_dir}/screenshots/{self.filename}' alive_endpoints_file = f'{self.results_dir}/endpoints_alive.txt' config = self.yaml_configuration.get(SCREENSHOT) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT + 5) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) # If intensity is normal, grab only the root endpoints of each subdomain strict = True if intensity == 'normal' else False # Get URLs to take screenshot of get_http_urls( is_alive=enable_http_crawl, strict=strict, write_filepath=alive_endpoints_file, get_only_default_urls=True, ctx=ctx ) # Send start notif notification = Notification.objects.first() send_output_file = notification.send_scan_output_file if notification else False # Run cmd cmd = f'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py -f {alive_endpoints_file} -d {screenshots_path} --no-prompt' cmd += f' --timeout {timeout}' if timeout > 0 else '' cmd += f' --threads {threads}' if threads > 0 else '' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(output_path): logger.error(f'Could not load EyeWitness results at {output_path} for {self.domain.name}.') return # Loop through results and save objects in DB screenshot_paths = [] with open(output_path, 'r') as file: reader = csv.reader(file) for row in reader: "Protocol,Port,Domain,Request Status,Screenshot Path, Source Path" protocol, port, subdomain_name, status, screenshot_path, source_path = tuple(row) logger.info(f'{protocol}:{port}:{subdomain_name}:{status}') subdomain_query = Subdomain.objects.filter(name=subdomain_name) if self.scan: subdomain_query = subdomain_query.filter(scan_history=self.scan) if status == 'Successful' and subdomain_query.exists(): subdomain = subdomain_query.first() screenshot_paths.append(screenshot_path) subdomain.screenshot_path = screenshot_path.replace('/usr/src/scan_results/', '') subdomain.save() logger.warning(f'Added screenshot for {subdomain.name} to DB') # Remove all db, html extra files in screenshot results run_command( 'rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format(screenshots_path), shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'rm -rf {screenshots_path}/source', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Send finish notifs screenshots_str = '• ' + '\n• '.join([f'`{path}`' for path in screenshot_paths]) self.notify(fields={'Screenshots': screenshots_str}) if send_output_file: for path in screenshot_paths: title = get_output_file_name( self.scan_id, self.subscan_id, self.filename) send_file_to_discord.delay(path, title) @app.task(name='port_scan', queue='main_scan_queue', base=RengineTask, bind=True) def port_scan(self, hosts=[], ctx={}, description=None): """Run port scan. Args: hosts (list, optional): Hosts to run port scan on. description (str, optional): Task description shown in UI. Returns: list: List of open ports (dict). """ input_file = f'{self.results_dir}/input_subdomains_port_scan.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(PORT_SCAN) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) exclude_ports = config.get(NAABU_EXCLUDE_PORTS, []) exclude_subdomains = config.get(NAABU_EXCLUDE_SUBDOMAINS, False) ports = config.get(PORTS, NAABU_DEFAULT_PORTS) ports = [str(port) for port in ports] rate_limit = config.get(NAABU_RATE) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) passive = config.get(NAABU_PASSIVE, False) use_naabu_config = config.get(USE_NAABU_CONFIG, False) exclude_ports_str = ','.join(return_iterable(exclude_ports)) # nmap args nmap_enabled = config.get(ENABLE_NMAP, False) nmap_cmd = config.get(NMAP_COMMAND, '') nmap_script = config.get(NMAP_SCRIPT, '') nmap_script = ','.join(return_iterable(nmap_script)) nmap_script_args = config.get(NMAP_SCRIPT_ARGS) if hosts: with open(input_file, 'w') as f: f.write('\n'.join(hosts)) else: hosts = get_subdomains( write_filepath=input_file, exclude_subdomains=exclude_subdomains, ctx=ctx) # Build cmd cmd = 'naabu -json -exclude-cdn' cmd += f' -list {input_file}' if len(hosts) > 0 else f' -host {hosts[0]}' if 'full' in ports or 'all' in ports: ports_str = ' -p "-"' elif 'top-100' in ports: ports_str = ' -top-ports 100' elif 'top-1000' in ports: ports_str = ' -top-ports 1000' else: ports_str = ','.join(ports) ports_str = f' -p {ports_str}' cmd += ports_str cmd += ' -config /root/.config/naabu/config.yaml' if use_naabu_config else '' cmd += f' -proxy "{proxy}"' if proxy else '' cmd += f' -c {threads}' if threads else '' cmd += f' -rate {rate_limit}' if rate_limit > 0 else '' cmd += f' -timeout {timeout*1000}' if timeout > 0 else '' cmd += f' -passive' if passive else '' cmd += f' -exclude-ports {exclude_ports_str}' if exclude_ports else '' cmd += f' -silent' # Execute cmd and gather results results = [] urls = [] ports_data = {} for line in stream_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) port_number = line['port'] ip_address = line['ip'] host = line.get('host') or ip_address if port_number == 0: continue # Grab subdomain subdomain = Subdomain.objects.filter( name=host, target_domain=self.domain, scan_history=self.scan ).first() # Add IP DB ip, _ = save_ip_address(ip_address, subdomain, subscan=self.subscan) if self.subscan: ip.ip_subscan_ids.add(self.subscan) ip.save() # Add endpoint to DB # port 80 and 443 not needed as http crawl already does that. if port_number not in [80, 443]: http_url = f'{host}:{port_number}' endpoint, _ = save_endpoint( http_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint: http_url = endpoint.http_url urls.append(http_url) # Add Port in DB port_details = whatportis.get_ports(str(port_number)) service_name = port_details[0].name if len(port_details) > 0 else 'unknown' description = port_details[0].description if len(port_details) > 0 else '' # get or create port port, created = Port.objects.get_or_create( number=port_number, service_name=service_name, description=description ) if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port.save() ip.ports.add(port) ip.save() if host in ports_data: ports_data[host].append(port_number) else: ports_data[host] = [port_number] # Send notification logger.warning(f'Found opened port {port_number} on {ip_address} ({host})') if len(ports_data) == 0: logger.info('Finished running naabu port scan - No open ports found.') if nmap_enabled: logger.info('Nmap scans skipped') return ports_data # Send notification fields_str = '' for host, ports in ports_data.items(): ports_str = ', '.join([f'`{port}`' for port in ports]) fields_str += f'• `{host}`: {ports_str}\n' self.notify(fields={'Ports discovered': fields_str}) # Save output to file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) logger.info('Finished running naabu port scan.') # Process nmap results: 1 process per host sigs = [] if nmap_enabled: logger.warning(f'Starting nmap scans ...') logger.warning(ports_data) for host, port_list in ports_data.items(): ports_str = '_'.join([str(p) for p in port_list]) ctx_nmap = ctx.copy() ctx_nmap['description'] = get_task_title(f'nmap_{host}', self.scan_id, self.subscan_id) ctx_nmap['track'] = False sig = nmap.si( cmd=nmap_cmd, ports=port_list, host=host, script=nmap_script, script_args=nmap_script_args, max_rate=rate_limit, ctx=ctx_nmap) sigs.append(sig) task = group(sigs).apply_async() with allow_join_result(): results = task.get() return ports_data @app.task(name='nmap', queue='main_scan_queue', base=RengineTask, bind=True) def nmap( self, cmd=None, ports=[], host=None, input_file=None, script=None, script_args=None, max_rate=None, ctx={}, description=None): """Run nmap on a host. Args: cmd (str, optional): Existing nmap command to complete. ports (list, optional): List of ports to scan. host (str, optional): Host to scan. input_file (str, optional): Input hosts file. script (str, optional): NSE script to run. script_args (str, optional): NSE script args. max_rate (int): Max rate. description (str, optional): Task description shown in UI. """ notif = Notification.objects.first() ports_str = ','.join(str(port) for port in ports) self.filename = self.filename.replace('.txt', '.xml') filename_vulns = self.filename.replace('.xml', '_vulns.json') output_file = self.output_path output_file_xml = f'{self.results_dir}/{host}_{self.filename}' vulns_file = f'{self.results_dir}/{host}_{filename_vulns}' logger.warning(f'Running nmap on {host}:{ports}') # Build cmd nmap_cmd = get_nmap_cmd( cmd=cmd, ports=ports_str, script=script, script_args=script_args, max_rate=max_rate, host=host, input_file=input_file, output_file=output_file_xml) # Run cmd run_command( nmap_cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Get nmap XML results and convert to JSON vulns = parse_nmap_results(output_file_xml, output_file) with open(vulns_file, 'w') as f: json.dump(vulns, f, indent=4) # Save vulnerabilities found by nmap vulns_str = '' for vuln_data in vulns: # URL is not necessarily an HTTP URL when running nmap (can be any # other vulnerable protocols). Look for existing endpoint and use its # URL as vulnerability.http_url if it exists. url = vuln_data['http_url'] endpoint = EndPoint.objects.filter(http_url__contains=url).first() if endpoint: vuln_data['http_url'] = endpoint.http_url vuln, created = save_vulnerability( target_domain=self.domain, subdomain=self.subdomain, scan_history=self.scan, subscan=self.subscan, endpoint=endpoint, **vuln_data) vulns_str += f'• {str(vuln)}\n' if created: logger.warning(str(vuln)) # Send only 1 notif for all vulns to reduce number of notifs if notif and notif.send_vuln_notif and vulns_str: logger.warning(vulns_str) self.notify(fields={'CVEs': vulns_str}) return vulns @app.task(name='waf_detection', queue='main_scan_queue', base=RengineTask, bind=True) def waf_detection(self, ctx={}, description=None): """ Uses wafw00f to check for the presence of a WAF. Args: description (str, optional): Task description shown in UI. Returns: list: List of startScan.models.Waf objects. """ input_path = f'{self.results_dir}/input_endpoints_waf_detection.txt' config = self.yaml_configuration.get(WAF_DETECTION) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Get alive endpoints from DB get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) cmd = f'wafw00f -i {input_path} -o {self.output_path}' run_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(self.output_path): logger.error(f'Could not find {self.output_path}') return with open(self.output_path) as file: wafs = file.readlines() for line in wafs: line = " ".join(line.split()) splitted = line.split(' ', 1) waf_info = splitted[1].strip() waf_name = waf_info[:waf_info.find('(')].strip() waf_manufacturer = waf_info[waf_info.find('(')+1:waf_info.find(')')].strip().replace('.', '') http_url = sanitize_url(splitted[0].strip()) if not waf_name or waf_name == 'None': continue # Add waf to db waf, _ = Waf.objects.get_or_create( name=waf_name, manufacturer=waf_manufacturer ) # Add waf info to Subdomain in DB subdomain = get_subdomain_from_url(http_url) logger.info(f'Wafw00f Subdomain : {subdomain}') subdomain_query, _ = Subdomain.objects.get_or_create(scan_history=self.scan, name=subdomain) subdomain_query.waf.add(waf) subdomain_query.save() return wafs @app.task(name='dir_file_fuzz', queue='main_scan_queue', base=RengineTask, bind=True) def dir_file_fuzz(self, ctx={}, description=None): """Perform directory scan, and currently uses `ffuf` as a default tool. Args: description (str, optional): Task description shown in UI. Returns: list: List of URLs discovered. """ # Config cmd = 'ffuf' config = self.yaml_configuration.get(DIR_FILE_FUZZ) or {} custom_header = self.yaml_configuration.get(CUSTOM_HEADER) auto_calibration = config.get(AUTO_CALIBRATION, True) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) extensions = config.get(EXTENSIONS, DEFAULT_DIR_FILE_FUZZ_EXTENSIONS) extensions_str = ','.join(map(str, extensions)) follow_redirect = config.get(FOLLOW_REDIRECT, FFUF_DEFAULT_FOLLOW_REDIRECT) max_time = config.get(MAX_TIME, 0) match_http_status = config.get(MATCH_HTTP_STATUS, FFUF_DEFAULT_MATCH_HTTP_STATUS) mc = ','.join([str(c) for c in match_http_status]) recursive_level = config.get(RECURSIVE_LEVEL, FFUF_DEFAULT_RECURSIVE_LEVEL) stop_on_error = config.get(STOP_ON_ERROR, False) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) wordlist_name = config.get(WORDLIST, 'dicc') delay = rate_limit / (threads * 100) # calculate request pause delay from rate_limit and number of threads input_path = f'{self.results_dir}/input_dir_file_fuzz.txt' # Get wordlist wordlist_name = 'dicc' if wordlist_name == 'default' else wordlist_name wordlist_path = f'/usr/src/wordlist/{wordlist_name}.txt' # Build command cmd += f' -w {wordlist_path}' cmd += f' -e {extensions_str}' if extensions else '' cmd += f' -maxtime {max_time}' if max_time > 0 else '' cmd += f' -p {delay}' if delay > 0 else '' cmd += f' -recursion -recursion-depth {recursive_level} ' if recursive_level > 0 else '' cmd += f' -t {threads}' if threads and threads > 0 else '' cmd += f' -timeout {timeout}' if timeout and timeout > 0 else '' cmd += ' -se' if stop_on_error else '' cmd += ' -fr' if follow_redirect else '' cmd += ' -ac' if auto_calibration else '' cmd += f' -mc {mc}' if mc else '' cmd += f' -H "{custom_header}"' if custom_header else '' # Grab URLs to fuzz urls = get_http_urls( is_alive=True, ignore_files=False, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) logger.warning(urls) # Loop through URLs and run command results = [] for url in urls: ''' Above while fetching urls, we are not ignoring files, because some default urls may redirect to https://example.com/login.php so, ignore_files is set to False but, during fuzzing, we will only need part of the path, in above example it is still a good idea to ffuf base url https://example.com so files from base url ''' url_parse = urlparse(url) url = url_parse.scheme + '://' + url_parse.netloc url += '/FUZZ' # TODO: fuzz not only URL but also POST / PUT / headers proxy = get_random_proxy() # Build final cmd fcmd = cmd fcmd += f' -x {proxy}' if proxy else '' fcmd += f' -u {url} -json' # Initialize DirectoryScan object dirscan = DirectoryScan() dirscan.scanned_date = timezone.now() dirscan.command_line = fcmd dirscan.save() # Loop through results and populate EndPoint and DirectoryFile in DB results = [] for line in stream_command( fcmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) name = line['input'].get('FUZZ') length = line['length'] status = line['status'] words = line['words'] url = line['url'] lines = line['lines'] content_type = line['content-type'] duration = line['duration'] if not name: logger.error(f'FUZZ not found for "{url}"') continue endpoint, created = save_endpoint(url, crawl=False, ctx=ctx) # endpoint.is_default = False endpoint.http_status = status endpoint.content_length = length endpoint.response_time = duration / 1000000000 endpoint.save() if created: urls.append(endpoint.http_url) endpoint.status = status endpoint.content_type = content_type endpoint.content_length = length dfile, created = DirectoryFile.objects.get_or_create( name=name, length=length, words=words, lines=lines, content_type=content_type, url=url) dfile.http_status = status dfile.save() # if created: # logger.warning(f'Found new directory or file {url}') dirscan.directory_files.add(dfile) dirscan.save() if self.subscan: dirscan.dir_subscan_ids.add(self.subscan) subdomain_name = get_subdomain_from_url(endpoint.http_url) subdomain = Subdomain.objects.get(name=subdomain_name, scan_history=self.scan) subdomain.directories.add(dirscan) subdomain.save() # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl(urls, ctx=ctx) return results @app.task(name='fetch_url', queue='main_scan_queue', base=RengineTask, bind=True) def fetch_url(self, urls=[], ctx={}, description=None): """Fetch URLs using different tools like gauplus, gau, gospider, waybackurls ... Args: urls (list): List of URLs to start from. description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/input_endpoints_fetch_url.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(FETCH_URL) or {} should_remove_duplicate_endpoints = config.get(REMOVE_DUPLICATE_ENDPOINTS, True) duplicate_removal_fields = config.get(DUPLICATE_REMOVAL_FIELDS, ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, DEFAULT_GF_PATTERNS) ignore_file_extension = config.get(IGNORE_FILE_EXTENSION, DEFAULT_IGNORE_FILE_EXTENSIONS) tools = config.get(USES_TOOLS, ENDPOINT_SCAN_DEFAULT_TOOLS) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) domain_request_headers = self.domain.request_headers if self.domain else None custom_header = domain_request_headers or self.yaml_configuration.get(CUSTOM_HEADER) exclude_subdomains = config.get(EXCLUDED_SUBDOMAINS, False) # Get URLs to scan and save to input file if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, exclude_subdomains=exclude_subdomains, get_only_default_urls=True, ctx=ctx ) # Domain regex host = self.domain.name if self.domain else urlparse(urls[0]).netloc host_regex = f"\'https?://([a-z0-9]+[.])*{host}.*\'" # Tools cmds cmd_map = { 'gau': f'gau', 'gauplus': f'gauplus -random-agent', 'hakrawler': 'hakrawler -subs -u', 'waybackurls': 'waybackurls', 'gospider': f'gospider -S {input_path} --js -d 2 --sitemap --robots -w -r', 'katana': f'katana -list {input_path} -silent -jc -kf all -d 3 -fs rdn', } if proxy: cmd_map['gau'] += f' --proxy "{proxy}"' cmd_map['gauplus'] += f' -p "{proxy}"' cmd_map['gospider'] += f' -p {proxy}' cmd_map['hakrawler'] += f' -proxy {proxy}' cmd_map['katana'] += f' -proxy {proxy}' if threads > 0: cmd_map['gau'] += f' --threads {threads}' cmd_map['gauplus'] += f' -t {threads}' cmd_map['gospider'] += f' -t {threads}' cmd_map['katana'] += f' -c {threads}' if custom_header: header_string = ';;'.join([ f'{key}: {value}' for key, value in custom_header.items() ]) cmd_map['hakrawler'] += f' -h {header_string}' cmd_map['katana'] += f' -H {header_string}' header_flags = [':'.join(h) for h in header_string.split(';;')] for flag in header_flags: cmd_map['gospider'] += f' -H {flag}' cat_input = f'cat {input_path}' grep_output = f'grep -Eo {host_regex}' cmd_map = { tool: f'{cat_input} | {cmd} | {grep_output} > {self.results_dir}/urls_{tool}.txt' for tool, cmd in cmd_map.items() } tasks = group( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for tool, cmd in cmd_map.items() if tool in tools ) # Cleanup task sort_output = [ f'cat {self.results_dir}/urls_* > {self.output_path}', f'cat {input_path} >> {self.output_path}', f'sort -u {self.output_path} -o {self.output_path}', ] if ignore_file_extension: ignore_exts = '|'.join(ignore_file_extension) grep_ext_filtered_output = [ f'cat {self.output_path} | grep -Eiv "\\.({ignore_exts}).*" > {self.results_dir}/urls_filtered.txt', f'mv {self.results_dir}/urls_filtered.txt {self.output_path}' ] sort_output.extend(grep_ext_filtered_output) cleanup = chain( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for cmd in sort_output ) # Run all commands task = chord(tasks)(cleanup) with allow_join_result(): task.get() # Store all the endpoints and run httpx with open(self.output_path) as f: discovered_urls = f.readlines() self.notify(fields={'Discovered URLs': len(discovered_urls)}) # Some tools can have an URL in the format <URL>] - <PATH> or <URL> - <PATH>, add them # to the final URL list all_urls = [] for url in discovered_urls: url = url.strip() urlpath = None base_url = None if '] ' in url: # found JS scraped endpoint e.g from gospider split = tuple(url.split('] ')) if not len(split) == 2: logger.warning(f'URL format not recognized for "{url}". Skipping.') continue base_url, urlpath = split urlpath = urlpath.lstrip('- ') elif ' - ' in url: # found JS scraped endpoint e.g from gospider base_url, urlpath = tuple(url.split(' - ')) if base_url and urlpath: subdomain = urlparse(base_url) url = f'{subdomain.scheme}://{subdomain.netloc}{self.url_filter}' if not validators.url(url): logger.warning(f'Invalid URL "{url}". Skipping.') if url not in all_urls: all_urls.append(url) # Filter out URLs if a path filter was passed if self.url_filter: all_urls = [url for url in all_urls if self.url_filter in url] # Write result to output path with open(self.output_path, 'w') as f: f.write('\n'.join(all_urls)) logger.warning(f'Found {len(all_urls)} usable URLs') # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl( all_urls, ctx=ctx, should_remove_duplicate_endpoints=should_remove_duplicate_endpoints, duplicate_removal_fields=duplicate_removal_fields ) #-------------------# # GF PATTERNS MATCH # #-------------------# # Combine old gf patterns with new ones if gf_patterns: self.scan.used_gf_patterns = ','.join(gf_patterns) self.scan.save() # Run gf patterns on saved endpoints # TODO: refactor to Celery task for gf_pattern in gf_patterns: # TODO: js var is causing issues, removing for now if gf_pattern == 'jsvar': logger.info('Ignoring jsvar as it is causing issues.') continue # Run gf on current pattern logger.warning(f'Running gf on pattern "{gf_pattern}"') gf_output_file = f'{self.results_dir}/gf_patterns_{gf_pattern}.txt' cmd = f'cat {self.output_path} | gf {gf_pattern} | grep -Eo {host_regex} >> {gf_output_file}' run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Check output file if not os.path.exists(gf_output_file): logger.error(f'Could not find GF output file {gf_output_file}. Skipping GF pattern "{gf_pattern}"') continue # Read output file line by line and with open(gf_output_file, 'r') as f: lines = f.readlines() # Add endpoints / subdomains to DB for url in lines: http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue endpoint, created = save_endpoint( http_url, crawl=False, subdomain=subdomain, ctx=ctx) if not endpoint: continue earlier_pattern = None if not created: earlier_pattern = endpoint.matched_gf_patterns pattern = f'{earlier_pattern},{gf_pattern}' if earlier_pattern else gf_pattern endpoint.matched_gf_patterns = pattern endpoint.save() return all_urls def parse_curl_output(response): # TODO: Enrich from other cURL fields. CURL_REGEX_HTTP_STATUS = f'HTTP\/(?:(?:\d\.?)+)\s(\d+)\s(?:\w+)' http_status = 0 if response: failed = False regex = re.compile(CURL_REGEX_HTTP_STATUS, re.MULTILINE) try: http_status = int(regex.findall(response)[0]) except (KeyError, TypeError, IndexError): pass return { 'http_status': http_status, } @app.task(name='vulnerability_scan', queue='main_scan_queue', bind=True, base=RengineTask) def vulnerability_scan(self, urls=[], ctx={}, description=None): """ This function will serve as an entrypoint to vulnerability scan. All other vulnerability scan will be run from here including nuclei, crlfuzz, etc """ logger.info('Running Vulnerability Scan Queue') config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_run_nuclei = config.get(RUN_NUCLEI, True) should_run_crlfuzz = config.get(RUN_CRLFUZZ, False) should_run_dalfox = config.get(RUN_DALFOX, False) should_run_s3scanner = config.get(RUN_S3SCANNER, True) grouped_tasks = [] if should_run_nuclei: _task = nuclei_scan.si( urls=urls, ctx=ctx, description=f'Nuclei Scan' ) grouped_tasks.append(_task) if should_run_crlfuzz: _task = crlfuzz_scan.si( urls=urls, ctx=ctx, description=f'CRLFuzz Scan' ) grouped_tasks.append(_task) if should_run_dalfox: _task = dalfox_xss_scan.si( urls=urls, ctx=ctx, description=f'Dalfox XSS Scan' ) grouped_tasks.append(_task) if should_run_s3scanner: _task = s3scanner.si( ctx=ctx, description=f'Misconfigured S3 Buckets Scanner' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan completed...') # return results return None @app.task(name='nuclei_individual_severity_module', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_individual_severity_module(self, cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx={}, description=None): ''' This celery task will run vulnerability scan in parallel. All severities supplied should run in parallel as grouped tasks. ''' results = [] logger.info(f'Running vulnerability scan with severity: {severity}') cmd += f' -severity {severity}' # Send start notification notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) # Gather nuclei results vuln_data = parse_nuclei_result(line) # Get corresponding subdomain http_url = sanitize_url(line.get('matched-at')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) # Look for duplicate vulnerabilities by excluding records that might change but are irrelevant. object_comparison_exclude = ['response', 'curl_command', 'tags', 'references', 'cve_ids', 'cwe_ids'] # Add subdomain and target domain to the duplicate check vuln_data_copy = vuln_data.copy() vuln_data_copy['subdomain'] = subdomain vuln_data_copy['target_domain'] = self.domain # Check if record exists, if exists do not save it if record_exists(Vulnerability, data=vuln_data_copy, exclude_keys=object_comparison_exclude): logger.warning(f'Nuclei vulnerability of severity {severity} : {vuln_data_copy["name"]} for {subdomain_name} already exists') continue # Get or create EndPoint object response = line.get('response') httpx_crawl = False if response else enable_http_crawl # avoid yet another httpx crawl endpoint, _ = save_endpoint( http_url, crawl=httpx_crawl, subdomain=subdomain, ctx=ctx) if endpoint: http_url = endpoint.http_url if not httpx_crawl: output = parse_curl_output(response) endpoint.http_status = output['http_status'] endpoint.save() # Get or create Vulnerability object vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, subdomain=subdomain, **vuln_data) if not vuln: continue # Print vuln severity = line['info'].get('severity', 'unknown') logger.warning(str(vuln)) # Send notification for all vulnerabilities except info url = vuln.http_url or vuln.subdomain send_vuln = ( notif and notif.send_vuln_notif and vuln and severity in ['low', 'medium', 'high', 'critical']) if send_vuln: fields = { 'Severity': f'**{severity.upper()}**', 'URL': http_url, 'Subdomain': subdomain_name, 'Name': vuln.name, 'Type': vuln.type, 'Description': vuln.description, 'Template': vuln.template_url, 'Tags': vuln.get_tags_str(), 'CVEs': vuln.get_cve_str(), 'CWEs': vuln.get_cwe_str(), 'References': vuln.get_refs_str() } severity_map = { 'low': 'info', 'medium': 'warning', 'high': 'error', 'critical': 'error' } self.notify( f'vulnerability_scan_#{vuln.id}', severity_map[severity], fields, add_meta_info=False) # Send report to hackerone hackerone_query = Hackerone.objects.all() send_report = ( hackerone_query.exists() and severity not in ('info', 'low') and vuln.target_domain.h1_team_handle ) if send_report: hackerone = hackerone_query.first() if hackerone.send_critical and severity == 'critical': send_hackerone_report.delay(vuln.id) elif hackerone.send_high and severity == 'high': send_hackerone_report.delay(vuln.id) elif hackerone.send_medium and severity == 'medium': send_hackerone_report.delay(vuln.id) # Write results to JSON file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) # Send finish notif if send_status: vulns = Vulnerability.objects.filter(scan_history__id=self.scan_id) info_count = vulns.filter(severity=0).count() low_count = vulns.filter(severity=1).count() medium_count = vulns.filter(severity=2).count() high_count = vulns.filter(severity=3).count() critical_count = vulns.filter(severity=4).count() unknown_count = vulns.filter(severity=-1).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count + unknown_count fields = { 'Total': vulnerability_count, 'Critical': critical_count, 'High': high_count, 'Medium': medium_count, 'Low': low_count, 'Info': info_count, 'Unknown': unknown_count } self.notify(fields=fields) # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=NUCLEI ).exclude( severity=0 ) # find all unique vulnerabilities based on path and title # all unique vulnerability will go thru gpt function and get report # once report is got, it will be matched with other vulnerabilities and saved unique_vulns = set() for vuln in vulns: unique_vulns.add((vuln.name, vuln.get_path())) unique_vulns = list(unique_vulns) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in unique_vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return None def get_vulnerability_gpt_report(vuln): title = vuln[0] path = vuln[1] logger.info(f'Getting GPT Report for {title}, PATH: {path}') # check if in db already exists stored = GPTVulnerabilityReport.objects.filter( url_path=path ).filter( title=title ).first() if stored: response = { 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: report = GPTVulnerabilityReportGenerator() vulnerability_description = get_gpt_vuln_input_description( title, path ) response = report.get_vulnerability_description(vulnerability_description) add_gpt_description_db( title, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) for vuln in Vulnerability.objects.filter(name=title, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() def add_gpt_description_db(title, path, description, impact, remediation, references): gpt_report = GPTVulnerabilityReport() gpt_report.url_path = path gpt_report.title = title gpt_report.description = description gpt_report.impact = impact gpt_report.remediation = remediation gpt_report.save() for url in references: ref, created = VulnerabilityReference.objects.get_or_create(url=url) gpt_report.references.add(ref) gpt_report.save() @app.task(name='nuclei_scan', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_scan(self, urls=[], ctx={}, description=None): """HTTP vulnerability scan using Nuclei Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. Notes: Unfurl the urls to keep only domain and path, will be sent to vuln scan and ignore certain file extensions. Thanks: https://github.com/six2dez/reconftw """ # Config config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} input_path = f'{self.results_dir}/input_endpoints_vulnerability_scan.txt' enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) concurrency = config.get(NUCLEI_CONCURRENCY) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) retries = config.get(RETRIES) or self.yaml_configuration.get(RETRIES, DEFAULT_RETRIES) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) custom_header = config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) should_fetch_gpt_report = config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) proxy = get_random_proxy() nuclei_specific_config = config.get('nuclei', {}) use_nuclei_conf = nuclei_specific_config.get(USE_CONFIG, False) severities = nuclei_specific_config.get(NUCLEI_SEVERITY, NUCLEI_DEFAULT_SEVERITIES) tags = nuclei_specific_config.get(NUCLEI_TAGS, []) tags = ','.join(tags) nuclei_templates = nuclei_specific_config.get(NUCLEI_TEMPLATE) custom_nuclei_templates = nuclei_specific_config.get(NUCLEI_CUSTOM_TEMPLATE) # severities_str = ','.join(severities) # Get alive endpoints if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=enable_http_crawl, ignore_files=True, write_filepath=input_path, ctx=ctx ) if intensity == 'normal': # reduce number of endpoints to scan unfurl_filter = f'{self.results_dir}/urls_unfurled.txt' run_command( f"cat {input_path} | unfurl -u format %s://%d%p |uro > {unfurl_filter}", shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {unfurl_filter} -o {unfurl_filter}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) input_path = unfurl_filter # Build templates # logger.info('Updating Nuclei templates ...') run_command( 'nuclei -update-templates', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) templates = [] if not (nuclei_templates or custom_nuclei_templates): templates.append(NUCLEI_DEFAULT_TEMPLATES_PATH) if nuclei_templates: if ALL in nuclei_templates: template = NUCLEI_DEFAULT_TEMPLATES_PATH templates.append(template) else: templates.extend(nuclei_templates) if custom_nuclei_templates: custom_nuclei_template_paths = [f'{str(elem)}.yaml' for elem in custom_nuclei_templates] template = templates.extend(custom_nuclei_template_paths) # Build CMD cmd = 'nuclei -j' cmd += ' -config /root/.config/nuclei/config.yaml' if use_nuclei_conf else '' cmd += f' -irr' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -l {input_path}' cmd += f' -c {str(concurrency)}' if concurrency > 0 else '' cmd += f' -proxy {proxy} ' if proxy else '' cmd += f' -retries {retries}' if retries > 0 else '' cmd += f' -rl {rate_limit}' if rate_limit > 0 else '' # cmd += f' -severity {severities_str}' cmd += f' -timeout {str(timeout)}' if timeout and timeout > 0 else '' cmd += f' -tags {tags}' if tags else '' cmd += f' -silent' for tpl in templates: cmd += f' -t {tpl}' grouped_tasks = [] custom_ctx = ctx for severity in severities: custom_ctx['track'] = True _task = nuclei_individual_severity_module.si( cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx=custom_ctx, description=f'Nuclei Scan with severity {severity}' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan with all severities completed...') return None @app.task(name='dalfox_xss_scan', queue='main_scan_queue', base=RengineTask, bind=True) def dalfox_xss_scan(self, urls=[], ctx={}, description=None): """XSS Scan using dalfox Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) dalfox_config = vuln_config.get(DALFOX) or {} custom_header = dalfox_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() is_waf_evasion = dalfox_config.get(WAF_EVASION, False) blind_xss_server = dalfox_config.get(BLIND_XSS_SERVER) user_agent = dalfox_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) timeout = dalfox_config.get(TIMEOUT) delay = dalfox_config.get(DELAY) threads = dalfox_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_dalfox_xss.txt' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=False, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'dalfox --silence --no-color --no-spinner' cmd += f' --only-poc r ' cmd += f' --ignore-return 302,404,403' cmd += f' --skip-bav' cmd += f' file {input_path}' cmd += f' --proxy {proxy}' if proxy else '' cmd += f' --waf-evasion' if is_waf_evasion else '' cmd += f' -b {blind_xss_server}' if blind_xss_server else '' cmd += f' --delay {delay}' if delay else '' cmd += f' --timeout {timeout}' if timeout else '' cmd += f' --user-agent {user_agent}' if user_agent else '' cmd += f' --header {custom_header}' if custom_header else '' cmd += f' --worker {threads}' if threads else '' cmd += f' --format json' results = [] for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id, trunc_char=',' ): if not isinstance(line, dict): continue results.append(line) vuln_data = parse_dalfox_result(line) http_url = sanitize_url(line.get('data')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Dalfox Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=DALFOX ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='crlfuzz_scan', queue='main_scan_queue', base=RengineTask, bind=True) def crlfuzz_scan(self, urls=[], ctx={}, description=None): """CRLF Fuzzing with CRLFuzz Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) custom_header = vuln_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() user_agent = vuln_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) threads = vuln_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_crlf.txt' output_path = f'{self.results_dir}/{self.filename}' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=True, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'crlfuzz -s' cmd += f' -l {input_path}' cmd += f' -x {proxy}' if proxy else '' cmd += f' --H {custom_header}' if custom_header else '' cmd += f' -o {output_path}' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id ) if not os.path.isfile(output_path): logger.info('No Results from CRLFuzz') return crlfs = [] results = [] with open(output_path, 'r') as file: crlfs = file.readlines() for crlf in crlfs: url = crlf.strip() vuln_data = parse_crlfuzz_result(url) http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting CRLFuzz Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=CRLFUZZ ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='s3scanner', queue='main_scan_queue', base=RengineTask, bind=True) def s3scanner(self, ctx={}, description=None): """Bucket Scanner Args: ctx (dict): Context description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/#{self.scan_id}_subdomain_discovery.txt' vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} s3_config = vuln_config.get(S3SCANNER) or {} threads = s3_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) providers = s3_config.get(PROVIDERS, S3SCANNER_DEFAULT_PROVIDERS) scan_history = ScanHistory.objects.filter(pk=self.scan_id).first() for provider in providers: cmd = f's3scanner -bucket-file {input_path} -enumerate -provider {provider} -threads {threads} -json' for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue if line.get('bucket', {}).get('exists', 0) == 1: result = parse_s3scanner_result(line) s3bucket, created = S3Bucket.objects.get_or_create(**result) scan_history.buckets.add(s3bucket) logger.info(f"s3 bucket added {result['provider']}-{result['name']}-{result['region']}") @app.task(name='http_crawl', queue='main_scan_queue', base=RengineTask, bind=True) def http_crawl( self, urls=[], method=None, recrawl=False, ctx={}, track=True, description=None, is_ran_from_subdomain_scan=False, should_remove_duplicate_endpoints=True, duplicate_removal_fields=[]): """Use httpx to query HTTP URLs for important info like page titles, http status, etc... Args: urls (list, optional): A set of URLs to check. Overrides default behavior which queries all endpoints related to this scan. method (str): HTTP method to use (GET, HEAD, POST, PUT, DELETE). recrawl (bool, optional): If False, filter out URLs that have already been crawled. should_remove_duplicate_endpoints (bool): Whether to remove duplicate endpoints duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates Returns: list: httpx results. """ logger.info('Initiating HTTP Crawl') if is_ran_from_subdomain_scan: logger.info('Running From Subdomain Scan...') cmd = '/go/bin/httpx' cfg = self.yaml_configuration.get(HTTP_CRAWL) or {} custom_header = cfg.get(CUSTOM_HEADER, '') threads = cfg.get(THREADS, DEFAULT_THREADS) follow_redirect = cfg.get(FOLLOW_REDIRECT, True) self.output_path = None input_path = f'{self.results_dir}/httpx_input.txt' history_file = f'{self.results_dir}/commands.txt' if urls: # direct passing URLs to check if self.url_filter: urls = [u for u in urls if self.url_filter in u] with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_uncrawled=not recrawl, write_filepath=input_path, ctx=ctx ) # logger.debug(urls) # If no URLs found, skip it if not urls: return # Re-adjust thread number if few URLs to avoid spinning up a monster to # kill a fly. if len(urls) < threads: threads = len(urls) # Get random proxy proxy = get_random_proxy() # Run command cmd += f' -cl -ct -rt -location -td -websocket -cname -asn -cdn -probe -random-agent' cmd += f' -t {threads}' if threads > 0 else '' cmd += f' --http-proxy {proxy}' if proxy else '' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -json' cmd += f' -u {urls[0]}' if len(urls) == 1 else f' -l {input_path}' cmd += f' -x {method}' if method else '' cmd += f' -silent' if follow_redirect: cmd += ' -fr' results = [] endpoint_ids = [] for line in stream_command( cmd, history_file=history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not line or not isinstance(line, dict): continue logger.debug(line) # No response from endpoint if line.get('failed', False): continue # Parse httpx output host = line.get('host', '') content_length = line.get('content_length', 0) http_status = line.get('status_code') http_url, is_redirect = extract_httpx_url(line) page_title = line.get('title') webserver = line.get('webserver') cdn = line.get('cdn', False) rt = line.get('time') techs = line.get('tech', []) cname = line.get('cname', '') content_type = line.get('content_type', '') response_time = -1 if rt: response_time = float(''.join(ch for ch in rt if not ch.isalpha())) if rt[-2:] == 'ms': response_time = response_time / 1000 # Create Subdomain object in DB subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue # Save default HTTP URL to endpoint object in DB endpoint, created = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain, is_default=is_ran_from_subdomain_scan ) if not endpoint: continue endpoint.http_status = http_status endpoint.page_title = page_title endpoint.content_length = content_length endpoint.webserver = webserver endpoint.response_time = response_time endpoint.content_type = content_type endpoint.save() endpoint_str = f'{http_url} [{http_status}] `{content_length}B` `{webserver}` `{rt}`' logger.warning(endpoint_str) if endpoint and endpoint.is_alive and endpoint.http_status != 403: self.notify( fields={'Alive endpoint': f'• {endpoint_str}'}, add_meta_info=False) # Add endpoint to results line['_cmd'] = cmd line['final_url'] = http_url line['endpoint_id'] = endpoint.id line['endpoint_created'] = created line['is_redirect'] = is_redirect results.append(line) # Add technology objects to DB for technology in techs: tech, _ = Technology.objects.get_or_create(name=technology) endpoint.techs.add(tech) if is_ran_from_subdomain_scan: subdomain.technologies.add(tech) subdomain.save() endpoint.save() techs_str = ', '.join([f'`{tech}`' for tech in techs]) self.notify( fields={'Technologies': techs_str}, add_meta_info=False) # Add IP objects for 'a' records to DB a_records = line.get('a', []) for ip_address in a_records: ip, created = save_ip_address( ip_address, subdomain, subscan=self.subscan, cdn=cdn) ips_str = '• ' + '\n• '.join([f'`{ip}`' for ip in a_records]) self.notify( fields={'IPs': ips_str}, add_meta_info=False) # Add IP object for host in DB if host: ip, created = save_ip_address( host, subdomain, subscan=self.subscan, cdn=cdn) self.notify( fields={'IPs': f'• `{ip.address}`'}, add_meta_info=False) # Save subdomain and endpoint if is_ran_from_subdomain_scan: # save subdomain stuffs subdomain.http_url = http_url subdomain.http_status = http_status subdomain.page_title = page_title subdomain.content_length = content_length subdomain.webserver = webserver subdomain.response_time = response_time subdomain.content_type = content_type subdomain.cname = ','.join(cname) subdomain.is_cdn = cdn if cdn: subdomain.cdn_name = line.get('cdn_name') subdomain.save() endpoint.save() endpoint_ids.append(endpoint.id) if should_remove_duplicate_endpoints: # Remove 'fake' alive endpoints that are just redirects to the same page remove_duplicate_endpoints( self.scan_id, self.domain_id, self.subdomain_id, filter_ids=endpoint_ids ) # Remove input file run_command( f'rm {input_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) return results #---------------------# # Notifications tasks # #---------------------# @app.task(name='send_notif', bind=False, queue='send_notif_queue') def send_notif( message, scan_history_id=None, subscan_id=None, **options): if not 'title' in options: message = enrich_notification(message, scan_history_id, subscan_id) send_discord_message(message, **options) send_slack_message(message) send_telegram_message(message) @app.task(name='send_scan_notif', bind=False, queue='send_scan_notif_queue') def send_scan_notif( scan_history_id, subscan_id=None, engine_id=None, status='RUNNING'): """Send scan status notification. Works for scan or a subscan if subscan_id is passed. Args: scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Get domain, engine, scan_history objects engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() tasks = ScanActivity.objects.filter(scan_of=scan) if scan else 0 # Build notif options url = get_scan_url(scan_history_id, subscan_id) title = get_scan_title(scan_history_id, subscan_id) fields = get_scan_fields(engine, scan, subscan, status, tasks) severity = None msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) if status: severity = STATUS_TO_SEVERITIES.get(status) opts = { 'title': title, 'url': url, 'fields': fields, 'severity': severity } logger.warning(f'Sending notification "{title}" [{severity}]') # Send notification send_notif( msg, scan_history_id, subscan_id, **opts) @app.task(name='send_task_notif', bind=False, queue='send_task_notif_queue') def send_task_notif( task_name, status=None, result=None, output_path=None, traceback=None, scan_history_id=None, engine_id=None, subscan_id=None, severity=None, add_meta_info=True, update_fields={}): """Send task status notification. Args: task_name (str): Task name. status (str, optional): Task status. result (str, optional): Task result. output_path (str, optional): Task output path. traceback (str, optional): Task traceback. scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. severity (str, optional): Severity (will be mapped to notif colors) add_meta_info (bool, optional): Wheter to add scan / subscan info to notif. update_fields (dict, optional): Fields key / value to update. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Build fields url = None fields = {} if add_meta_info: engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() url = get_scan_url(scan_history_id) if status: fields['Status'] = f'**{status}**' if engine: fields['Engine'] = engine.engine_name if scan: fields['Scan ID'] = f'[#{scan.id}]({url})' if subscan: url = get_scan_url(scan_history_id, subscan_id) fields['Subscan ID'] = f'[#{subscan.id}]({url})' title = get_task_title(task_name, scan_history_id, subscan_id) if status: severity = STATUS_TO_SEVERITIES.get(status) msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) # Add fields to update for k, v in update_fields.items(): fields[k] = v # Add traceback to notif if traceback and notif.send_scan_tracebacks: fields['Traceback'] = f'```\n{traceback}\n```' # Add files to notif files = [] attach_file = ( notif.send_scan_output_file and output_path and result and not traceback ) if attach_file: output_title = output_path.split('/')[-1] files = [(output_path, output_title)] # Send notif opts = { 'title': title, 'url': url, 'files': files, 'severity': severity, 'fields': fields, 'fields_append': update_fields.keys() } send_notif( msg, scan_history_id=scan_history_id, subscan_id=subscan_id, **opts) @app.task(name='send_file_to_discord', bind=False, queue='send_file_to_discord_queue') def send_file_to_discord(file_path, title=None): notif = Notification.objects.first() do_send = notif and notif.send_to_discord and notif.discord_hook_url if not do_send: return False webhook = DiscordWebhook( url=notif.discord_hook_url, rate_limit_retry=True, username=title or "reNgine Discord Plugin" ) with open(file_path, "rb") as f: head, tail = os.path.split(file_path) webhook.add_file(file=f.read(), filename=tail) webhook.execute() @app.task(name='send_hackerone_report', bind=False, queue='send_hackerone_report_queue') def send_hackerone_report(vulnerability_id): """Send HackerOne vulnerability report. Args: vulnerability_id (int): Vulnerability id. Returns: int: HTTP response status code. """ vulnerability = Vulnerability.objects.get(id=vulnerability_id) severities = {v: k for k,v in NUCLEI_SEVERITY_MAP.items()} headers = { 'Content-Type': 'application/json', 'Accept': 'application/json' } # can only send vulnerability report if team_handle exists if len(vulnerability.target_domain.h1_team_handle) !=0: hackerone_query = Hackerone.objects.all() if hackerone_query.exists(): hackerone = Hackerone.objects.first() severity_value = severities[vulnerability.severity] tpl = hackerone.report_template # Replace syntax of report template with actual content tpl = tpl.replace('{vulnerability_name}', vulnerability.name) tpl = tpl.replace('{vulnerable_url}', vulnerability.http_url) tpl = tpl.replace('{vulnerability_severity}', severity_value) tpl = tpl.replace('{vulnerability_description}', vulnerability.description if vulnerability.description else '') tpl = tpl.replace('{vulnerability_extracted_results}', vulnerability.extracted_results if vulnerability.extracted_results else '') tpl = tpl.replace('{vulnerability_reference}', vulnerability.reference if vulnerability.reference else '') data = { "data": { "type": "report", "attributes": { "team_handle": vulnerability.target_domain.h1_team_handle, "title": '{} found in {}'.format(vulnerability.name, vulnerability.http_url), "vulnerability_information": tpl, "severity_rating": severity_value, "impact": "More information about the impact and vulnerability can be found here: \n" + vulnerability.reference if vulnerability.reference else "NA", } } } r = requests.post( 'https://api.hackerone.com/v1/hackers/reports', auth=(hackerone.username, hackerone.api_key), json=data, headers=headers ) response = r.json() status_code = r.status_code if status_code == 201: vulnerability.hackerone_report_id = response['data']["id"] vulnerability.open_status = False vulnerability.save() return status_code else: logger.error('No team handle found.') status_code = 111 return status_code #-------------# # Utils tasks # #-------------# @app.task(name='parse_nmap_results', bind=False, queue='parse_nmap_results_queue') def parse_nmap_results(xml_file, output_file=None): """Parse results from nmap output file. Args: xml_file (str): nmap XML report file path. Returns: list: List of vulnerabilities found from nmap results. """ with open(xml_file, encoding='utf8') as f: content = f.read() try: nmap_results = xmltodict.parse(content) # parse XML to dict except Exception as e: logger.exception(e) logger.error(f'Cannot parse {xml_file} to valid JSON. Skipping.') return [] # Write JSON to output file if output_file: with open(output_file, 'w') as f: json.dump(nmap_results, f, indent=4) logger.warning(json.dumps(nmap_results, indent=4)) hosts = ( nmap_results .get('nmaprun', {}) .get('host', {}) ) all_vulns = [] if isinstance(hosts, dict): hosts = [hosts] for host in hosts: # Grab hostname / IP from output hostnames_dict = host.get('hostnames', {}) if hostnames_dict: # Ensure that hostnames['hostname'] is a list for consistency hostnames_list = hostnames_dict['hostname'] if isinstance(hostnames_dict['hostname'], list) else [hostnames_dict['hostname']] # Extract all the @name values from the list of dictionaries hostnames = [entry.get('@name') for entry in hostnames_list] else: hostnames = [host.get('address')['@addr']] # Iterate over each hostname for each port for hostname in hostnames: # Grab ports from output ports = host.get('ports', {}).get('port', []) if isinstance(ports, dict): ports = [ports] for port in ports: url_vulns = [] port_number = port['@portid'] url = sanitize_url(f'{hostname}:{port_number}') logger.info(f'Parsing nmap results for {hostname}:{port_number} ...') if not port_number or not port_number.isdigit(): continue port_protocol = port['@protocol'] scripts = port.get('script', []) if isinstance(scripts, dict): scripts = [scripts] for script in scripts: script_id = script['@id'] script_output = script['@output'] script_output_table = script.get('table', []) logger.debug(f'Ran nmap script "{script_id}" on {port_number}/{port_protocol}:\n{script_output}\n') if script_id == 'vulscan': vulns = parse_nmap_vulscan_output(script_output) url_vulns.extend(vulns) elif script_id == 'vulners': vulns = parse_nmap_vulners_output(script_output) url_vulns.extend(vulns) # elif script_id == 'http-server-header': # TODO: nmap can help find technologies as well using the http-server-header script # regex = r'(\w+)/([\d.]+)\s?(?:\((\w+)\))?' # tech_name, tech_version, tech_os = re.match(regex, test_string).groups() # Technology.objects.get_or_create(...) # elif script_id == 'http_csrf': # vulns = parse_nmap_http_csrf_output(script_output) # url_vulns.extend(vulns) else: logger.warning(f'Script output parsing for script "{script_id}" is not supported yet.') # Add URL to vuln for vuln in url_vulns: # TODO: This should extend to any URL, not just HTTP vuln['http_url'] = url if 'http_path' in vuln: vuln['http_url'] += vuln['http_path'] all_vulns.append(vuln) return all_vulns def parse_nmap_http_csrf_output(script_output): pass def parse_nmap_vulscan_output(script_output): """Parse nmap vulscan script output. Args: script_output (str): Vulscan script output. Returns: list: List of Vulnerability dicts. """ data = {} vulns = [] provider_name = '' # Sort all vulns found by provider so that we can match each provider with # a function that pulls from its API to get more info about the # vulnerability. for line in script_output.splitlines(): if not line: continue if not line.startswith('['): # provider line if "No findings" in line: logger.info(f"No findings: {line}") continue elif ' - ' in line: provider_name, provider_url = tuple(line.split(' - ')) data[provider_name] = {'url': provider_url.rstrip(':'), 'entries': []} continue else: # Log a warning logger.warning(f"Unexpected line format: {line}") continue reg = r'\[(.*)\] (.*)' matches = re.match(reg, line) id, title = matches.groups() entry = {'id': id, 'title': title} data[provider_name]['entries'].append(entry) logger.warning('Vulscan parsed output:') logger.warning(pprint.pformat(data)) for provider_name in data: if provider_name == 'Exploit-DB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'IBM X-Force': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'MITRE CVE': logger.error(f'Provider {provider_name} is not supported YET.') for entry in data[provider_name]['entries']: cve_id = entry['id'] vuln = cve_to_vuln(cve_id) vulns.append(vuln) elif provider_name == 'OSVDB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'OpenVAS (Nessus)': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'SecurityFocus': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'VulDB': logger.error(f'Provider {provider_name} is not supported YET.') pass else: logger.error(f'Provider {provider_name} is not supported.') return vulns def parse_nmap_vulners_output(script_output, url=''): """Parse nmap vulners script output. TODO: Rework this as it's currently matching all CVEs no matter the confidence. Args: script_output (str): Script output. Returns: list: List of found vulnerabilities. """ vulns = [] # Check for CVE in script output CVE_REGEX = re.compile(r'.*(CVE-\d\d\d\d-\d+).*') matches = CVE_REGEX.findall(script_output) matches = list(dict.fromkeys(matches)) for cve_id in matches: # get CVE info vuln = cve_to_vuln(cve_id, vuln_type='nmap-vulners-nse') if vuln: vulns.append(vuln) return vulns def cve_to_vuln(cve_id, vuln_type=''): """Search for a CVE using CVESearch and return Vulnerability data. Args: cve_id (str): CVE ID in the form CVE-* Returns: dict: Vulnerability dict. """ cve_info = CVESearch('https://cve.circl.lu').id(cve_id) if not cve_info: logger.error(f'Could not fetch CVE info for cve {cve_id}. Skipping.') return None vuln_cve_id = cve_info['id'] vuln_name = vuln_cve_id vuln_description = cve_info.get('summary', 'none').replace(vuln_cve_id, '').strip() try: vuln_cvss = float(cve_info.get('cvss', -1)) except (ValueError, TypeError): vuln_cvss = -1 vuln_cwe_id = cve_info.get('cwe', '') exploit_ids = cve_info.get('refmap', {}).get('exploit-db', []) osvdb_ids = cve_info.get('refmap', {}).get('osvdb', []) references = cve_info.get('references', []) capec_objects = cve_info.get('capec', []) # Parse ovals for a better vuln name / type ovals = cve_info.get('oval', []) if ovals: vuln_name = ovals[0]['title'] vuln_type = ovals[0]['family'] # Set vulnerability severity based on CVSS score vuln_severity = 'info' if vuln_cvss < 4: vuln_severity = 'low' elif vuln_cvss < 7: vuln_severity = 'medium' elif vuln_cvss < 9: vuln_severity = 'high' else: vuln_severity = 'critical' # Build console warning message msg = f'{vuln_name} | {vuln_severity.upper()} | {vuln_cve_id} | {vuln_cwe_id} | {vuln_cvss}' for id in osvdb_ids: msg += f'\n\tOSVDB: {id}' for exploit_id in exploit_ids: msg += f'\n\tEXPLOITDB: {exploit_id}' logger.warning(msg) vuln = { 'name': vuln_name, 'type': vuln_type, 'severity': NUCLEI_SEVERITY_MAP[vuln_severity], 'description': vuln_description, 'cvss_score': vuln_cvss, 'references': references, 'cve_ids': [vuln_cve_id], 'cwe_ids': [vuln_cwe_id] } return vuln def parse_s3scanner_result(line): ''' Parses and returns s3Scanner Data ''' bucket = line['bucket'] return { 'name': bucket['name'], 'region': bucket['region'], 'provider': bucket['provider'], 'owner_display_name': bucket['owner_display_name'], 'owner_id': bucket['owner_id'], 'perm_auth_users_read': bucket['perm_auth_users_read'], 'perm_auth_users_write': bucket['perm_auth_users_write'], 'perm_auth_users_read_acl': bucket['perm_auth_users_read_acl'], 'perm_auth_users_write_acl': bucket['perm_auth_users_write_acl'], 'perm_auth_users_full_control': bucket['perm_auth_users_full_control'], 'perm_all_users_read': bucket['perm_all_users_read'], 'perm_all_users_write': bucket['perm_all_users_write'], 'perm_all_users_read_acl': bucket['perm_all_users_read_acl'], 'perm_all_users_write_acl': bucket['perm_all_users_write_acl'], 'perm_all_users_full_control': bucket['perm_all_users_full_control'], 'num_objects': bucket['num_objects'], 'size': bucket['bucket_size'] } def parse_nuclei_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ return { 'name': line['info'].get('name', ''), 'type': line['type'], 'severity': NUCLEI_SEVERITY_MAP[line['info'].get('severity', 'unknown')], 'template': line['template'], 'template_url': line['template-url'], 'template_id': line['template-id'], 'description': line['info'].get('description', ''), 'matcher_name': line.get('matcher-name', ''), 'curl_command': line.get('curl-command'), 'request': line.get('request'), 'response': line.get('response'), 'extracted_results': line.get('extracted-results', []), 'cvss_metrics': line['info'].get('classification', {}).get('cvss-metrics', ''), 'cvss_score': line['info'].get('classification', {}).get('cvss-score'), 'cve_ids': line['info'].get('classification', {}).get('cve_id', []) or [], 'cwe_ids': line['info'].get('classification', {}).get('cwe_id', []) or [], 'references': line['info'].get('reference', []) or [], 'tags': line['info'].get('tags', []), 'source': NUCLEI, } def parse_dalfox_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ description = '' description += f" Evidence: {line.get('evidence')} <br>" if line.get('evidence') else '' description += f" Message: {line.get('message')} <br>" if line.get('message') else '' description += f" Payload: {line.get('message_str')} <br>" if line.get('message_str') else '' description += f" Vulnerable Parameter: {line.get('param')} <br>" if line.get('param') else '' return { 'name': 'XSS (Cross Site Scripting)', 'type': 'XSS', 'severity': DALFOX_SEVERITY_MAP[line.get('severity', 'unknown')], 'description': description, 'source': DALFOX, 'cwe_ids': [line.get('cwe')] } def parse_crlfuzz_result(url): """Parse CRLF results Args: url (str): CRLF Vulnerable URL Returns: dict: Vulnerability data. """ return { 'name': 'CRLF (HTTP Response Splitting)', 'type': 'CRLF', 'severity': 2, 'description': 'A CRLF (HTTP Response Splitting) vulnerability has been discovered.', 'source': CRLFUZZ, } def record_exists(model, data, exclude_keys=[]): """ Check if a record already exists in the database based on the given data. Args: model (django.db.models.Model): The Django model to check against. data (dict): Data dictionary containing fields and values. exclude_keys (list): List of keys to exclude from the lookup. Returns: bool: True if the record exists, False otherwise. """ # Extract the keys that will be used for the lookup lookup_fields = {key: data[key] for key in data if key not in exclude_keys} # Return True if a record exists based on the lookup fields, False otherwise return model.objects.filter(**lookup_fields).exists() @app.task(name='geo_localize', bind=False, queue='geo_localize_queue') def geo_localize(host, ip_id=None): """Uses geoiplookup to find location associated with host. Args: host (str): Hostname. ip_id (int): IpAddress object id. Returns: startScan.models.CountryISO: CountryISO object from DB or None. """ if validators.ipv6(host): logger.info(f'Ipv6 "{host}" is not supported by geoiplookup. Skipping.') return None cmd = f'geoiplookup {host}' _, out = run_command(cmd) if 'IP Address not found' not in out and "can't resolve hostname" not in out: country_iso = out.split(':')[1].strip().split(',')[0] country_name = out.split(':')[1].strip().split(',')[1].strip() geo_object, _ = CountryISO.objects.get_or_create( iso=country_iso, name=country_name ) geo_json = { 'iso': country_iso, 'name': country_name } if ip_id: ip = IpAddress.objects.get(pk=ip_id) ip.geo_iso = geo_object ip.save() return geo_json logger.info(f'Geo IP lookup failed for host "{host}"') return None @app.task(name='query_whois', bind=False, queue='query_whois_queue') def query_whois(ip_domain, force_reload_whois=False): """Query WHOIS information for an IP or a domain name. Args: ip_domain (str): IP address or domain name. save_domain (bool): Whether to save domain or not, default False Returns: dict: WHOIS information. """ if not force_reload_whois and Domain.objects.filter(name=ip_domain).exists() and Domain.objects.get(name=ip_domain).domain_info: domain = Domain.objects.get(name=ip_domain) if not domain.insert_date: domain.insert_date = timezone.now() domain.save() domain_info_db = domain.domain_info domain_info = DottedDict( dnssec=domain_info_db.dnssec, created=domain_info_db.created, updated=domain_info_db.updated, expires=domain_info_db.expires, geolocation_iso=domain_info_db.geolocation_iso, status=[status['name'] for status in DomainWhoisStatusSerializer(domain_info_db.status, many=True).data], whois_server=domain_info_db.whois_server, ns_records=[ns['name'] for ns in NameServersSerializer(domain_info_db.name_servers, many=True).data], registrar_name=domain_info_db.registrar.name, registrar_phone=domain_info_db.registrar.phone, registrar_email=domain_info_db.registrar.email, registrar_url=domain_info_db.registrar.url, registrant_name=domain_info_db.registrant.name, registrant_id=domain_info_db.registrant.id_str, registrant_organization=domain_info_db.registrant.organization, registrant_city=domain_info_db.registrant.city, registrant_state=domain_info_db.registrant.state, registrant_zip_code=domain_info_db.registrant.zip_code, registrant_country=domain_info_db.registrant.country, registrant_phone=domain_info_db.registrant.phone, registrant_fax=domain_info_db.registrant.fax, registrant_email=domain_info_db.registrant.email, registrant_address=domain_info_db.registrant.address, admin_name=domain_info_db.admin.name, admin_id=domain_info_db.admin.id_str, admin_organization=domain_info_db.admin.organization, admin_city=domain_info_db.admin.city, admin_state=domain_info_db.admin.state, admin_zip_code=domain_info_db.admin.zip_code, admin_country=domain_info_db.admin.country, admin_phone=domain_info_db.admin.phone, admin_fax=domain_info_db.admin.fax, admin_email=domain_info_db.admin.email, admin_address=domain_info_db.admin.address, tech_name=domain_info_db.tech.name, tech_id=domain_info_db.tech.id_str, tech_organization=domain_info_db.tech.organization, tech_city=domain_info_db.tech.city, tech_state=domain_info_db.tech.state, tech_zip_code=domain_info_db.tech.zip_code, tech_country=domain_info_db.tech.country, tech_phone=domain_info_db.tech.phone, tech_fax=domain_info_db.tech.fax, tech_email=domain_info_db.tech.email, tech_address=domain_info_db.tech.address, related_tlds=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_tlds, many=True).data], related_domains=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_domains, many=True).data], historical_ips=[ip for ip in HistoricalIPSerializer(domain_info_db.historical_ips, many=True).data], ) if domain_info_db.dns_records: a_records = [] txt_records = [] mx_records = [] dns_records = [{'name': dns['name'], 'type': dns['type']} for dns in DomainDNSRecordSerializer(domain_info_db.dns_records, many=True).data] for dns in dns_records: if dns['type'] == 'a': a_records.append(dns['name']) elif dns['type'] == 'txt': txt_records.append(dns['name']) elif dns['type'] == 'mx': mx_records.append(dns['name']) domain_info.a_records = a_records domain_info.txt_records = txt_records domain_info.mx_records = mx_records else: logger.info(f'Domain info for "{ip_domain}" not found in DB, querying whois') domain_info = DottedDict() # find domain historical ip try: historical_ips = get_domain_historical_ip_address(ip_domain) domain_info.historical_ips = historical_ips except Exception as e: logger.error(f'HistoricalIP for {ip_domain} not found!\nError: {str(e)}') historical_ips = [] # find associated domains using ip_domain try: related_domains = reverse_whois(ip_domain.split('.')[0]) except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] # find related tlds using TLSx try: related_tlds = [] output_path = '/tmp/ip_domain_tlsx.txt' tlsx_command = f'tlsx -san -cn -silent -ro -host {ip_domain} -o {output_path}' run_command( tlsx_command, shell=True, ) tlsx_output = [] with open(output_path) as f: tlsx_output = f.readlines() tldextract_target = tldextract.extract(ip_domain) for doms in tlsx_output: doms = doms.strip() tldextract_res = tldextract.extract(doms) if ip_domain != doms and tldextract_res.domain == tldextract_target.domain and tldextract_res.subdomain == '': related_tlds.append(doms) related_tlds = list(set(related_tlds)) domain_info.related_tlds = related_tlds except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] related_domains_list = [] if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_related = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_related) related_domains_list.append(_domain['name']) for _domain in related_tlds: domain_related = RelatedDomain.objects.get_or_create( name=_domain, )[0] db_domain_info.related_tlds.add(domain_related) for _ip in historical_ips: historical_ip = HistoricalIP.objects.get_or_create( ip=_ip['ip'], owner=_ip['owner'], location=_ip['location'], last_seen=_ip['last_seen'], )[0] db_domain_info.historical_ips.add(historical_ip) domain.domain_info = db_domain_info domain.save() command = f'netlas host {ip_domain} -f json' # check if netlas key is provided netlas_key = get_netlas_key() command += f' -a {netlas_key}' if netlas_key else '' result = subprocess.check_output(command.split()).decode('utf-8') if 'Failed to parse response data' in result: # do fallback return { 'status': False, 'ip_domain': ip_domain, 'result': "Netlas limit exceeded.", 'message': 'Netlas limit exceeded.' } try: result = json.loads(result) logger.info(result) whois = result.get('whois') if result.get('whois') else {} domain_info.created = whois.get('created_date') domain_info.expires = whois.get('expiration_date') domain_info.updated = whois.get('updated_date') domain_info.whois_server = whois.get('whois_server') if 'registrant' in whois: registrant = whois.get('registrant') domain_info.registrant_name = registrant.get('name') domain_info.registrant_country = registrant.get('country') domain_info.registrant_id = registrant.get('id') domain_info.registrant_state = registrant.get('province') domain_info.registrant_city = registrant.get('city') domain_info.registrant_phone = registrant.get('phone') domain_info.registrant_address = registrant.get('street') domain_info.registrant_organization = registrant.get('organization') domain_info.registrant_fax = registrant.get('fax') domain_info.registrant_zip_code = registrant.get('postal_code') email_search = EMAIL_REGEX.search(str(registrant.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.registrant_email = field_content if 'administrative' in whois: administrative = whois.get('administrative') domain_info.admin_name = administrative.get('name') domain_info.admin_country = administrative.get('country') domain_info.admin_id = administrative.get('id') domain_info.admin_state = administrative.get('province') domain_info.admin_city = administrative.get('city') domain_info.admin_phone = administrative.get('phone') domain_info.admin_address = administrative.get('street') domain_info.admin_organization = administrative.get('organization') domain_info.admin_fax = administrative.get('fax') domain_info.admin_zip_code = administrative.get('postal_code') mail_search = EMAIL_REGEX.search(str(administrative.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.admin_email = field_content if 'technical' in whois: technical = whois.get('technical') domain_info.tech_name = technical.get('name') domain_info.tech_country = technical.get('country') domain_info.tech_state = technical.get('province') domain_info.tech_id = technical.get('id') domain_info.tech_city = technical.get('city') domain_info.tech_phone = technical.get('phone') domain_info.tech_address = technical.get('street') domain_info.tech_organization = technical.get('organization') domain_info.tech_fax = technical.get('fax') domain_info.tech_zip_code = technical.get('postal_code') mail_search = EMAIL_REGEX.search(str(technical.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.tech_email = field_content if 'dns' in result: dns = result.get('dns') domain_info.mx_records = dns.get('mx') domain_info.txt_records = dns.get('txt') domain_info.a_records = dns.get('a') domain_info.ns_records = whois.get('name_servers') domain_info.dnssec = True if whois.get('dnssec') else False domain_info.status = whois.get('status') if 'registrar' in whois: registrar = whois.get('registrar') domain_info.registrar_name = registrar.get('name') domain_info.registrar_email = registrar.get('email') domain_info.registrar_phone = registrar.get('phone') domain_info.registrar_url = registrar.get('url') # find associated domains if registrant email is found related_domains = reverse_whois(domain_info.get('registrant_email')) if domain_info.get('registrant_email') else [] for _domain in related_domains: related_domains_list.append(_domain['name']) # remove duplicate domains from related domains list related_domains_list = list(set(related_domains_list)) domain_info.related_domains = related_domains_list # save to db if domain exists if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_rel = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_rel) db_domain_info.dnssec = domain_info.get('dnssec') #dates db_domain_info.created = domain_info.get('created') db_domain_info.updated = domain_info.get('updated') db_domain_info.expires = domain_info.get('expires') #registrar db_domain_info.registrar = Registrar.objects.get_or_create( name=domain_info.get('registrar_name'), email=domain_info.get('registrar_email'), phone=domain_info.get('registrar_phone'), url=domain_info.get('registrar_url'), )[0] db_domain_info.registrant = DomainRegistration.objects.get_or_create( name=domain_info.get('registrant_name'), organization=domain_info.get('registrant_organization'), address=domain_info.get('registrant_address'), city=domain_info.get('registrant_city'), state=domain_info.get('registrant_state'), zip_code=domain_info.get('registrant_zip_code'), country=domain_info.get('registrant_country'), email=domain_info.get('registrant_email'), phone=domain_info.get('registrant_phone'), fax=domain_info.get('registrant_fax'), id_str=domain_info.get('registrant_id'), )[0] db_domain_info.admin = DomainRegistration.objects.get_or_create( name=domain_info.get('admin_name'), organization=domain_info.get('admin_organization'), address=domain_info.get('admin_address'), city=domain_info.get('admin_city'), state=domain_info.get('admin_state'), zip_code=domain_info.get('admin_zip_code'), country=domain_info.get('admin_country'), email=domain_info.get('admin_email'), phone=domain_info.get('admin_phone'), fax=domain_info.get('admin_fax'), id_str=domain_info.get('admin_id'), )[0] db_domain_info.tech = DomainRegistration.objects.get_or_create( name=domain_info.get('tech_name'), organization=domain_info.get('tech_organization'), address=domain_info.get('tech_address'), city=domain_info.get('tech_city'), state=domain_info.get('tech_state'), zip_code=domain_info.get('tech_zip_code'), country=domain_info.get('tech_country'), email=domain_info.get('tech_email'), phone=domain_info.get('tech_phone'), fax=domain_info.get('tech_fax'), id_str=domain_info.get('tech_id'), )[0] for status in domain_info.get('status') or []: _status = WhoisStatus.objects.get_or_create( name=status )[0] _status.save() db_domain_info.status.add(_status) for ns in domain_info.get('ns_records') or []: _ns = NameServer.objects.get_or_create( name=ns )[0] _ns.save() db_domain_info.name_servers.add(_ns) for a in domain_info.get('a_records') or []: _a = DNSRecord.objects.get_or_create( name=a, type='a' )[0] _a.save() db_domain_info.dns_records.add(_a) for mx in domain_info.get('mx_records') or []: _mx = DNSRecord.objects.get_or_create( name=mx, type='mx' )[0] _mx.save() db_domain_info.dns_records.add(_mx) for txt in domain_info.get('txt_records') or []: _txt = DNSRecord.objects.get_or_create( name=txt, type='txt' )[0] _txt.save() db_domain_info.dns_records.add(_txt) db_domain_info.geolocation_iso = domain_info.get('registrant_country') db_domain_info.whois_server = domain_info.get('whois_server') db_domain_info.save() domain.domain_info = db_domain_info domain.save() except Exception as e: return { 'status': False, 'ip_domain': ip_domain, 'result': "unable to fetch records from WHOIS database.", 'message': str(e) } return { 'status': True, 'ip_domain': ip_domain, 'dnssec': domain_info.get('dnssec'), 'created': domain_info.get('created'), 'updated': domain_info.get('updated'), 'expires': domain_info.get('expires'), 'geolocation_iso': domain_info.get('registrant_country'), 'domain_statuses': domain_info.get('status'), 'whois_server': domain_info.get('whois_server'), 'dns': { 'a': domain_info.get('a_records'), 'mx': domain_info.get('mx_records'), 'txt': domain_info.get('txt_records'), }, 'registrar': { 'name': domain_info.get('registrar_name'), 'phone': domain_info.get('registrar_phone'), 'email': domain_info.get('registrar_email'), 'url': domain_info.get('registrar_url'), }, 'registrant': { 'name': domain_info.get('registrant_name'), 'id': domain_info.get('registrant_id'), 'organization': domain_info.get('registrant_organization'), 'address': domain_info.get('registrant_address'), 'city': domain_info.get('registrant_city'), 'state': domain_info.get('registrant_state'), 'zipcode': domain_info.get('registrant_zip_code'), 'country': domain_info.get('registrant_country'), 'phone': domain_info.get('registrant_phone'), 'fax': domain_info.get('registrant_fax'), 'email': domain_info.get('registrant_email'), }, 'admin': { 'name': domain_info.get('admin_name'), 'id': domain_info.get('admin_id'), 'organization': domain_info.get('admin_organization'), 'address':domain_info.get('admin_address'), 'city': domain_info.get('admin_city'), 'state': domain_info.get('admin_state'), 'zipcode': domain_info.get('admin_zip_code'), 'country': domain_info.get('admin_country'), 'phone': domain_info.get('admin_phone'), 'fax': domain_info.get('admin_fax'), 'email': domain_info.get('admin_email'), }, 'technical_contact': { 'name': domain_info.get('tech_name'), 'id': domain_info.get('tech_id'), 'organization': domain_info.get('tech_organization'), 'address': domain_info.get('tech_address'), 'city': domain_info.get('tech_city'), 'state': domain_info.get('tech_state'), 'zipcode': domain_info.get('tech_zip_code'), 'country': domain_info.get('tech_country'), 'phone': domain_info.get('tech_phone'), 'fax': domain_info.get('tech_fax'), 'email': domain_info.get('tech_email'), }, 'nameservers': domain_info.get('ns_records'), # 'similar_domains': domain_info.get('similar_domains'), 'related_domains': domain_info.get('related_domains'), 'related_tlds': domain_info.get('related_tlds'), 'historical_ips': domain_info.get('historical_ips'), } @app.task(name='remove_duplicate_endpoints', bind=False, queue='remove_duplicate_endpoints_queue') def remove_duplicate_endpoints( scan_history_id, domain_id, subdomain_id=None, filter_ids=[], filter_status=[200, 301, 404], duplicate_removal_fields=ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS ): """Remove duplicate endpoints. Check for implicit redirections by comparing endpoints: - [x] `content_length` similarities indicating redirections - [x] `page_title` (check for same page title) - [ ] Sign-in / login page (check for endpoints with the same words) Args: scan_history_id: ScanHistory id. domain_id (int): Domain id. subdomain_id (int, optional): Subdomain id. filter_ids (list): List of endpoint ids to filter on. filter_status (list): List of HTTP status codes to filter on. duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates """ logger.info(f'Removing duplicate endpoints based on {duplicate_removal_fields}') endpoints = ( EndPoint.objects .filter(scan_history__id=scan_history_id) .filter(target_domain__id=domain_id) ) if filter_status: endpoints = endpoints.filter(http_status__in=filter_status) if subdomain_id: endpoints = endpoints.filter(subdomain__id=subdomain_id) if filter_ids: endpoints = endpoints.filter(id__in=filter_ids) for field_name in duplicate_removal_fields: cl_query = ( endpoints .values_list(field_name) .annotate(mc=Count(field_name)) .order_by('-mc') ) for (field_value, count) in cl_query: if count > DELETE_DUPLICATES_THRESHOLD: eps_to_delete = ( endpoints .filter(**{field_name: field_value}) .order_by('discovered_date') .all()[1:] ) msg = f'Deleting {len(eps_to_delete)} endpoints [reason: same {field_name} {field_value}]' for ep in eps_to_delete: url = urlparse(ep.http_url) if url.path in ['', '/', '/login']: # try do not delete the original page that other pages redirect to continue msg += f'\n\t {ep.http_url} [{ep.http_status}] [{field_name}={field_value}]' ep.delete() logger.warning(msg) @app.task(name='run_command', bind=False, queue='run_command_queue') def run_command(cmd, cwd=None, shell=False, history_file=None, scan_id=None, activity_id=None): """Run a given command using subprocess module. Args: cmd (str): Command to run. cwd (str): Current working directory. echo (bool): Log command. shell (bool): Run within separate shell if True. history_file (str): Write command + output to history file. Returns: tuple: Tuple with return_code, output. """ logger.info(cmd) logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Run the command using subprocess popen = subprocess.Popen( cmd if shell else cmd.split(), shell=shell, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, cwd=cwd, universal_newlines=True) output = '' for stdout_line in iter(popen.stdout.readline, ""): item = stdout_line.strip() output += '\n' + item logger.debug(item) popen.stdout.close() popen.wait() return_code = popen.returncode command_obj.output = output command_obj.return_code = return_code command_obj.save() if history_file: mode = 'a' if not os.path.exists(history_file): mode = 'w' with open(history_file, mode) as f: f.write(f'\n{cmd}\n{return_code}\n{output}\n------------------\n') return return_code, output #-------------# # Other utils # #-------------# def stream_command(cmd, cwd=None, shell=False, history_file=None, encoding='utf-8', scan_id=None, activity_id=None, trunc_char=None): # Log cmd logger.info(cmd) # logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Sanitize the cmd command = cmd if shell else cmd.split() # Run the command using subprocess process = subprocess.Popen( command, stdout=subprocess.PIPE, stderr=subprocess.PIPE, shell=shell) # Log the output in real-time to the database output = "" # Process the output for line in iter(lambda: process.stdout.readline() or process.stderr.readline(), b''): line = re.sub(r'\x1b[^m]*m', '', line.decode('utf-8').strip()) if trunc_char and line.endswith(trunc_char): line = line[:-1] item = line # Try to parse the line as JSON try: item = json.loads(line) except json.JSONDecodeError: pass # Yield the line #logger.debug(item) yield item # Add the log line to the output output += line + "\n" # Update the command record in the database command_obj.output = output command_obj.save() # Retrieve the return code and output process.wait() return_code = process.returncode # Update the return code and final output in the database command_obj.return_code = return_code command_obj.save() # Append the command, return code and output to the history file if history_file is not None: with open(history_file, "a") as f: f.write(f"{cmd}\n{return_code}\n{output}\n") def process_httpx_response(line): """TODO: implement this""" def extract_httpx_url(line): """Extract final URL from httpx results. Always follow redirects to find the last URL. Args: line (dict): URL data output by httpx. Returns: tuple: (final_url, redirect_bool) tuple. """ status_code = line.get('status_code', 0) final_url = line.get('final_url') location = line.get('location') chain_status_codes = line.get('chain_status_codes', []) # Final URL is already looking nice, if it exists return it if final_url: return final_url, False http_url = line['url'] # fallback to url field # Handle redirects manually REDIRECT_STATUS_CODES = [301, 302] is_redirect = ( status_code in REDIRECT_STATUS_CODES or any(x in REDIRECT_STATUS_CODES for x in chain_status_codes) ) if is_redirect and location: if location.startswith(('http', 'https')): http_url = location else: http_url = f'{http_url}/{location.lstrip("/")}' # Sanitize URL http_url = sanitize_url(http_url) return http_url, is_redirect #-------------# # OSInt utils # #-------------# def get_and_save_dork_results(lookup_target, results_dir, type, lookup_keywords=None, lookup_extensions=None, delay=3, page_count=2, scan_history=None): """ Uses gofuzz to dork and store information Args: lookup_target (str): target to look into such as stackoverflow or even the target itself results_dir (str): Results directory type (str): Dork Type Title lookup_keywords (str): comma separated keywords or paths to look for lookup_extensions (str): comma separated extensions to look for delay (int): delay between each requests page_count (int): pages in google to extract information scan_history (startScan.ScanHistory): Scan History Object """ results = [] gofuzz_command = f'{GOFUZZ_EXEC_PATH} -t {lookup_target} -d {delay} -p {page_count}' if lookup_extensions: gofuzz_command += f' -e {lookup_extensions}' elif lookup_keywords: gofuzz_command += f' -w {lookup_keywords}' output_file = f'{results_dir}/gofuzz.txt' gofuzz_command += f' -o {output_file}' history_file = f'{results_dir}/commands.txt' try: run_command( gofuzz_command, shell=False, history_file=history_file, scan_id=scan_history.id, ) if not os.path.isfile(output_file): return with open(output_file) as f: for line in f.readlines(): url = line.strip() if url: results.append(url) dork, created = Dork.objects.get_or_create( type=type, url=url ) if scan_history: scan_history.dorks.add(dork) # remove output file os.remove(output_file) except Exception as e: logger.exception(e) return results def get_and_save_emails(scan_history, activity_id, results_dir): """Get and save emails from Google, Bing and Baidu. Args: scan_history (startScan.ScanHistory): Scan history object. activity_id: ScanActivity Object results_dir (str): Results directory. Returns: list: List of emails found. """ emails = [] # Proxy settings # get_random_proxy() # Gather emails from Google, Bing and Baidu output_file = f'{results_dir}/emails_tmp.txt' history_file = f'{results_dir}/commands.txt' command = f'python3 /usr/src/github/Infoga/infoga.py --domain {scan_history.domain.name} --source all --report {output_file}' try: run_command( command, shell=False, history_file=history_file, scan_id=scan_history.id, activity_id=activity_id) if not os.path.isfile(output_file): logger.info('No Email results') return [] with open(output_file) as f: for line in f.readlines(): if 'Email' in line: split_email = line.split(' ')[2] emails.append(split_email) output_path = f'{results_dir}/emails.txt' with open(output_path, 'w') as output_file: for email_address in emails: save_email(email_address, scan_history) output_file.write(f'{email_address}\n') except Exception as e: logger.exception(e) return emails def save_metadata_info(meta_dict): """Extract metadata from Google Search. Args: meta_dict (dict): Info dict. Returns: list: List of startScan.MetaFinderDocument objects. """ logger.warning(f'Getting metadata for {meta_dict.osint_target}') scan_history = ScanHistory.objects.get(id=meta_dict.scan_id) # Proxy settings get_random_proxy() # Get metadata result = extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if not result: logger.error(f'No metadata result from Google Search for {meta_dict.osint_target}.') return [] # Add metadata info to DB results = [] for metadata_name, data in result.get_metadata().items(): subdomain = Subdomain.objects.get( scan_history=meta_dict.scan_id, name=meta_dict.osint_target) metadata = DottedDict({k: v for k, v in data.items()}) meta_finder_document = MetaFinderDocument( subdomain=subdomain, target_domain=meta_dict.domain, scan_history=scan_history, url=metadata.url, doc_name=metadata_name, http_status=metadata.status_code, producer=metadata.metadata.get('Producer'), creator=metadata.metadata.get('Creator'), creation_date=metadata.metadata.get('CreationDate'), modified_date=metadata.metadata.get('ModDate'), author=metadata.metadata.get('Author'), title=metadata.metadata.get('Title'), os=metadata.metadata.get('OSInfo')) meta_finder_document.save() results.append(data) return results #-----------------# # Utils functions # #-----------------# def create_scan_activity(scan_history_id, message, status): scan_activity = ScanActivity() scan_activity.scan_of = ScanHistory.objects.get(pk=scan_history_id) scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id #--------------------# # Database functions # #--------------------# def save_vulnerability(**vuln_data): references = vuln_data.pop('references', []) cve_ids = vuln_data.pop('cve_ids', []) cwe_ids = vuln_data.pop('cwe_ids', []) tags = vuln_data.pop('tags', []) subscan = vuln_data.pop('subscan', None) # remove nulls vuln_data = replace_nulls(vuln_data) # Create vulnerability vuln, created = Vulnerability.objects.get_or_create(**vuln_data) if created: vuln.discovered_date = timezone.now() vuln.open_status = True vuln.save() # Save vuln tags for tag_name in tags or []: tag, created = VulnerabilityTags.objects.get_or_create(name=tag_name) if tag: vuln.tags.add(tag) vuln.save() # Save CVEs for cve_id in cve_ids or []: cve, created = CveId.objects.get_or_create(name=cve_id) if cve: vuln.cve_ids.add(cve) vuln.save() # Save CWEs for cve_id in cwe_ids or []: cwe, created = CweId.objects.get_or_create(name=cve_id) if cwe: vuln.cwe_ids.add(cwe) vuln.save() # Save vuln reference for url in references or []: ref, created = VulnerabilityReference.objects.get_or_create(url=url) if created: vuln.references.add(ref) vuln.save() # Save subscan id in vuln object if subscan: vuln.vuln_subscan_ids.add(subscan) vuln.save() return vuln, created def save_endpoint( http_url, ctx={}, crawl=False, is_default=False, **endpoint_data): """Get or create EndPoint object. If crawl is True, also crawl the endpoint HTTP URL with httpx. Args: http_url (str): Input HTTP URL. is_default (bool): If the url is a default url for SubDomains. scan_history (startScan.models.ScanHistory): ScanHistory object. domain (startScan.models.Domain): Domain object. subdomain (starScan.models.Subdomain): Subdomain object. results_dir (str, optional): Results directory. crawl (bool, optional): Run httpx on endpoint if True. Default: False. force (bool, optional): Force crawl even if ENABLE_HTTP_CRAWL mode is on. subscan (startScan.models.SubScan, optional): SubScan object. Returns: tuple: (startScan.models.EndPoint, created) where `created` is a boolean indicating if the object is new or already existed. """ # remove nulls endpoint_data = replace_nulls(endpoint_data) scheme = urlparse(http_url).scheme endpoint = None created = False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in http_url: logger.error(f"{http_url} is not a URL of domain {domain.name}. Skipping.") return None, False if crawl: ctx['track'] = False results = http_crawl( urls=[http_url], method='HEAD', ctx=ctx) if results: endpoint_data = results[0] endpoint_id = endpoint_data['endpoint_id'] created = endpoint_data['endpoint_created'] endpoint = EndPoint.objects.get(pk=endpoint_id) elif not scheme: return None, False else: # add dumb endpoint without probing it scan = ScanHistory.objects.filter(pk=ctx.get('scan_history_id')).first() domain = Domain.objects.filter(pk=ctx.get('domain_id')).first() if not validators.url(http_url): return None, False http_url = sanitize_url(http_url) endpoint, created = EndPoint.objects.get_or_create( scan_history=scan, target_domain=domain, http_url=http_url, **endpoint_data) if created: endpoint.is_default = is_default endpoint.discovered_date = timezone.now() endpoint.save() subscan_id = ctx.get('subscan_id') if subscan_id: endpoint.endpoint_subscan_ids.add(subscan_id) endpoint.save() return endpoint, created def save_subdomain(subdomain_name, ctx={}): """Get or create Subdomain object. Args: subdomain_name (str): Subdomain name. scan_history (startScan.models.ScanHistory): ScanHistory object. Returns: tuple: (startScan.models.Subdomain, created) where `created` is a boolean indicating if the object has been created in DB. """ scan_id = ctx.get('scan_history_id') subscan_id = ctx.get('subscan_id') out_of_scope_subdomains = ctx.get('out_of_scope_subdomains', []) valid_domain = ( validators.domain(subdomain_name) or validators.ipv4(subdomain_name) or validators.ipv6(subdomain_name) ) if not valid_domain: logger.error(f'{subdomain_name} is not an invalid domain. Skipping.') return None, False if subdomain_name in out_of_scope_subdomains: logger.error(f'{subdomain_name} is out-of-scope. Skipping.') return None, False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in subdomain_name: logger.error(f"{subdomain_name} is not a subdomain of domain {domain.name}. Skipping.") return None, False scan = ScanHistory.objects.filter(pk=scan_id).first() domain = scan.domain if scan else None subdomain, created = Subdomain.objects.get_or_create( scan_history=scan, target_domain=domain, name=subdomain_name) if created: # logger.warning(f'Found new subdomain {subdomain_name}') subdomain.discovered_date = timezone.now() if subscan_id: subdomain.subdomain_subscan_ids.add(subscan_id) subdomain.save() return subdomain, created def save_email(email_address, scan_history=None): if not validators.email(email_address): logger.info(f'Email {email_address} is invalid. Skipping.') return None, False email, created = Email.objects.get_or_create(address=email_address) # if created: # logger.warning(f'Found new email address {email_address}') # Add email to ScanHistory if scan_history: scan_history.emails.add(email) scan_history.save() return email, created def save_employee(name, designation, scan_history=None): employee, created = Employee.objects.get_or_create( name=name, designation=designation) # if created: # logger.warning(f'Found new employee {name}') # Add employee to ScanHistory if scan_history: scan_history.employees.add(employee) scan_history.save() return employee, created def save_ip_address(ip_address, subdomain=None, subscan=None, **kwargs): if not (validators.ipv4(ip_address) or validators.ipv6(ip_address)): logger.info(f'IP {ip_address} is not a valid IP. Skipping.') return None, False ip, created = IpAddress.objects.get_or_create(address=ip_address) # if created: # logger.warning(f'Found new IP {ip_address}') # Set extra attributes for key, value in kwargs.items(): setattr(ip, key, value) ip.save() # Add IP to subdomain if subdomain: subdomain.ip_addresses.add(ip) subdomain.save() # Add subscan to IP if subscan: ip.ip_subscan_ids.add(subscan) # Geo-localize IP asynchronously if created: geo_localize.delay(ip_address, ip.id) return ip, created def save_imported_subdomains(subdomains, ctx={}): """Take a list of subdomains imported and write them to from_imported.txt. Args: subdomains (list): List of subdomain names. scan_history (startScan.models.ScanHistory): ScanHistory instance. domain (startScan.models.Domain): Domain instance. results_dir (str): Results directory. """ domain_id = ctx['domain_id'] domain = Domain.objects.get(pk=domain_id) results_dir = ctx.get('results_dir', RENGINE_RESULTS) # Validate each subdomain and de-duplicate entries subdomains = list(set([ subdomain for subdomain in subdomains if validators.domain(subdomain) and domain.name == get_domain_from_subdomain(subdomain) ])) if not subdomains: return logger.warning(f'Found {len(subdomains)} imported subdomains.') with open(f'{results_dir}/from_imported.txt', 'w+') as output_file: for name in subdomains: subdomain_name = name.strip() subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain.is_imported_subdomain = True subdomain.save() output_file.write(f'{subdomain}\n') @app.task(name='query_reverse_whois', bind=False, queue='query_reverse_whois_queue') def query_reverse_whois(lookup_keyword): """Queries Reverse WHOIS information for an organization or email address. Args: lookup_keyword (str): Registrar Name or email Returns: dict: Reverse WHOIS information. """ return get_associated_domains(lookup_keyword) @app.task(name='query_ip_history', bind=False, queue='query_ip_history_queue') def query_ip_history(domain): """Queries the IP history for a domain Args: domain (str): domain_name Returns: list: list of historical ip addresses """ return get_domain_historical_ip_address(domain) @app.task(name='gpt_vulnerability_description', bind=False, queue='gpt_queue') def gpt_vulnerability_description(vulnerability_id): """Generate and store Vulnerability Description using GPT. Args: vulnerability_id (Vulnerability Model ID): Vulnerability ID to fetch Description. """ logger.info('Getting GPT Vulnerability Description') try: lookup_vulnerability = Vulnerability.objects.get(id=vulnerability_id) lookup_url = urlparse(lookup_vulnerability.http_url) path = lookup_url.path except Exception as e: return { 'status': False, 'error': str(e) } # check in db GPTVulnerabilityReport model if vulnerability description and path matches stored = GPTVulnerabilityReport.objects.filter(url_path=path).filter(title=lookup_vulnerability.name).first() if stored: response = { 'status': True, 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: vulnerability_description = get_gpt_vuln_input_description( lookup_vulnerability.name, path ) # one can add more description here later gpt_generator = GPTVulnerabilityReportGenerator() response = gpt_generator.get_vulnerability_description(vulnerability_description) add_gpt_description_db( lookup_vulnerability.name, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) # for all vulnerabilities with the same vulnerability name this description has to be stored. # also the consition is that the url must contain a part of this. for vuln in Vulnerability.objects.filter(name=lookup_vulnerability.name, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() return response

import csv import json import os import pprint import subprocess import time import validators import whatportis import xmltodict import yaml import tldextract import concurrent.futures from datetime import datetime from urllib.parse import urlparse from api.serializers import SubdomainSerializer from celery import chain, chord, group from celery.result import allow_join_result from celery.utils.log import get_task_logger from django.db.models import Count from dotted_dict import DottedDict from django.utils import timezone from pycvesearch import CVESearch from metafinder.extractor import extract_metadata_from_google_search from reNgine.celery import app from reNgine.gpt import GPTVulnerabilityReportGenerator from reNgine.celery_custom_task import RengineTask from reNgine.common_func import * from reNgine.definitions import * from reNgine.settings import * from reNgine.gpt import * from reNgine.utilities import * from scanEngine.models import (EngineType, InstalledExternalTool, Notification, Proxy) from startScan.models import * from startScan.models import EndPoint, Subdomain, Vulnerability from targetApp.models import Domain """ Celery tasks. """ logger = get_task_logger(__name__) #----------------------# # Scan / Subscan tasks # #----------------------# @app.task(name='initiate_scan', bind=False, queue='initiate_scan_queue') def initiate_scan( scan_history_id, domain_id, engine_id=None, scan_type=LIVE_SCAN, results_dir=RENGINE_RESULTS, imported_subdomains=[], out_of_scope_subdomains=[], url_filter=''): """Initiate a new scan. Args: scan_history_id (int): ScanHistory id. domain_id (int): Domain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. imported_subdomains (list): Imported subdomains. out_of_scope_subdomains (list): Out-of-scope subdomains. url_filter (str): URL path. Default: '' """ # Get scan history scan = ScanHistory.objects.get(pk=scan_history_id) # Get scan engine engine_id = engine_id or scan.scan_type.id # scan history engine_id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, []) # Get domain and set last_scan_date domain = Domain.objects.get(pk=domain_id) domain.last_scan_date = timezone.now() domain.save() # Get path filter url_filter = url_filter.rstrip('/') # Get or create ScanHistory() object if scan_type == LIVE_SCAN: # immediate scan = ScanHistory.objects.get(pk=scan_history_id) scan.scan_status = RUNNING_TASK elif scan_type == SCHEDULED_SCAN: # scheduled scan = ScanHistory() scan.scan_status = INITIATED_TASK scan.scan_type = engine scan.celery_ids = [initiate_scan.request.id] scan.domain = domain scan.start_scan_date = timezone.now() scan.tasks = engine.tasks scan.results_dir = f'{results_dir}/{domain.name}_{scan.id}' add_gf_patterns = gf_patterns and 'fetch_url' in engine.tasks if add_gf_patterns: scan.used_gf_patterns = ','.join(gf_patterns) scan.save() # Create scan results dir os.makedirs(scan.results_dir) # Build task context ctx = { 'scan_history_id': scan_history_id, 'engine_id': engine_id, 'domain_id': domain.id, 'results_dir': scan.results_dir, 'url_filter': url_filter, 'yaml_configuration': config, 'out_of_scope_subdomains': out_of_scope_subdomains } ctx_str = json.dumps(ctx, indent=2) # Send start notif logger.warning(f'Starting scan {scan_history_id} with context:\n{ctx_str}') send_scan_notif.delay( scan_history_id, subscan_id=None, engine_id=engine_id, status=CELERY_TASK_STATUS_MAP[scan.scan_status]) # Save imported subdomains in DB save_imported_subdomains(imported_subdomains, ctx=ctx) # Create initial subdomain in DB: make a copy of domain as a subdomain so # that other tasks using subdomains can use it. subdomain_name = domain.name subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) # If enable_http_crawl is set, create an initial root HTTP endpoint so that # HTTP crawling can start somewhere http_url = f'{domain.name}{url_filter}' if url_filter else domain.name endpoint, _ = save_endpoint( http_url, ctx=ctx, crawl=enable_http_crawl, is_default=True, subdomain=subdomain ) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build Celery tasks, crafted according to the dependency graph below: # subdomain_discovery --> port_scan --> fetch_url --> dir_file_fuzz # osint vulnerability_scan # osint dalfox xss scan # screenshot # waf_detection workflow = chain( group( subdomain_discovery.si(ctx=ctx, description='Subdomain discovery'), osint.si(ctx=ctx, description='OS Intelligence') ), port_scan.si(ctx=ctx, description='Port scan'), fetch_url.si(ctx=ctx, description='Fetch URL'), group( dir_file_fuzz.si(ctx=ctx, description='Directories & files fuzz'), vulnerability_scan.si(ctx=ctx, description='Vulnerability scan'), screenshot.si(ctx=ctx, description='Screenshot'), waf_detection.si(ctx=ctx, description='WAF detection') ) ) # Build callback callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery chord logger.info(f'Running Celery workflow with {len(workflow.tasks) + 1} tasks') task = chain(workflow, callback).on_error(callback).delay() scan.celery_ids.append(task.id) scan.save() return { 'success': True, 'task_id': task.id } @app.task(name='initiate_subscan', bind=False, queue='subscan_queue') def initiate_subscan( scan_history_id, subdomain_id, engine_id=None, scan_type=None, results_dir=RENGINE_RESULTS, url_filter=''): """Initiate a new subscan. Args: scan_history_id (int): ScanHistory id. subdomain_id (int): Subdomain id. engine_id (int): Engine ID. scan_type (int): Scan type (periodic, live). results_dir (str): Results directory. url_filter (str): URL path. Default: '' """ # Get Subdomain, Domain and ScanHistory subdomain = Subdomain.objects.get(pk=subdomain_id) scan = ScanHistory.objects.get(pk=subdomain.scan_history.id) domain = Domain.objects.get(pk=subdomain.target_domain.id) # Get EngineType engine_id = engine_id or scan.scan_type.id engine = EngineType.objects.get(pk=engine_id) # Get YAML config config = yaml.safe_load(engine.yaml_configuration) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Create scan activity of SubScan Model subscan = SubScan( start_scan_date=timezone.now(), celery_ids=[initiate_subscan.request.id], scan_history=scan, subdomain=subdomain, type=scan_type, status=RUNNING_TASK, engine=engine) subscan.save() # Get YAML configuration config = yaml.safe_load(engine.yaml_configuration) # Create results directory results_dir = f'{scan.results_dir}/subscans/{subscan.id}' os.makedirs(results_dir, exist_ok=True) # Run task method = globals().get(scan_type) if not method: logger.warning(f'Task {scan_type} is not supported by reNgine. Skipping') return scan.tasks.append(scan_type) scan.save() # Send start notif send_scan_notif.delay( scan.id, subscan_id=subscan.id, engine_id=engine_id, status='RUNNING') # Build context ctx = { 'scan_history_id': scan.id, 'subscan_id': subscan.id, 'engine_id': engine_id, 'domain_id': domain.id, 'subdomain_id': subdomain.id, 'yaml_configuration': config, 'results_dir': results_dir, 'url_filter': url_filter } # Create initial endpoints in DB: find domain HTTP endpoint so that HTTP # crawling can start somewhere base_url = f'{subdomain.name}{url_filter}' if url_filter else subdomain.name endpoint, _ = save_endpoint( base_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint and endpoint.is_alive: # TODO: add `root_endpoint` property to subdomain and simply do # subdomain.root_endpoint = endpoint instead logger.warning(f'Found subdomain root HTTP URL {endpoint.http_url}') subdomain.http_url = endpoint.http_url subdomain.http_status = endpoint.http_status subdomain.response_time = endpoint.response_time subdomain.page_title = endpoint.page_title subdomain.content_type = endpoint.content_type subdomain.content_length = endpoint.content_length for tech in endpoint.techs.all(): subdomain.technologies.add(tech) subdomain.save() # Build header + callback workflow = method.si(ctx=ctx) callback = report.si(ctx=ctx).set(link_error=[report.si(ctx=ctx)]) # Run Celery tasks task = chain(workflow, callback).on_error(callback).delay() subscan.celery_ids.append(task.id) subscan.save() return { 'success': True, 'task_id': task.id } @app.task(name='report', bind=False, queue='report_queue') def report(ctx={}, description=None): """Report task running after all other tasks. Mark ScanHistory or SubScan object as completed and update with final status, log run details and send notification. Args: description (str, optional): Task description shown in UI. """ # Get objects subscan_id = ctx.get('subscan_id') scan_id = ctx.get('scan_history_id') engine_id = ctx.get('engine_id') scan = ScanHistory.objects.filter(pk=scan_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() # Get failed tasks tasks = ScanActivity.objects.filter(scan_of=scan).all() if subscan: tasks = tasks.filter(celery_id__in=subscan.celery_ids) failed_tasks = tasks.filter(status=FAILED_TASK) # Get task status failed_count = failed_tasks.count() status = SUCCESS_TASK if failed_count == 0 else FAILED_TASK status_h = 'SUCCESS' if failed_count == 0 else 'FAILED' # Update scan / subscan status if subscan: subscan.stop_scan_date = timezone.now() subscan.status = status subscan.save() else: scan.scan_status = status scan.stop_scan_date = timezone.now() scan.save() # Send scan status notif send_scan_notif.delay( scan_history_id=scan_id, subscan_id=subscan_id, engine_id=engine_id, status=status_h) #------------------------- # # Tracked reNgine tasks # #--------------------------# @app.task(name='subdomain_discovery', queue='main_scan_queue', base=RengineTask, bind=True) def subdomain_discovery( self, host=None, ctx=None, description=None): """Uses a set of tools (see SUBDOMAIN_SCAN_DEFAULT_TOOLS) to scan all subdomains associated with a domain. Args: host (str): Hostname to scan. Returns: subdomains (list): List of subdomain names. """ if not host: host = self.subdomain.name if self.subdomain else self.domain.name if self.url_filter: logger.warning(f'Ignoring subdomains scan as an URL path filter was passed ({self.url_filter}).') return # Config config = self.yaml_configuration.get(SUBDOMAIN_DISCOVERY) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL) or self.yaml_configuration.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) tools = config.get(USES_TOOLS, SUBDOMAIN_SCAN_DEFAULT_TOOLS) default_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=True).filter(is_subdomain_gathering=True)] custom_subdomain_tools = [tool.name.lower() for tool in InstalledExternalTool.objects.filter(is_default=False).filter(is_subdomain_gathering=True)] send_subdomain_changes, send_interesting = False, False notif = Notification.objects.first() if notif: send_subdomain_changes = notif.send_subdomain_changes_notif send_interesting = notif.send_interesting_notif # Gather tools to run for subdomain scan if ALL in tools: tools = SUBDOMAIN_SCAN_DEFAULT_TOOLS + custom_subdomain_tools tools = [t.lower() for t in tools] # Make exception for amass since tool name is amass, but command is amass-active/passive default_subdomain_tools.append('amass-passive') default_subdomain_tools.append('amass-active') # Run tools for tool in tools: cmd = None logger.info(f'Scanning subdomains for {host} with {tool}') proxy = get_random_proxy() if tool in default_subdomain_tools: if tool == 'amass-passive': cmd = f'amass enum -passive -d {host} -o {self.results_dir}/subdomains_amass.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' elif tool == 'amass-active': use_amass_config = config.get(USE_AMASS_CONFIG, False) amass_wordlist_name = config.get(AMASS_WORDLIST, 'deepmagic.com-prefixes-top50000') wordlist_path = f'/usr/src/wordlist/{amass_wordlist_name}.txt' cmd = f'amass enum -active -d {host} -o {self.results_dir}/subdomains_amass_active.txt' cmd += ' -config /root/.config/amass.ini' if use_amass_config else '' cmd += f' -brute -w {wordlist_path}' elif tool == 'sublist3r': cmd = f'python3 /usr/src/github/Sublist3r/sublist3r.py -d {host} -t {threads} -o {self.results_dir}/subdomains_sublister.txt' elif tool == 'subfinder': cmd = f'subfinder -d {host} -o {self.results_dir}/subdomains_subfinder.txt' use_subfinder_config = config.get(USE_SUBFINDER_CONFIG, False) cmd += ' -config /root/.config/subfinder/config.yaml' if use_subfinder_config else '' cmd += f' -proxy {proxy}' if proxy else '' cmd += f' -timeout {timeout}' if timeout else '' cmd += f' -t {threads}' if threads else '' cmd += f' -silent' elif tool == 'oneforall': cmd = f'python3 /usr/src/github/OneForAll/oneforall.py --target {host} run' cmd_extract = f'cut -d\',\' -f6 /usr/src/github/OneForAll/results/{host}.csv > {self.results_dir}/subdomains_oneforall.txt' cmd_rm = f'rm -rf /usr/src/github/OneForAll/results/{host}.csv' cmd += f' && {cmd_extract} && {cmd_rm}' elif tool == 'ctfr': results_file = self.results_dir + '/subdomains_ctfr.txt' cmd = f'python3 /usr/src/github/ctfr/ctfr.py -d {host} -o {results_file}' cmd_extract = f"cat {results_file} | sed 's/\*.//g' | tail -n +12 | uniq | sort > {results_file}" cmd += f' && {cmd_extract}' elif tool == 'tlsx': results_file = self.results_dir + '/subdomains_tlsx.txt' cmd = f'tlsx -san -cn -silent -ro -host {host}' cmd += f" | sed -n '/^\([a-zA-Z0-9]\([-a-zA-Z0-9]*[a-zA-Z0-9]\)\?\.\)\+{host}$/p' | uniq | sort" cmd += f' > {results_file}' elif tool == 'netlas': results_file = self.results_dir + '/subdomains_netlas.txt' cmd = f'netlas search -d domain -i domain domain:"*.{host}" -f json' netlas_key = get_netlas_key() cmd += f' -a {netlas_key}' if netlas_key else '' cmd_extract = f"grep -oE '([a-zA-Z0-9]([-a-zA-Z0-9]*[a-zA-Z0-9])?\.)+{host}'" cmd += f' | {cmd_extract} > {results_file}' elif tool in custom_subdomain_tools: tool_query = InstalledExternalTool.objects.filter(name__icontains=tool.lower()) if not tool_query.exists(): logger.error(f'Missing {{TARGET}} and {{OUTPUT}} placeholders in {tool} configuration. Skipping.') continue custom_tool = tool_query.first() cmd = custom_tool.subdomain_gathering_command if '{TARGET}' in cmd and '{OUTPUT}' in cmd: cmd = cmd.replace('{TARGET}', host) cmd = cmd.replace('{OUTPUT}', f'{self.results_dir}/subdomains_{tool}.txt') cmd = cmd.replace('{PATH}', custom_tool.github_clone_path) if '{PATH}' in cmd else cmd else: logger.warning( f'Subdomain discovery tool "{tool}" is not supported by reNgine. Skipping.') continue # Run tool try: run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) except Exception as e: logger.error( f'Subdomain discovery tool "{tool}" raised an exception') logger.exception(e) # Gather all the tools' results in one single file. Write subdomains into # separate files, and sort all subdomains. run_command( f'cat {self.results_dir}/subdomains_*.txt > {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {self.output_path} -o {self.output_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) with open(self.output_path) as f: lines = f.readlines() # Parse the output_file file and store Subdomain and EndPoint objects found # in db. subdomain_count = 0 subdomains = [] urls = [] for line in lines: subdomain_name = line.strip() valid_url = bool(validators.url(subdomain_name)) valid_domain = ( bool(validators.domain(subdomain_name)) or bool(validators.ipv4(subdomain_name)) or bool(validators.ipv6(subdomain_name)) or valid_url ) if not valid_domain: logger.error(f'Subdomain {subdomain_name} is not a valid domain, IP or URL. Skipping.') continue if valid_url: subdomain_name = urlparse(subdomain_name).netloc if subdomain_name in self.out_of_scope_subdomains: logger.error(f'Subdomain {subdomain_name} is out of scope. Skipping.') continue # Add subdomain subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain_count += 1 subdomains.append(subdomain) urls.append(subdomain.name) # Bulk crawl subdomains if enable_http_crawl: ctx['track'] = True http_crawl(urls, ctx=ctx, is_ran_from_subdomain_scan=True) # Find root subdomain endpoints for subdomain in subdomains: pass # Send notifications subdomains_str = '\n'.join([f'• `{subdomain.name}`' for subdomain in subdomains]) self.notify(fields={ 'Subdomain count': len(subdomains), 'Subdomains': subdomains_str, }) if send_subdomain_changes and self.scan_id and self.domain_id: added = get_new_added_subdomain(self.scan_id, self.domain_id) removed = get_removed_subdomain(self.scan_id, self.domain_id) if added: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in added]) self.notify(fields={'Added subdomains': subdomains_str}) if removed: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in removed]) self.notify(fields={'Removed subdomains': subdomains_str}) if send_interesting and self.scan_id and self.domain_id: interesting_subdomains = get_interesting_subdomains(self.scan_id, self.domain_id) if interesting_subdomains: subdomains_str = '\n'.join([f'• `{subdomain}`' for subdomain in interesting_subdomains]) self.notify(fields={'Interesting subdomains': subdomains_str}) return SubdomainSerializer(subdomains, many=True).data @app.task(name='osint', queue='main_scan_queue', base=RengineTask, bind=True) def osint(self, host=None, ctx={}, description=None): """Run Open-Source Intelligence tools on selected domain. Args: host (str): Hostname to scan. Returns: dict: Results from osint discovery and dorking. """ config = self.yaml_configuration.get(OSINT) or OSINT_DEFAULT_CONFIG results = {} grouped_tasks = [] if 'discover' in config: ctx['track'] = False # results = osint_discovery(host=host, ctx=ctx) _task = osint_discovery.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, activity_id=self.activity_id, results_dir=self.results_dir, ctx=ctx ) grouped_tasks.append(_task) if OSINT_DORK in config or OSINT_CUSTOM_DORK in config: _task = dorking.si( config=config, host=self.scan.domain.name, scan_history_id=self.scan.id, results_dir=self.results_dir ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('OSINT Tasks finished...') # with open(self.output_path, 'w') as f: # json.dump(results, f, indent=4) # # return results @app.task(name='osint_discovery', queue='osint_discovery_queue', bind=False) def osint_discovery(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run OSINT discovery. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: dict: osint metadat and theHarvester and h8mail results. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) osint_lookup = config.get(OSINT_DISCOVER, []) osint_intensity = config.get(INTENSITY, 'normal') documents_limit = config.get(OSINT_DOCUMENTS_LIMIT, 50) results = {} meta_info = [] emails = [] creds = [] # Get and save meta info if 'metainfo' in osint_lookup: if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': host, 'domain': host, 'scan_id': scan_history_id, 'documents_limit': documents_limit }) meta_info.append(save_metadata_info(meta_dict)) # TODO: disabled for now # elif osint_intensity == 'deep': # subdomains = Subdomain.objects # if self.scan: # subdomains = subdomains.filter(scan_history=self.scan) # for subdomain in subdomains: # meta_dict = DottedDict({ # 'osint_target': subdomain.name, # 'domain': self.domain, # 'scan_id': self.scan_id, # 'documents_limit': documents_limit # }) # meta_info.append(save_metadata_info(meta_dict)) grouped_tasks = [] if 'emails' in osint_lookup: emails = get_and_save_emails(scan_history, activity_id, results_dir) emails_str = '\n'.join([f'• `{email}`' for email in emails]) # self.notify(fields={'Emails': emails_str}) # ctx['track'] = False _task = h8mail.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) if 'employees' in osint_lookup: ctx['track'] = False _task = theHarvester.si( config=config, host=host, scan_history_id=scan_history_id, activity_id=activity_id, results_dir=results_dir, ctx=ctx ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) # results['emails'] = results.get('emails', []) + emails # results['creds'] = creds # results['meta_info'] = meta_info return results @app.task(name='dorking', bind=False, queue='dorking_queue') def dorking(config, host, scan_history_id, results_dir): """Run Google dorks. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID results_dir (str): Path to store scan results Returns: list: Dorking results for each dork ran. """ # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh scan_history = ScanHistory.objects.get(pk=scan_history_id) dorks = config.get(OSINT_DORK, []) custom_dorks = config.get(OSINT_CUSTOM_DORK, []) results = [] # custom dorking has higher priority try: for custom_dork in custom_dorks: lookup_target = custom_dork.get('lookup_site') # replace with original host if _target_ lookup_target = host if lookup_target == '_target_' else lookup_target if 'lookup_extensions' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_extensions=custom_dork.get('lookup_extensions'), scan_history=scan_history ) elif 'lookup_keywords' in custom_dork: results = get_and_save_dork_results( lookup_target=lookup_target, results_dir=results_dir, type='custom_dork', lookup_keywords=custom_dork.get('lookup_keywords'), scan_history=scan_history ) except Exception as e: logger.exception(e) # default dorking try: for dork in dorks: logger.info(f'Getting dork information for {dork}') if dork == 'stackoverflow': results = get_and_save_dork_results( lookup_target='stackoverflow.com', results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'login_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/login/,login.html', page_count=5, scan_history=scan_history ) elif dork == 'admin_panels': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/admin/,admin.html', page_count=5, scan_history=scan_history ) elif dork == 'dashboard_pages': results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords='/dashboard/,dashboard.html', page_count=5, scan_history=scan_history ) elif dork == 'social_media' : social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'reddit.com' ] for site in social_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'project_management' : project_websites = [ 'trello.com', 'atlassian.net' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'code_sharing' : project_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] for site in project_websites: results = get_and_save_dork_results( lookup_target=site, results_dir=results_dir, type=dork, lookup_keywords=host, scan_history=scan_history ) elif dork == 'config_files' : config_file_exts = [ 'env', 'xml', 'conf', 'toml', 'yml', 'yaml', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(config_file_exts), page_count=4, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keyword = 'Jenkins' results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=lookup_keyword, page_count=1, scan_history=scan_history ) elif dork == 'wordpress_files' : lookup_keywords = [ '/wp-content/', '/wp-includes/' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'php_error' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'jenkins' : lookup_keywords = [ 'PHP Parse error', 'PHP Warning', 'PHP Error' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_keywords=','.join(lookup_keywords), page_count=5, scan_history=scan_history ) elif dork == 'exposed_documents' : docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(docs_file_ext), page_count=7, scan_history=scan_history ) elif dork == 'db_files' : file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) elif dork == 'git_exposed' : file_ext = [ 'git', ] results = get_and_save_dork_results( lookup_target=host, results_dir=results_dir, type=dork, lookup_extensions=','.join(file_ext), page_count=1, scan_history=scan_history ) except Exception as e: logger.exception(e) return results @app.task(name='theHarvester', queue='theHarvester_queue', bind=False) def theHarvester(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run theHarvester to get save emails, hosts, employees found in domain. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: dict: Dict of emails, employees, hosts and ips found during crawling. """ scan_history = ScanHistory.objects.get(pk=scan_history_id) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) output_path_json = f'{results_dir}/theHarvester.json' theHarvester_dir = '/usr/src/github/theHarvester' history_file = f'{results_dir}/commands.txt' cmd = f'python3 {theHarvester_dir}/theHarvester.py -d {host} -b all -f {output_path_json}' # Update proxies.yaml proxy_query = Proxy.objects.all() if proxy_query.exists(): proxy = proxy_query.first() if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(f'{theHarvester_dir}/proxies.yaml', 'w') as file: yaml.dump(yaml_data, file) # Run cmd run_command( cmd, shell=False, cwd=theHarvester_dir, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) # Get file location if not os.path.isfile(output_path_json): logger.error(f'Could not open {output_path_json}') return {} # Load theHarvester results with open(output_path_json, 'r') as f: data = json.load(f) # Re-indent theHarvester JSON with open(output_path_json, 'w') as f: json.dump(data, f, indent=4) emails = data.get('emails', []) for email_address in emails: email, _ = save_email(email_address, scan_history=scan_history) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) linkedin_people = data.get('linkedin_people', []) for people in linkedin_people: employee, _ = save_employee( people, designation='linkedin', scan_history=scan_history) # if employee: # self.notify(fields={'LinkedIn people': f'• {employee.name}'}) twitter_people = data.get('twitter_people', []) for people in twitter_people: employee, _ = save_employee( people, designation='twitter', scan_history=scan_history) # if employee: # self.notify(fields={'Twitter people': f'• {employee.name}'}) hosts = data.get('hosts', []) urls = [] for host in hosts: split = tuple(host.split(':')) http_url = split[0] subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) endpoint, _ = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain) # if endpoint: # urls.append(endpoint.http_url) # self.notify(fields={'Hosts': f'• {endpoint.http_url}'}) # if enable_http_crawl: # ctx['track'] = False # http_crawl(urls, ctx=ctx) # TODO: Lots of ips unrelated with our domain are found, disabling # this for now. # ips = data.get('ips', []) # for ip_address in ips: # ip, created = save_ip_address( # ip_address, # subscan=subscan) # if ip: # send_task_notif.delay( # 'osint', # scan_history_id=scan_history_id, # subscan_id=subscan_id, # severity='success', # update_fields={'IPs': f'{ip.address}'}) return data @app.task(name='h8mail', queue='h8mail_queue', bind=False) def h8mail(config, host, scan_history_id, activity_id, results_dir, ctx={}): """Run h8mail. Args: config (dict): yaml_configuration host (str): target name scan_history_id (startScan.ScanHistory): Scan History ID activity_id: ScanActivity ID results_dir (str): Path to store scan results ctx (dict): context of scan Returns: list[dict]: List of credentials info. """ logger.warning('Getting leaked credentials') scan_history = ScanHistory.objects.get(pk=scan_history_id) input_path = f'{results_dir}/emails.txt' output_file = f'{results_dir}/h8mail.json' cmd = f'h8mail -t {input_path} --json {output_file}' history_file = f'{results_dir}/commands.txt' run_command( cmd, history_file=history_file, scan_id=scan_history_id, activity_id=activity_id) with open(output_file) as f: data = json.load(f) creds = data.get('targets', []) # TODO: go through h8mail output and save emails to DB for cred in creds: logger.warning(cred) email_address = cred['target'] pwn_num = cred['pwn_num'] pwn_data = cred.get('data', []) email, created = save_email(email_address, scan_history=scan) # if email: # self.notify(fields={'Emails': f'• `{email.address}`'}) return creds @app.task(name='screenshot', queue='main_scan_queue', base=RengineTask, bind=True) def screenshot(self, ctx={}, description=None): """Uses EyeWitness to gather screenshot of a domain and/or url. Args: description (str, optional): Task description shown in UI. """ # Config screenshots_path = f'{self.results_dir}/screenshots' output_path = f'{self.results_dir}/screenshots/{self.filename}' alive_endpoints_file = f'{self.results_dir}/endpoints_alive.txt' config = self.yaml_configuration.get(SCREENSHOT) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT + 5) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) # If intensity is normal, grab only the root endpoints of each subdomain strict = True if intensity == 'normal' else False # Get URLs to take screenshot of get_http_urls( is_alive=enable_http_crawl, strict=strict, write_filepath=alive_endpoints_file, get_only_default_urls=True, ctx=ctx ) # Send start notif notification = Notification.objects.first() send_output_file = notification.send_scan_output_file if notification else False # Run cmd cmd = f'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py -f {alive_endpoints_file} -d {screenshots_path} --no-prompt' cmd += f' --timeout {timeout}' if timeout > 0 else '' cmd += f' --threads {threads}' if threads > 0 else '' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(output_path): logger.error(f'Could not load EyeWitness results at {output_path} for {self.domain.name}.') return # Loop through results and save objects in DB screenshot_paths = [] with open(output_path, 'r') as file: reader = csv.reader(file) for row in reader: "Protocol,Port,Domain,Request Status,Screenshot Path, Source Path" protocol, port, subdomain_name, status, screenshot_path, source_path = tuple(row) logger.info(f'{protocol}:{port}:{subdomain_name}:{status}') subdomain_query = Subdomain.objects.filter(name=subdomain_name) if self.scan: subdomain_query = subdomain_query.filter(scan_history=self.scan) if status == 'Successful' and subdomain_query.exists(): subdomain = subdomain_query.first() screenshot_paths.append(screenshot_path) subdomain.screenshot_path = screenshot_path.replace('/usr/src/scan_results/', '') subdomain.save() logger.warning(f'Added screenshot for {subdomain.name} to DB') # Remove all db, html extra files in screenshot results run_command( 'rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format(screenshots_path), shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'rm -rf {screenshots_path}/source', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Send finish notifs screenshots_str = '• ' + '\n• '.join([f'`{path}`' for path in screenshot_paths]) self.notify(fields={'Screenshots': screenshots_str}) if send_output_file: for path in screenshot_paths: title = get_output_file_name( self.scan_id, self.subscan_id, self.filename) send_file_to_discord.delay(path, title) @app.task(name='port_scan', queue='main_scan_queue', base=RengineTask, bind=True) def port_scan(self, hosts=[], ctx={}, description=None): """Run port scan. Args: hosts (list, optional): Hosts to run port scan on. description (str, optional): Task description shown in UI. Returns: list: List of open ports (dict). """ input_file = f'{self.results_dir}/input_subdomains_port_scan.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(PORT_SCAN) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) exclude_ports = config.get(NAABU_EXCLUDE_PORTS, []) exclude_subdomains = config.get(NAABU_EXCLUDE_SUBDOMAINS, False) ports = config.get(PORTS, NAABU_DEFAULT_PORTS) ports = [str(port) for port in ports] rate_limit = config.get(NAABU_RATE) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) passive = config.get(NAABU_PASSIVE, False) use_naabu_config = config.get(USE_NAABU_CONFIG, False) exclude_ports_str = ','.join(return_iterable(exclude_ports)) # nmap args nmap_enabled = config.get(ENABLE_NMAP, False) nmap_cmd = config.get(NMAP_COMMAND, '') nmap_script = config.get(NMAP_SCRIPT, '') nmap_script = ','.join(return_iterable(nmap_script)) nmap_script_args = config.get(NMAP_SCRIPT_ARGS) if hosts: with open(input_file, 'w') as f: f.write('\n'.join(hosts)) else: hosts = get_subdomains( write_filepath=input_file, exclude_subdomains=exclude_subdomains, ctx=ctx) # Build cmd cmd = 'naabu -json -exclude-cdn' cmd += f' -list {input_file}' if len(hosts) > 0 else f' -host {hosts[0]}' if 'full' in ports or 'all' in ports: ports_str = ' -p "-"' elif 'top-100' in ports: ports_str = ' -top-ports 100' elif 'top-1000' in ports: ports_str = ' -top-ports 1000' else: ports_str = ','.join(ports) ports_str = f' -p {ports_str}' cmd += ports_str cmd += ' -config /root/.config/naabu/config.yaml' if use_naabu_config else '' cmd += f' -proxy "{proxy}"' if proxy else '' cmd += f' -c {threads}' if threads else '' cmd += f' -rate {rate_limit}' if rate_limit > 0 else '' cmd += f' -timeout {timeout*1000}' if timeout > 0 else '' cmd += f' -passive' if passive else '' cmd += f' -exclude-ports {exclude_ports_str}' if exclude_ports else '' cmd += f' -silent' # Execute cmd and gather results results = [] urls = [] ports_data = {} for line in stream_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) port_number = line['port'] ip_address = line['ip'] host = line.get('host') or ip_address if port_number == 0: continue # Grab subdomain subdomain = Subdomain.objects.filter( name=host, target_domain=self.domain, scan_history=self.scan ).first() # Add IP DB ip, _ = save_ip_address(ip_address, subdomain, subscan=self.subscan) if self.subscan: ip.ip_subscan_ids.add(self.subscan) ip.save() # Add endpoint to DB # port 80 and 443 not needed as http crawl already does that. if port_number not in [80, 443]: http_url = f'{host}:{port_number}' endpoint, _ = save_endpoint( http_url, crawl=enable_http_crawl, ctx=ctx, subdomain=subdomain) if endpoint: http_url = endpoint.http_url urls.append(http_url) # Add Port in DB port_details = whatportis.get_ports(str(port_number)) service_name = port_details[0].name if len(port_details) > 0 else 'unknown' description = port_details[0].description if len(port_details) > 0 else '' # get or create port port, created = Port.objects.get_or_create( number=port_number, service_name=service_name, description=description ) if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port.save() ip.ports.add(port) ip.save() if host in ports_data: ports_data[host].append(port_number) else: ports_data[host] = [port_number] # Send notification logger.warning(f'Found opened port {port_number} on {ip_address} ({host})') if len(ports_data) == 0: logger.info('Finished running naabu port scan - No open ports found.') if nmap_enabled: logger.info('Nmap scans skipped') return ports_data # Send notification fields_str = '' for host, ports in ports_data.items(): ports_str = ', '.join([f'`{port}`' for port in ports]) fields_str += f'• `{host}`: {ports_str}\n' self.notify(fields={'Ports discovered': fields_str}) # Save output to file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) logger.info('Finished running naabu port scan.') # Process nmap results: 1 process per host sigs = [] if nmap_enabled: logger.warning(f'Starting nmap scans ...') logger.warning(ports_data) for host, port_list in ports_data.items(): ports_str = '_'.join([str(p) for p in port_list]) ctx_nmap = ctx.copy() ctx_nmap['description'] = get_task_title(f'nmap_{host}', self.scan_id, self.subscan_id) ctx_nmap['track'] = False sig = nmap.si( cmd=nmap_cmd, ports=port_list, host=host, script=nmap_script, script_args=nmap_script_args, max_rate=rate_limit, ctx=ctx_nmap) sigs.append(sig) task = group(sigs).apply_async() with allow_join_result(): results = task.get() return ports_data @app.task(name='nmap', queue='main_scan_queue', base=RengineTask, bind=True) def nmap( self, cmd=None, ports=[], host=None, input_file=None, script=None, script_args=None, max_rate=None, ctx={}, description=None): """Run nmap on a host. Args: cmd (str, optional): Existing nmap command to complete. ports (list, optional): List of ports to scan. host (str, optional): Host to scan. input_file (str, optional): Input hosts file. script (str, optional): NSE script to run. script_args (str, optional): NSE script args. max_rate (int): Max rate. description (str, optional): Task description shown in UI. """ notif = Notification.objects.first() ports_str = ','.join(str(port) for port in ports) self.filename = self.filename.replace('.txt', '.xml') filename_vulns = self.filename.replace('.xml', '_vulns.json') output_file = self.output_path output_file_xml = f'{self.results_dir}/{host}_{self.filename}' vulns_file = f'{self.results_dir}/{host}_{filename_vulns}' logger.warning(f'Running nmap on {host}:{ports}') # Build cmd nmap_cmd = get_nmap_cmd( cmd=cmd, ports=ports_str, script=script, script_args=script_args, max_rate=max_rate, host=host, input_file=input_file, output_file=output_file_xml) # Run cmd run_command( nmap_cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Get nmap XML results and convert to JSON vulns = parse_nmap_results(output_file_xml, output_file) with open(vulns_file, 'w') as f: json.dump(vulns, f, indent=4) # Save vulnerabilities found by nmap vulns_str = '' for vuln_data in vulns: # URL is not necessarily an HTTP URL when running nmap (can be any # other vulnerable protocols). Look for existing endpoint and use its # URL as vulnerability.http_url if it exists. url = vuln_data['http_url'] endpoint = EndPoint.objects.filter(http_url__contains=url).first() if endpoint: vuln_data['http_url'] = endpoint.http_url vuln, created = save_vulnerability( target_domain=self.domain, subdomain=self.subdomain, scan_history=self.scan, subscan=self.subscan, endpoint=endpoint, **vuln_data) vulns_str += f'• {str(vuln)}\n' if created: logger.warning(str(vuln)) # Send only 1 notif for all vulns to reduce number of notifs if notif and notif.send_vuln_notif and vulns_str: logger.warning(vulns_str) self.notify(fields={'CVEs': vulns_str}) return vulns @app.task(name='waf_detection', queue='main_scan_queue', base=RengineTask, bind=True) def waf_detection(self, ctx={}, description=None): """ Uses wafw00f to check for the presence of a WAF. Args: description (str, optional): Task description shown in UI. Returns: list: List of startScan.models.Waf objects. """ input_path = f'{self.results_dir}/input_endpoints_waf_detection.txt' config = self.yaml_configuration.get(WAF_DETECTION) or {} enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) # Get alive endpoints from DB get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) cmd = f'wafw00f -i {input_path} -o {self.output_path}' run_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) if not os.path.isfile(self.output_path): logger.error(f'Could not find {self.output_path}') return with open(self.output_path) as file: wafs = file.readlines() for line in wafs: line = " ".join(line.split()) splitted = line.split(' ', 1) waf_info = splitted[1].strip() waf_name = waf_info[:waf_info.find('(')].strip() waf_manufacturer = waf_info[waf_info.find('(')+1:waf_info.find(')')].strip().replace('.', '') http_url = sanitize_url(splitted[0].strip()) if not waf_name or waf_name == 'None': continue # Add waf to db waf, _ = Waf.objects.get_or_create( name=waf_name, manufacturer=waf_manufacturer ) # Add waf info to Subdomain in DB subdomain = get_subdomain_from_url(http_url) logger.info(f'Wafw00f Subdomain : {subdomain}') subdomain_query, _ = Subdomain.objects.get_or_create(scan_history=self.scan, name=subdomain) subdomain_query.waf.add(waf) subdomain_query.save() return wafs @app.task(name='dir_file_fuzz', queue='main_scan_queue', base=RengineTask, bind=True) def dir_file_fuzz(self, ctx={}, description=None): """Perform directory scan, and currently uses `ffuf` as a default tool. Args: description (str, optional): Task description shown in UI. Returns: list: List of URLs discovered. """ # Config cmd = 'ffuf' config = self.yaml_configuration.get(DIR_FILE_FUZZ) or {} custom_header = self.yaml_configuration.get(CUSTOM_HEADER) auto_calibration = config.get(AUTO_CALIBRATION, True) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) extensions = config.get(EXTENSIONS, DEFAULT_DIR_FILE_FUZZ_EXTENSIONS) # prepend . on extensions extensions = [ext if ext.startswith('.') else '.' + ext for ext in extensions] extensions_str = ','.join(map(str, extensions)) follow_redirect = config.get(FOLLOW_REDIRECT, FFUF_DEFAULT_FOLLOW_REDIRECT) max_time = config.get(MAX_TIME, 0) match_http_status = config.get(MATCH_HTTP_STATUS, FFUF_DEFAULT_MATCH_HTTP_STATUS) mc = ','.join([str(c) for c in match_http_status]) recursive_level = config.get(RECURSIVE_LEVEL, FFUF_DEFAULT_RECURSIVE_LEVEL) stop_on_error = config.get(STOP_ON_ERROR, False) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) wordlist_name = config.get(WORDLIST, 'dicc') delay = rate_limit / (threads * 100) # calculate request pause delay from rate_limit and number of threads input_path = f'{self.results_dir}/input_dir_file_fuzz.txt' # Get wordlist wordlist_name = 'dicc' if wordlist_name == 'default' else wordlist_name wordlist_path = f'/usr/src/wordlist/{wordlist_name}.txt' # Build command cmd += f' -w {wordlist_path}' cmd += f' -e {extensions_str}' if extensions else '' cmd += f' -maxtime {max_time}' if max_time > 0 else '' cmd += f' -p {delay}' if delay > 0 else '' cmd += f' -recursion -recursion-depth {recursive_level} ' if recursive_level > 0 else '' cmd += f' -t {threads}' if threads and threads > 0 else '' cmd += f' -timeout {timeout}' if timeout and timeout > 0 else '' cmd += ' -se' if stop_on_error else '' cmd += ' -fr' if follow_redirect else '' cmd += ' -ac' if auto_calibration else '' cmd += f' -mc {mc}' if mc else '' cmd += f' -H "{custom_header}"' if custom_header else '' # Grab URLs to fuzz urls = get_http_urls( is_alive=True, ignore_files=False, write_filepath=input_path, get_only_default_urls=True, ctx=ctx ) logger.warning(urls) # Loop through URLs and run command results = [] for url in urls: ''' Above while fetching urls, we are not ignoring files, because some default urls may redirect to https://example.com/login.php so, ignore_files is set to False but, during fuzzing, we will only need part of the path, in above example it is still a good idea to ffuf base url https://example.com so files from base url ''' url_parse = urlparse(url) url = url_parse.scheme + '://' + url_parse.netloc url += '/FUZZ' # TODO: fuzz not only URL but also POST / PUT / headers proxy = get_random_proxy() # Build final cmd fcmd = cmd fcmd += f' -x {proxy}' if proxy else '' fcmd += f' -u {url} -json' # Initialize DirectoryScan object dirscan = DirectoryScan() dirscan.scanned_date = timezone.now() dirscan.command_line = fcmd dirscan.save() # Loop through results and populate EndPoint and DirectoryFile in DB results = [] for line in stream_command( fcmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) name = line['input'].get('FUZZ') length = line['length'] status = line['status'] words = line['words'] url = line['url'] lines = line['lines'] content_type = line['content-type'] duration = line['duration'] if not name: logger.error(f'FUZZ not found for "{url}"') continue endpoint, created = save_endpoint(url, crawl=False, ctx=ctx) # endpoint.is_default = False endpoint.http_status = status endpoint.content_length = length endpoint.response_time = duration / 1000000000 endpoint.save() if created: urls.append(endpoint.http_url) endpoint.status = status endpoint.content_type = content_type endpoint.content_length = length dfile, created = DirectoryFile.objects.get_or_create( name=name, length=length, words=words, lines=lines, content_type=content_type, url=url) dfile.http_status = status dfile.save() # if created: # logger.warning(f'Found new directory or file {url}') dirscan.directory_files.add(dfile) dirscan.save() if self.subscan: dirscan.dir_subscan_ids.add(self.subscan) subdomain_name = get_subdomain_from_url(endpoint.http_url) subdomain = Subdomain.objects.get(name=subdomain_name, scan_history=self.scan) subdomain.directories.add(dirscan) subdomain.save() # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl(urls, ctx=ctx) return results @app.task(name='fetch_url', queue='main_scan_queue', base=RengineTask, bind=True) def fetch_url(self, urls=[], ctx={}, description=None): """Fetch URLs using different tools like gauplus, gau, gospider, waybackurls ... Args: urls (list): List of URLs to start from. description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/input_endpoints_fetch_url.txt' proxy = get_random_proxy() # Config config = self.yaml_configuration.get(FETCH_URL) or {} should_remove_duplicate_endpoints = config.get(REMOVE_DUPLICATE_ENDPOINTS, True) duplicate_removal_fields = config.get(DUPLICATE_REMOVAL_FIELDS, ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS) enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) gf_patterns = config.get(GF_PATTERNS, DEFAULT_GF_PATTERNS) ignore_file_extension = config.get(IGNORE_FILE_EXTENSION, DEFAULT_IGNORE_FILE_EXTENSIONS) tools = config.get(USES_TOOLS, ENDPOINT_SCAN_DEFAULT_TOOLS) threads = config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) domain_request_headers = self.domain.request_headers if self.domain else None custom_header = domain_request_headers or self.yaml_configuration.get(CUSTOM_HEADER) exclude_subdomains = config.get(EXCLUDED_SUBDOMAINS, False) # Get URLs to scan and save to input file if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_alive=enable_http_crawl, write_filepath=input_path, exclude_subdomains=exclude_subdomains, get_only_default_urls=True, ctx=ctx ) # Domain regex host = self.domain.name if self.domain else urlparse(urls[0]).netloc host_regex = f"\'https?://([a-z0-9]+[.])*{host}.*\'" # Tools cmds cmd_map = { 'gau': f'gau', 'gauplus': f'gauplus -random-agent', 'hakrawler': 'hakrawler -subs -u', 'waybackurls': 'waybackurls', 'gospider': f'gospider -S {input_path} --js -d 2 --sitemap --robots -w -r', 'katana': f'katana -list {input_path} -silent -jc -kf all -d 3 -fs rdn', } if proxy: cmd_map['gau'] += f' --proxy "{proxy}"' cmd_map['gauplus'] += f' -p "{proxy}"' cmd_map['gospider'] += f' -p {proxy}' cmd_map['hakrawler'] += f' -proxy {proxy}' cmd_map['katana'] += f' -proxy {proxy}' if threads > 0: cmd_map['gau'] += f' --threads {threads}' cmd_map['gauplus'] += f' -t {threads}' cmd_map['gospider'] += f' -t {threads}' cmd_map['katana'] += f' -c {threads}' if custom_header: header_string = ';;'.join([ f'{key}: {value}' for key, value in custom_header.items() ]) cmd_map['hakrawler'] += f' -h {header_string}' cmd_map['katana'] += f' -H {header_string}' header_flags = [':'.join(h) for h in header_string.split(';;')] for flag in header_flags: cmd_map['gospider'] += f' -H {flag}' cat_input = f'cat {input_path}' grep_output = f'grep -Eo {host_regex}' cmd_map = { tool: f'{cat_input} | {cmd} | {grep_output} > {self.results_dir}/urls_{tool}.txt' for tool, cmd in cmd_map.items() } tasks = group( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for tool, cmd in cmd_map.items() if tool in tools ) # Cleanup task sort_output = [ f'cat {self.results_dir}/urls_* > {self.output_path}', f'cat {input_path} >> {self.output_path}', f'sort -u {self.output_path} -o {self.output_path}', ] if ignore_file_extension: ignore_exts = '|'.join(ignore_file_extension) grep_ext_filtered_output = [ f'cat {self.output_path} | grep -Eiv "\\.({ignore_exts}).*" > {self.results_dir}/urls_filtered.txt', f'mv {self.results_dir}/urls_filtered.txt {self.output_path}' ] sort_output.extend(grep_ext_filtered_output) cleanup = chain( run_command.si( cmd, shell=True, scan_id=self.scan_id, activity_id=self.activity_id) for cmd in sort_output ) # Run all commands task = chord(tasks)(cleanup) with allow_join_result(): task.get() # Store all the endpoints and run httpx with open(self.output_path) as f: discovered_urls = f.readlines() self.notify(fields={'Discovered URLs': len(discovered_urls)}) # Some tools can have an URL in the format <URL>] - <PATH> or <URL> - <PATH>, add them # to the final URL list all_urls = [] for url in discovered_urls: url = url.strip() urlpath = None base_url = None if '] ' in url: # found JS scraped endpoint e.g from gospider split = tuple(url.split('] ')) if not len(split) == 2: logger.warning(f'URL format not recognized for "{url}". Skipping.') continue base_url, urlpath = split urlpath = urlpath.lstrip('- ') elif ' - ' in url: # found JS scraped endpoint e.g from gospider base_url, urlpath = tuple(url.split(' - ')) if base_url and urlpath: subdomain = urlparse(base_url) url = f'{subdomain.scheme}://{subdomain.netloc}{self.url_filter}' if not validators.url(url): logger.warning(f'Invalid URL "{url}". Skipping.') if url not in all_urls: all_urls.append(url) # Filter out URLs if a path filter was passed if self.url_filter: all_urls = [url for url in all_urls if self.url_filter in url] # Write result to output path with open(self.output_path, 'w') as f: f.write('\n'.join(all_urls)) logger.warning(f'Found {len(all_urls)} usable URLs') # Crawl discovered URLs if enable_http_crawl: ctx['track'] = False http_crawl( all_urls, ctx=ctx, should_remove_duplicate_endpoints=should_remove_duplicate_endpoints, duplicate_removal_fields=duplicate_removal_fields ) #-------------------# # GF PATTERNS MATCH # #-------------------# # Combine old gf patterns with new ones if gf_patterns: self.scan.used_gf_patterns = ','.join(gf_patterns) self.scan.save() # Run gf patterns on saved endpoints # TODO: refactor to Celery task for gf_pattern in gf_patterns: # TODO: js var is causing issues, removing for now if gf_pattern == 'jsvar': logger.info('Ignoring jsvar as it is causing issues.') continue # Run gf on current pattern logger.warning(f'Running gf on pattern "{gf_pattern}"') gf_output_file = f'{self.results_dir}/gf_patterns_{gf_pattern}.txt' cmd = f'cat {self.output_path} | gf {gf_pattern} | grep -Eo {host_regex} >> {gf_output_file}' run_command( cmd, shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) # Check output file if not os.path.exists(gf_output_file): logger.error(f'Could not find GF output file {gf_output_file}. Skipping GF pattern "{gf_pattern}"') continue # Read output file line by line and with open(gf_output_file, 'r') as f: lines = f.readlines() # Add endpoints / subdomains to DB for url in lines: http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue endpoint, created = save_endpoint( http_url, crawl=False, subdomain=subdomain, ctx=ctx) if not endpoint: continue earlier_pattern = None if not created: earlier_pattern = endpoint.matched_gf_patterns pattern = f'{earlier_pattern},{gf_pattern}' if earlier_pattern else gf_pattern endpoint.matched_gf_patterns = pattern endpoint.save() return all_urls def parse_curl_output(response): # TODO: Enrich from other cURL fields. CURL_REGEX_HTTP_STATUS = f'HTTP\/(?:(?:\d\.?)+)\s(\d+)\s(?:\w+)' http_status = 0 if response: failed = False regex = re.compile(CURL_REGEX_HTTP_STATUS, re.MULTILINE) try: http_status = int(regex.findall(response)[0]) except (KeyError, TypeError, IndexError): pass return { 'http_status': http_status, } @app.task(name='vulnerability_scan', queue='main_scan_queue', bind=True, base=RengineTask) def vulnerability_scan(self, urls=[], ctx={}, description=None): """ This function will serve as an entrypoint to vulnerability scan. All other vulnerability scan will be run from here including nuclei, crlfuzz, etc """ logger.info('Running Vulnerability Scan Queue') config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_run_nuclei = config.get(RUN_NUCLEI, True) should_run_crlfuzz = config.get(RUN_CRLFUZZ, False) should_run_dalfox = config.get(RUN_DALFOX, False) should_run_s3scanner = config.get(RUN_S3SCANNER, True) grouped_tasks = [] if should_run_nuclei: _task = nuclei_scan.si( urls=urls, ctx=ctx, description=f'Nuclei Scan' ) grouped_tasks.append(_task) if should_run_crlfuzz: _task = crlfuzz_scan.si( urls=urls, ctx=ctx, description=f'CRLFuzz Scan' ) grouped_tasks.append(_task) if should_run_dalfox: _task = dalfox_xss_scan.si( urls=urls, ctx=ctx, description=f'Dalfox XSS Scan' ) grouped_tasks.append(_task) if should_run_s3scanner: _task = s3scanner.si( ctx=ctx, description=f'Misconfigured S3 Buckets Scanner' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan completed...') # return results return None @app.task(name='nuclei_individual_severity_module', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_individual_severity_module(self, cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx={}, description=None): ''' This celery task will run vulnerability scan in parallel. All severities supplied should run in parallel as grouped tasks. ''' results = [] logger.info(f'Running vulnerability scan with severity: {severity}') cmd += f' -severity {severity}' # Send start notification notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue results.append(line) # Gather nuclei results vuln_data = parse_nuclei_result(line) # Get corresponding subdomain http_url = sanitize_url(line.get('matched-at')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) # Look for duplicate vulnerabilities by excluding records that might change but are irrelevant. object_comparison_exclude = ['response', 'curl_command', 'tags', 'references', 'cve_ids', 'cwe_ids'] # Add subdomain and target domain to the duplicate check vuln_data_copy = vuln_data.copy() vuln_data_copy['subdomain'] = subdomain vuln_data_copy['target_domain'] = self.domain # Check if record exists, if exists do not save it if record_exists(Vulnerability, data=vuln_data_copy, exclude_keys=object_comparison_exclude): logger.warning(f'Nuclei vulnerability of severity {severity} : {vuln_data_copy["name"]} for {subdomain_name} already exists') continue # Get or create EndPoint object response = line.get('response') httpx_crawl = False if response else enable_http_crawl # avoid yet another httpx crawl endpoint, _ = save_endpoint( http_url, crawl=httpx_crawl, subdomain=subdomain, ctx=ctx) if endpoint: http_url = endpoint.http_url if not httpx_crawl: output = parse_curl_output(response) endpoint.http_status = output['http_status'] endpoint.save() # Get or create Vulnerability object vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, subdomain=subdomain, **vuln_data) if not vuln: continue # Print vuln severity = line['info'].get('severity', 'unknown') logger.warning(str(vuln)) # Send notification for all vulnerabilities except info url = vuln.http_url or vuln.subdomain send_vuln = ( notif and notif.send_vuln_notif and vuln and severity in ['low', 'medium', 'high', 'critical']) if send_vuln: fields = { 'Severity': f'**{severity.upper()}**', 'URL': http_url, 'Subdomain': subdomain_name, 'Name': vuln.name, 'Type': vuln.type, 'Description': vuln.description, 'Template': vuln.template_url, 'Tags': vuln.get_tags_str(), 'CVEs': vuln.get_cve_str(), 'CWEs': vuln.get_cwe_str(), 'References': vuln.get_refs_str() } severity_map = { 'low': 'info', 'medium': 'warning', 'high': 'error', 'critical': 'error' } self.notify( f'vulnerability_scan_#{vuln.id}', severity_map[severity], fields, add_meta_info=False) # Send report to hackerone hackerone_query = Hackerone.objects.all() send_report = ( hackerone_query.exists() and severity not in ('info', 'low') and vuln.target_domain.h1_team_handle ) if send_report: hackerone = hackerone_query.first() if hackerone.send_critical and severity == 'critical': send_hackerone_report.delay(vuln.id) elif hackerone.send_high and severity == 'high': send_hackerone_report.delay(vuln.id) elif hackerone.send_medium and severity == 'medium': send_hackerone_report.delay(vuln.id) # Write results to JSON file with open(self.output_path, 'w') as f: json.dump(results, f, indent=4) # Send finish notif if send_status: vulns = Vulnerability.objects.filter(scan_history__id=self.scan_id) info_count = vulns.filter(severity=0).count() low_count = vulns.filter(severity=1).count() medium_count = vulns.filter(severity=2).count() high_count = vulns.filter(severity=3).count() critical_count = vulns.filter(severity=4).count() unknown_count = vulns.filter(severity=-1).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count + unknown_count fields = { 'Total': vulnerability_count, 'Critical': critical_count, 'High': high_count, 'Medium': medium_count, 'Low': low_count, 'Info': info_count, 'Unknown': unknown_count } self.notify(fields=fields) # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=NUCLEI ).exclude( severity=0 ) # find all unique vulnerabilities based on path and title # all unique vulnerability will go thru gpt function and get report # once report is got, it will be matched with other vulnerabilities and saved unique_vulns = set() for vuln in vulns: unique_vulns.add((vuln.name, vuln.get_path())) unique_vulns = list(unique_vulns) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in unique_vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return None def get_vulnerability_gpt_report(vuln): title = vuln[0] path = vuln[1] logger.info(f'Getting GPT Report for {title}, PATH: {path}') # check if in db already exists stored = GPTVulnerabilityReport.objects.filter( url_path=path ).filter( title=title ).first() if stored: response = { 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: report = GPTVulnerabilityReportGenerator() vulnerability_description = get_gpt_vuln_input_description( title, path ) response = report.get_vulnerability_description(vulnerability_description) add_gpt_description_db( title, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) for vuln in Vulnerability.objects.filter(name=title, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() def add_gpt_description_db(title, path, description, impact, remediation, references): gpt_report = GPTVulnerabilityReport() gpt_report.url_path = path gpt_report.title = title gpt_report.description = description gpt_report.impact = impact gpt_report.remediation = remediation gpt_report.save() for url in references: ref, created = VulnerabilityReference.objects.get_or_create(url=url) gpt_report.references.add(ref) gpt_report.save() @app.task(name='nuclei_scan', queue='main_scan_queue', base=RengineTask, bind=True) def nuclei_scan(self, urls=[], ctx={}, description=None): """HTTP vulnerability scan using Nuclei Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. Notes: Unfurl the urls to keep only domain and path, will be sent to vuln scan and ignore certain file extensions. Thanks: https://github.com/six2dez/reconftw """ # Config config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} input_path = f'{self.results_dir}/input_endpoints_vulnerability_scan.txt' enable_http_crawl = config.get(ENABLE_HTTP_CRAWL, DEFAULT_ENABLE_HTTP_CRAWL) concurrency = config.get(NUCLEI_CONCURRENCY) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) intensity = config.get(INTENSITY) or self.yaml_configuration.get(INTENSITY, DEFAULT_SCAN_INTENSITY) rate_limit = config.get(RATE_LIMIT) or self.yaml_configuration.get(RATE_LIMIT, DEFAULT_RATE_LIMIT) retries = config.get(RETRIES) or self.yaml_configuration.get(RETRIES, DEFAULT_RETRIES) timeout = config.get(TIMEOUT) or self.yaml_configuration.get(TIMEOUT, DEFAULT_HTTP_TIMEOUT) custom_header = config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) should_fetch_gpt_report = config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) proxy = get_random_proxy() nuclei_specific_config = config.get('nuclei', {}) use_nuclei_conf = nuclei_specific_config.get(USE_CONFIG, False) severities = nuclei_specific_config.get(NUCLEI_SEVERITY, NUCLEI_DEFAULT_SEVERITIES) tags = nuclei_specific_config.get(NUCLEI_TAGS, []) tags = ','.join(tags) nuclei_templates = nuclei_specific_config.get(NUCLEI_TEMPLATE) custom_nuclei_templates = nuclei_specific_config.get(NUCLEI_CUSTOM_TEMPLATE) # severities_str = ','.join(severities) # Get alive endpoints if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=enable_http_crawl, ignore_files=True, write_filepath=input_path, ctx=ctx ) if intensity == 'normal': # reduce number of endpoints to scan unfurl_filter = f'{self.results_dir}/urls_unfurled.txt' run_command( f"cat {input_path} | unfurl -u format %s://%d%p |uro > {unfurl_filter}", shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) run_command( f'sort -u {unfurl_filter} -o {unfurl_filter}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) input_path = unfurl_filter # Build templates # logger.info('Updating Nuclei templates ...') run_command( 'nuclei -update-templates', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) templates = [] if not (nuclei_templates or custom_nuclei_templates): templates.append(NUCLEI_DEFAULT_TEMPLATES_PATH) if nuclei_templates: if ALL in nuclei_templates: template = NUCLEI_DEFAULT_TEMPLATES_PATH templates.append(template) else: templates.extend(nuclei_templates) if custom_nuclei_templates: custom_nuclei_template_paths = [f'{str(elem)}.yaml' for elem in custom_nuclei_templates] template = templates.extend(custom_nuclei_template_paths) # Build CMD cmd = 'nuclei -j' cmd += ' -config /root/.config/nuclei/config.yaml' if use_nuclei_conf else '' cmd += f' -irr' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -l {input_path}' cmd += f' -c {str(concurrency)}' if concurrency > 0 else '' cmd += f' -proxy {proxy} ' if proxy else '' cmd += f' -retries {retries}' if retries > 0 else '' cmd += f' -rl {rate_limit}' if rate_limit > 0 else '' # cmd += f' -severity {severities_str}' cmd += f' -timeout {str(timeout)}' if timeout and timeout > 0 else '' cmd += f' -tags {tags}' if tags else '' cmd += f' -silent' for tpl in templates: cmd += f' -t {tpl}' grouped_tasks = [] custom_ctx = ctx for severity in severities: custom_ctx['track'] = True _task = nuclei_individual_severity_module.si( cmd, severity, enable_http_crawl, should_fetch_gpt_report, ctx=custom_ctx, description=f'Nuclei Scan with severity {severity}' ) grouped_tasks.append(_task) celery_group = group(grouped_tasks) job = celery_group.apply_async() while not job.ready(): # wait for all jobs to complete time.sleep(5) logger.info('Vulnerability scan with all severities completed...') return None @app.task(name='dalfox_xss_scan', queue='main_scan_queue', base=RengineTask, bind=True) def dalfox_xss_scan(self, urls=[], ctx={}, description=None): """XSS Scan using dalfox Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) dalfox_config = vuln_config.get(DALFOX) or {} custom_header = dalfox_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() is_waf_evasion = dalfox_config.get(WAF_EVASION, False) blind_xss_server = dalfox_config.get(BLIND_XSS_SERVER) user_agent = dalfox_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) timeout = dalfox_config.get(TIMEOUT) delay = dalfox_config.get(DELAY) threads = dalfox_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_dalfox_xss.txt' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=False, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'dalfox --silence --no-color --no-spinner' cmd += f' --only-poc r ' cmd += f' --ignore-return 302,404,403' cmd += f' --skip-bav' cmd += f' file {input_path}' cmd += f' --proxy {proxy}' if proxy else '' cmd += f' --waf-evasion' if is_waf_evasion else '' cmd += f' -b {blind_xss_server}' if blind_xss_server else '' cmd += f' --delay {delay}' if delay else '' cmd += f' --timeout {timeout}' if timeout else '' cmd += f' --user-agent {user_agent}' if user_agent else '' cmd += f' --header {custom_header}' if custom_header else '' cmd += f' --worker {threads}' if threads else '' cmd += f' --format json' results = [] for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id, trunc_char=',' ): if not isinstance(line, dict): continue results.append(line) vuln_data = parse_dalfox_result(line) http_url = sanitize_url(line.get('data')) subdomain_name = get_subdomain_from_url(http_url) # TODO: this should be get only subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting Dalfox Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=DALFOX ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='crlfuzz_scan', queue='main_scan_queue', base=RengineTask, bind=True) def crlfuzz_scan(self, urls=[], ctx={}, description=None): """CRLF Fuzzing with CRLFuzz Args: urls (list, optional): If passed, filter on those URLs. description (str, optional): Task description shown in UI. """ vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} should_fetch_gpt_report = vuln_config.get(FETCH_GPT_REPORT, DEFAULT_GET_GPT_REPORT) custom_header = vuln_config.get(CUSTOM_HEADER) or self.yaml_configuration.get(CUSTOM_HEADER) proxy = get_random_proxy() user_agent = vuln_config.get(USER_AGENT) or self.yaml_configuration.get(USER_AGENT) threads = vuln_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) input_path = f'{self.results_dir}/input_endpoints_crlf.txt' output_path = f'{self.results_dir}/{self.filename}' if urls: with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: get_http_urls( is_alive=False, ignore_files=True, write_filepath=input_path, ctx=ctx ) notif = Notification.objects.first() send_status = notif.send_scan_status_notif if notif else False # command builder cmd = 'crlfuzz -s' cmd += f' -l {input_path}' cmd += f' -x {proxy}' if proxy else '' cmd += f' --H {custom_header}' if custom_header else '' cmd += f' -o {output_path}' run_command( cmd, shell=False, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id ) if not os.path.isfile(output_path): logger.info('No Results from CRLFuzz') return crlfs = [] results = [] with open(output_path, 'r') as file: crlfs = file.readlines() for crlf in crlfs: url = crlf.strip() vuln_data = parse_crlfuzz_result(url) http_url = sanitize_url(url) subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = Subdomain.objects.get_or_create( name=subdomain_name, scan_history=self.scan, target_domain=self.domain ) endpoint, _ = save_endpoint( http_url, crawl=True, subdomain=subdomain, ctx=ctx ) if endpoint: http_url = endpoint.http_url endpoint.save() vuln, _ = save_vulnerability( target_domain=self.domain, http_url=http_url, scan_history=self.scan, subscan=self.subscan, **vuln_data ) if not vuln: continue # after vulnerability scan is done, we need to run gpt if # should_fetch_gpt_report and openapi key exists if should_fetch_gpt_report and OpenAiAPIKey.objects.all().first(): logger.info('Getting CRLFuzz Vulnerability GPT Report') vulns = Vulnerability.objects.filter( scan_history__id=self.scan_id ).filter( source=CRLFUZZ ).exclude( severity=0 ) _vulns = [] for vuln in vulns: _vulns.append((vuln.name, vuln.http_url)) with concurrent.futures.ThreadPoolExecutor(max_workers=DEFAULT_THREADS) as executor: future_to_gpt = {executor.submit(get_vulnerability_gpt_report, vuln): vuln for vuln in _vulns} # Wait for all tasks to complete for future in concurrent.futures.as_completed(future_to_gpt): gpt = future_to_gpt[future] try: future.result() except Exception as e: logger.error(f"Exception for Vulnerability {vuln}: {e}") return results @app.task(name='s3scanner', queue='main_scan_queue', base=RengineTask, bind=True) def s3scanner(self, ctx={}, description=None): """Bucket Scanner Args: ctx (dict): Context description (str, optional): Task description shown in UI. """ input_path = f'{self.results_dir}/#{self.scan_id}_subdomain_discovery.txt' vuln_config = self.yaml_configuration.get(VULNERABILITY_SCAN) or {} s3_config = vuln_config.get(S3SCANNER) or {} threads = s3_config.get(THREADS) or self.yaml_configuration.get(THREADS, DEFAULT_THREADS) providers = s3_config.get(PROVIDERS, S3SCANNER_DEFAULT_PROVIDERS) scan_history = ScanHistory.objects.filter(pk=self.scan_id).first() for provider in providers: cmd = f's3scanner -bucket-file {input_path} -enumerate -provider {provider} -threads {threads} -json' for line in stream_command( cmd, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not isinstance(line, dict): continue if line.get('bucket', {}).get('exists', 0) == 1: result = parse_s3scanner_result(line) s3bucket, created = S3Bucket.objects.get_or_create(**result) scan_history.buckets.add(s3bucket) logger.info(f"s3 bucket added {result['provider']}-{result['name']}-{result['region']}") @app.task(name='http_crawl', queue='main_scan_queue', base=RengineTask, bind=True) def http_crawl( self, urls=[], method=None, recrawl=False, ctx={}, track=True, description=None, is_ran_from_subdomain_scan=False, should_remove_duplicate_endpoints=True, duplicate_removal_fields=[]): """Use httpx to query HTTP URLs for important info like page titles, http status, etc... Args: urls (list, optional): A set of URLs to check. Overrides default behavior which queries all endpoints related to this scan. method (str): HTTP method to use (GET, HEAD, POST, PUT, DELETE). recrawl (bool, optional): If False, filter out URLs that have already been crawled. should_remove_duplicate_endpoints (bool): Whether to remove duplicate endpoints duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates Returns: list: httpx results. """ logger.info('Initiating HTTP Crawl') if is_ran_from_subdomain_scan: logger.info('Running From Subdomain Scan...') cmd = '/go/bin/httpx' cfg = self.yaml_configuration.get(HTTP_CRAWL) or {} custom_header = cfg.get(CUSTOM_HEADER, '') threads = cfg.get(THREADS, DEFAULT_THREADS) follow_redirect = cfg.get(FOLLOW_REDIRECT, True) self.output_path = None input_path = f'{self.results_dir}/httpx_input.txt' history_file = f'{self.results_dir}/commands.txt' if urls: # direct passing URLs to check if self.url_filter: urls = [u for u in urls if self.url_filter in u] with open(input_path, 'w') as f: f.write('\n'.join(urls)) else: urls = get_http_urls( is_uncrawled=not recrawl, write_filepath=input_path, ctx=ctx ) # logger.debug(urls) # If no URLs found, skip it if not urls: return # Re-adjust thread number if few URLs to avoid spinning up a monster to # kill a fly. if len(urls) < threads: threads = len(urls) # Get random proxy proxy = get_random_proxy() # Run command cmd += f' -cl -ct -rt -location -td -websocket -cname -asn -cdn -probe -random-agent' cmd += f' -t {threads}' if threads > 0 else '' cmd += f' --http-proxy {proxy}' if proxy else '' cmd += f' -H "{custom_header}"' if custom_header else '' cmd += f' -json' cmd += f' -u {urls[0]}' if len(urls) == 1 else f' -l {input_path}' cmd += f' -x {method}' if method else '' cmd += f' -silent' if follow_redirect: cmd += ' -fr' results = [] endpoint_ids = [] for line in stream_command( cmd, history_file=history_file, scan_id=self.scan_id, activity_id=self.activity_id): if not line or not isinstance(line, dict): continue logger.debug(line) # No response from endpoint if line.get('failed', False): continue # Parse httpx output host = line.get('host', '') content_length = line.get('content_length', 0) http_status = line.get('status_code') http_url, is_redirect = extract_httpx_url(line) page_title = line.get('title') webserver = line.get('webserver') cdn = line.get('cdn', False) rt = line.get('time') techs = line.get('tech', []) cname = line.get('cname', '') content_type = line.get('content_type', '') response_time = -1 if rt: response_time = float(''.join(ch for ch in rt if not ch.isalpha())) if rt[-2:] == 'ms': response_time = response_time / 1000 # Create Subdomain object in DB subdomain_name = get_subdomain_from_url(http_url) subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) if not subdomain: continue # Save default HTTP URL to endpoint object in DB endpoint, created = save_endpoint( http_url, crawl=False, ctx=ctx, subdomain=subdomain, is_default=is_ran_from_subdomain_scan ) if not endpoint: continue endpoint.http_status = http_status endpoint.page_title = page_title endpoint.content_length = content_length endpoint.webserver = webserver endpoint.response_time = response_time endpoint.content_type = content_type endpoint.save() endpoint_str = f'{http_url} [{http_status}] `{content_length}B` `{webserver}` `{rt}`' logger.warning(endpoint_str) if endpoint and endpoint.is_alive and endpoint.http_status != 403: self.notify( fields={'Alive endpoint': f'• {endpoint_str}'}, add_meta_info=False) # Add endpoint to results line['_cmd'] = cmd line['final_url'] = http_url line['endpoint_id'] = endpoint.id line['endpoint_created'] = created line['is_redirect'] = is_redirect results.append(line) # Add technology objects to DB for technology in techs: tech, _ = Technology.objects.get_or_create(name=technology) endpoint.techs.add(tech) if is_ran_from_subdomain_scan: subdomain.technologies.add(tech) subdomain.save() endpoint.save() techs_str = ', '.join([f'`{tech}`' for tech in techs]) self.notify( fields={'Technologies': techs_str}, add_meta_info=False) # Add IP objects for 'a' records to DB a_records = line.get('a', []) for ip_address in a_records: ip, created = save_ip_address( ip_address, subdomain, subscan=self.subscan, cdn=cdn) ips_str = '• ' + '\n• '.join([f'`{ip}`' for ip in a_records]) self.notify( fields={'IPs': ips_str}, add_meta_info=False) # Add IP object for host in DB if host: ip, created = save_ip_address( host, subdomain, subscan=self.subscan, cdn=cdn) self.notify( fields={'IPs': f'• `{ip.address}`'}, add_meta_info=False) # Save subdomain and endpoint if is_ran_from_subdomain_scan: # save subdomain stuffs subdomain.http_url = http_url subdomain.http_status = http_status subdomain.page_title = page_title subdomain.content_length = content_length subdomain.webserver = webserver subdomain.response_time = response_time subdomain.content_type = content_type subdomain.cname = ','.join(cname) subdomain.is_cdn = cdn if cdn: subdomain.cdn_name = line.get('cdn_name') subdomain.save() endpoint.save() endpoint_ids.append(endpoint.id) if should_remove_duplicate_endpoints: # Remove 'fake' alive endpoints that are just redirects to the same page remove_duplicate_endpoints( self.scan_id, self.domain_id, self.subdomain_id, filter_ids=endpoint_ids ) # Remove input file run_command( f'rm {input_path}', shell=True, history_file=self.history_file, scan_id=self.scan_id, activity_id=self.activity_id) return results #---------------------# # Notifications tasks # #---------------------# @app.task(name='send_notif', bind=False, queue='send_notif_queue') def send_notif( message, scan_history_id=None, subscan_id=None, **options): if not 'title' in options: message = enrich_notification(message, scan_history_id, subscan_id) send_discord_message(message, **options) send_slack_message(message) send_telegram_message(message) @app.task(name='send_scan_notif', bind=False, queue='send_scan_notif_queue') def send_scan_notif( scan_history_id, subscan_id=None, engine_id=None, status='RUNNING'): """Send scan status notification. Works for scan or a subscan if subscan_id is passed. Args: scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Get domain, engine, scan_history objects engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() tasks = ScanActivity.objects.filter(scan_of=scan) if scan else 0 # Build notif options url = get_scan_url(scan_history_id, subscan_id) title = get_scan_title(scan_history_id, subscan_id) fields = get_scan_fields(engine, scan, subscan, status, tasks) severity = None msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) if status: severity = STATUS_TO_SEVERITIES.get(status) opts = { 'title': title, 'url': url, 'fields': fields, 'severity': severity } logger.warning(f'Sending notification "{title}" [{severity}]') # Send notification send_notif( msg, scan_history_id, subscan_id, **opts) @app.task(name='send_task_notif', bind=False, queue='send_task_notif_queue') def send_task_notif( task_name, status=None, result=None, output_path=None, traceback=None, scan_history_id=None, engine_id=None, subscan_id=None, severity=None, add_meta_info=True, update_fields={}): """Send task status notification. Args: task_name (str): Task name. status (str, optional): Task status. result (str, optional): Task result. output_path (str, optional): Task output path. traceback (str, optional): Task traceback. scan_history_id (int, optional): ScanHistory id. subscan_id (int, optional): SuScan id. engine_id (int, optional): EngineType id. severity (str, optional): Severity (will be mapped to notif colors) add_meta_info (bool, optional): Wheter to add scan / subscan info to notif. update_fields (dict, optional): Fields key / value to update. """ # Skip send if notification settings are not configured notif = Notification.objects.first() if not (notif and notif.send_scan_status_notif): return # Build fields url = None fields = {} if add_meta_info: engine = EngineType.objects.filter(pk=engine_id).first() scan = ScanHistory.objects.filter(pk=scan_history_id).first() subscan = SubScan.objects.filter(pk=subscan_id).first() url = get_scan_url(scan_history_id) if status: fields['Status'] = f'**{status}**' if engine: fields['Engine'] = engine.engine_name if scan: fields['Scan ID'] = f'[#{scan.id}]({url})' if subscan: url = get_scan_url(scan_history_id, subscan_id) fields['Subscan ID'] = f'[#{subscan.id}]({url})' title = get_task_title(task_name, scan_history_id, subscan_id) if status: severity = STATUS_TO_SEVERITIES.get(status) msg = f'{title} {status}\n' msg += '\n🡆 '.join(f'**{k}:** {v}' for k, v in fields.items()) # Add fields to update for k, v in update_fields.items(): fields[k] = v # Add traceback to notif if traceback and notif.send_scan_tracebacks: fields['Traceback'] = f'```\n{traceback}\n```' # Add files to notif files = [] attach_file = ( notif.send_scan_output_file and output_path and result and not traceback ) if attach_file: output_title = output_path.split('/')[-1] files = [(output_path, output_title)] # Send notif opts = { 'title': title, 'url': url, 'files': files, 'severity': severity, 'fields': fields, 'fields_append': update_fields.keys() } send_notif( msg, scan_history_id=scan_history_id, subscan_id=subscan_id, **opts) @app.task(name='send_file_to_discord', bind=False, queue='send_file_to_discord_queue') def send_file_to_discord(file_path, title=None): notif = Notification.objects.first() do_send = notif and notif.send_to_discord and notif.discord_hook_url if not do_send: return False webhook = DiscordWebhook( url=notif.discord_hook_url, rate_limit_retry=True, username=title or "reNgine Discord Plugin" ) with open(file_path, "rb") as f: head, tail = os.path.split(file_path) webhook.add_file(file=f.read(), filename=tail) webhook.execute() @app.task(name='send_hackerone_report', bind=False, queue='send_hackerone_report_queue') def send_hackerone_report(vulnerability_id): """Send HackerOne vulnerability report. Args: vulnerability_id (int): Vulnerability id. Returns: int: HTTP response status code. """ vulnerability = Vulnerability.objects.get(id=vulnerability_id) severities = {v: k for k,v in NUCLEI_SEVERITY_MAP.items()} headers = { 'Content-Type': 'application/json', 'Accept': 'application/json' } # can only send vulnerability report if team_handle exists if len(vulnerability.target_domain.h1_team_handle) !=0: hackerone_query = Hackerone.objects.all() if hackerone_query.exists(): hackerone = Hackerone.objects.first() severity_value = severities[vulnerability.severity] tpl = hackerone.report_template # Replace syntax of report template with actual content tpl = tpl.replace('{vulnerability_name}', vulnerability.name) tpl = tpl.replace('{vulnerable_url}', vulnerability.http_url) tpl = tpl.replace('{vulnerability_severity}', severity_value) tpl = tpl.replace('{vulnerability_description}', vulnerability.description if vulnerability.description else '') tpl = tpl.replace('{vulnerability_extracted_results}', vulnerability.extracted_results if vulnerability.extracted_results else '') tpl = tpl.replace('{vulnerability_reference}', vulnerability.reference if vulnerability.reference else '') data = { "data": { "type": "report", "attributes": { "team_handle": vulnerability.target_domain.h1_team_handle, "title": '{} found in {}'.format(vulnerability.name, vulnerability.http_url), "vulnerability_information": tpl, "severity_rating": severity_value, "impact": "More information about the impact and vulnerability can be found here: \n" + vulnerability.reference if vulnerability.reference else "NA", } } } r = requests.post( 'https://api.hackerone.com/v1/hackers/reports', auth=(hackerone.username, hackerone.api_key), json=data, headers=headers ) response = r.json() status_code = r.status_code if status_code == 201: vulnerability.hackerone_report_id = response['data']["id"] vulnerability.open_status = False vulnerability.save() return status_code else: logger.error('No team handle found.') status_code = 111 return status_code #-------------# # Utils tasks # #-------------# @app.task(name='parse_nmap_results', bind=False, queue='parse_nmap_results_queue') def parse_nmap_results(xml_file, output_file=None): """Parse results from nmap output file. Args: xml_file (str): nmap XML report file path. Returns: list: List of vulnerabilities found from nmap results. """ with open(xml_file, encoding='utf8') as f: content = f.read() try: nmap_results = xmltodict.parse(content) # parse XML to dict except Exception as e: logger.exception(e) logger.error(f'Cannot parse {xml_file} to valid JSON. Skipping.') return [] # Write JSON to output file if output_file: with open(output_file, 'w') as f: json.dump(nmap_results, f, indent=4) logger.warning(json.dumps(nmap_results, indent=4)) hosts = ( nmap_results .get('nmaprun', {}) .get('host', {}) ) all_vulns = [] if isinstance(hosts, dict): hosts = [hosts] for host in hosts: # Grab hostname / IP from output hostnames_dict = host.get('hostnames', {}) if hostnames_dict: # Ensure that hostnames['hostname'] is a list for consistency hostnames_list = hostnames_dict['hostname'] if isinstance(hostnames_dict['hostname'], list) else [hostnames_dict['hostname']] # Extract all the @name values from the list of dictionaries hostnames = [entry.get('@name') for entry in hostnames_list] else: hostnames = [host.get('address')['@addr']] # Iterate over each hostname for each port for hostname in hostnames: # Grab ports from output ports = host.get('ports', {}).get('port', []) if isinstance(ports, dict): ports = [ports] for port in ports: url_vulns = [] port_number = port['@portid'] url = sanitize_url(f'{hostname}:{port_number}') logger.info(f'Parsing nmap results for {hostname}:{port_number} ...') if not port_number or not port_number.isdigit(): continue port_protocol = port['@protocol'] scripts = port.get('script', []) if isinstance(scripts, dict): scripts = [scripts] for script in scripts: script_id = script['@id'] script_output = script['@output'] script_output_table = script.get('table', []) logger.debug(f'Ran nmap script "{script_id}" on {port_number}/{port_protocol}:\n{script_output}\n') if script_id == 'vulscan': vulns = parse_nmap_vulscan_output(script_output) url_vulns.extend(vulns) elif script_id == 'vulners': vulns = parse_nmap_vulners_output(script_output) url_vulns.extend(vulns) # elif script_id == 'http-server-header': # TODO: nmap can help find technologies as well using the http-server-header script # regex = r'(\w+)/([\d.]+)\s?(?:\((\w+)\))?' # tech_name, tech_version, tech_os = re.match(regex, test_string).groups() # Technology.objects.get_or_create(...) # elif script_id == 'http_csrf': # vulns = parse_nmap_http_csrf_output(script_output) # url_vulns.extend(vulns) else: logger.warning(f'Script output parsing for script "{script_id}" is not supported yet.') # Add URL to vuln for vuln in url_vulns: # TODO: This should extend to any URL, not just HTTP vuln['http_url'] = url if 'http_path' in vuln: vuln['http_url'] += vuln['http_path'] all_vulns.append(vuln) return all_vulns def parse_nmap_http_csrf_output(script_output): pass def parse_nmap_vulscan_output(script_output): """Parse nmap vulscan script output. Args: script_output (str): Vulscan script output. Returns: list: List of Vulnerability dicts. """ data = {} vulns = [] provider_name = '' # Sort all vulns found by provider so that we can match each provider with # a function that pulls from its API to get more info about the # vulnerability. for line in script_output.splitlines(): if not line: continue if not line.startswith('['): # provider line if "No findings" in line: logger.info(f"No findings: {line}") continue elif ' - ' in line: provider_name, provider_url = tuple(line.split(' - ')) data[provider_name] = {'url': provider_url.rstrip(':'), 'entries': []} continue else: # Log a warning logger.warning(f"Unexpected line format: {line}") continue reg = r'\[(.*)\] (.*)' matches = re.match(reg, line) id, title = matches.groups() entry = {'id': id, 'title': title} data[provider_name]['entries'].append(entry) logger.warning('Vulscan parsed output:') logger.warning(pprint.pformat(data)) for provider_name in data: if provider_name == 'Exploit-DB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'IBM X-Force': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'MITRE CVE': logger.error(f'Provider {provider_name} is not supported YET.') for entry in data[provider_name]['entries']: cve_id = entry['id'] vuln = cve_to_vuln(cve_id) vulns.append(vuln) elif provider_name == 'OSVDB': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'OpenVAS (Nessus)': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'SecurityFocus': logger.error(f'Provider {provider_name} is not supported YET.') pass elif provider_name == 'VulDB': logger.error(f'Provider {provider_name} is not supported YET.') pass else: logger.error(f'Provider {provider_name} is not supported.') return vulns def parse_nmap_vulners_output(script_output, url=''): """Parse nmap vulners script output. TODO: Rework this as it's currently matching all CVEs no matter the confidence. Args: script_output (str): Script output. Returns: list: List of found vulnerabilities. """ vulns = [] # Check for CVE in script output CVE_REGEX = re.compile(r'.*(CVE-\d\d\d\d-\d+).*') matches = CVE_REGEX.findall(script_output) matches = list(dict.fromkeys(matches)) for cve_id in matches: # get CVE info vuln = cve_to_vuln(cve_id, vuln_type='nmap-vulners-nse') if vuln: vulns.append(vuln) return vulns def cve_to_vuln(cve_id, vuln_type=''): """Search for a CVE using CVESearch and return Vulnerability data. Args: cve_id (str): CVE ID in the form CVE-* Returns: dict: Vulnerability dict. """ cve_info = CVESearch('https://cve.circl.lu').id(cve_id) if not cve_info: logger.error(f'Could not fetch CVE info for cve {cve_id}. Skipping.') return None vuln_cve_id = cve_info['id'] vuln_name = vuln_cve_id vuln_description = cve_info.get('summary', 'none').replace(vuln_cve_id, '').strip() try: vuln_cvss = float(cve_info.get('cvss', -1)) except (ValueError, TypeError): vuln_cvss = -1 vuln_cwe_id = cve_info.get('cwe', '') exploit_ids = cve_info.get('refmap', {}).get('exploit-db', []) osvdb_ids = cve_info.get('refmap', {}).get('osvdb', []) references = cve_info.get('references', []) capec_objects = cve_info.get('capec', []) # Parse ovals for a better vuln name / type ovals = cve_info.get('oval', []) if ovals: vuln_name = ovals[0]['title'] vuln_type = ovals[0]['family'] # Set vulnerability severity based on CVSS score vuln_severity = 'info' if vuln_cvss < 4: vuln_severity = 'low' elif vuln_cvss < 7: vuln_severity = 'medium' elif vuln_cvss < 9: vuln_severity = 'high' else: vuln_severity = 'critical' # Build console warning message msg = f'{vuln_name} | {vuln_severity.upper()} | {vuln_cve_id} | {vuln_cwe_id} | {vuln_cvss}' for id in osvdb_ids: msg += f'\n\tOSVDB: {id}' for exploit_id in exploit_ids: msg += f'\n\tEXPLOITDB: {exploit_id}' logger.warning(msg) vuln = { 'name': vuln_name, 'type': vuln_type, 'severity': NUCLEI_SEVERITY_MAP[vuln_severity], 'description': vuln_description, 'cvss_score': vuln_cvss, 'references': references, 'cve_ids': [vuln_cve_id], 'cwe_ids': [vuln_cwe_id] } return vuln def parse_s3scanner_result(line): ''' Parses and returns s3Scanner Data ''' bucket = line['bucket'] return { 'name': bucket['name'], 'region': bucket['region'], 'provider': bucket['provider'], 'owner_display_name': bucket['owner_display_name'], 'owner_id': bucket['owner_id'], 'perm_auth_users_read': bucket['perm_auth_users_read'], 'perm_auth_users_write': bucket['perm_auth_users_write'], 'perm_auth_users_read_acl': bucket['perm_auth_users_read_acl'], 'perm_auth_users_write_acl': bucket['perm_auth_users_write_acl'], 'perm_auth_users_full_control': bucket['perm_auth_users_full_control'], 'perm_all_users_read': bucket['perm_all_users_read'], 'perm_all_users_write': bucket['perm_all_users_write'], 'perm_all_users_read_acl': bucket['perm_all_users_read_acl'], 'perm_all_users_write_acl': bucket['perm_all_users_write_acl'], 'perm_all_users_full_control': bucket['perm_all_users_full_control'], 'num_objects': bucket['num_objects'], 'size': bucket['bucket_size'] } def parse_nuclei_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ return { 'name': line['info'].get('name', ''), 'type': line['type'], 'severity': NUCLEI_SEVERITY_MAP[line['info'].get('severity', 'unknown')], 'template': line['template'], 'template_url': line['template-url'], 'template_id': line['template-id'], 'description': line['info'].get('description', ''), 'matcher_name': line.get('matcher-name', ''), 'curl_command': line.get('curl-command'), 'request': line.get('request'), 'response': line.get('response'), 'extracted_results': line.get('extracted-results', []), 'cvss_metrics': line['info'].get('classification', {}).get('cvss-metrics', ''), 'cvss_score': line['info'].get('classification', {}).get('cvss-score'), 'cve_ids': line['info'].get('classification', {}).get('cve_id', []) or [], 'cwe_ids': line['info'].get('classification', {}).get('cwe_id', []) or [], 'references': line['info'].get('reference', []) or [], 'tags': line['info'].get('tags', []), 'source': NUCLEI, } def parse_dalfox_result(line): """Parse results from nuclei JSON output. Args: line (dict): Nuclei JSON line output. Returns: dict: Vulnerability data. """ description = '' description += f" Evidence: {line.get('evidence')} <br>" if line.get('evidence') else '' description += f" Message: {line.get('message')} <br>" if line.get('message') else '' description += f" Payload: {line.get('message_str')} <br>" if line.get('message_str') else '' description += f" Vulnerable Parameter: {line.get('param')} <br>" if line.get('param') else '' return { 'name': 'XSS (Cross Site Scripting)', 'type': 'XSS', 'severity': DALFOX_SEVERITY_MAP[line.get('severity', 'unknown')], 'description': description, 'source': DALFOX, 'cwe_ids': [line.get('cwe')] } def parse_crlfuzz_result(url): """Parse CRLF results Args: url (str): CRLF Vulnerable URL Returns: dict: Vulnerability data. """ return { 'name': 'CRLF (HTTP Response Splitting)', 'type': 'CRLF', 'severity': 2, 'description': 'A CRLF (HTTP Response Splitting) vulnerability has been discovered.', 'source': CRLFUZZ, } def record_exists(model, data, exclude_keys=[]): """ Check if a record already exists in the database based on the given data. Args: model (django.db.models.Model): The Django model to check against. data (dict): Data dictionary containing fields and values. exclude_keys (list): List of keys to exclude from the lookup. Returns: bool: True if the record exists, False otherwise. """ # Extract the keys that will be used for the lookup lookup_fields = {key: data[key] for key in data if key not in exclude_keys} # Return True if a record exists based on the lookup fields, False otherwise return model.objects.filter(**lookup_fields).exists() @app.task(name='geo_localize', bind=False, queue='geo_localize_queue') def geo_localize(host, ip_id=None): """Uses geoiplookup to find location associated with host. Args: host (str): Hostname. ip_id (int): IpAddress object id. Returns: startScan.models.CountryISO: CountryISO object from DB or None. """ if validators.ipv6(host): logger.info(f'Ipv6 "{host}" is not supported by geoiplookup. Skipping.') return None cmd = f'geoiplookup {host}' _, out = run_command(cmd) if 'IP Address not found' not in out and "can't resolve hostname" not in out: country_iso = out.split(':')[1].strip().split(',')[0] country_name = out.split(':')[1].strip().split(',')[1].strip() geo_object, _ = CountryISO.objects.get_or_create( iso=country_iso, name=country_name ) geo_json = { 'iso': country_iso, 'name': country_name } if ip_id: ip = IpAddress.objects.get(pk=ip_id) ip.geo_iso = geo_object ip.save() return geo_json logger.info(f'Geo IP lookup failed for host "{host}"') return None @app.task(name='query_whois', bind=False, queue='query_whois_queue') def query_whois(ip_domain, force_reload_whois=False): """Query WHOIS information for an IP or a domain name. Args: ip_domain (str): IP address or domain name. save_domain (bool): Whether to save domain or not, default False Returns: dict: WHOIS information. """ if not force_reload_whois and Domain.objects.filter(name=ip_domain).exists() and Domain.objects.get(name=ip_domain).domain_info: domain = Domain.objects.get(name=ip_domain) if not domain.insert_date: domain.insert_date = timezone.now() domain.save() domain_info_db = domain.domain_info domain_info = DottedDict( dnssec=domain_info_db.dnssec, created=domain_info_db.created, updated=domain_info_db.updated, expires=domain_info_db.expires, geolocation_iso=domain_info_db.geolocation_iso, status=[status['name'] for status in DomainWhoisStatusSerializer(domain_info_db.status, many=True).data], whois_server=domain_info_db.whois_server, ns_records=[ns['name'] for ns in NameServersSerializer(domain_info_db.name_servers, many=True).data], registrar_name=domain_info_db.registrar.name, registrar_phone=domain_info_db.registrar.phone, registrar_email=domain_info_db.registrar.email, registrar_url=domain_info_db.registrar.url, registrant_name=domain_info_db.registrant.name, registrant_id=domain_info_db.registrant.id_str, registrant_organization=domain_info_db.registrant.organization, registrant_city=domain_info_db.registrant.city, registrant_state=domain_info_db.registrant.state, registrant_zip_code=domain_info_db.registrant.zip_code, registrant_country=domain_info_db.registrant.country, registrant_phone=domain_info_db.registrant.phone, registrant_fax=domain_info_db.registrant.fax, registrant_email=domain_info_db.registrant.email, registrant_address=domain_info_db.registrant.address, admin_name=domain_info_db.admin.name, admin_id=domain_info_db.admin.id_str, admin_organization=domain_info_db.admin.organization, admin_city=domain_info_db.admin.city, admin_state=domain_info_db.admin.state, admin_zip_code=domain_info_db.admin.zip_code, admin_country=domain_info_db.admin.country, admin_phone=domain_info_db.admin.phone, admin_fax=domain_info_db.admin.fax, admin_email=domain_info_db.admin.email, admin_address=domain_info_db.admin.address, tech_name=domain_info_db.tech.name, tech_id=domain_info_db.tech.id_str, tech_organization=domain_info_db.tech.organization, tech_city=domain_info_db.tech.city, tech_state=domain_info_db.tech.state, tech_zip_code=domain_info_db.tech.zip_code, tech_country=domain_info_db.tech.country, tech_phone=domain_info_db.tech.phone, tech_fax=domain_info_db.tech.fax, tech_email=domain_info_db.tech.email, tech_address=domain_info_db.tech.address, related_tlds=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_tlds, many=True).data], related_domains=[domain['name'] for domain in RelatedDomainSerializer(domain_info_db.related_domains, many=True).data], historical_ips=[ip for ip in HistoricalIPSerializer(domain_info_db.historical_ips, many=True).data], ) if domain_info_db.dns_records: a_records = [] txt_records = [] mx_records = [] dns_records = [{'name': dns['name'], 'type': dns['type']} for dns in DomainDNSRecordSerializer(domain_info_db.dns_records, many=True).data] for dns in dns_records: if dns['type'] == 'a': a_records.append(dns['name']) elif dns['type'] == 'txt': txt_records.append(dns['name']) elif dns['type'] == 'mx': mx_records.append(dns['name']) domain_info.a_records = a_records domain_info.txt_records = txt_records domain_info.mx_records = mx_records else: logger.info(f'Domain info for "{ip_domain}" not found in DB, querying whois') domain_info = DottedDict() # find domain historical ip try: historical_ips = get_domain_historical_ip_address(ip_domain) domain_info.historical_ips = historical_ips except Exception as e: logger.error(f'HistoricalIP for {ip_domain} not found!\nError: {str(e)}') historical_ips = [] # find associated domains using ip_domain try: related_domains = reverse_whois(ip_domain.split('.')[0]) except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] # find related tlds using TLSx try: related_tlds = [] output_path = '/tmp/ip_domain_tlsx.txt' tlsx_command = f'tlsx -san -cn -silent -ro -host {ip_domain} -o {output_path}' run_command( tlsx_command, shell=True, ) tlsx_output = [] with open(output_path) as f: tlsx_output = f.readlines() tldextract_target = tldextract.extract(ip_domain) for doms in tlsx_output: doms = doms.strip() tldextract_res = tldextract.extract(doms) if ip_domain != doms and tldextract_res.domain == tldextract_target.domain and tldextract_res.subdomain == '': related_tlds.append(doms) related_tlds = list(set(related_tlds)) domain_info.related_tlds = related_tlds except Exception as e: logger.error(f'Associated domain not found for {ip_domain}\nError: {str(e)}') similar_domains = [] related_domains_list = [] if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_related = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_related) related_domains_list.append(_domain['name']) for _domain in related_tlds: domain_related = RelatedDomain.objects.get_or_create( name=_domain, )[0] db_domain_info.related_tlds.add(domain_related) for _ip in historical_ips: historical_ip = HistoricalIP.objects.get_or_create( ip=_ip['ip'], owner=_ip['owner'], location=_ip['location'], last_seen=_ip['last_seen'], )[0] db_domain_info.historical_ips.add(historical_ip) domain.domain_info = db_domain_info domain.save() command = f'netlas host {ip_domain} -f json' # check if netlas key is provided netlas_key = get_netlas_key() command += f' -a {netlas_key}' if netlas_key else '' result = subprocess.check_output(command.split()).decode('utf-8') if 'Failed to parse response data' in result: # do fallback return { 'status': False, 'ip_domain': ip_domain, 'result': "Netlas limit exceeded.", 'message': 'Netlas limit exceeded.' } try: result = json.loads(result) logger.info(result) whois = result.get('whois') if result.get('whois') else {} domain_info.created = whois.get('created_date') domain_info.expires = whois.get('expiration_date') domain_info.updated = whois.get('updated_date') domain_info.whois_server = whois.get('whois_server') if 'registrant' in whois: registrant = whois.get('registrant') domain_info.registrant_name = registrant.get('name') domain_info.registrant_country = registrant.get('country') domain_info.registrant_id = registrant.get('id') domain_info.registrant_state = registrant.get('province') domain_info.registrant_city = registrant.get('city') domain_info.registrant_phone = registrant.get('phone') domain_info.registrant_address = registrant.get('street') domain_info.registrant_organization = registrant.get('organization') domain_info.registrant_fax = registrant.get('fax') domain_info.registrant_zip_code = registrant.get('postal_code') email_search = EMAIL_REGEX.search(str(registrant.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.registrant_email = field_content if 'administrative' in whois: administrative = whois.get('administrative') domain_info.admin_name = administrative.get('name') domain_info.admin_country = administrative.get('country') domain_info.admin_id = administrative.get('id') domain_info.admin_state = administrative.get('province') domain_info.admin_city = administrative.get('city') domain_info.admin_phone = administrative.get('phone') domain_info.admin_address = administrative.get('street') domain_info.admin_organization = administrative.get('organization') domain_info.admin_fax = administrative.get('fax') domain_info.admin_zip_code = administrative.get('postal_code') mail_search = EMAIL_REGEX.search(str(administrative.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.admin_email = field_content if 'technical' in whois: technical = whois.get('technical') domain_info.tech_name = technical.get('name') domain_info.tech_country = technical.get('country') domain_info.tech_state = technical.get('province') domain_info.tech_id = technical.get('id') domain_info.tech_city = technical.get('city') domain_info.tech_phone = technical.get('phone') domain_info.tech_address = technical.get('street') domain_info.tech_organization = technical.get('organization') domain_info.tech_fax = technical.get('fax') domain_info.tech_zip_code = technical.get('postal_code') mail_search = EMAIL_REGEX.search(str(technical.get('email'))) field_content = email_search.group(0) if email_search else None domain_info.tech_email = field_content if 'dns' in result: dns = result.get('dns') domain_info.mx_records = dns.get('mx') domain_info.txt_records = dns.get('txt') domain_info.a_records = dns.get('a') domain_info.ns_records = whois.get('name_servers') domain_info.dnssec = True if whois.get('dnssec') else False domain_info.status = whois.get('status') if 'registrar' in whois: registrar = whois.get('registrar') domain_info.registrar_name = registrar.get('name') domain_info.registrar_email = registrar.get('email') domain_info.registrar_phone = registrar.get('phone') domain_info.registrar_url = registrar.get('url') # find associated domains if registrant email is found related_domains = reverse_whois(domain_info.get('registrant_email')) if domain_info.get('registrant_email') else [] for _domain in related_domains: related_domains_list.append(_domain['name']) # remove duplicate domains from related domains list related_domains_list = list(set(related_domains_list)) domain_info.related_domains = related_domains_list # save to db if domain exists if Domain.objects.filter(name=ip_domain).exists(): domain = Domain.objects.get(name=ip_domain) db_domain_info = domain.domain_info if domain.domain_info else DomainInfo() db_domain_info.save() for _domain in related_domains: domain_rel = RelatedDomain.objects.get_or_create( name=_domain['name'], )[0] db_domain_info.related_domains.add(domain_rel) db_domain_info.dnssec = domain_info.get('dnssec') #dates db_domain_info.created = domain_info.get('created') db_domain_info.updated = domain_info.get('updated') db_domain_info.expires = domain_info.get('expires') #registrar db_domain_info.registrar = Registrar.objects.get_or_create( name=domain_info.get('registrar_name'), email=domain_info.get('registrar_email'), phone=domain_info.get('registrar_phone'), url=domain_info.get('registrar_url'), )[0] db_domain_info.registrant = DomainRegistration.objects.get_or_create( name=domain_info.get('registrant_name'), organization=domain_info.get('registrant_organization'), address=domain_info.get('registrant_address'), city=domain_info.get('registrant_city'), state=domain_info.get('registrant_state'), zip_code=domain_info.get('registrant_zip_code'), country=domain_info.get('registrant_country'), email=domain_info.get('registrant_email'), phone=domain_info.get('registrant_phone'), fax=domain_info.get('registrant_fax'), id_str=domain_info.get('registrant_id'), )[0] db_domain_info.admin = DomainRegistration.objects.get_or_create( name=domain_info.get('admin_name'), organization=domain_info.get('admin_organization'), address=domain_info.get('admin_address'), city=domain_info.get('admin_city'), state=domain_info.get('admin_state'), zip_code=domain_info.get('admin_zip_code'), country=domain_info.get('admin_country'), email=domain_info.get('admin_email'), phone=domain_info.get('admin_phone'), fax=domain_info.get('admin_fax'), id_str=domain_info.get('admin_id'), )[0] db_domain_info.tech = DomainRegistration.objects.get_or_create( name=domain_info.get('tech_name'), organization=domain_info.get('tech_organization'), address=domain_info.get('tech_address'), city=domain_info.get('tech_city'), state=domain_info.get('tech_state'), zip_code=domain_info.get('tech_zip_code'), country=domain_info.get('tech_country'), email=domain_info.get('tech_email'), phone=domain_info.get('tech_phone'), fax=domain_info.get('tech_fax'), id_str=domain_info.get('tech_id'), )[0] for status in domain_info.get('status') or []: _status = WhoisStatus.objects.get_or_create( name=status )[0] _status.save() db_domain_info.status.add(_status) for ns in domain_info.get('ns_records') or []: _ns = NameServer.objects.get_or_create( name=ns )[0] _ns.save() db_domain_info.name_servers.add(_ns) for a in domain_info.get('a_records') or []: _a = DNSRecord.objects.get_or_create( name=a, type='a' )[0] _a.save() db_domain_info.dns_records.add(_a) for mx in domain_info.get('mx_records') or []: _mx = DNSRecord.objects.get_or_create( name=mx, type='mx' )[0] _mx.save() db_domain_info.dns_records.add(_mx) for txt in domain_info.get('txt_records') or []: _txt = DNSRecord.objects.get_or_create( name=txt, type='txt' )[0] _txt.save() db_domain_info.dns_records.add(_txt) db_domain_info.geolocation_iso = domain_info.get('registrant_country') db_domain_info.whois_server = domain_info.get('whois_server') db_domain_info.save() domain.domain_info = db_domain_info domain.save() except Exception as e: return { 'status': False, 'ip_domain': ip_domain, 'result': "unable to fetch records from WHOIS database.", 'message': str(e) } return { 'status': True, 'ip_domain': ip_domain, 'dnssec': domain_info.get('dnssec'), 'created': domain_info.get('created'), 'updated': domain_info.get('updated'), 'expires': domain_info.get('expires'), 'geolocation_iso': domain_info.get('registrant_country'), 'domain_statuses': domain_info.get('status'), 'whois_server': domain_info.get('whois_server'), 'dns': { 'a': domain_info.get('a_records'), 'mx': domain_info.get('mx_records'), 'txt': domain_info.get('txt_records'), }, 'registrar': { 'name': domain_info.get('registrar_name'), 'phone': domain_info.get('registrar_phone'), 'email': domain_info.get('registrar_email'), 'url': domain_info.get('registrar_url'), }, 'registrant': { 'name': domain_info.get('registrant_name'), 'id': domain_info.get('registrant_id'), 'organization': domain_info.get('registrant_organization'), 'address': domain_info.get('registrant_address'), 'city': domain_info.get('registrant_city'), 'state': domain_info.get('registrant_state'), 'zipcode': domain_info.get('registrant_zip_code'), 'country': domain_info.get('registrant_country'), 'phone': domain_info.get('registrant_phone'), 'fax': domain_info.get('registrant_fax'), 'email': domain_info.get('registrant_email'), }, 'admin': { 'name': domain_info.get('admin_name'), 'id': domain_info.get('admin_id'), 'organization': domain_info.get('admin_organization'), 'address':domain_info.get('admin_address'), 'city': domain_info.get('admin_city'), 'state': domain_info.get('admin_state'), 'zipcode': domain_info.get('admin_zip_code'), 'country': domain_info.get('admin_country'), 'phone': domain_info.get('admin_phone'), 'fax': domain_info.get('admin_fax'), 'email': domain_info.get('admin_email'), }, 'technical_contact': { 'name': domain_info.get('tech_name'), 'id': domain_info.get('tech_id'), 'organization': domain_info.get('tech_organization'), 'address': domain_info.get('tech_address'), 'city': domain_info.get('tech_city'), 'state': domain_info.get('tech_state'), 'zipcode': domain_info.get('tech_zip_code'), 'country': domain_info.get('tech_country'), 'phone': domain_info.get('tech_phone'), 'fax': domain_info.get('tech_fax'), 'email': domain_info.get('tech_email'), }, 'nameservers': domain_info.get('ns_records'), # 'similar_domains': domain_info.get('similar_domains'), 'related_domains': domain_info.get('related_domains'), 'related_tlds': domain_info.get('related_tlds'), 'historical_ips': domain_info.get('historical_ips'), } @app.task(name='remove_duplicate_endpoints', bind=False, queue='remove_duplicate_endpoints_queue') def remove_duplicate_endpoints( scan_history_id, domain_id, subdomain_id=None, filter_ids=[], filter_status=[200, 301, 404], duplicate_removal_fields=ENDPOINT_SCAN_DEFAULT_DUPLICATE_FIELDS ): """Remove duplicate endpoints. Check for implicit redirections by comparing endpoints: - [x] `content_length` similarities indicating redirections - [x] `page_title` (check for same page title) - [ ] Sign-in / login page (check for endpoints with the same words) Args: scan_history_id: ScanHistory id. domain_id (int): Domain id. subdomain_id (int, optional): Subdomain id. filter_ids (list): List of endpoint ids to filter on. filter_status (list): List of HTTP status codes to filter on. duplicate_removal_fields (list): List of Endpoint model fields to check for duplicates """ logger.info(f'Removing duplicate endpoints based on {duplicate_removal_fields}') endpoints = ( EndPoint.objects .filter(scan_history__id=scan_history_id) .filter(target_domain__id=domain_id) ) if filter_status: endpoints = endpoints.filter(http_status__in=filter_status) if subdomain_id: endpoints = endpoints.filter(subdomain__id=subdomain_id) if filter_ids: endpoints = endpoints.filter(id__in=filter_ids) for field_name in duplicate_removal_fields: cl_query = ( endpoints .values_list(field_name) .annotate(mc=Count(field_name)) .order_by('-mc') ) for (field_value, count) in cl_query: if count > DELETE_DUPLICATES_THRESHOLD: eps_to_delete = ( endpoints .filter(**{field_name: field_value}) .order_by('discovered_date') .all()[1:] ) msg = f'Deleting {len(eps_to_delete)} endpoints [reason: same {field_name} {field_value}]' for ep in eps_to_delete: url = urlparse(ep.http_url) if url.path in ['', '/', '/login']: # try do not delete the original page that other pages redirect to continue msg += f'\n\t {ep.http_url} [{ep.http_status}] [{field_name}={field_value}]' ep.delete() logger.warning(msg) @app.task(name='run_command', bind=False, queue='run_command_queue') def run_command(cmd, cwd=None, shell=False, history_file=None, scan_id=None, activity_id=None): """Run a given command using subprocess module. Args: cmd (str): Command to run. cwd (str): Current working directory. echo (bool): Log command. shell (bool): Run within separate shell if True. history_file (str): Write command + output to history file. Returns: tuple: Tuple with return_code, output. """ logger.info(cmd) logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Run the command using subprocess popen = subprocess.Popen( cmd if shell else cmd.split(), shell=shell, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, cwd=cwd, universal_newlines=True) output = '' for stdout_line in iter(popen.stdout.readline, ""): item = stdout_line.strip() output += '\n' + item logger.debug(item) popen.stdout.close() popen.wait() return_code = popen.returncode command_obj.output = output command_obj.return_code = return_code command_obj.save() if history_file: mode = 'a' if not os.path.exists(history_file): mode = 'w' with open(history_file, mode) as f: f.write(f'\n{cmd}\n{return_code}\n{output}\n------------------\n') return return_code, output #-------------# # Other utils # #-------------# def stream_command(cmd, cwd=None, shell=False, history_file=None, encoding='utf-8', scan_id=None, activity_id=None, trunc_char=None): # Log cmd logger.info(cmd) # logger.warning(activity_id) # Create a command record in the database command_obj = Command.objects.create( command=cmd, time=timezone.now(), scan_history_id=scan_id, activity_id=activity_id) # Sanitize the cmd command = cmd if shell else cmd.split() # Run the command using subprocess process = subprocess.Popen( command, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, universal_newlines=True, shell=shell) # Log the output in real-time to the database output = "" # Process the output for line in iter(lambda: process.stdout.readline(), b''): if not line: break line = line.strip() ansi_escape = re.compile(r'\x1B(?:[@-Z\\-_]|\[[0-?]*[ -/]*[@-~])') line = ansi_escape.sub('', line) line = line.replace('\\x0d\\x0a', '\n') if trunc_char and line.endswith(trunc_char): line = line[:-1] item = line # Try to parse the line as JSON try: item = json.loads(line) except json.JSONDecodeError: pass # Yield the line #logger.debug(item) yield item # Add the log line to the output output += line + "\n" # Update the command record in the database command_obj.output = output command_obj.save() # Retrieve the return code and output process.wait() return_code = process.returncode # Update the return code and final output in the database command_obj.return_code = return_code command_obj.save() # Append the command, return code and output to the history file if history_file is not None: with open(history_file, "a") as f: f.write(f"{cmd}\n{return_code}\n{output}\n") def process_httpx_response(line): """TODO: implement this""" def extract_httpx_url(line): """Extract final URL from httpx results. Always follow redirects to find the last URL. Args: line (dict): URL data output by httpx. Returns: tuple: (final_url, redirect_bool) tuple. """ status_code = line.get('status_code', 0) final_url = line.get('final_url') location = line.get('location') chain_status_codes = line.get('chain_status_codes', []) # Final URL is already looking nice, if it exists return it if final_url: return final_url, False http_url = line['url'] # fallback to url field # Handle redirects manually REDIRECT_STATUS_CODES = [301, 302] is_redirect = ( status_code in REDIRECT_STATUS_CODES or any(x in REDIRECT_STATUS_CODES for x in chain_status_codes) ) if is_redirect and location: if location.startswith(('http', 'https')): http_url = location else: http_url = f'{http_url}/{location.lstrip("/")}' # Sanitize URL http_url = sanitize_url(http_url) return http_url, is_redirect #-------------# # OSInt utils # #-------------# def get_and_save_dork_results(lookup_target, results_dir, type, lookup_keywords=None, lookup_extensions=None, delay=3, page_count=2, scan_history=None): """ Uses gofuzz to dork and store information Args: lookup_target (str): target to look into such as stackoverflow or even the target itself results_dir (str): Results directory type (str): Dork Type Title lookup_keywords (str): comma separated keywords or paths to look for lookup_extensions (str): comma separated extensions to look for delay (int): delay between each requests page_count (int): pages in google to extract information scan_history (startScan.ScanHistory): Scan History Object """ results = [] gofuzz_command = f'{GOFUZZ_EXEC_PATH} -t {lookup_target} -d {delay} -p {page_count}' if lookup_extensions: gofuzz_command += f' -e {lookup_extensions}' elif lookup_keywords: gofuzz_command += f' -w {lookup_keywords}' output_file = f'{results_dir}/gofuzz.txt' gofuzz_command += f' -o {output_file}' history_file = f'{results_dir}/commands.txt' try: run_command( gofuzz_command, shell=False, history_file=history_file, scan_id=scan_history.id, ) if not os.path.isfile(output_file): return with open(output_file) as f: for line in f.readlines(): url = line.strip() if url: results.append(url) dork, created = Dork.objects.get_or_create( type=type, url=url ) if scan_history: scan_history.dorks.add(dork) # remove output file os.remove(output_file) except Exception as e: logger.exception(e) return results def get_and_save_emails(scan_history, activity_id, results_dir): """Get and save emails from Google, Bing and Baidu. Args: scan_history (startScan.ScanHistory): Scan history object. activity_id: ScanActivity Object results_dir (str): Results directory. Returns: list: List of emails found. """ emails = [] # Proxy settings # get_random_proxy() # Gather emails from Google, Bing and Baidu output_file = f'{results_dir}/emails_tmp.txt' history_file = f'{results_dir}/commands.txt' command = f'python3 /usr/src/github/Infoga/infoga.py --domain {scan_history.domain.name} --source all --report {output_file}' try: run_command( command, shell=False, history_file=history_file, scan_id=scan_history.id, activity_id=activity_id) if not os.path.isfile(output_file): logger.info('No Email results') return [] with open(output_file) as f: for line in f.readlines(): if 'Email' in line: split_email = line.split(' ')[2] emails.append(split_email) output_path = f'{results_dir}/emails.txt' with open(output_path, 'w') as output_file: for email_address in emails: save_email(email_address, scan_history) output_file.write(f'{email_address}\n') except Exception as e: logger.exception(e) return emails def save_metadata_info(meta_dict): """Extract metadata from Google Search. Args: meta_dict (dict): Info dict. Returns: list: List of startScan.MetaFinderDocument objects. """ logger.warning(f'Getting metadata for {meta_dict.osint_target}') scan_history = ScanHistory.objects.get(id=meta_dict.scan_id) # Proxy settings get_random_proxy() # Get metadata result = extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if not result: logger.error(f'No metadata result from Google Search for {meta_dict.osint_target}.') return [] # Add metadata info to DB results = [] for metadata_name, data in result.get_metadata().items(): subdomain = Subdomain.objects.get( scan_history=meta_dict.scan_id, name=meta_dict.osint_target) metadata = DottedDict({k: v for k, v in data.items()}) meta_finder_document = MetaFinderDocument( subdomain=subdomain, target_domain=meta_dict.domain, scan_history=scan_history, url=metadata.url, doc_name=metadata_name, http_status=metadata.status_code, producer=metadata.metadata.get('Producer'), creator=metadata.metadata.get('Creator'), creation_date=metadata.metadata.get('CreationDate'), modified_date=metadata.metadata.get('ModDate'), author=metadata.metadata.get('Author'), title=metadata.metadata.get('Title'), os=metadata.metadata.get('OSInfo')) meta_finder_document.save() results.append(data) return results #-----------------# # Utils functions # #-----------------# def create_scan_activity(scan_history_id, message, status): scan_activity = ScanActivity() scan_activity.scan_of = ScanHistory.objects.get(pk=scan_history_id) scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id #--------------------# # Database functions # #--------------------# def save_vulnerability(**vuln_data): references = vuln_data.pop('references', []) cve_ids = vuln_data.pop('cve_ids', []) cwe_ids = vuln_data.pop('cwe_ids', []) tags = vuln_data.pop('tags', []) subscan = vuln_data.pop('subscan', None) # remove nulls vuln_data = replace_nulls(vuln_data) # Create vulnerability vuln, created = Vulnerability.objects.get_or_create(**vuln_data) if created: vuln.discovered_date = timezone.now() vuln.open_status = True vuln.save() # Save vuln tags for tag_name in tags or []: tag, created = VulnerabilityTags.objects.get_or_create(name=tag_name) if tag: vuln.tags.add(tag) vuln.save() # Save CVEs for cve_id in cve_ids or []: cve, created = CveId.objects.get_or_create(name=cve_id) if cve: vuln.cve_ids.add(cve) vuln.save() # Save CWEs for cve_id in cwe_ids or []: cwe, created = CweId.objects.get_or_create(name=cve_id) if cwe: vuln.cwe_ids.add(cwe) vuln.save() # Save vuln reference for url in references or []: ref, created = VulnerabilityReference.objects.get_or_create(url=url) if created: vuln.references.add(ref) vuln.save() # Save subscan id in vuln object if subscan: vuln.vuln_subscan_ids.add(subscan) vuln.save() return vuln, created def save_endpoint( http_url, ctx={}, crawl=False, is_default=False, **endpoint_data): """Get or create EndPoint object. If crawl is True, also crawl the endpoint HTTP URL with httpx. Args: http_url (str): Input HTTP URL. is_default (bool): If the url is a default url for SubDomains. scan_history (startScan.models.ScanHistory): ScanHistory object. domain (startScan.models.Domain): Domain object. subdomain (starScan.models.Subdomain): Subdomain object. results_dir (str, optional): Results directory. crawl (bool, optional): Run httpx on endpoint if True. Default: False. force (bool, optional): Force crawl even if ENABLE_HTTP_CRAWL mode is on. subscan (startScan.models.SubScan, optional): SubScan object. Returns: tuple: (startScan.models.EndPoint, created) where `created` is a boolean indicating if the object is new or already existed. """ # remove nulls endpoint_data = replace_nulls(endpoint_data) scheme = urlparse(http_url).scheme endpoint = None created = False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in http_url: logger.error(f"{http_url} is not a URL of domain {domain.name}. Skipping.") return None, False if crawl: ctx['track'] = False results = http_crawl( urls=[http_url], method='HEAD', ctx=ctx) if results: endpoint_data = results[0] endpoint_id = endpoint_data['endpoint_id'] created = endpoint_data['endpoint_created'] endpoint = EndPoint.objects.get(pk=endpoint_id) elif not scheme: return None, False else: # add dumb endpoint without probing it scan = ScanHistory.objects.filter(pk=ctx.get('scan_history_id')).first() domain = Domain.objects.filter(pk=ctx.get('domain_id')).first() if not validators.url(http_url): return None, False http_url = sanitize_url(http_url) endpoint, created = EndPoint.objects.get_or_create( scan_history=scan, target_domain=domain, http_url=http_url, **endpoint_data) if created: endpoint.is_default = is_default endpoint.discovered_date = timezone.now() endpoint.save() subscan_id = ctx.get('subscan_id') if subscan_id: endpoint.endpoint_subscan_ids.add(subscan_id) endpoint.save() return endpoint, created def save_subdomain(subdomain_name, ctx={}): """Get or create Subdomain object. Args: subdomain_name (str): Subdomain name. scan_history (startScan.models.ScanHistory): ScanHistory object. Returns: tuple: (startScan.models.Subdomain, created) where `created` is a boolean indicating if the object has been created in DB. """ scan_id = ctx.get('scan_history_id') subscan_id = ctx.get('subscan_id') out_of_scope_subdomains = ctx.get('out_of_scope_subdomains', []) valid_domain = ( validators.domain(subdomain_name) or validators.ipv4(subdomain_name) or validators.ipv6(subdomain_name) ) if not valid_domain: logger.error(f'{subdomain_name} is not an invalid domain. Skipping.') return None, False if subdomain_name in out_of_scope_subdomains: logger.error(f'{subdomain_name} is out-of-scope. Skipping.') return None, False if ctx.get('domain_id'): domain = Domain.objects.get(id=ctx.get('domain_id')) if domain.name not in subdomain_name: logger.error(f"{subdomain_name} is not a subdomain of domain {domain.name}. Skipping.") return None, False scan = ScanHistory.objects.filter(pk=scan_id).first() domain = scan.domain if scan else None subdomain, created = Subdomain.objects.get_or_create( scan_history=scan, target_domain=domain, name=subdomain_name) if created: # logger.warning(f'Found new subdomain {subdomain_name}') subdomain.discovered_date = timezone.now() if subscan_id: subdomain.subdomain_subscan_ids.add(subscan_id) subdomain.save() return subdomain, created def save_email(email_address, scan_history=None): if not validators.email(email_address): logger.info(f'Email {email_address} is invalid. Skipping.') return None, False email, created = Email.objects.get_or_create(address=email_address) # if created: # logger.warning(f'Found new email address {email_address}') # Add email to ScanHistory if scan_history: scan_history.emails.add(email) scan_history.save() return email, created def save_employee(name, designation, scan_history=None): employee, created = Employee.objects.get_or_create( name=name, designation=designation) # if created: # logger.warning(f'Found new employee {name}') # Add employee to ScanHistory if scan_history: scan_history.employees.add(employee) scan_history.save() return employee, created def save_ip_address(ip_address, subdomain=None, subscan=None, **kwargs): if not (validators.ipv4(ip_address) or validators.ipv6(ip_address)): logger.info(f'IP {ip_address} is not a valid IP. Skipping.') return None, False ip, created = IpAddress.objects.get_or_create(address=ip_address) # if created: # logger.warning(f'Found new IP {ip_address}') # Set extra attributes for key, value in kwargs.items(): setattr(ip, key, value) ip.save() # Add IP to subdomain if subdomain: subdomain.ip_addresses.add(ip) subdomain.save() # Add subscan to IP if subscan: ip.ip_subscan_ids.add(subscan) # Geo-localize IP asynchronously if created: geo_localize.delay(ip_address, ip.id) return ip, created def save_imported_subdomains(subdomains, ctx={}): """Take a list of subdomains imported and write them to from_imported.txt. Args: subdomains (list): List of subdomain names. scan_history (startScan.models.ScanHistory): ScanHistory instance. domain (startScan.models.Domain): Domain instance. results_dir (str): Results directory. """ domain_id = ctx['domain_id'] domain = Domain.objects.get(pk=domain_id) results_dir = ctx.get('results_dir', RENGINE_RESULTS) # Validate each subdomain and de-duplicate entries subdomains = list(set([ subdomain for subdomain in subdomains if validators.domain(subdomain) and domain.name == get_domain_from_subdomain(subdomain) ])) if not subdomains: return logger.warning(f'Found {len(subdomains)} imported subdomains.') with open(f'{results_dir}/from_imported.txt', 'w+') as output_file: for name in subdomains: subdomain_name = name.strip() subdomain, _ = save_subdomain(subdomain_name, ctx=ctx) subdomain.is_imported_subdomain = True subdomain.save() output_file.write(f'{subdomain}\n') @app.task(name='query_reverse_whois', bind=False, queue='query_reverse_whois_queue') def query_reverse_whois(lookup_keyword): """Queries Reverse WHOIS information for an organization or email address. Args: lookup_keyword (str): Registrar Name or email Returns: dict: Reverse WHOIS information. """ return get_associated_domains(lookup_keyword) @app.task(name='query_ip_history', bind=False, queue='query_ip_history_queue') def query_ip_history(domain): """Queries the IP history for a domain Args: domain (str): domain_name Returns: list: list of historical ip addresses """ return get_domain_historical_ip_address(domain) @app.task(name='gpt_vulnerability_description', bind=False, queue='gpt_queue') def gpt_vulnerability_description(vulnerability_id): """Generate and store Vulnerability Description using GPT. Args: vulnerability_id (Vulnerability Model ID): Vulnerability ID to fetch Description. """ logger.info('Getting GPT Vulnerability Description') try: lookup_vulnerability = Vulnerability.objects.get(id=vulnerability_id) lookup_url = urlparse(lookup_vulnerability.http_url) path = lookup_url.path except Exception as e: return { 'status': False, 'error': str(e) } # check in db GPTVulnerabilityReport model if vulnerability description and path matches stored = GPTVulnerabilityReport.objects.filter(url_path=path).filter(title=lookup_vulnerability.name).first() if stored: response = { 'status': True, 'description': stored.description, 'impact': stored.impact, 'remediation': stored.remediation, 'references': [url.url for url in stored.references.all()] } else: vulnerability_description = get_gpt_vuln_input_description( lookup_vulnerability.name, path ) # one can add more description here later gpt_generator = GPTVulnerabilityReportGenerator() response = gpt_generator.get_vulnerability_description(vulnerability_description) add_gpt_description_db( lookup_vulnerability.name, path, response.get('description'), response.get('impact'), response.get('remediation'), response.get('references', []) ) # for all vulnerabilities with the same vulnerability name this description has to be stored. # also the consition is that the url must contain a part of this. for vuln in Vulnerability.objects.filter(name=lookup_vulnerability.name, http_url__icontains=path): vuln.description = response.get('description', vuln.description) vuln.impact = response.get('impact') vuln.remediation = response.get('remediation') vuln.is_gpt_used = True vuln.save() for url in response.get('references', []): ref, created = VulnerabilityReference.objects.get_or_create(url=url) vuln.references.add(ref) vuln.save() return response

ocervell

b557c6b8b70ea554c232095bf2fbb213e6d3648f

0ded32c1bee7852e7fc5daea0fb6de999097400b

## Overly permissive regular expression range Suspicious character range that is equivalent to \[0-9:;<=>?\]. [Show more details](https://github.com/yogeshojha/rengine/security/code-scanning/168)

github-advanced-security[bot]

yogeshojha/rengine

Add non-interactive installation parameter

Add a non-interactive installation method via a new parameter to be passed to the install.sh script. Essential for automated/industrialized systems (e.g. via Ansible or another automated environment creation system).

2023-10-12 01:09:15+00:00

2023-11-21 12:49:22+00:00

Makefile

.DEFAULT_GOAL:=help # Credits: https://github.com/sherifabdlnaby/elastdocker/ # This for future release of Compose that will use Docker Buildkit, which is much efficient. COMPOSE_PREFIX_CMD := COMPOSE_DOCKER_CLI_BUILD=1 COMPOSE_ALL_FILES := -f docker-compose.yml SERVICES := db web proxy redis celery celery-beat # -------------------------- .PHONY: setup certs up build username pull down stop restart rm logs certs: ## Generate certificates. @${COMPOSE_PREFIX_CMD} docker-compose -f docker-compose.setup.yml run --rm certs setup: ## Generate certificates. @make certs up: ## Build and start all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} up -d --build ${SERVICES} build: ## Build all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} build ${SERVICES} username: ## Generate Username (Use only after make up). ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} exec web python3 manage.py createsuperuser pull: ## Pull Docker images. docker login docker.pkg.github.com ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} pull down: ## Down all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} down stop: ## Stop all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} stop ${SERVICES} restart: ## Restart all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} restart ${SERVICES} rm: ## Remove all services containers. ${COMPOSE_PREFIX_CMD} docker-compose $(COMPOSE_ALL_FILES) rm -f ${SERVICES} test: ${COMPOSE_PREFIX_CMD} docker-compose $(COMPOSE_ALL_FILES) exec celery python3 -m unittest tests/test_scan.py logs: ## Tail all logs with -n 1000. ${COMPOSE_PREFIX_CMD} docker-compose $(COMPOSE_ALL_FILES) logs --follow --tail=1000 ${SERVICES} images: ## Show all Docker images. ${COMPOSE_PREFIX_CMD} docker-compose $(COMPOSE_ALL_FILES) images ${SERVICES} prune: ## Remove containers and delete volume data. @make stop && make rm && docker volume prune -f help: ## Show this help. @echo "Make application docker images and manage containers using docker-compose files." @awk 'BEGIN {FS = ":.*##"; printf "\nUsage:\n make \033[36m<target>\033[0m (default: help)\n\nTargets:\n"} /^[a-zA-Z_-]+:.*?##/ { printf " \033[36m%-12s\033[0m %s\n", $$1, $$2 }' $(MAKEFILE_LIST)

include .env .DEFAULT_GOAL:=help # Credits: https://github.com/sherifabdlnaby/elastdocker/ # This for future release of Compose that will use Docker Buildkit, which is much efficient. COMPOSE_PREFIX_CMD := COMPOSE_DOCKER_CLI_BUILD=1 COMPOSE_ALL_FILES := -f docker-compose.yml SERVICES := db web proxy redis celery celery-beat # -------------------------- .PHONY: setup certs up build username pull down stop restart rm logs certs: ## Generate certificates. @${COMPOSE_PREFIX_CMD} docker-compose -f docker-compose.setup.yml run --rm certs setup: ## Generate certificates. @make certs up: ## Build and start all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} up -d --build ${SERVICES} build: ## Build all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} build ${SERVICES} username: ## Generate Username (Use only after make up). ifeq ($(isNonInteractive), true) ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} exec web python3 manage.py createsuperuser --username ${DJANGO_SUPERUSER_USERNAME} --email ${DJANGO_SUPERUSER_EMAIL} --noinput else ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} exec web python3 manage.py createsuperuser endif pull: ## Pull Docker images. docker login docker.pkg.github.com ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} pull down: ## Down all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} down stop: ## Stop all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} stop ${SERVICES} restart: ## Restart all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} restart ${SERVICES} rm: ## Remove all services containers. ${COMPOSE_PREFIX_CMD} docker-compose $(COMPOSE_ALL_FILES) rm -f ${SERVICES} test: ${COMPOSE_PREFIX_CMD} docker-compose $(COMPOSE_ALL_FILES) exec celery python3 -m unittest tests/test_scan.py logs: ## Tail all logs with -n 1000. ${COMPOSE_PREFIX_CMD} docker-compose $(COMPOSE_ALL_FILES) logs --follow --tail=1000 ${SERVICES} images: ## Show all Docker images. ${COMPOSE_PREFIX_CMD} docker-compose $(COMPOSE_ALL_FILES) images ${SERVICES} prune: ## Remove containers and delete volume data. @make stop && make rm && docker volume prune -f help: ## Show this help. @echo "Make application docker images and manage containers using docker-compose files." @awk 'BEGIN {FS = ":.*##"; printf "\nUsage:\n make \033[36m<target>\033[0m (default: help)\n\nTargets:\n"} /^[a-zA-Z_-]+:.*?##/ { printf " \033[36m%-12s\033[0m %s\n", $$1, $$2 }' $(MAKEFILE_LIST)

C0wnuts

3dd700357a4bd5701b07ede4511f66042655be00

64b7f291240b3b8853e3cec7ee6230827c97b907

What's the addition/benefit of adding the .env file to `Makefile`?

AnonymousWP

yogeshojha/rengine

Add non-interactive installation parameter

Add a non-interactive installation method via a new parameter to be passed to the install.sh script. Essential for automated/industrialized systems (e.g. via Ansible or another automated environment creation system).

2023-10-12 01:09:15+00:00

2023-11-21 12:49:22+00:00

Makefile

.DEFAULT_GOAL:=help # Credits: https://github.com/sherifabdlnaby/elastdocker/ # This for future release of Compose that will use Docker Buildkit, which is much efficient. COMPOSE_PREFIX_CMD := COMPOSE_DOCKER_CLI_BUILD=1 COMPOSE_ALL_FILES := -f docker-compose.yml SERVICES := db web proxy redis celery celery-beat # -------------------------- .PHONY: setup certs up build username pull down stop restart rm logs certs: ## Generate certificates. @${COMPOSE_PREFIX_CMD} docker-compose -f docker-compose.setup.yml run --rm certs setup: ## Generate certificates. @make certs up: ## Build and start all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} up -d --build ${SERVICES} build: ## Build all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} build ${SERVICES} username: ## Generate Username (Use only after make up). ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} exec web python3 manage.py createsuperuser pull: ## Pull Docker images. docker login docker.pkg.github.com ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} pull down: ## Down all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} down stop: ## Stop all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} stop ${SERVICES} restart: ## Restart all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} restart ${SERVICES} rm: ## Remove all services containers. ${COMPOSE_PREFIX_CMD} docker-compose $(COMPOSE_ALL_FILES) rm -f ${SERVICES} test: ${COMPOSE_PREFIX_CMD} docker-compose $(COMPOSE_ALL_FILES) exec celery python3 -m unittest tests/test_scan.py logs: ## Tail all logs with -n 1000. ${COMPOSE_PREFIX_CMD} docker-compose $(COMPOSE_ALL_FILES) logs --follow --tail=1000 ${SERVICES} images: ## Show all Docker images. ${COMPOSE_PREFIX_CMD} docker-compose $(COMPOSE_ALL_FILES) images ${SERVICES} prune: ## Remove containers and delete volume data. @make stop && make rm && docker volume prune -f help: ## Show this help. @echo "Make application docker images and manage containers using docker-compose files." @awk 'BEGIN {FS = ":.*##"; printf "\nUsage:\n make \033[36m<target>\033[0m (default: help)\n\nTargets:\n"} /^[a-zA-Z_-]+:.*?##/ { printf " \033[36m%-12s\033[0m %s\n", $$1, $$2 }' $(MAKEFILE_LIST)

include .env .DEFAULT_GOAL:=help # Credits: https://github.com/sherifabdlnaby/elastdocker/ # This for future release of Compose that will use Docker Buildkit, which is much efficient. COMPOSE_PREFIX_CMD := COMPOSE_DOCKER_CLI_BUILD=1 COMPOSE_ALL_FILES := -f docker-compose.yml SERVICES := db web proxy redis celery celery-beat # -------------------------- .PHONY: setup certs up build username pull down stop restart rm logs certs: ## Generate certificates. @${COMPOSE_PREFIX_CMD} docker-compose -f docker-compose.setup.yml run --rm certs setup: ## Generate certificates. @make certs up: ## Build and start all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} up -d --build ${SERVICES} build: ## Build all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} build ${SERVICES} username: ## Generate Username (Use only after make up). ifeq ($(isNonInteractive), true) ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} exec web python3 manage.py createsuperuser --username ${DJANGO_SUPERUSER_USERNAME} --email ${DJANGO_SUPERUSER_EMAIL} --noinput else ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} exec web python3 manage.py createsuperuser endif pull: ## Pull Docker images. docker login docker.pkg.github.com ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} pull down: ## Down all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} down stop: ## Stop all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} stop ${SERVICES} restart: ## Restart all services. ${COMPOSE_PREFIX_CMD} docker-compose ${COMPOSE_ALL_FILES} restart ${SERVICES} rm: ## Remove all services containers. ${COMPOSE_PREFIX_CMD} docker-compose $(COMPOSE_ALL_FILES) rm -f ${SERVICES} test: ${COMPOSE_PREFIX_CMD} docker-compose $(COMPOSE_ALL_FILES) exec celery python3 -m unittest tests/test_scan.py logs: ## Tail all logs with -n 1000. ${COMPOSE_PREFIX_CMD} docker-compose $(COMPOSE_ALL_FILES) logs --follow --tail=1000 ${SERVICES} images: ## Show all Docker images. ${COMPOSE_PREFIX_CMD} docker-compose $(COMPOSE_ALL_FILES) images ${SERVICES} prune: ## Remove containers and delete volume data. @make stop && make rm && docker volume prune -f help: ## Show this help. @echo "Make application docker images and manage containers using docker-compose files." @awk 'BEGIN {FS = ":.*##"; printf "\nUsage:\n make \033[36m<target>\033[0m (default: help)\n\nTargets:\n"} /^[a-zA-Z_-]+:.*?##/ { printf " \033[36m%-12s\033[0m %s\n", $$1, $$2 }' $(MAKEFILE_LIST)

C0wnuts

3dd700357a4bd5701b07ede4511f66042655be00

64b7f291240b3b8853e3cec7ee6230827c97b907

To give it access to variables located in the .env and avoid writing the username/email directly in the Makefile. This way, all configuration elements remain centralized in the .env file. Following the Django documentation, for non-interactive createsuperuser process, you need to specify --username and --email argument (With the desired values) inside the docker-compose command

C0wnuts

yogeshojha/rengine

Add non-interactive installation parameter

Add a non-interactive installation method via a new parameter to be passed to the install.sh script. Essential for automated/industrialized systems (e.g. via Ansible or another automated environment creation system).

2023-10-12 01:09:15+00:00

2023-11-21 12:49:22+00:00

README.md

<p align="center"> <a href="https://rengine.wiki"><img src=".github/screenshots/banner.gif" alt=""/></a> </p> <p align="center"><a href="https://github.com/yogeshojha/rengine/releases" target="_blank"><img src="https://img.shields.io/badge/version-v2.0.0-informational?&logo=none" alt="reNgine Latest Version" /></a>&nbsp;<a href="https://www.gnu.org/licenses/gpl-3.0" target="_blank"><img src="https://img.shields.io/badge/License-GPLv3-red.svg?&logo=none" alt="License" /></a>&nbsp;<a href="#" target="_blank"><img src="https://img.shields.io/badge/first--timers--only-friendly-blue.svg?&logo=none" alt="" /></a>&nbsp;<a href="https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine" target="_blank"><img src="https://cdn.huntr.dev/huntr_security_badge_mono.svg" alt="" /></a>&nbsp;</p> <p align="center"> <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Asia-2023-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/Open--Source--Summit-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://cyberweek.ae/2021/hitb-armory/" target="_blank"><img src="https://img.shields.io/badge/HITB--Armory-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=7uvP6MaQOX0" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://drive.google.com/file/d/1Bh8lbf-Dztt5ViHJVACyrXMiglyICPQ2/view?usp=sharing" target="_blank"><img src="https://img.shields.io/badge/Defcon--Demolabs--29-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=A1oNOIc0h5A" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Europe-2020-blue.svg?&logo=none" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml/badge.svg" alt="" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/actions/workflows/build.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/build.yml/badge.svg" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://discord.gg/H6WzebwX3H" target="_blank"><img src="https://img.shields.io/discord/880363103689277461" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://opensourcesecurityindex.io/" target="_blank" rel="noopener"> <img style="width: 282px; height: 56px" src="https://opensourcesecurityindex.io/badge.svg" alt="Open Source Security Index - Fastest Growing Open Source Security Projects" width="282" height="56" /> </a> </p> <h3>reNgine 2.0-jasper<br>Redefining the future of reconnaissance!</h3> <h4>What is reNgine?</h4> <p align="left">reNgine is your go-to web application reconnaissance suite that's designed to simplify and streamline the reconnaissance process for security professionals, penetration testers, and bug bounty hunters. With its highly configurable engines, data correlation capabilities, continuous monitoring, database-backed reconnaissance data, and an intuitive user interface, reNgine redefines how you gather critical information about your target web applications. Traditional reconnaissance tools often fall short in terms of configurability and efficiency. reNgine addresses these shortcomings and emerges as a excellent alternative to existing commercial tools. reNgine was created to address the limitations of traditional reconnaissance tools and provide a better alternative, even surpassing some commercial offerings. Whether you're a bug bounty hunter, a penetration tester, or a corporate security team, reNgine is your go-to solution for automating and enhancing your information-gathering efforts. </p> reNgine 2.0-jasper is out now, you can [watch reNgine 2.0-jasper release trailer here!](https://youtu.be/VwkOWqiWW5g) reNgine 2.0-Jasper would not have been possible without [@ocervell](https://github.com/ocervell) valuable contributions. [@ocervell](https://github.com/ocervell) did majority of the refactoring if not all and also added a ton of features. Together, we wish to shape the future of web application reconnaissance, and it's developers like [@ocervell](https://github.com/ocervell) and a [ton of other developers and hackers from our community](https://github.com/yogeshojha/rengine/graphs/contributors) who inspire and drive us forward. Thank you, [@ocervell](https://github.com/ocervell), for your outstanding work and unwavering commitment to reNgine. Checkout our contributers here: [Contributers](https://github.com/yogeshojha/rengine/graphs/contributors)  ### Documentation You can find detailed documentation at [https://rengine.wiki](https://rengine.wiki)  ### Table of Contents * [About reNgine](#about-rengine) * [Workflow](#workflow) * [Features](#features) * [Scan Engine](#scan-engine) * [Quick Installation](#quick-installation) * [What's new in reNgine 2.0](#changelog) * [Screenshots](#screenshots) * [Contributing](#contributing) * [reNgine Support](#rengine-support) * [Support and Sponsoring](#support-and-sponsoring) * [reNgine Bug Bounty Program](#rengine-bug-bounty-program) * [License](#license)  ### About reNgine reNgine is not an ordinary reconnaissance suite; it's a game-changer! We've turbocharged the traditional workflow with groundbreaking features that is sure to ease your reconnaissance game. reNgine redefines the art of reconnaissance with highly configurable scan engines, recon data correlation, continuous monitoring, GPT powered Vulnerability Report, Project Management and role based access control etc. 🦾&nbsp;&nbsp; reNgine has advanced reconnaissance capabilities, harnessing a range of open-source tools to deliver a comprehensive web application reconnaissance experience. With it's intuitive User Interface, it excels in subdomain discovery, pinpointing IP addresses and open ports, collecting endpoints, conducting directory and file fuzzing, capturing screenshots, and performing vulnerability scans. To summarize, it does end-to-end reconnaissance. With WHOIS identification and WAF detection, it offers deep insights into target domains. Additionally, reNgine also identifies misconfigured S3 buckets and find interesting subdomains and URLS, based on specific keywords to helps you identify your next target, making it an go to tool for efficient reconnaissance. 🗃️&nbsp; &nbsp; Say goodbye to recon data chaos! reNgine seamlessly integrates with a database, providing you with unmatched data correlation and organization. Forgot the hassle of grepping through json, txt or csv files. Plus, our custom query language lets you filter reconnaissance data effortlessly using natural language like operators such as filtering all alive subdomains with `http_status=200` and also filter all subdomains that are alive and has admin in name `http_status=200&name=admin` 🔧&nbsp;&nbsp; reNgine offers unparalleled flexibility through its highly configurable scan engines, based on a YAML-based configuration. It offers the freedom to create and customize recon scan engines based on any kind of requirement, users can tailor them to their specific objectives and preferences, from thread management to timeout settings and rate-limit configurations, everything is customizable. Additionally, reNgine offers a range of pre-configured scan engines right out of the box, including Full Scan, Passive Scan, Screenshot Gathering, and the OSINT Scan Engine. These ready-to-use engines eliminate the need for extensive manual setup, aligning perfectly with reNgine's core mission of simplifying the reconnaissance process and enabling users to effortlessly access the right reconnaissance data with minimal effort. 💎&nbsp;&nbsp;Subscans: Subscan is a game-changing feature in reNgine, setting it apart as the only open-source tool of its kind to offer this capability. With Subscan, waiting for the entire pipeline to complete is a thing of the past. Now, users can swiftly respond to newfound discoveries during reconnaissance. Whether you've stumbled upon an intriguing subdomain and wish to conduct a focused port scan or want to delve deeper with a vulnerability assessment, reNgine has you covered. 📃&nbsp;&nbsp; PDF Reports: In addition to its robust reconnaissance capabilities, reNgine goes the extra mile by simplifying the report generation process, recognizing the crucial role that PDF reports play in the realm of end-to-end reconnaissance. Users can effortlessly generate and customize PDF reports to suit their exact needs. Whether it's a Full Scan Report, Vulnerability Report, or a concise reconnaissance report, reNgine provides the flexibility to choose the report type that best communicates your findings. Moreover, the level of customization is unparalleled, allowing users to select report colors, fine-tune executive summaries, and even add personalized touches like company names and footers. With GPT integration, your reports aren't just a report, with remediation steps, and impacts, you get 360-degree view of the vulnerabilities you've uncovered. 🔖&nbsp; &nbsp; Say Hello to Projects! reNgine 2.0 introduces a powerful addition that enables you to efficiently organize your web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. Each projects will have separate dashboard and all the scan results will be separated from each projects, while scan engines and configuration will be shared across all the projects. ⚙️&nbsp; &nbsp; Roles and Permissions! Begining reNgine 2.0, we've taken your web application reconnaissance to a whole new level of control and security. Now, you can assign distinct roles to your team members—Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - 🔐 Sys Admin: Sys Admin is a super user that has permission to modify system and scan related configurations, scan engines, create new users, add new tools etc. Super user can initiate scans and subscans effortlessly. - 🔍 Penetration Tester: Penetration Tester will be allowed to modify and initiate scans and subscans, add or update targets, etc. A penetration tester will not be allowed to modify system configurations. - 📊 Auditor: Auditor can only view and download the report. An auditor can not change any system or scan related configurations nor can initiate any scans or subscans. 🚀&nbsp;&nbsp; GPT Vulnerability Report Generation: Get ready for the future of penetration testing reports with reNgine's groundbreaking feature: "GPT-Powered Report Generation"! With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments that read like they were written by a human security expert! **But that's not all!** Our GPT-driven reports go the extra mile by scouring the web for related news articles, blogs, and references, so you have a 360-degree view of the vulnerabilities you've uncovered. With reNgine 2.0 revolutionize your penetration testing game and impress your clients with reports that are not just informative but engaging and comprehensive with detailed analysis on impact assessment and remediation strategies. 🥷&nbsp;&nbsp; GPT-Powered Attack Surface Generation: With reNgine 2.0, reNgine seamlessly integrates with GPT to identify the attacks that you can likely perform on a subdomain. By making use of reconnaissance data such as page title, open ports, subdomain name etc, reNgine can advice you the attacks you could perform on a target. reNgine will also provide you the rationale on why the specific attack is likely to be successful. 🧭&nbsp;&nbsp;Continuous monitoring: Continuous monitoring is at the core of reNgine's mission, and it's robust continuous monitoring feature ensures that their targets are under constant scrutiny. With the flexibility to schedule scans at regular intervals, penetration testers can effortlessly stay informed about their targets. What sets reNgine apart is its seamless integration with popular notification channels such as Discord, Slack, and Telegram, delivering real-time alerts for newly discovered subdomains, vulnerabilities, or any changes in reconnaissance data.  ### Workflow <img src="https://github.com/yogeshojha/rengine/assets/17223002/10c475b8-b4a8-440d-9126-77fe2038a386">  ### Features * Reconnaissance: * Subdomain Discovery * IP and Open Ports Identification * Endpoints Discovery * Directory/Files fuzzing * Screenshot Gathering * Vulnerability Scan * Nuclei * Dalfox XSS Scanner * CRLFuzzer * Misconfigured S3 Scanner * WHOIS Identification * WAF Detection * OSINT Capabilities * Meta info Gathering * Employees Gathering * Email Address gathering * Google Dorking for sensitive info and urls * Projects, create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. * Perform Advanced Query lookup using natural language alike and, or, not operations * Highly configurable YAML-based Scan Engines * Support for Parallel Scans * Support for Subscans * Recon Data visualization * GPT Vulnerability Description, Impact and Remediation generation * GPT Attack Surface Generator * Multiple Roles and Permissions to cater a team's need * Customizable Alerts/Notifications on Slack, Discord, and Telegram * Automatically report Vulnerabilities to HackerOne * Recon Notes and Todos * Clocked Scans (Run reconnaissance exactly at X Hours and Y minutes) and Periodic Scans (Runs reconnaissance every X minutes/- hours/days/week) * Proxy Support * Screenshot Gallery with Filters * Powerful recon data filtering with autosuggestions * Recon Data changes, find new/removed subdomains/endpoints * Tag targets into the Organization * Smart Duplicate endpoint removal based on page title and content length to cleanup the reconnaissance data * Identify Interesting Subdomains * Custom GF patterns and custom Nuclei Templates * Edit tool-related configuration files (Nuclei, Subfinder, Naabu, amass) * Add external tools from Github/Go * Interoperable with other tools, Import/Export Subdomains/Endpoints * Import Targets via IP and/or CIDRs * Report Generation * Toolbox: Comes bundled with most commonly used tools during penetration testing such as whois lookup, CMS detector, CVE lookup, etc. * Identification of related domains and related TLDs for targets * Find actionable insights such as Most Common Vulnerability, Most Common CVE ID, Most Vulnerable Target/Subdomain, etc.  ### Scan Engine ```yaml subdomain_discovery: { 'uses_tools': [ 'subfinder', 'ctfr', 'sublist3r', 'tlsx', 'oneforall', 'netlas' ], 'enable_http_crawl': true, 'threads': 30, 'timeout': 5, } http_crawl: {} port_scan: { 'enable_http_crawl': true, 'timeout': 5, # 'exclude_ports': [], # 'exclude_subdomains': true, 'ports': ['top-100'], 'rate_limit': 150, 'threads': 30, 'passive': false, # 'use_naabu_config': false, # 'enable_nmap': true, # 'nmap_cmd': '', # 'nmap_script': '', # 'nmap_script_args': '' } osint: { 'discover': [ 'emails', 'metainfo', 'employees' ], 'dorks': [ 'login_pages', 'admin_panels', 'dashboard_pages', 'stackoverflow', 'social_media', 'project_management', 'code_sharing', 'config_files', 'jenkins', 'wordpress_files', 'php_error', 'exposed_documents', 'db_files', 'git_exposed' ], 'custom_dorks': [ { 'lookup_site': 'google.com', 'lookup_keywords': '/home/' }, { 'lookup_site': '_target_', 'lookup_extensions': 'jpg,png' } ], 'intensity': 'normal', 'documents_limit': 50 } dir_file_fuzz: { 'auto_calibration': true, 'enable_http_crawl': true, 'rate_limit': 150, 'extensions': ['html', 'php','git','yaml','conf','cnf','config','gz','env','log','db','mysql','bak','asp','aspx','txt','conf','sql','json','yml','pdf'], 'follow_redirect': false, 'max_time': 0, 'match_http_status': [200, 204], 'recursive_level': 2, 'stop_on_error': false, 'timeout': 5, 'threads': 30, 'wordlist_name': 'dicc' } fetch_url: { 'uses_tools': [ 'gospider', 'hakrawler', 'waybackurls', 'gospider', 'katana' ], 'remove_duplicate_endpoints': true, 'duplicate_fields': [ 'content_length', 'page_title' ], 'enable_http_crawl': true, 'gf_patterns': ['debug_logic', 'idor', 'interestingEXT', 'interestingparams', 'interestingsubs', 'lfi', 'rce', 'redirect', 'sqli', 'ssrf', 'ssti', 'xss'], 'ignore_file_extensions': ['png', 'jpg', 'jpeg', 'gif', 'mp4', 'mpeg', 'mp3'] # 'exclude_subdomains': true } vulnerability_scan: { 'run_nuclei': false, 'run_dalfox': false, 'run_crlfuzz': false, 'run_s3scanner': true, 'enable_http_crawl': true, 'concurrency': 50, 'intensity': 'normal', 'rate_limit': 150, 'retries': 1, 'timeout': 5, 'fetch_gpt_report': true, 'nuclei': { 'use_conf': false, 'severities': [ 'unknown', 'info', 'low', 'medium', 'high', 'critical' ], # 'tags': [], # 'templates': [], # 'custom_templates': [], }, 's3scanner': { 'threads': 100, 'providers': [ 'aws', 'gcp', 'digitalocean', 'dreamhost', 'linode' ] } } waf_detection: {} screenshot: { 'enable_http_crawl': true, 'intensity': 'normal', 'timeout': 10, 'threads': 40 } # custom_header: "Cookie: Test" ```  ### Quick Installation **Note:** Only Ubuntu/VPS 1. Clone this repo ```bash git clone https://github.com/yogeshojha/rengine && cd rengine ``` 1. Edit the dotenv file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```bash nano .env ``` 1. In the dotenv file, you may also modify the Scaling Configurations ```bash MAX_CONCURRENCY=80 MIN_CONCURRENCY=10 ``` MAX_CONCURRENCY: This parameter specifies the maximum number of reNgine's concurrent Celery worker processes that can be spawned. In this case, it's set to 80, meaning that the application can utilize up to 80 concurrent worker processes to execute tasks concurrently. This is useful for handling a high volume of scans or when you want to scale up processing power during periods of high demand. If you have more CPU cores, you will need to increase this for maximised performance. MIN_CONCURRENCY: On the other hand, MIN_CONCURRENCY specifies the minimum number of concurrent worker processes that should be maintained, even during periods of lower demand. In this example, it's set to 10, which means that even when there are fewer tasks to process, at least 10 worker processes will be kept running. This helps ensure that the application can respond promptly to incoming tasks without the overhead of repeatedly starting and stopping worker processes. These settings allow for dynamic scaling of Celery workers, ensuring that the application efficiently manages its workload by adjusting the number of concurrent workers based on the workload's size and complexity 1. Run the installation script, Please keep an eye for any prompt, you will also be asked for username and password for reNgine. ```bash sudo ./install.sh ``` If `install.sh` does not have install permission, please change it, `chmod +x install.sh` **reNgine can now be accessed from <https://127.0.0.1> or if you're on the VPS <https://your_vps_ip_address>** **Unless you are on development branch, please do not access reNgine via any ports** ### Installation (Mac/Windows/Other) Installation instructions can be found at [https://reNgine.wiki/install/detailed/](https://reNgine.wiki/2.0/install/detailed/) ### Updating 1. Updating is as simple as running the following command: ```bash cd rengine && sudo ./update.sh ``` If `update.sh` does not have execution permissions, please change it, `sudo chmod +x update.sh` **NOTE:** if you're updating from 1.3.6 and you're getting a 'password authentication failed' error, consider uninstalling 1.3.6 first, then install 2.x.x as you'd normally do. ### Changelog [Please find the latest release notes and changelog here.](https://rengine.wiki/changelog/)  ### Screenshots #### Scan Results  #### General Usage <img src="https://user-images.githubusercontent.com/17223002/164993781-b6012995-522b-480a-a8bf-911193d35894.gif"> #### Initiating Subscan <img src="https://user-images.githubusercontent.com/17223002/164993749-1ad343d6-8ce7-43d6-aee7-b3add0321da7.gif"> #### Recon Data filtering <img src="https://user-images.githubusercontent.com/17223002/164993687-b63f3de8-e033-4ac0-808e-a2aa377d3cf8.gif"> #### Report Generation <img src="https://user-images.githubusercontent.com/17223002/164993689-c796c6cd-eb61-43f4-800d-08aba9740088.gif"> #### Toolbox <img src="https://user-images.githubusercontent.com/17223002/164993751-d687e88a-eb79-440f-9dc0-0ad006901620.gif"> #### Adding Custom tool in Tools Arsenal <img src="https://user-images.githubusercontent.com/17223002/164993670-466f6459-9499-498b-a9bd-526476d735a7.gif">  ### Contributing Contributions are what make the open-source community such an amazing place to learn, inspire and create. Every contributions you make is **greatly appreciated**. Your contributions can be as simple as fixing the indentation or UI, or as complex as adding new modules and features. See the [Contributing Guide](.github/CONTRIBUTING.md) to get started. You can also [join our Discord channel #development](https://discord.gg/JuhHdHTtwd) for any development related questions.  ### First-time Open Source contributors Please note that reNgine is beginner friendly. If you have never done open-source before, we encourage you to do so. **We will be happy and proud of your first PR ever.** You can start by resolving any [open issues](https://github.com/yogeshojha/rengine/issues).  ### reNgine Support Please do not use GitHub for support requests. Instead, [join our Discord channel #support](https://discord.gg/azv6fzhNCE).  ### Support and Sponsoring Over the past few years, I have been working hard on reNgine to add new features with the sole aim of making it the de facto standard for reconnaissance. I spend most of my free time and weekends working on reNgine. I do this in addition to my day job. I am happy to have received such overwhelming support from the community. But to keep this project alive, I am looking for financial support. | Paypal | Bitcoin | Ethereum | | :-------------------------------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | |[https://www.paypal.com/paypalme/yogeshojha11](https://www.paypal.com/paypalme/yogeshojha11) | `35AiKyNswNZ4TZUSdriHopSCjNMPi63BCX` | `0xe7A337Da6ff98A28513C26A7Fec8C9b42A63d346` OR * Add a [GitHub Star](https://github.com/yogeshojha/rengine) to the project. * Tweet about this project, or maybe blogs? * Maybe nominate me for [GitHub Stars?](https://stars.github.com/nominate/) * Join DigitalOcean using my [referral link](https://m.do.co/c/e353502d19fc) your profit is **$100** and I get $25 DO credit. This will help me test reNgine on VPS before I release any major features. It takes a considerable amount of time to add new features and make sure everything works. Donating is your way of saying: **reNgine is awesome**. Any support is greatly appreciated! Thank you!  ### reNgine Bug Bounty Program [](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine) Security researchers, welcome aboard! I'm excited to announce the reNgine bug bounty programme in collaboration with [huntr.dev](https://huntr.dev), which means that you will be rewarded for any vulnerabilities you find in reNgine. Thank you for your interest in reporting reNgine vulnerabilities! If you are aware of any potential security vulnerabilities in reNgine, we encourage you to report them immediately via [huntr.dev](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine). **Please do not disclose vulnerabilities via Github issues/blogs/tweets after/before reporting to huntr.dev as this is explicitly against the disclosure policy of huntr.dev and reNgine and will not be considered for monetary rewards.** Please note that the reNgine maintainer does not set the bounty amount. The bounty reward is determined by an industry-first equation developed by huntr.dev to understand the popularity, impact and value of repositories to the open-source community. **What do I expect from security researchers?** * Patience: Please note that I am currently the only maintainer in reNgine and it will take some time to validate your report. I ask for your patience during this process. * Respect for privacy and security reports: Please do not publicly disclose any vulnerabilities (including GitHub issues) before or after reporting them on huntr.dev! This is against the disclosure policy and will not be rewarded. * Respect the rules **What do you get in return?** * Thanks from the maintainer * Monetary rewards * CVE ID(s) Please find the [FAQ](https://www.huntr.dev/faq) and [Responsible disclosure policy](https://www.huntr.dev/policy/) from huntr.dev.  ### License Distributed under the GNU GPL v3 License. See [LICENSE](LICENSE) for more information.  <p align="right">(ChatGPT was used to write some or most part of this README section.)</p>

<p align="center"> <a href="https://rengine.wiki"><img src=".github/screenshots/banner.gif" alt=""/></a> </p> <p align="center"><a href="https://github.com/yogeshojha/rengine/releases" target="_blank"><img src="https://img.shields.io/badge/version-v2.0.0-informational?&logo=none" alt="reNgine Latest Version" /></a>&nbsp;<a href="https://www.gnu.org/licenses/gpl-3.0" target="_blank"><img src="https://img.shields.io/badge/License-GPLv3-red.svg?&logo=none" alt="License" /></a>&nbsp;<a href="#" target="_blank"><img src="https://img.shields.io/badge/first--timers--only-friendly-blue.svg?&logo=none" alt="" /></a>&nbsp;<a href="https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine" target="_blank"><img src="https://cdn.huntr.dev/huntr_security_badge_mono.svg" alt="" /></a>&nbsp;</p> <p align="center"> <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Asia-2023-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/Open--Source--Summit-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://cyberweek.ae/2021/hitb-armory/" target="_blank"><img src="https://img.shields.io/badge/HITB--Armory-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=7uvP6MaQOX0" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://drive.google.com/file/d/1Bh8lbf-Dztt5ViHJVACyrXMiglyICPQ2/view?usp=sharing" target="_blank"><img src="https://img.shields.io/badge/Defcon--Demolabs--29-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=A1oNOIc0h5A" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Europe-2020-blue.svg?&logo=none" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml/badge.svg" alt="" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/actions/workflows/build.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/build.yml/badge.svg" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://discord.gg/H6WzebwX3H" target="_blank"><img src="https://img.shields.io/discord/880363103689277461" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://opensourcesecurityindex.io/" target="_blank" rel="noopener"> <img style="width: 282px; height: 56px" src="https://opensourcesecurityindex.io/badge.svg" alt="Open Source Security Index - Fastest Growing Open Source Security Projects" width="282" height="56" /> </a> </p> <h3>reNgine 2.0-jasper<br>Redefining the future of reconnaissance!</h3> <h4>What is reNgine?</h4> <p align="left">reNgine is your go-to web application reconnaissance suite that's designed to simplify and streamline the reconnaissance process for security professionals, penetration testers, and bug bounty hunters. With its highly configurable engines, data correlation capabilities, continuous monitoring, database-backed reconnaissance data, and an intuitive user interface, reNgine redefines how you gather critical information about your target web applications. Traditional reconnaissance tools often fall short in terms of configurability and efficiency. reNgine addresses these shortcomings and emerges as a excellent alternative to existing commercial tools. reNgine was created to address the limitations of traditional reconnaissance tools and provide a better alternative, even surpassing some commercial offerings. Whether you're a bug bounty hunter, a penetration tester, or a corporate security team, reNgine is your go-to solution for automating and enhancing your information-gathering efforts. </p> reNgine 2.0-jasper is out now, you can [watch reNgine 2.0-jasper release trailer here!](https://youtu.be/VwkOWqiWW5g) reNgine 2.0-Jasper would not have been possible without [@ocervell](https://github.com/ocervell) valuable contributions. [@ocervell](https://github.com/ocervell) did majority of the refactoring if not all and also added a ton of features. Together, we wish to shape the future of web application reconnaissance, and it's developers like [@ocervell](https://github.com/ocervell) and a [ton of other developers and hackers from our community](https://github.com/yogeshojha/rengine/graphs/contributors) who inspire and drive us forward. Thank you, [@ocervell](https://github.com/ocervell), for your outstanding work and unwavering commitment to reNgine. Checkout our contributers here: [Contributers](https://github.com/yogeshojha/rengine/graphs/contributors)  ### Documentation You can find detailed documentation at [https://rengine.wiki](https://rengine.wiki)  ### Table of Contents * [About reNgine](#about-rengine) * [Workflow](#workflow) * [Features](#features) * [Scan Engine](#scan-engine) * [Quick Installation](#quick-installation) * [What's new in reNgine 2.0](#changelog) * [Screenshots](#screenshots) * [Contributing](#contributing) * [reNgine Support](#rengine-support) * [Support and Sponsoring](#support-and-sponsoring) * [reNgine Bug Bounty Program](#rengine-bug-bounty-program) * [License](#license)  ### About reNgine reNgine is not an ordinary reconnaissance suite; it's a game-changer! We've turbocharged the traditional workflow with groundbreaking features that is sure to ease your reconnaissance game. reNgine redefines the art of reconnaissance with highly configurable scan engines, recon data correlation, continuous monitoring, GPT powered Vulnerability Report, Project Management and role based access control etc. 🦾&nbsp;&nbsp; reNgine has advanced reconnaissance capabilities, harnessing a range of open-source tools to deliver a comprehensive web application reconnaissance experience. With it's intuitive User Interface, it excels in subdomain discovery, pinpointing IP addresses and open ports, collecting endpoints, conducting directory and file fuzzing, capturing screenshots, and performing vulnerability scans. To summarize, it does end-to-end reconnaissance. With WHOIS identification and WAF detection, it offers deep insights into target domains. Additionally, reNgine also identifies misconfigured S3 buckets and find interesting subdomains and URLS, based on specific keywords to helps you identify your next target, making it an go to tool for efficient reconnaissance. 🗃️&nbsp; &nbsp; Say goodbye to recon data chaos! reNgine seamlessly integrates with a database, providing you with unmatched data correlation and organization. Forgot the hassle of grepping through json, txt or csv files. Plus, our custom query language lets you filter reconnaissance data effortlessly using natural language like operators such as filtering all alive subdomains with `http_status=200` and also filter all subdomains that are alive and has admin in name `http_status=200&name=admin` 🔧&nbsp;&nbsp; reNgine offers unparalleled flexibility through its highly configurable scan engines, based on a YAML-based configuration. It offers the freedom to create and customize recon scan engines based on any kind of requirement, users can tailor them to their specific objectives and preferences, from thread management to timeout settings and rate-limit configurations, everything is customizable. Additionally, reNgine offers a range of pre-configured scan engines right out of the box, including Full Scan, Passive Scan, Screenshot Gathering, and the OSINT Scan Engine. These ready-to-use engines eliminate the need for extensive manual setup, aligning perfectly with reNgine's core mission of simplifying the reconnaissance process and enabling users to effortlessly access the right reconnaissance data with minimal effort. 💎&nbsp;&nbsp;Subscans: Subscan is a game-changing feature in reNgine, setting it apart as the only open-source tool of its kind to offer this capability. With Subscan, waiting for the entire pipeline to complete is a thing of the past. Now, users can swiftly respond to newfound discoveries during reconnaissance. Whether you've stumbled upon an intriguing subdomain and wish to conduct a focused port scan or want to delve deeper with a vulnerability assessment, reNgine has you covered. 📃&nbsp;&nbsp; PDF Reports: In addition to its robust reconnaissance capabilities, reNgine goes the extra mile by simplifying the report generation process, recognizing the crucial role that PDF reports play in the realm of end-to-end reconnaissance. Users can effortlessly generate and customize PDF reports to suit their exact needs. Whether it's a Full Scan Report, Vulnerability Report, or a concise reconnaissance report, reNgine provides the flexibility to choose the report type that best communicates your findings. Moreover, the level of customization is unparalleled, allowing users to select report colors, fine-tune executive summaries, and even add personalized touches like company names and footers. With GPT integration, your reports aren't just a report, with remediation steps, and impacts, you get 360-degree view of the vulnerabilities you've uncovered. 🔖&nbsp; &nbsp; Say Hello to Projects! reNgine 2.0 introduces a powerful addition that enables you to efficiently organize your web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. Each projects will have separate dashboard and all the scan results will be separated from each projects, while scan engines and configuration will be shared across all the projects. ⚙️&nbsp; &nbsp; Roles and Permissions! Begining reNgine 2.0, we've taken your web application reconnaissance to a whole new level of control and security. Now, you can assign distinct roles to your team members—Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - 🔐 Sys Admin: Sys Admin is a super user that has permission to modify system and scan related configurations, scan engines, create new users, add new tools etc. Super user can initiate scans and subscans effortlessly. - 🔍 Penetration Tester: Penetration Tester will be allowed to modify and initiate scans and subscans, add or update targets, etc. A penetration tester will not be allowed to modify system configurations. - 📊 Auditor: Auditor can only view and download the report. An auditor can not change any system or scan related configurations nor can initiate any scans or subscans. 🚀&nbsp;&nbsp; GPT Vulnerability Report Generation: Get ready for the future of penetration testing reports with reNgine's groundbreaking feature: "GPT-Powered Report Generation"! With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments that read like they were written by a human security expert! **But that's not all!** Our GPT-driven reports go the extra mile by scouring the web for related news articles, blogs, and references, so you have a 360-degree view of the vulnerabilities you've uncovered. With reNgine 2.0 revolutionize your penetration testing game and impress your clients with reports that are not just informative but engaging and comprehensive with detailed analysis on impact assessment and remediation strategies. 🥷&nbsp;&nbsp; GPT-Powered Attack Surface Generation: With reNgine 2.0, reNgine seamlessly integrates with GPT to identify the attacks that you can likely perform on a subdomain. By making use of reconnaissance data such as page title, open ports, subdomain name etc, reNgine can advice you the attacks you could perform on a target. reNgine will also provide you the rationale on why the specific attack is likely to be successful. 🧭&nbsp;&nbsp;Continuous monitoring: Continuous monitoring is at the core of reNgine's mission, and it's robust continuous monitoring feature ensures that their targets are under constant scrutiny. With the flexibility to schedule scans at regular intervals, penetration testers can effortlessly stay informed about their targets. What sets reNgine apart is its seamless integration with popular notification channels such as Discord, Slack, and Telegram, delivering real-time alerts for newly discovered subdomains, vulnerabilities, or any changes in reconnaissance data.  ### Workflow <img src="https://github.com/yogeshojha/rengine/assets/17223002/10c475b8-b4a8-440d-9126-77fe2038a386">  ### Features * Reconnaissance: * Subdomain Discovery * IP and Open Ports Identification * Endpoints Discovery * Directory/Files fuzzing * Screenshot Gathering * Vulnerability Scan * Nuclei * Dalfox XSS Scanner * CRLFuzzer * Misconfigured S3 Scanner * WHOIS Identification * WAF Detection * OSINT Capabilities * Meta info Gathering * Employees Gathering * Email Address gathering * Google Dorking for sensitive info and urls * Projects, create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. * Perform Advanced Query lookup using natural language alike and, or, not operations * Highly configurable YAML-based Scan Engines * Support for Parallel Scans * Support for Subscans * Recon Data visualization * GPT Vulnerability Description, Impact and Remediation generation * GPT Attack Surface Generator * Multiple Roles and Permissions to cater a team's need * Customizable Alerts/Notifications on Slack, Discord, and Telegram * Automatically report Vulnerabilities to HackerOne * Recon Notes and Todos * Clocked Scans (Run reconnaissance exactly at X Hours and Y minutes) and Periodic Scans (Runs reconnaissance every X minutes/- hours/days/week) * Proxy Support * Screenshot Gallery with Filters * Powerful recon data filtering with autosuggestions * Recon Data changes, find new/removed subdomains/endpoints * Tag targets into the Organization * Smart Duplicate endpoint removal based on page title and content length to cleanup the reconnaissance data * Identify Interesting Subdomains * Custom GF patterns and custom Nuclei Templates * Edit tool-related configuration files (Nuclei, Subfinder, Naabu, amass) * Add external tools from Github/Go * Interoperable with other tools, Import/Export Subdomains/Endpoints * Import Targets via IP and/or CIDRs * Report Generation * Toolbox: Comes bundled with most commonly used tools during penetration testing such as whois lookup, CMS detector, CVE lookup, etc. * Identification of related domains and related TLDs for targets * Find actionable insights such as Most Common Vulnerability, Most Common CVE ID, Most Vulnerable Target/Subdomain, etc.  ### Scan Engine ```yaml subdomain_discovery: { 'uses_tools': [ 'subfinder', 'ctfr', 'sublist3r', 'tlsx', 'oneforall', 'netlas' ], 'enable_http_crawl': true, 'threads': 30, 'timeout': 5, } http_crawl: {} port_scan: { 'enable_http_crawl': true, 'timeout': 5, # 'exclude_ports': [], # 'exclude_subdomains': true, 'ports': ['top-100'], 'rate_limit': 150, 'threads': 30, 'passive': false, # 'use_naabu_config': false, # 'enable_nmap': true, # 'nmap_cmd': '', # 'nmap_script': '', # 'nmap_script_args': '' } osint: { 'discover': [ 'emails', 'metainfo', 'employees' ], 'dorks': [ 'login_pages', 'admin_panels', 'dashboard_pages', 'stackoverflow', 'social_media', 'project_management', 'code_sharing', 'config_files', 'jenkins', 'wordpress_files', 'php_error', 'exposed_documents', 'db_files', 'git_exposed' ], 'custom_dorks': [ { 'lookup_site': 'google.com', 'lookup_keywords': '/home/' }, { 'lookup_site': '_target_', 'lookup_extensions': 'jpg,png' } ], 'intensity': 'normal', 'documents_limit': 50 } dir_file_fuzz: { 'auto_calibration': true, 'enable_http_crawl': true, 'rate_limit': 150, 'extensions': ['html', 'php','git','yaml','conf','cnf','config','gz','env','log','db','mysql','bak','asp','aspx','txt','conf','sql','json','yml','pdf'], 'follow_redirect': false, 'max_time': 0, 'match_http_status': [200, 204], 'recursive_level': 2, 'stop_on_error': false, 'timeout': 5, 'threads': 30, 'wordlist_name': 'dicc' } fetch_url: { 'uses_tools': [ 'gospider', 'hakrawler', 'waybackurls', 'gospider', 'katana' ], 'remove_duplicate_endpoints': true, 'duplicate_fields': [ 'content_length', 'page_title' ], 'enable_http_crawl': true, 'gf_patterns': ['debug_logic', 'idor', 'interestingEXT', 'interestingparams', 'interestingsubs', 'lfi', 'rce', 'redirect', 'sqli', 'ssrf', 'ssti', 'xss'], 'ignore_file_extensions': ['png', 'jpg', 'jpeg', 'gif', 'mp4', 'mpeg', 'mp3'] # 'exclude_subdomains': true } vulnerability_scan: { 'run_nuclei': false, 'run_dalfox': false, 'run_crlfuzz': false, 'run_s3scanner': true, 'enable_http_crawl': true, 'concurrency': 50, 'intensity': 'normal', 'rate_limit': 150, 'retries': 1, 'timeout': 5, 'fetch_gpt_report': true, 'nuclei': { 'use_conf': false, 'severities': [ 'unknown', 'info', 'low', 'medium', 'high', 'critical' ], # 'tags': [], # 'templates': [], # 'custom_templates': [], }, 's3scanner': { 'threads': 100, 'providers': [ 'aws', 'gcp', 'digitalocean', 'dreamhost', 'linode' ] } } waf_detection: {} screenshot: { 'enable_http_crawl': true, 'intensity': 'normal', 'timeout': 10, 'threads': 40 } # custom_header: "Cookie: Test" ```  ### Quick Installation **Note:** Only Ubuntu/VPS 1. Clone this repo ```bash git clone https://github.com/yogeshojha/rengine && cd rengine ``` 1. Edit the `.env` file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```bash nano .env ``` 1. **Optional, only for non-interactive install**: In the `.env` file, **please make sure to change the super admin values!** ```bash DJANGO_SUPERUSER_USERNAME=yourUsername DJANGO_SUPERUSER_EMAIL=YourMail@example.com DJANGO_SUPERUSER_PASSWORD=yourStrongPassword ``` If you need to carry out a non-interactive installation, you can setup the login, email and password of the web interface admin directly from the .env file (instead of manually setting them from prompts during the installation process). This option can be interesting for automated installation (via ansible, vagrant, etc.). `DJANGO_SUPERUSER_USERNAME`: web interface admin username (used to login to the web interface). `DJANGO_SUPERUSER_EMAIL`: web interface admin email. `DJANGO_SUPERUSER_PASSWORD`: web interface admin password (used to login to the web interface). 1. In the dotenv file, you may also modify the Scaling Configurations ```bash MAX_CONCURRENCY=80 MIN_CONCURRENCY=10 ``` `MAX_CONCURRENCY`: This parameter specifies the maximum number of reNgine's concurrent Celery worker processes that can be spawned. In this case, it's set to 80, meaning that the application can utilize up to 80 concurrent worker processes to execute tasks concurrently. This is useful for handling a high volume of scans or when you want to scale up processing power during periods of high demand. If you have more CPU cores, you will need to increase this for maximised performance. `MIN_CONCURRENCY`: On the other hand, MIN_CONCURRENCY specifies the minimum number of concurrent worker processes that should be maintained, even during periods of lower demand. In this example, it's set to 10, which means that even when there are fewer tasks to process, at least 10 worker processes will be kept running. This helps ensure that the application can respond promptly to incoming tasks without the overhead of repeatedly starting and stopping worker processes. These settings allow for dynamic scaling of Celery workers, ensuring that the application efficiently manages its workload by adjusting the number of concurrent workers based on the workload's size and complexity 1. Run the installation script, Please keep an eye for any prompt, you will also be asked for username and password for reNgine. ```bash sudo ./install.sh ``` Or for a non-interactive installation, use `-n` argument (make sure you've modified the `.env` file before launching the installation). ```bash sudo ./install.sh -n ``` If `install.sh` does not have install permission, please change it, `chmod +x install.sh` **reNgine can now be accessed from <https://127.0.0.1> or if you're on the VPS <https://your_vps_ip_address>** **Unless you are on development branch, please do not access reNgine via any ports** ### Installation (Mac/Windows/Other) Installation instructions can be found at [https://reNgine.wiki/install/detailed/](https://reNgine.wiki/2.0/install/detailed/) ### Updating 1. Updating is as simple as running the following command: ```bash cd rengine && sudo ./update.sh ``` If `update.sh` does not have execution permissions, please change it, `sudo chmod +x update.sh` **NOTE:** if you're updating from 1.3.6 and you're getting a 'password authentication failed' error, consider uninstalling 1.3.6 first, then install 2.x.x as you'd normally do. ### Changelog [Please find the latest release notes and changelog here.](https://rengine.wiki/changelog/)  ### Screenshots #### Scan Results  #### General Usage <img src="https://user-images.githubusercontent.com/17223002/164993781-b6012995-522b-480a-a8bf-911193d35894.gif"> #### Initiating Subscan <img src="https://user-images.githubusercontent.com/17223002/164993749-1ad343d6-8ce7-43d6-aee7-b3add0321da7.gif"> #### Recon Data filtering <img src="https://user-images.githubusercontent.com/17223002/164993687-b63f3de8-e033-4ac0-808e-a2aa377d3cf8.gif"> #### Report Generation <img src="https://user-images.githubusercontent.com/17223002/164993689-c796c6cd-eb61-43f4-800d-08aba9740088.gif"> #### Toolbox <img src="https://user-images.githubusercontent.com/17223002/164993751-d687e88a-eb79-440f-9dc0-0ad006901620.gif"> #### Adding Custom tool in Tools Arsenal <img src="https://user-images.githubusercontent.com/17223002/164993670-466f6459-9499-498b-a9bd-526476d735a7.gif">  ### Contributing Contributions are what make the open-source community such an amazing place to learn, inspire and create. Every contributions you make is **greatly appreciated**. Your contributions can be as simple as fixing the indentation or UI, or as complex as adding new modules and features. See the [Contributing Guide](.github/CONTRIBUTING.md) to get started. You can also [join our Discord channel #development](https://discord.gg/JuhHdHTtwd) for any development related questions.  ### First-time Open Source contributors Please note that reNgine is beginner friendly. If you have never done open-source before, we encourage you to do so. **We will be happy and proud of your first PR ever.** You can start by resolving any [open issues](https://github.com/yogeshojha/rengine/issues).  ### reNgine Support Please do not use GitHub for support requests. Instead, [join our Discord channel #support](https://discord.gg/azv6fzhNCE).  ### Support and Sponsoring Over the past few years, I have been working hard on reNgine to add new features with the sole aim of making it the de facto standard for reconnaissance. I spend most of my free time and weekends working on reNgine. I do this in addition to my day job. I am happy to have received such overwhelming support from the community. But to keep this project alive, I am looking for financial support. | Paypal | Bitcoin | Ethereum | | :-------------------------------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | |[https://www.paypal.com/paypalme/yogeshojha11](https://www.paypal.com/paypalme/yogeshojha11) | `35AiKyNswNZ4TZUSdriHopSCjNMPi63BCX` | `0xe7A337Da6ff98A28513C26A7Fec8C9b42A63d346` OR * Add a [GitHub Star](https://github.com/yogeshojha/rengine) to the project. * Tweet about this project, or maybe blogs? * Maybe nominate me for [GitHub Stars?](https://stars.github.com/nominate/) * Join DigitalOcean using my [referral link](https://m.do.co/c/e353502d19fc) your profit is **$100** and I get $25 DO credit. This will help me test reNgine on VPS before I release any major features. It takes a considerable amount of time to add new features and make sure everything works. Donating is your way of saying: **reNgine is awesome**. Any support is greatly appreciated! Thank you!  ### reNgine Bug Bounty Program [](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine) Security researchers, welcome aboard! I'm excited to announce the reNgine bug bounty programme in collaboration with [huntr.dev](https://huntr.dev), which means that you will be rewarded for any vulnerabilities you find in reNgine. Thank you for your interest in reporting reNgine vulnerabilities! If you are aware of any potential security vulnerabilities in reNgine, we encourage you to report them immediately via [huntr.dev](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine). **Please do not disclose vulnerabilities via Github issues/blogs/tweets after/before reporting to huntr.dev as this is explicitly against the disclosure policy of huntr.dev and reNgine and will not be considered for monetary rewards.** Please note that the reNgine maintainer does not set the bounty amount. The bounty reward is determined by an industry-first equation developed by huntr.dev to understand the popularity, impact and value of repositories to the open-source community. **What do I expect from security researchers?** * Patience: Please note that I am currently the only maintainer in reNgine and it will take some time to validate your report. I ask for your patience during this process. * Respect for privacy and security reports: Please do not publicly disclose any vulnerabilities (including GitHub issues) before or after reporting them on huntr.dev! This is against the disclosure policy and will not be rewarded. * Respect the rules **What do you get in return?** * Thanks from the maintainer * Monetary rewards * CVE ID(s) Please find the [FAQ](https://www.huntr.dev/faq) and [Responsible disclosure policy](https://www.huntr.dev/policy/) from huntr.dev.  ### License Distributed under the GNU GPL v3 License. See [LICENSE](LICENSE) for more information.  <p align="right">(ChatGPT was used to write some or most part of this README section.)</p>

C0wnuts

3dd700357a4bd5701b07ede4511f66042655be00

64b7f291240b3b8853e3cec7ee6230827c97b907

```suggestion 1. Edit the `.env` file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```

AnonymousWP

yogeshojha/rengine

Add non-interactive installation parameter

Add a non-interactive installation method via a new parameter to be passed to the install.sh script. Essential for automated/industrialized systems (e.g. via Ansible or another automated environment creation system).

2023-10-12 01:09:15+00:00

2023-11-21 12:49:22+00:00

README.md

<p align="center"> <a href="https://rengine.wiki"><img src=".github/screenshots/banner.gif" alt=""/></a> </p> <p align="center"><a href="https://github.com/yogeshojha/rengine/releases" target="_blank"><img src="https://img.shields.io/badge/version-v2.0.0-informational?&logo=none" alt="reNgine Latest Version" /></a>&nbsp;<a href="https://www.gnu.org/licenses/gpl-3.0" target="_blank"><img src="https://img.shields.io/badge/License-GPLv3-red.svg?&logo=none" alt="License" /></a>&nbsp;<a href="#" target="_blank"><img src="https://img.shields.io/badge/first--timers--only-friendly-blue.svg?&logo=none" alt="" /></a>&nbsp;<a href="https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine" target="_blank"><img src="https://cdn.huntr.dev/huntr_security_badge_mono.svg" alt="" /></a>&nbsp;</p> <p align="center"> <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Asia-2023-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/Open--Source--Summit-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://cyberweek.ae/2021/hitb-armory/" target="_blank"><img src="https://img.shields.io/badge/HITB--Armory-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=7uvP6MaQOX0" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://drive.google.com/file/d/1Bh8lbf-Dztt5ViHJVACyrXMiglyICPQ2/view?usp=sharing" target="_blank"><img src="https://img.shields.io/badge/Defcon--Demolabs--29-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=A1oNOIc0h5A" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Europe-2020-blue.svg?&logo=none" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml/badge.svg" alt="" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/actions/workflows/build.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/build.yml/badge.svg" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://discord.gg/H6WzebwX3H" target="_blank"><img src="https://img.shields.io/discord/880363103689277461" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://opensourcesecurityindex.io/" target="_blank" rel="noopener"> <img style="width: 282px; height: 56px" src="https://opensourcesecurityindex.io/badge.svg" alt="Open Source Security Index - Fastest Growing Open Source Security Projects" width="282" height="56" /> </a> </p> <h3>reNgine 2.0-jasper<br>Redefining the future of reconnaissance!</h3> <h4>What is reNgine?</h4> <p align="left">reNgine is your go-to web application reconnaissance suite that's designed to simplify and streamline the reconnaissance process for security professionals, penetration testers, and bug bounty hunters. With its highly configurable engines, data correlation capabilities, continuous monitoring, database-backed reconnaissance data, and an intuitive user interface, reNgine redefines how you gather critical information about your target web applications. Traditional reconnaissance tools often fall short in terms of configurability and efficiency. reNgine addresses these shortcomings and emerges as a excellent alternative to existing commercial tools. reNgine was created to address the limitations of traditional reconnaissance tools and provide a better alternative, even surpassing some commercial offerings. Whether you're a bug bounty hunter, a penetration tester, or a corporate security team, reNgine is your go-to solution for automating and enhancing your information-gathering efforts. </p> reNgine 2.0-jasper is out now, you can [watch reNgine 2.0-jasper release trailer here!](https://youtu.be/VwkOWqiWW5g) reNgine 2.0-Jasper would not have been possible without [@ocervell](https://github.com/ocervell) valuable contributions. [@ocervell](https://github.com/ocervell) did majority of the refactoring if not all and also added a ton of features. Together, we wish to shape the future of web application reconnaissance, and it's developers like [@ocervell](https://github.com/ocervell) and a [ton of other developers and hackers from our community](https://github.com/yogeshojha/rengine/graphs/contributors) who inspire and drive us forward. Thank you, [@ocervell](https://github.com/ocervell), for your outstanding work and unwavering commitment to reNgine. Checkout our contributers here: [Contributers](https://github.com/yogeshojha/rengine/graphs/contributors)  ### Documentation You can find detailed documentation at [https://rengine.wiki](https://rengine.wiki)  ### Table of Contents * [About reNgine](#about-rengine) * [Workflow](#workflow) * [Features](#features) * [Scan Engine](#scan-engine) * [Quick Installation](#quick-installation) * [What's new in reNgine 2.0](#changelog) * [Screenshots](#screenshots) * [Contributing](#contributing) * [reNgine Support](#rengine-support) * [Support and Sponsoring](#support-and-sponsoring) * [reNgine Bug Bounty Program](#rengine-bug-bounty-program) * [License](#license)  ### About reNgine reNgine is not an ordinary reconnaissance suite; it's a game-changer! We've turbocharged the traditional workflow with groundbreaking features that is sure to ease your reconnaissance game. reNgine redefines the art of reconnaissance with highly configurable scan engines, recon data correlation, continuous monitoring, GPT powered Vulnerability Report, Project Management and role based access control etc. 🦾&nbsp;&nbsp; reNgine has advanced reconnaissance capabilities, harnessing a range of open-source tools to deliver a comprehensive web application reconnaissance experience. With it's intuitive User Interface, it excels in subdomain discovery, pinpointing IP addresses and open ports, collecting endpoints, conducting directory and file fuzzing, capturing screenshots, and performing vulnerability scans. To summarize, it does end-to-end reconnaissance. With WHOIS identification and WAF detection, it offers deep insights into target domains. Additionally, reNgine also identifies misconfigured S3 buckets and find interesting subdomains and URLS, based on specific keywords to helps you identify your next target, making it an go to tool for efficient reconnaissance. 🗃️&nbsp; &nbsp; Say goodbye to recon data chaos! reNgine seamlessly integrates with a database, providing you with unmatched data correlation and organization. Forgot the hassle of grepping through json, txt or csv files. Plus, our custom query language lets you filter reconnaissance data effortlessly using natural language like operators such as filtering all alive subdomains with `http_status=200` and also filter all subdomains that are alive and has admin in name `http_status=200&name=admin` 🔧&nbsp;&nbsp; reNgine offers unparalleled flexibility through its highly configurable scan engines, based on a YAML-based configuration. It offers the freedom to create and customize recon scan engines based on any kind of requirement, users can tailor them to their specific objectives and preferences, from thread management to timeout settings and rate-limit configurations, everything is customizable. Additionally, reNgine offers a range of pre-configured scan engines right out of the box, including Full Scan, Passive Scan, Screenshot Gathering, and the OSINT Scan Engine. These ready-to-use engines eliminate the need for extensive manual setup, aligning perfectly with reNgine's core mission of simplifying the reconnaissance process and enabling users to effortlessly access the right reconnaissance data with minimal effort. 💎&nbsp;&nbsp;Subscans: Subscan is a game-changing feature in reNgine, setting it apart as the only open-source tool of its kind to offer this capability. With Subscan, waiting for the entire pipeline to complete is a thing of the past. Now, users can swiftly respond to newfound discoveries during reconnaissance. Whether you've stumbled upon an intriguing subdomain and wish to conduct a focused port scan or want to delve deeper with a vulnerability assessment, reNgine has you covered. 📃&nbsp;&nbsp; PDF Reports: In addition to its robust reconnaissance capabilities, reNgine goes the extra mile by simplifying the report generation process, recognizing the crucial role that PDF reports play in the realm of end-to-end reconnaissance. Users can effortlessly generate and customize PDF reports to suit their exact needs. Whether it's a Full Scan Report, Vulnerability Report, or a concise reconnaissance report, reNgine provides the flexibility to choose the report type that best communicates your findings. Moreover, the level of customization is unparalleled, allowing users to select report colors, fine-tune executive summaries, and even add personalized touches like company names and footers. With GPT integration, your reports aren't just a report, with remediation steps, and impacts, you get 360-degree view of the vulnerabilities you've uncovered. 🔖&nbsp; &nbsp; Say Hello to Projects! reNgine 2.0 introduces a powerful addition that enables you to efficiently organize your web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. Each projects will have separate dashboard and all the scan results will be separated from each projects, while scan engines and configuration will be shared across all the projects. ⚙️&nbsp; &nbsp; Roles and Permissions! Begining reNgine 2.0, we've taken your web application reconnaissance to a whole new level of control and security. Now, you can assign distinct roles to your team members—Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - 🔐 Sys Admin: Sys Admin is a super user that has permission to modify system and scan related configurations, scan engines, create new users, add new tools etc. Super user can initiate scans and subscans effortlessly. - 🔍 Penetration Tester: Penetration Tester will be allowed to modify and initiate scans and subscans, add or update targets, etc. A penetration tester will not be allowed to modify system configurations. - 📊 Auditor: Auditor can only view and download the report. An auditor can not change any system or scan related configurations nor can initiate any scans or subscans. 🚀&nbsp;&nbsp; GPT Vulnerability Report Generation: Get ready for the future of penetration testing reports with reNgine's groundbreaking feature: "GPT-Powered Report Generation"! With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments that read like they were written by a human security expert! **But that's not all!** Our GPT-driven reports go the extra mile by scouring the web for related news articles, blogs, and references, so you have a 360-degree view of the vulnerabilities you've uncovered. With reNgine 2.0 revolutionize your penetration testing game and impress your clients with reports that are not just informative but engaging and comprehensive with detailed analysis on impact assessment and remediation strategies. 🥷&nbsp;&nbsp; GPT-Powered Attack Surface Generation: With reNgine 2.0, reNgine seamlessly integrates with GPT to identify the attacks that you can likely perform on a subdomain. By making use of reconnaissance data such as page title, open ports, subdomain name etc, reNgine can advice you the attacks you could perform on a target. reNgine will also provide you the rationale on why the specific attack is likely to be successful. 🧭&nbsp;&nbsp;Continuous monitoring: Continuous monitoring is at the core of reNgine's mission, and it's robust continuous monitoring feature ensures that their targets are under constant scrutiny. With the flexibility to schedule scans at regular intervals, penetration testers can effortlessly stay informed about their targets. What sets reNgine apart is its seamless integration with popular notification channels such as Discord, Slack, and Telegram, delivering real-time alerts for newly discovered subdomains, vulnerabilities, or any changes in reconnaissance data.  ### Workflow <img src="https://github.com/yogeshojha/rengine/assets/17223002/10c475b8-b4a8-440d-9126-77fe2038a386">  ### Features * Reconnaissance: * Subdomain Discovery * IP and Open Ports Identification * Endpoints Discovery * Directory/Files fuzzing * Screenshot Gathering * Vulnerability Scan * Nuclei * Dalfox XSS Scanner * CRLFuzzer * Misconfigured S3 Scanner * WHOIS Identification * WAF Detection * OSINT Capabilities * Meta info Gathering * Employees Gathering * Email Address gathering * Google Dorking for sensitive info and urls * Projects, create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. * Perform Advanced Query lookup using natural language alike and, or, not operations * Highly configurable YAML-based Scan Engines * Support for Parallel Scans * Support for Subscans * Recon Data visualization * GPT Vulnerability Description, Impact and Remediation generation * GPT Attack Surface Generator * Multiple Roles and Permissions to cater a team's need * Customizable Alerts/Notifications on Slack, Discord, and Telegram * Automatically report Vulnerabilities to HackerOne * Recon Notes and Todos * Clocked Scans (Run reconnaissance exactly at X Hours and Y minutes) and Periodic Scans (Runs reconnaissance every X minutes/- hours/days/week) * Proxy Support * Screenshot Gallery with Filters * Powerful recon data filtering with autosuggestions * Recon Data changes, find new/removed subdomains/endpoints * Tag targets into the Organization * Smart Duplicate endpoint removal based on page title and content length to cleanup the reconnaissance data * Identify Interesting Subdomains * Custom GF patterns and custom Nuclei Templates * Edit tool-related configuration files (Nuclei, Subfinder, Naabu, amass) * Add external tools from Github/Go * Interoperable with other tools, Import/Export Subdomains/Endpoints * Import Targets via IP and/or CIDRs * Report Generation * Toolbox: Comes bundled with most commonly used tools during penetration testing such as whois lookup, CMS detector, CVE lookup, etc. * Identification of related domains and related TLDs for targets * Find actionable insights such as Most Common Vulnerability, Most Common CVE ID, Most Vulnerable Target/Subdomain, etc.  ### Scan Engine ```yaml subdomain_discovery: { 'uses_tools': [ 'subfinder', 'ctfr', 'sublist3r', 'tlsx', 'oneforall', 'netlas' ], 'enable_http_crawl': true, 'threads': 30, 'timeout': 5, } http_crawl: {} port_scan: { 'enable_http_crawl': true, 'timeout': 5, # 'exclude_ports': [], # 'exclude_subdomains': true, 'ports': ['top-100'], 'rate_limit': 150, 'threads': 30, 'passive': false, # 'use_naabu_config': false, # 'enable_nmap': true, # 'nmap_cmd': '', # 'nmap_script': '', # 'nmap_script_args': '' } osint: { 'discover': [ 'emails', 'metainfo', 'employees' ], 'dorks': [ 'login_pages', 'admin_panels', 'dashboard_pages', 'stackoverflow', 'social_media', 'project_management', 'code_sharing', 'config_files', 'jenkins', 'wordpress_files', 'php_error', 'exposed_documents', 'db_files', 'git_exposed' ], 'custom_dorks': [ { 'lookup_site': 'google.com', 'lookup_keywords': '/home/' }, { 'lookup_site': '_target_', 'lookup_extensions': 'jpg,png' } ], 'intensity': 'normal', 'documents_limit': 50 } dir_file_fuzz: { 'auto_calibration': true, 'enable_http_crawl': true, 'rate_limit': 150, 'extensions': ['html', 'php','git','yaml','conf','cnf','config','gz','env','log','db','mysql','bak','asp','aspx','txt','conf','sql','json','yml','pdf'], 'follow_redirect': false, 'max_time': 0, 'match_http_status': [200, 204], 'recursive_level': 2, 'stop_on_error': false, 'timeout': 5, 'threads': 30, 'wordlist_name': 'dicc' } fetch_url: { 'uses_tools': [ 'gospider', 'hakrawler', 'waybackurls', 'gospider', 'katana' ], 'remove_duplicate_endpoints': true, 'duplicate_fields': [ 'content_length', 'page_title' ], 'enable_http_crawl': true, 'gf_patterns': ['debug_logic', 'idor', 'interestingEXT', 'interestingparams', 'interestingsubs', 'lfi', 'rce', 'redirect', 'sqli', 'ssrf', 'ssti', 'xss'], 'ignore_file_extensions': ['png', 'jpg', 'jpeg', 'gif', 'mp4', 'mpeg', 'mp3'] # 'exclude_subdomains': true } vulnerability_scan: { 'run_nuclei': false, 'run_dalfox': false, 'run_crlfuzz': false, 'run_s3scanner': true, 'enable_http_crawl': true, 'concurrency': 50, 'intensity': 'normal', 'rate_limit': 150, 'retries': 1, 'timeout': 5, 'fetch_gpt_report': true, 'nuclei': { 'use_conf': false, 'severities': [ 'unknown', 'info', 'low', 'medium', 'high', 'critical' ], # 'tags': [], # 'templates': [], # 'custom_templates': [], }, 's3scanner': { 'threads': 100, 'providers': [ 'aws', 'gcp', 'digitalocean', 'dreamhost', 'linode' ] } } waf_detection: {} screenshot: { 'enable_http_crawl': true, 'intensity': 'normal', 'timeout': 10, 'threads': 40 } # custom_header: "Cookie: Test" ```  ### Quick Installation **Note:** Only Ubuntu/VPS 1. Clone this repo ```bash git clone https://github.com/yogeshojha/rengine && cd rengine ``` 1. Edit the dotenv file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```bash nano .env ``` 1. In the dotenv file, you may also modify the Scaling Configurations ```bash MAX_CONCURRENCY=80 MIN_CONCURRENCY=10 ``` MAX_CONCURRENCY: This parameter specifies the maximum number of reNgine's concurrent Celery worker processes that can be spawned. In this case, it's set to 80, meaning that the application can utilize up to 80 concurrent worker processes to execute tasks concurrently. This is useful for handling a high volume of scans or when you want to scale up processing power during periods of high demand. If you have more CPU cores, you will need to increase this for maximised performance. MIN_CONCURRENCY: On the other hand, MIN_CONCURRENCY specifies the minimum number of concurrent worker processes that should be maintained, even during periods of lower demand. In this example, it's set to 10, which means that even when there are fewer tasks to process, at least 10 worker processes will be kept running. This helps ensure that the application can respond promptly to incoming tasks without the overhead of repeatedly starting and stopping worker processes. These settings allow for dynamic scaling of Celery workers, ensuring that the application efficiently manages its workload by adjusting the number of concurrent workers based on the workload's size and complexity 1. Run the installation script, Please keep an eye for any prompt, you will also be asked for username and password for reNgine. ```bash sudo ./install.sh ``` If `install.sh` does not have install permission, please change it, `chmod +x install.sh` **reNgine can now be accessed from <https://127.0.0.1> or if you're on the VPS <https://your_vps_ip_address>** **Unless you are on development branch, please do not access reNgine via any ports** ### Installation (Mac/Windows/Other) Installation instructions can be found at [https://reNgine.wiki/install/detailed/](https://reNgine.wiki/2.0/install/detailed/) ### Updating 1. Updating is as simple as running the following command: ```bash cd rengine && sudo ./update.sh ``` If `update.sh` does not have execution permissions, please change it, `sudo chmod +x update.sh` **NOTE:** if you're updating from 1.3.6 and you're getting a 'password authentication failed' error, consider uninstalling 1.3.6 first, then install 2.x.x as you'd normally do. ### Changelog [Please find the latest release notes and changelog here.](https://rengine.wiki/changelog/)  ### Screenshots #### Scan Results  #### General Usage <img src="https://user-images.githubusercontent.com/17223002/164993781-b6012995-522b-480a-a8bf-911193d35894.gif"> #### Initiating Subscan <img src="https://user-images.githubusercontent.com/17223002/164993749-1ad343d6-8ce7-43d6-aee7-b3add0321da7.gif"> #### Recon Data filtering <img src="https://user-images.githubusercontent.com/17223002/164993687-b63f3de8-e033-4ac0-808e-a2aa377d3cf8.gif"> #### Report Generation <img src="https://user-images.githubusercontent.com/17223002/164993689-c796c6cd-eb61-43f4-800d-08aba9740088.gif"> #### Toolbox <img src="https://user-images.githubusercontent.com/17223002/164993751-d687e88a-eb79-440f-9dc0-0ad006901620.gif"> #### Adding Custom tool in Tools Arsenal <img src="https://user-images.githubusercontent.com/17223002/164993670-466f6459-9499-498b-a9bd-526476d735a7.gif">  ### Contributing Contributions are what make the open-source community such an amazing place to learn, inspire and create. Every contributions you make is **greatly appreciated**. Your contributions can be as simple as fixing the indentation or UI, or as complex as adding new modules and features. See the [Contributing Guide](.github/CONTRIBUTING.md) to get started. You can also [join our Discord channel #development](https://discord.gg/JuhHdHTtwd) for any development related questions.  ### First-time Open Source contributors Please note that reNgine is beginner friendly. If you have never done open-source before, we encourage you to do so. **We will be happy and proud of your first PR ever.** You can start by resolving any [open issues](https://github.com/yogeshojha/rengine/issues).  ### reNgine Support Please do not use GitHub for support requests. Instead, [join our Discord channel #support](https://discord.gg/azv6fzhNCE).  ### Support and Sponsoring Over the past few years, I have been working hard on reNgine to add new features with the sole aim of making it the de facto standard for reconnaissance. I spend most of my free time and weekends working on reNgine. I do this in addition to my day job. I am happy to have received such overwhelming support from the community. But to keep this project alive, I am looking for financial support. | Paypal | Bitcoin | Ethereum | | :-------------------------------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | |[https://www.paypal.com/paypalme/yogeshojha11](https://www.paypal.com/paypalme/yogeshojha11) | `35AiKyNswNZ4TZUSdriHopSCjNMPi63BCX` | `0xe7A337Da6ff98A28513C26A7Fec8C9b42A63d346` OR * Add a [GitHub Star](https://github.com/yogeshojha/rengine) to the project. * Tweet about this project, or maybe blogs? * Maybe nominate me for [GitHub Stars?](https://stars.github.com/nominate/) * Join DigitalOcean using my [referral link](https://m.do.co/c/e353502d19fc) your profit is **$100** and I get $25 DO credit. This will help me test reNgine on VPS before I release any major features. It takes a considerable amount of time to add new features and make sure everything works. Donating is your way of saying: **reNgine is awesome**. Any support is greatly appreciated! Thank you!  ### reNgine Bug Bounty Program [](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine) Security researchers, welcome aboard! I'm excited to announce the reNgine bug bounty programme in collaboration with [huntr.dev](https://huntr.dev), which means that you will be rewarded for any vulnerabilities you find in reNgine. Thank you for your interest in reporting reNgine vulnerabilities! If you are aware of any potential security vulnerabilities in reNgine, we encourage you to report them immediately via [huntr.dev](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine). **Please do not disclose vulnerabilities via Github issues/blogs/tweets after/before reporting to huntr.dev as this is explicitly against the disclosure policy of huntr.dev and reNgine and will not be considered for monetary rewards.** Please note that the reNgine maintainer does not set the bounty amount. The bounty reward is determined by an industry-first equation developed by huntr.dev to understand the popularity, impact and value of repositories to the open-source community. **What do I expect from security researchers?** * Patience: Please note that I am currently the only maintainer in reNgine and it will take some time to validate your report. I ask for your patience during this process. * Respect for privacy and security reports: Please do not publicly disclose any vulnerabilities (including GitHub issues) before or after reporting them on huntr.dev! This is against the disclosure policy and will not be rewarded. * Respect the rules **What do you get in return?** * Thanks from the maintainer * Monetary rewards * CVE ID(s) Please find the [FAQ](https://www.huntr.dev/faq) and [Responsible disclosure policy](https://www.huntr.dev/policy/) from huntr.dev.  ### License Distributed under the GNU GPL v3 License. See [LICENSE](LICENSE) for more information.  <p align="right">(ChatGPT was used to write some or most part of this README section.)</p>

<p align="center"> <a href="https://rengine.wiki"><img src=".github/screenshots/banner.gif" alt=""/></a> </p> <p align="center"><a href="https://github.com/yogeshojha/rengine/releases" target="_blank"><img src="https://img.shields.io/badge/version-v2.0.0-informational?&logo=none" alt="reNgine Latest Version" /></a>&nbsp;<a href="https://www.gnu.org/licenses/gpl-3.0" target="_blank"><img src="https://img.shields.io/badge/License-GPLv3-red.svg?&logo=none" alt="License" /></a>&nbsp;<a href="#" target="_blank"><img src="https://img.shields.io/badge/first--timers--only-friendly-blue.svg?&logo=none" alt="" /></a>&nbsp;<a href="https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine" target="_blank"><img src="https://cdn.huntr.dev/huntr_security_badge_mono.svg" alt="" /></a>&nbsp;</p> <p align="center"> <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Asia-2023-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/Open--Source--Summit-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://cyberweek.ae/2021/hitb-armory/" target="_blank"><img src="https://img.shields.io/badge/HITB--Armory-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=7uvP6MaQOX0" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://drive.google.com/file/d/1Bh8lbf-Dztt5ViHJVACyrXMiglyICPQ2/view?usp=sharing" target="_blank"><img src="https://img.shields.io/badge/Defcon--Demolabs--29-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=A1oNOIc0h5A" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Europe-2020-blue.svg?&logo=none" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml/badge.svg" alt="" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/actions/workflows/build.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/build.yml/badge.svg" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://discord.gg/H6WzebwX3H" target="_blank"><img src="https://img.shields.io/discord/880363103689277461" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://opensourcesecurityindex.io/" target="_blank" rel="noopener"> <img style="width: 282px; height: 56px" src="https://opensourcesecurityindex.io/badge.svg" alt="Open Source Security Index - Fastest Growing Open Source Security Projects" width="282" height="56" /> </a> </p> <h3>reNgine 2.0-jasper<br>Redefining the future of reconnaissance!</h3> <h4>What is reNgine?</h4> <p align="left">reNgine is your go-to web application reconnaissance suite that's designed to simplify and streamline the reconnaissance process for security professionals, penetration testers, and bug bounty hunters. With its highly configurable engines, data correlation capabilities, continuous monitoring, database-backed reconnaissance data, and an intuitive user interface, reNgine redefines how you gather critical information about your target web applications. Traditional reconnaissance tools often fall short in terms of configurability and efficiency. reNgine addresses these shortcomings and emerges as a excellent alternative to existing commercial tools. reNgine was created to address the limitations of traditional reconnaissance tools and provide a better alternative, even surpassing some commercial offerings. Whether you're a bug bounty hunter, a penetration tester, or a corporate security team, reNgine is your go-to solution for automating and enhancing your information-gathering efforts. </p> reNgine 2.0-jasper is out now, you can [watch reNgine 2.0-jasper release trailer here!](https://youtu.be/VwkOWqiWW5g) reNgine 2.0-Jasper would not have been possible without [@ocervell](https://github.com/ocervell) valuable contributions. [@ocervell](https://github.com/ocervell) did majority of the refactoring if not all and also added a ton of features. Together, we wish to shape the future of web application reconnaissance, and it's developers like [@ocervell](https://github.com/ocervell) and a [ton of other developers and hackers from our community](https://github.com/yogeshojha/rengine/graphs/contributors) who inspire and drive us forward. Thank you, [@ocervell](https://github.com/ocervell), for your outstanding work and unwavering commitment to reNgine. Checkout our contributers here: [Contributers](https://github.com/yogeshojha/rengine/graphs/contributors)  ### Documentation You can find detailed documentation at [https://rengine.wiki](https://rengine.wiki)  ### Table of Contents * [About reNgine](#about-rengine) * [Workflow](#workflow) * [Features](#features) * [Scan Engine](#scan-engine) * [Quick Installation](#quick-installation) * [What's new in reNgine 2.0](#changelog) * [Screenshots](#screenshots) * [Contributing](#contributing) * [reNgine Support](#rengine-support) * [Support and Sponsoring](#support-and-sponsoring) * [reNgine Bug Bounty Program](#rengine-bug-bounty-program) * [License](#license)  ### About reNgine reNgine is not an ordinary reconnaissance suite; it's a game-changer! We've turbocharged the traditional workflow with groundbreaking features that is sure to ease your reconnaissance game. reNgine redefines the art of reconnaissance with highly configurable scan engines, recon data correlation, continuous monitoring, GPT powered Vulnerability Report, Project Management and role based access control etc. 🦾&nbsp;&nbsp; reNgine has advanced reconnaissance capabilities, harnessing a range of open-source tools to deliver a comprehensive web application reconnaissance experience. With it's intuitive User Interface, it excels in subdomain discovery, pinpointing IP addresses and open ports, collecting endpoints, conducting directory and file fuzzing, capturing screenshots, and performing vulnerability scans. To summarize, it does end-to-end reconnaissance. With WHOIS identification and WAF detection, it offers deep insights into target domains. Additionally, reNgine also identifies misconfigured S3 buckets and find interesting subdomains and URLS, based on specific keywords to helps you identify your next target, making it an go to tool for efficient reconnaissance. 🗃️&nbsp; &nbsp; Say goodbye to recon data chaos! reNgine seamlessly integrates with a database, providing you with unmatched data correlation and organization. Forgot the hassle of grepping through json, txt or csv files. Plus, our custom query language lets you filter reconnaissance data effortlessly using natural language like operators such as filtering all alive subdomains with `http_status=200` and also filter all subdomains that are alive and has admin in name `http_status=200&name=admin` 🔧&nbsp;&nbsp; reNgine offers unparalleled flexibility through its highly configurable scan engines, based on a YAML-based configuration. It offers the freedom to create and customize recon scan engines based on any kind of requirement, users can tailor them to their specific objectives and preferences, from thread management to timeout settings and rate-limit configurations, everything is customizable. Additionally, reNgine offers a range of pre-configured scan engines right out of the box, including Full Scan, Passive Scan, Screenshot Gathering, and the OSINT Scan Engine. These ready-to-use engines eliminate the need for extensive manual setup, aligning perfectly with reNgine's core mission of simplifying the reconnaissance process and enabling users to effortlessly access the right reconnaissance data with minimal effort. 💎&nbsp;&nbsp;Subscans: Subscan is a game-changing feature in reNgine, setting it apart as the only open-source tool of its kind to offer this capability. With Subscan, waiting for the entire pipeline to complete is a thing of the past. Now, users can swiftly respond to newfound discoveries during reconnaissance. Whether you've stumbled upon an intriguing subdomain and wish to conduct a focused port scan or want to delve deeper with a vulnerability assessment, reNgine has you covered. 📃&nbsp;&nbsp; PDF Reports: In addition to its robust reconnaissance capabilities, reNgine goes the extra mile by simplifying the report generation process, recognizing the crucial role that PDF reports play in the realm of end-to-end reconnaissance. Users can effortlessly generate and customize PDF reports to suit their exact needs. Whether it's a Full Scan Report, Vulnerability Report, or a concise reconnaissance report, reNgine provides the flexibility to choose the report type that best communicates your findings. Moreover, the level of customization is unparalleled, allowing users to select report colors, fine-tune executive summaries, and even add personalized touches like company names and footers. With GPT integration, your reports aren't just a report, with remediation steps, and impacts, you get 360-degree view of the vulnerabilities you've uncovered. 🔖&nbsp; &nbsp; Say Hello to Projects! reNgine 2.0 introduces a powerful addition that enables you to efficiently organize your web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. Each projects will have separate dashboard and all the scan results will be separated from each projects, while scan engines and configuration will be shared across all the projects. ⚙️&nbsp; &nbsp; Roles and Permissions! Begining reNgine 2.0, we've taken your web application reconnaissance to a whole new level of control and security. Now, you can assign distinct roles to your team members—Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - 🔐 Sys Admin: Sys Admin is a super user that has permission to modify system and scan related configurations, scan engines, create new users, add new tools etc. Super user can initiate scans and subscans effortlessly. - 🔍 Penetration Tester: Penetration Tester will be allowed to modify and initiate scans and subscans, add or update targets, etc. A penetration tester will not be allowed to modify system configurations. - 📊 Auditor: Auditor can only view and download the report. An auditor can not change any system or scan related configurations nor can initiate any scans or subscans. 🚀&nbsp;&nbsp; GPT Vulnerability Report Generation: Get ready for the future of penetration testing reports with reNgine's groundbreaking feature: "GPT-Powered Report Generation"! With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments that read like they were written by a human security expert! **But that's not all!** Our GPT-driven reports go the extra mile by scouring the web for related news articles, blogs, and references, so you have a 360-degree view of the vulnerabilities you've uncovered. With reNgine 2.0 revolutionize your penetration testing game and impress your clients with reports that are not just informative but engaging and comprehensive with detailed analysis on impact assessment and remediation strategies. 🥷&nbsp;&nbsp; GPT-Powered Attack Surface Generation: With reNgine 2.0, reNgine seamlessly integrates with GPT to identify the attacks that you can likely perform on a subdomain. By making use of reconnaissance data such as page title, open ports, subdomain name etc, reNgine can advice you the attacks you could perform on a target. reNgine will also provide you the rationale on why the specific attack is likely to be successful. 🧭&nbsp;&nbsp;Continuous monitoring: Continuous monitoring is at the core of reNgine's mission, and it's robust continuous monitoring feature ensures that their targets are under constant scrutiny. With the flexibility to schedule scans at regular intervals, penetration testers can effortlessly stay informed about their targets. What sets reNgine apart is its seamless integration with popular notification channels such as Discord, Slack, and Telegram, delivering real-time alerts for newly discovered subdomains, vulnerabilities, or any changes in reconnaissance data.  ### Workflow <img src="https://github.com/yogeshojha/rengine/assets/17223002/10c475b8-b4a8-440d-9126-77fe2038a386">  ### Features * Reconnaissance: * Subdomain Discovery * IP and Open Ports Identification * Endpoints Discovery * Directory/Files fuzzing * Screenshot Gathering * Vulnerability Scan * Nuclei * Dalfox XSS Scanner * CRLFuzzer * Misconfigured S3 Scanner * WHOIS Identification * WAF Detection * OSINT Capabilities * Meta info Gathering * Employees Gathering * Email Address gathering * Google Dorking for sensitive info and urls * Projects, create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. * Perform Advanced Query lookup using natural language alike and, or, not operations * Highly configurable YAML-based Scan Engines * Support for Parallel Scans * Support for Subscans * Recon Data visualization * GPT Vulnerability Description, Impact and Remediation generation * GPT Attack Surface Generator * Multiple Roles and Permissions to cater a team's need * Customizable Alerts/Notifications on Slack, Discord, and Telegram * Automatically report Vulnerabilities to HackerOne * Recon Notes and Todos * Clocked Scans (Run reconnaissance exactly at X Hours and Y minutes) and Periodic Scans (Runs reconnaissance every X minutes/- hours/days/week) * Proxy Support * Screenshot Gallery with Filters * Powerful recon data filtering with autosuggestions * Recon Data changes, find new/removed subdomains/endpoints * Tag targets into the Organization * Smart Duplicate endpoint removal based on page title and content length to cleanup the reconnaissance data * Identify Interesting Subdomains * Custom GF patterns and custom Nuclei Templates * Edit tool-related configuration files (Nuclei, Subfinder, Naabu, amass) * Add external tools from Github/Go * Interoperable with other tools, Import/Export Subdomains/Endpoints * Import Targets via IP and/or CIDRs * Report Generation * Toolbox: Comes bundled with most commonly used tools during penetration testing such as whois lookup, CMS detector, CVE lookup, etc. * Identification of related domains and related TLDs for targets * Find actionable insights such as Most Common Vulnerability, Most Common CVE ID, Most Vulnerable Target/Subdomain, etc.  ### Scan Engine ```yaml subdomain_discovery: { 'uses_tools': [ 'subfinder', 'ctfr', 'sublist3r', 'tlsx', 'oneforall', 'netlas' ], 'enable_http_crawl': true, 'threads': 30, 'timeout': 5, } http_crawl: {} port_scan: { 'enable_http_crawl': true, 'timeout': 5, # 'exclude_ports': [], # 'exclude_subdomains': true, 'ports': ['top-100'], 'rate_limit': 150, 'threads': 30, 'passive': false, # 'use_naabu_config': false, # 'enable_nmap': true, # 'nmap_cmd': '', # 'nmap_script': '', # 'nmap_script_args': '' } osint: { 'discover': [ 'emails', 'metainfo', 'employees' ], 'dorks': [ 'login_pages', 'admin_panels', 'dashboard_pages', 'stackoverflow', 'social_media', 'project_management', 'code_sharing', 'config_files', 'jenkins', 'wordpress_files', 'php_error', 'exposed_documents', 'db_files', 'git_exposed' ], 'custom_dorks': [ { 'lookup_site': 'google.com', 'lookup_keywords': '/home/' }, { 'lookup_site': '_target_', 'lookup_extensions': 'jpg,png' } ], 'intensity': 'normal', 'documents_limit': 50 } dir_file_fuzz: { 'auto_calibration': true, 'enable_http_crawl': true, 'rate_limit': 150, 'extensions': ['html', 'php','git','yaml','conf','cnf','config','gz','env','log','db','mysql','bak','asp','aspx','txt','conf','sql','json','yml','pdf'], 'follow_redirect': false, 'max_time': 0, 'match_http_status': [200, 204], 'recursive_level': 2, 'stop_on_error': false, 'timeout': 5, 'threads': 30, 'wordlist_name': 'dicc' } fetch_url: { 'uses_tools': [ 'gospider', 'hakrawler', 'waybackurls', 'gospider', 'katana' ], 'remove_duplicate_endpoints': true, 'duplicate_fields': [ 'content_length', 'page_title' ], 'enable_http_crawl': true, 'gf_patterns': ['debug_logic', 'idor', 'interestingEXT', 'interestingparams', 'interestingsubs', 'lfi', 'rce', 'redirect', 'sqli', 'ssrf', 'ssti', 'xss'], 'ignore_file_extensions': ['png', 'jpg', 'jpeg', 'gif', 'mp4', 'mpeg', 'mp3'] # 'exclude_subdomains': true } vulnerability_scan: { 'run_nuclei': false, 'run_dalfox': false, 'run_crlfuzz': false, 'run_s3scanner': true, 'enable_http_crawl': true, 'concurrency': 50, 'intensity': 'normal', 'rate_limit': 150, 'retries': 1, 'timeout': 5, 'fetch_gpt_report': true, 'nuclei': { 'use_conf': false, 'severities': [ 'unknown', 'info', 'low', 'medium', 'high', 'critical' ], # 'tags': [], # 'templates': [], # 'custom_templates': [], }, 's3scanner': { 'threads': 100, 'providers': [ 'aws', 'gcp', 'digitalocean', 'dreamhost', 'linode' ] } } waf_detection: {} screenshot: { 'enable_http_crawl': true, 'intensity': 'normal', 'timeout': 10, 'threads': 40 } # custom_header: "Cookie: Test" ```  ### Quick Installation **Note:** Only Ubuntu/VPS 1. Clone this repo ```bash git clone https://github.com/yogeshojha/rengine && cd rengine ``` 1. Edit the `.env` file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```bash nano .env ``` 1. **Optional, only for non-interactive install**: In the `.env` file, **please make sure to change the super admin values!** ```bash DJANGO_SUPERUSER_USERNAME=yourUsername DJANGO_SUPERUSER_EMAIL=YourMail@example.com DJANGO_SUPERUSER_PASSWORD=yourStrongPassword ``` If you need to carry out a non-interactive installation, you can setup the login, email and password of the web interface admin directly from the .env file (instead of manually setting them from prompts during the installation process). This option can be interesting for automated installation (via ansible, vagrant, etc.). `DJANGO_SUPERUSER_USERNAME`: web interface admin username (used to login to the web interface). `DJANGO_SUPERUSER_EMAIL`: web interface admin email. `DJANGO_SUPERUSER_PASSWORD`: web interface admin password (used to login to the web interface). 1. In the dotenv file, you may also modify the Scaling Configurations ```bash MAX_CONCURRENCY=80 MIN_CONCURRENCY=10 ``` `MAX_CONCURRENCY`: This parameter specifies the maximum number of reNgine's concurrent Celery worker processes that can be spawned. In this case, it's set to 80, meaning that the application can utilize up to 80 concurrent worker processes to execute tasks concurrently. This is useful for handling a high volume of scans or when you want to scale up processing power during periods of high demand. If you have more CPU cores, you will need to increase this for maximised performance. `MIN_CONCURRENCY`: On the other hand, MIN_CONCURRENCY specifies the minimum number of concurrent worker processes that should be maintained, even during periods of lower demand. In this example, it's set to 10, which means that even when there are fewer tasks to process, at least 10 worker processes will be kept running. This helps ensure that the application can respond promptly to incoming tasks without the overhead of repeatedly starting and stopping worker processes. These settings allow for dynamic scaling of Celery workers, ensuring that the application efficiently manages its workload by adjusting the number of concurrent workers based on the workload's size and complexity 1. Run the installation script, Please keep an eye for any prompt, you will also be asked for username and password for reNgine. ```bash sudo ./install.sh ``` Or for a non-interactive installation, use `-n` argument (make sure you've modified the `.env` file before launching the installation). ```bash sudo ./install.sh -n ``` If `install.sh` does not have install permission, please change it, `chmod +x install.sh` **reNgine can now be accessed from <https://127.0.0.1> or if you're on the VPS <https://your_vps_ip_address>** **Unless you are on development branch, please do not access reNgine via any ports** ### Installation (Mac/Windows/Other) Installation instructions can be found at [https://reNgine.wiki/install/detailed/](https://reNgine.wiki/2.0/install/detailed/) ### Updating 1. Updating is as simple as running the following command: ```bash cd rengine && sudo ./update.sh ``` If `update.sh` does not have execution permissions, please change it, `sudo chmod +x update.sh` **NOTE:** if you're updating from 1.3.6 and you're getting a 'password authentication failed' error, consider uninstalling 1.3.6 first, then install 2.x.x as you'd normally do. ### Changelog [Please find the latest release notes and changelog here.](https://rengine.wiki/changelog/)  ### Screenshots #### Scan Results  #### General Usage <img src="https://user-images.githubusercontent.com/17223002/164993781-b6012995-522b-480a-a8bf-911193d35894.gif"> #### Initiating Subscan <img src="https://user-images.githubusercontent.com/17223002/164993749-1ad343d6-8ce7-43d6-aee7-b3add0321da7.gif"> #### Recon Data filtering <img src="https://user-images.githubusercontent.com/17223002/164993687-b63f3de8-e033-4ac0-808e-a2aa377d3cf8.gif"> #### Report Generation <img src="https://user-images.githubusercontent.com/17223002/164993689-c796c6cd-eb61-43f4-800d-08aba9740088.gif"> #### Toolbox <img src="https://user-images.githubusercontent.com/17223002/164993751-d687e88a-eb79-440f-9dc0-0ad006901620.gif"> #### Adding Custom tool in Tools Arsenal <img src="https://user-images.githubusercontent.com/17223002/164993670-466f6459-9499-498b-a9bd-526476d735a7.gif">  ### Contributing Contributions are what make the open-source community such an amazing place to learn, inspire and create. Every contributions you make is **greatly appreciated**. Your contributions can be as simple as fixing the indentation or UI, or as complex as adding new modules and features. See the [Contributing Guide](.github/CONTRIBUTING.md) to get started. You can also [join our Discord channel #development](https://discord.gg/JuhHdHTtwd) for any development related questions.  ### First-time Open Source contributors Please note that reNgine is beginner friendly. If you have never done open-source before, we encourage you to do so. **We will be happy and proud of your first PR ever.** You can start by resolving any [open issues](https://github.com/yogeshojha/rengine/issues).  ### reNgine Support Please do not use GitHub for support requests. Instead, [join our Discord channel #support](https://discord.gg/azv6fzhNCE).  ### Support and Sponsoring Over the past few years, I have been working hard on reNgine to add new features with the sole aim of making it the de facto standard for reconnaissance. I spend most of my free time and weekends working on reNgine. I do this in addition to my day job. I am happy to have received such overwhelming support from the community. But to keep this project alive, I am looking for financial support. | Paypal | Bitcoin | Ethereum | | :-------------------------------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | |[https://www.paypal.com/paypalme/yogeshojha11](https://www.paypal.com/paypalme/yogeshojha11) | `35AiKyNswNZ4TZUSdriHopSCjNMPi63BCX` | `0xe7A337Da6ff98A28513C26A7Fec8C9b42A63d346` OR * Add a [GitHub Star](https://github.com/yogeshojha/rengine) to the project. * Tweet about this project, or maybe blogs? * Maybe nominate me for [GitHub Stars?](https://stars.github.com/nominate/) * Join DigitalOcean using my [referral link](https://m.do.co/c/e353502d19fc) your profit is **$100** and I get $25 DO credit. This will help me test reNgine on VPS before I release any major features. It takes a considerable amount of time to add new features and make sure everything works. Donating is your way of saying: **reNgine is awesome**. Any support is greatly appreciated! Thank you!  ### reNgine Bug Bounty Program [](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine) Security researchers, welcome aboard! I'm excited to announce the reNgine bug bounty programme in collaboration with [huntr.dev](https://huntr.dev), which means that you will be rewarded for any vulnerabilities you find in reNgine. Thank you for your interest in reporting reNgine vulnerabilities! If you are aware of any potential security vulnerabilities in reNgine, we encourage you to report them immediately via [huntr.dev](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine). **Please do not disclose vulnerabilities via Github issues/blogs/tweets after/before reporting to huntr.dev as this is explicitly against the disclosure policy of huntr.dev and reNgine and will not be considered for monetary rewards.** Please note that the reNgine maintainer does not set the bounty amount. The bounty reward is determined by an industry-first equation developed by huntr.dev to understand the popularity, impact and value of repositories to the open-source community. **What do I expect from security researchers?** * Patience: Please note that I am currently the only maintainer in reNgine and it will take some time to validate your report. I ask for your patience during this process. * Respect for privacy and security reports: Please do not publicly disclose any vulnerabilities (including GitHub issues) before or after reporting them on huntr.dev! This is against the disclosure policy and will not be rewarded. * Respect the rules **What do you get in return?** * Thanks from the maintainer * Monetary rewards * CVE ID(s) Please find the [FAQ](https://www.huntr.dev/faq) and [Responsible disclosure policy](https://www.huntr.dev/policy/) from huntr.dev.  ### License Distributed under the GNU GPL v3 License. See [LICENSE](LICENSE) for more information.  <p align="right">(ChatGPT was used to write some or most part of this README section.)</p>

C0wnuts

3dd700357a4bd5701b07ede4511f66042655be00

64b7f291240b3b8853e3cec7ee6230827c97b907

```suggestion 1. **Optional, only for non-interactive install**: In the `.env` file, **please make sure to change the super admin values!** ```

AnonymousWP

yogeshojha/rengine

Add non-interactive installation parameter

Add a non-interactive installation method via a new parameter to be passed to the install.sh script. Essential for automated/industrialized systems (e.g. via Ansible or another automated environment creation system).

2023-10-12 01:09:15+00:00

2023-11-21 12:49:22+00:00

README.md

<p align="center"> <a href="https://rengine.wiki"><img src=".github/screenshots/banner.gif" alt=""/></a> </p> <p align="center"><a href="https://github.com/yogeshojha/rengine/releases" target="_blank"><img src="https://img.shields.io/badge/version-v2.0.0-informational?&logo=none" alt="reNgine Latest Version" /></a>&nbsp;<a href="https://www.gnu.org/licenses/gpl-3.0" target="_blank"><img src="https://img.shields.io/badge/License-GPLv3-red.svg?&logo=none" alt="License" /></a>&nbsp;<a href="#" target="_blank"><img src="https://img.shields.io/badge/first--timers--only-friendly-blue.svg?&logo=none" alt="" /></a>&nbsp;<a href="https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine" target="_blank"><img src="https://cdn.huntr.dev/huntr_security_badge_mono.svg" alt="" /></a>&nbsp;</p> <p align="center"> <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Asia-2023-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/Open--Source--Summit-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://cyberweek.ae/2021/hitb-armory/" target="_blank"><img src="https://img.shields.io/badge/HITB--Armory-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=7uvP6MaQOX0" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://drive.google.com/file/d/1Bh8lbf-Dztt5ViHJVACyrXMiglyICPQ2/view?usp=sharing" target="_blank"><img src="https://img.shields.io/badge/Defcon--Demolabs--29-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=A1oNOIc0h5A" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Europe-2020-blue.svg?&logo=none" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml/badge.svg" alt="" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/actions/workflows/build.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/build.yml/badge.svg" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://discord.gg/H6WzebwX3H" target="_blank"><img src="https://img.shields.io/discord/880363103689277461" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://opensourcesecurityindex.io/" target="_blank" rel="noopener"> <img style="width: 282px; height: 56px" src="https://opensourcesecurityindex.io/badge.svg" alt="Open Source Security Index - Fastest Growing Open Source Security Projects" width="282" height="56" /> </a> </p> <h3>reNgine 2.0-jasper<br>Redefining the future of reconnaissance!</h3> <h4>What is reNgine?</h4> <p align="left">reNgine is your go-to web application reconnaissance suite that's designed to simplify and streamline the reconnaissance process for security professionals, penetration testers, and bug bounty hunters. With its highly configurable engines, data correlation capabilities, continuous monitoring, database-backed reconnaissance data, and an intuitive user interface, reNgine redefines how you gather critical information about your target web applications. Traditional reconnaissance tools often fall short in terms of configurability and efficiency. reNgine addresses these shortcomings and emerges as a excellent alternative to existing commercial tools. reNgine was created to address the limitations of traditional reconnaissance tools and provide a better alternative, even surpassing some commercial offerings. Whether you're a bug bounty hunter, a penetration tester, or a corporate security team, reNgine is your go-to solution for automating and enhancing your information-gathering efforts. </p> reNgine 2.0-jasper is out now, you can [watch reNgine 2.0-jasper release trailer here!](https://youtu.be/VwkOWqiWW5g) reNgine 2.0-Jasper would not have been possible without [@ocervell](https://github.com/ocervell) valuable contributions. [@ocervell](https://github.com/ocervell) did majority of the refactoring if not all and also added a ton of features. Together, we wish to shape the future of web application reconnaissance, and it's developers like [@ocervell](https://github.com/ocervell) and a [ton of other developers and hackers from our community](https://github.com/yogeshojha/rengine/graphs/contributors) who inspire and drive us forward. Thank you, [@ocervell](https://github.com/ocervell), for your outstanding work and unwavering commitment to reNgine. Checkout our contributers here: [Contributers](https://github.com/yogeshojha/rengine/graphs/contributors)  ### Documentation You can find detailed documentation at [https://rengine.wiki](https://rengine.wiki)  ### Table of Contents * [About reNgine](#about-rengine) * [Workflow](#workflow) * [Features](#features) * [Scan Engine](#scan-engine) * [Quick Installation](#quick-installation) * [What's new in reNgine 2.0](#changelog) * [Screenshots](#screenshots) * [Contributing](#contributing) * [reNgine Support](#rengine-support) * [Support and Sponsoring](#support-and-sponsoring) * [reNgine Bug Bounty Program](#rengine-bug-bounty-program) * [License](#license)  ### About reNgine reNgine is not an ordinary reconnaissance suite; it's a game-changer! We've turbocharged the traditional workflow with groundbreaking features that is sure to ease your reconnaissance game. reNgine redefines the art of reconnaissance with highly configurable scan engines, recon data correlation, continuous monitoring, GPT powered Vulnerability Report, Project Management and role based access control etc. 🦾&nbsp;&nbsp; reNgine has advanced reconnaissance capabilities, harnessing a range of open-source tools to deliver a comprehensive web application reconnaissance experience. With it's intuitive User Interface, it excels in subdomain discovery, pinpointing IP addresses and open ports, collecting endpoints, conducting directory and file fuzzing, capturing screenshots, and performing vulnerability scans. To summarize, it does end-to-end reconnaissance. With WHOIS identification and WAF detection, it offers deep insights into target domains. Additionally, reNgine also identifies misconfigured S3 buckets and find interesting subdomains and URLS, based on specific keywords to helps you identify your next target, making it an go to tool for efficient reconnaissance. 🗃️&nbsp; &nbsp; Say goodbye to recon data chaos! reNgine seamlessly integrates with a database, providing you with unmatched data correlation and organization. Forgot the hassle of grepping through json, txt or csv files. Plus, our custom query language lets you filter reconnaissance data effortlessly using natural language like operators such as filtering all alive subdomains with `http_status=200` and also filter all subdomains that are alive and has admin in name `http_status=200&name=admin` 🔧&nbsp;&nbsp; reNgine offers unparalleled flexibility through its highly configurable scan engines, based on a YAML-based configuration. It offers the freedom to create and customize recon scan engines based on any kind of requirement, users can tailor them to their specific objectives and preferences, from thread management to timeout settings and rate-limit configurations, everything is customizable. Additionally, reNgine offers a range of pre-configured scan engines right out of the box, including Full Scan, Passive Scan, Screenshot Gathering, and the OSINT Scan Engine. These ready-to-use engines eliminate the need for extensive manual setup, aligning perfectly with reNgine's core mission of simplifying the reconnaissance process and enabling users to effortlessly access the right reconnaissance data with minimal effort. 💎&nbsp;&nbsp;Subscans: Subscan is a game-changing feature in reNgine, setting it apart as the only open-source tool of its kind to offer this capability. With Subscan, waiting for the entire pipeline to complete is a thing of the past. Now, users can swiftly respond to newfound discoveries during reconnaissance. Whether you've stumbled upon an intriguing subdomain and wish to conduct a focused port scan or want to delve deeper with a vulnerability assessment, reNgine has you covered. 📃&nbsp;&nbsp; PDF Reports: In addition to its robust reconnaissance capabilities, reNgine goes the extra mile by simplifying the report generation process, recognizing the crucial role that PDF reports play in the realm of end-to-end reconnaissance. Users can effortlessly generate and customize PDF reports to suit their exact needs. Whether it's a Full Scan Report, Vulnerability Report, or a concise reconnaissance report, reNgine provides the flexibility to choose the report type that best communicates your findings. Moreover, the level of customization is unparalleled, allowing users to select report colors, fine-tune executive summaries, and even add personalized touches like company names and footers. With GPT integration, your reports aren't just a report, with remediation steps, and impacts, you get 360-degree view of the vulnerabilities you've uncovered. 🔖&nbsp; &nbsp; Say Hello to Projects! reNgine 2.0 introduces a powerful addition that enables you to efficiently organize your web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. Each projects will have separate dashboard and all the scan results will be separated from each projects, while scan engines and configuration will be shared across all the projects. ⚙️&nbsp; &nbsp; Roles and Permissions! Begining reNgine 2.0, we've taken your web application reconnaissance to a whole new level of control and security. Now, you can assign distinct roles to your team members—Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - 🔐 Sys Admin: Sys Admin is a super user that has permission to modify system and scan related configurations, scan engines, create new users, add new tools etc. Super user can initiate scans and subscans effortlessly. - 🔍 Penetration Tester: Penetration Tester will be allowed to modify and initiate scans and subscans, add or update targets, etc. A penetration tester will not be allowed to modify system configurations. - 📊 Auditor: Auditor can only view and download the report. An auditor can not change any system or scan related configurations nor can initiate any scans or subscans. 🚀&nbsp;&nbsp; GPT Vulnerability Report Generation: Get ready for the future of penetration testing reports with reNgine's groundbreaking feature: "GPT-Powered Report Generation"! With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments that read like they were written by a human security expert! **But that's not all!** Our GPT-driven reports go the extra mile by scouring the web for related news articles, blogs, and references, so you have a 360-degree view of the vulnerabilities you've uncovered. With reNgine 2.0 revolutionize your penetration testing game and impress your clients with reports that are not just informative but engaging and comprehensive with detailed analysis on impact assessment and remediation strategies. 🥷&nbsp;&nbsp; GPT-Powered Attack Surface Generation: With reNgine 2.0, reNgine seamlessly integrates with GPT to identify the attacks that you can likely perform on a subdomain. By making use of reconnaissance data such as page title, open ports, subdomain name etc, reNgine can advice you the attacks you could perform on a target. reNgine will also provide you the rationale on why the specific attack is likely to be successful. 🧭&nbsp;&nbsp;Continuous monitoring: Continuous monitoring is at the core of reNgine's mission, and it's robust continuous monitoring feature ensures that their targets are under constant scrutiny. With the flexibility to schedule scans at regular intervals, penetration testers can effortlessly stay informed about their targets. What sets reNgine apart is its seamless integration with popular notification channels such as Discord, Slack, and Telegram, delivering real-time alerts for newly discovered subdomains, vulnerabilities, or any changes in reconnaissance data.  ### Workflow <img src="https://github.com/yogeshojha/rengine/assets/17223002/10c475b8-b4a8-440d-9126-77fe2038a386">  ### Features * Reconnaissance: * Subdomain Discovery * IP and Open Ports Identification * Endpoints Discovery * Directory/Files fuzzing * Screenshot Gathering * Vulnerability Scan * Nuclei * Dalfox XSS Scanner * CRLFuzzer * Misconfigured S3 Scanner * WHOIS Identification * WAF Detection * OSINT Capabilities * Meta info Gathering * Employees Gathering * Email Address gathering * Google Dorking for sensitive info and urls * Projects, create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. * Perform Advanced Query lookup using natural language alike and, or, not operations * Highly configurable YAML-based Scan Engines * Support for Parallel Scans * Support for Subscans * Recon Data visualization * GPT Vulnerability Description, Impact and Remediation generation * GPT Attack Surface Generator * Multiple Roles and Permissions to cater a team's need * Customizable Alerts/Notifications on Slack, Discord, and Telegram * Automatically report Vulnerabilities to HackerOne * Recon Notes and Todos * Clocked Scans (Run reconnaissance exactly at X Hours and Y minutes) and Periodic Scans (Runs reconnaissance every X minutes/- hours/days/week) * Proxy Support * Screenshot Gallery with Filters * Powerful recon data filtering with autosuggestions * Recon Data changes, find new/removed subdomains/endpoints * Tag targets into the Organization * Smart Duplicate endpoint removal based on page title and content length to cleanup the reconnaissance data * Identify Interesting Subdomains * Custom GF patterns and custom Nuclei Templates * Edit tool-related configuration files (Nuclei, Subfinder, Naabu, amass) * Add external tools from Github/Go * Interoperable with other tools, Import/Export Subdomains/Endpoints * Import Targets via IP and/or CIDRs * Report Generation * Toolbox: Comes bundled with most commonly used tools during penetration testing such as whois lookup, CMS detector, CVE lookup, etc. * Identification of related domains and related TLDs for targets * Find actionable insights such as Most Common Vulnerability, Most Common CVE ID, Most Vulnerable Target/Subdomain, etc.  ### Scan Engine ```yaml subdomain_discovery: { 'uses_tools': [ 'subfinder', 'ctfr', 'sublist3r', 'tlsx', 'oneforall', 'netlas' ], 'enable_http_crawl': true, 'threads': 30, 'timeout': 5, } http_crawl: {} port_scan: { 'enable_http_crawl': true, 'timeout': 5, # 'exclude_ports': [], # 'exclude_subdomains': true, 'ports': ['top-100'], 'rate_limit': 150, 'threads': 30, 'passive': false, # 'use_naabu_config': false, # 'enable_nmap': true, # 'nmap_cmd': '', # 'nmap_script': '', # 'nmap_script_args': '' } osint: { 'discover': [ 'emails', 'metainfo', 'employees' ], 'dorks': [ 'login_pages', 'admin_panels', 'dashboard_pages', 'stackoverflow', 'social_media', 'project_management', 'code_sharing', 'config_files', 'jenkins', 'wordpress_files', 'php_error', 'exposed_documents', 'db_files', 'git_exposed' ], 'custom_dorks': [ { 'lookup_site': 'google.com', 'lookup_keywords': '/home/' }, { 'lookup_site': '_target_', 'lookup_extensions': 'jpg,png' } ], 'intensity': 'normal', 'documents_limit': 50 } dir_file_fuzz: { 'auto_calibration': true, 'enable_http_crawl': true, 'rate_limit': 150, 'extensions': ['html', 'php','git','yaml','conf','cnf','config','gz','env','log','db','mysql','bak','asp','aspx','txt','conf','sql','json','yml','pdf'], 'follow_redirect': false, 'max_time': 0, 'match_http_status': [200, 204], 'recursive_level': 2, 'stop_on_error': false, 'timeout': 5, 'threads': 30, 'wordlist_name': 'dicc' } fetch_url: { 'uses_tools': [ 'gospider', 'hakrawler', 'waybackurls', 'gospider', 'katana' ], 'remove_duplicate_endpoints': true, 'duplicate_fields': [ 'content_length', 'page_title' ], 'enable_http_crawl': true, 'gf_patterns': ['debug_logic', 'idor', 'interestingEXT', 'interestingparams', 'interestingsubs', 'lfi', 'rce', 'redirect', 'sqli', 'ssrf', 'ssti', 'xss'], 'ignore_file_extensions': ['png', 'jpg', 'jpeg', 'gif', 'mp4', 'mpeg', 'mp3'] # 'exclude_subdomains': true } vulnerability_scan: { 'run_nuclei': false, 'run_dalfox': false, 'run_crlfuzz': false, 'run_s3scanner': true, 'enable_http_crawl': true, 'concurrency': 50, 'intensity': 'normal', 'rate_limit': 150, 'retries': 1, 'timeout': 5, 'fetch_gpt_report': true, 'nuclei': { 'use_conf': false, 'severities': [ 'unknown', 'info', 'low', 'medium', 'high', 'critical' ], # 'tags': [], # 'templates': [], # 'custom_templates': [], }, 's3scanner': { 'threads': 100, 'providers': [ 'aws', 'gcp', 'digitalocean', 'dreamhost', 'linode' ] } } waf_detection: {} screenshot: { 'enable_http_crawl': true, 'intensity': 'normal', 'timeout': 10, 'threads': 40 } # custom_header: "Cookie: Test" ```  ### Quick Installation **Note:** Only Ubuntu/VPS 1. Clone this repo ```bash git clone https://github.com/yogeshojha/rengine && cd rengine ``` 1. Edit the dotenv file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```bash nano .env ``` 1. In the dotenv file, you may also modify the Scaling Configurations ```bash MAX_CONCURRENCY=80 MIN_CONCURRENCY=10 ``` MAX_CONCURRENCY: This parameter specifies the maximum number of reNgine's concurrent Celery worker processes that can be spawned. In this case, it's set to 80, meaning that the application can utilize up to 80 concurrent worker processes to execute tasks concurrently. This is useful for handling a high volume of scans or when you want to scale up processing power during periods of high demand. If you have more CPU cores, you will need to increase this for maximised performance. MIN_CONCURRENCY: On the other hand, MIN_CONCURRENCY specifies the minimum number of concurrent worker processes that should be maintained, even during periods of lower demand. In this example, it's set to 10, which means that even when there are fewer tasks to process, at least 10 worker processes will be kept running. This helps ensure that the application can respond promptly to incoming tasks without the overhead of repeatedly starting and stopping worker processes. These settings allow for dynamic scaling of Celery workers, ensuring that the application efficiently manages its workload by adjusting the number of concurrent workers based on the workload's size and complexity 1. Run the installation script, Please keep an eye for any prompt, you will also be asked for username and password for reNgine. ```bash sudo ./install.sh ``` If `install.sh` does not have install permission, please change it, `chmod +x install.sh` **reNgine can now be accessed from <https://127.0.0.1> or if you're on the VPS <https://your_vps_ip_address>** **Unless you are on development branch, please do not access reNgine via any ports** ### Installation (Mac/Windows/Other) Installation instructions can be found at [https://reNgine.wiki/install/detailed/](https://reNgine.wiki/2.0/install/detailed/) ### Updating 1. Updating is as simple as running the following command: ```bash cd rengine && sudo ./update.sh ``` If `update.sh` does not have execution permissions, please change it, `sudo chmod +x update.sh` **NOTE:** if you're updating from 1.3.6 and you're getting a 'password authentication failed' error, consider uninstalling 1.3.6 first, then install 2.x.x as you'd normally do. ### Changelog [Please find the latest release notes and changelog here.](https://rengine.wiki/changelog/)  ### Screenshots #### Scan Results  #### General Usage <img src="https://user-images.githubusercontent.com/17223002/164993781-b6012995-522b-480a-a8bf-911193d35894.gif"> #### Initiating Subscan <img src="https://user-images.githubusercontent.com/17223002/164993749-1ad343d6-8ce7-43d6-aee7-b3add0321da7.gif"> #### Recon Data filtering <img src="https://user-images.githubusercontent.com/17223002/164993687-b63f3de8-e033-4ac0-808e-a2aa377d3cf8.gif"> #### Report Generation <img src="https://user-images.githubusercontent.com/17223002/164993689-c796c6cd-eb61-43f4-800d-08aba9740088.gif"> #### Toolbox <img src="https://user-images.githubusercontent.com/17223002/164993751-d687e88a-eb79-440f-9dc0-0ad006901620.gif"> #### Adding Custom tool in Tools Arsenal <img src="https://user-images.githubusercontent.com/17223002/164993670-466f6459-9499-498b-a9bd-526476d735a7.gif">  ### Contributing Contributions are what make the open-source community such an amazing place to learn, inspire and create. Every contributions you make is **greatly appreciated**. Your contributions can be as simple as fixing the indentation or UI, or as complex as adding new modules and features. See the [Contributing Guide](.github/CONTRIBUTING.md) to get started. You can also [join our Discord channel #development](https://discord.gg/JuhHdHTtwd) for any development related questions.  ### First-time Open Source contributors Please note that reNgine is beginner friendly. If you have never done open-source before, we encourage you to do so. **We will be happy and proud of your first PR ever.** You can start by resolving any [open issues](https://github.com/yogeshojha/rengine/issues).  ### reNgine Support Please do not use GitHub for support requests. Instead, [join our Discord channel #support](https://discord.gg/azv6fzhNCE).  ### Support and Sponsoring Over the past few years, I have been working hard on reNgine to add new features with the sole aim of making it the de facto standard for reconnaissance. I spend most of my free time and weekends working on reNgine. I do this in addition to my day job. I am happy to have received such overwhelming support from the community. But to keep this project alive, I am looking for financial support. | Paypal | Bitcoin | Ethereum | | :-------------------------------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | |[https://www.paypal.com/paypalme/yogeshojha11](https://www.paypal.com/paypalme/yogeshojha11) | `35AiKyNswNZ4TZUSdriHopSCjNMPi63BCX` | `0xe7A337Da6ff98A28513C26A7Fec8C9b42A63d346` OR * Add a [GitHub Star](https://github.com/yogeshojha/rengine) to the project. * Tweet about this project, or maybe blogs? * Maybe nominate me for [GitHub Stars?](https://stars.github.com/nominate/) * Join DigitalOcean using my [referral link](https://m.do.co/c/e353502d19fc) your profit is **$100** and I get $25 DO credit. This will help me test reNgine on VPS before I release any major features. It takes a considerable amount of time to add new features and make sure everything works. Donating is your way of saying: **reNgine is awesome**. Any support is greatly appreciated! Thank you!  ### reNgine Bug Bounty Program [](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine) Security researchers, welcome aboard! I'm excited to announce the reNgine bug bounty programme in collaboration with [huntr.dev](https://huntr.dev), which means that you will be rewarded for any vulnerabilities you find in reNgine. Thank you for your interest in reporting reNgine vulnerabilities! If you are aware of any potential security vulnerabilities in reNgine, we encourage you to report them immediately via [huntr.dev](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine). **Please do not disclose vulnerabilities via Github issues/blogs/tweets after/before reporting to huntr.dev as this is explicitly against the disclosure policy of huntr.dev and reNgine and will not be considered for monetary rewards.** Please note that the reNgine maintainer does not set the bounty amount. The bounty reward is determined by an industry-first equation developed by huntr.dev to understand the popularity, impact and value of repositories to the open-source community. **What do I expect from security researchers?** * Patience: Please note that I am currently the only maintainer in reNgine and it will take some time to validate your report. I ask for your patience during this process. * Respect for privacy and security reports: Please do not publicly disclose any vulnerabilities (including GitHub issues) before or after reporting them on huntr.dev! This is against the disclosure policy and will not be rewarded. * Respect the rules **What do you get in return?** * Thanks from the maintainer * Monetary rewards * CVE ID(s) Please find the [FAQ](https://www.huntr.dev/faq) and [Responsible disclosure policy](https://www.huntr.dev/policy/) from huntr.dev.  ### License Distributed under the GNU GPL v3 License. See [LICENSE](LICENSE) for more information.  <p align="right">(ChatGPT was used to write some or most part of this README section.)</p>

<p align="center"> <a href="https://rengine.wiki"><img src=".github/screenshots/banner.gif" alt=""/></a> </p> <p align="center"><a href="https://github.com/yogeshojha/rengine/releases" target="_blank"><img src="https://img.shields.io/badge/version-v2.0.0-informational?&logo=none" alt="reNgine Latest Version" /></a>&nbsp;<a href="https://www.gnu.org/licenses/gpl-3.0" target="_blank"><img src="https://img.shields.io/badge/License-GPLv3-red.svg?&logo=none" alt="License" /></a>&nbsp;<a href="#" target="_blank"><img src="https://img.shields.io/badge/first--timers--only-friendly-blue.svg?&logo=none" alt="" /></a>&nbsp;<a href="https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine" target="_blank"><img src="https://cdn.huntr.dev/huntr_security_badge_mono.svg" alt="" /></a>&nbsp;</p> <p align="center"> <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Asia-2023-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/Open--Source--Summit-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://cyberweek.ae/2021/hitb-armory/" target="_blank"><img src="https://img.shields.io/badge/HITB--Armory-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=7uvP6MaQOX0" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://drive.google.com/file/d/1Bh8lbf-Dztt5ViHJVACyrXMiglyICPQ2/view?usp=sharing" target="_blank"><img src="https://img.shields.io/badge/Defcon--Demolabs--29-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=A1oNOIc0h5A" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Europe-2020-blue.svg?&logo=none" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml/badge.svg" alt="" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/actions/workflows/build.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/build.yml/badge.svg" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://discord.gg/H6WzebwX3H" target="_blank"><img src="https://img.shields.io/discord/880363103689277461" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://opensourcesecurityindex.io/" target="_blank" rel="noopener"> <img style="width: 282px; height: 56px" src="https://opensourcesecurityindex.io/badge.svg" alt="Open Source Security Index - Fastest Growing Open Source Security Projects" width="282" height="56" /> </a> </p> <h3>reNgine 2.0-jasper<br>Redefining the future of reconnaissance!</h3> <h4>What is reNgine?</h4> <p align="left">reNgine is your go-to web application reconnaissance suite that's designed to simplify and streamline the reconnaissance process for security professionals, penetration testers, and bug bounty hunters. With its highly configurable engines, data correlation capabilities, continuous monitoring, database-backed reconnaissance data, and an intuitive user interface, reNgine redefines how you gather critical information about your target web applications. Traditional reconnaissance tools often fall short in terms of configurability and efficiency. reNgine addresses these shortcomings and emerges as a excellent alternative to existing commercial tools. reNgine was created to address the limitations of traditional reconnaissance tools and provide a better alternative, even surpassing some commercial offerings. Whether you're a bug bounty hunter, a penetration tester, or a corporate security team, reNgine is your go-to solution for automating and enhancing your information-gathering efforts. </p> reNgine 2.0-jasper is out now, you can [watch reNgine 2.0-jasper release trailer here!](https://youtu.be/VwkOWqiWW5g) reNgine 2.0-Jasper would not have been possible without [@ocervell](https://github.com/ocervell) valuable contributions. [@ocervell](https://github.com/ocervell) did majority of the refactoring if not all and also added a ton of features. Together, we wish to shape the future of web application reconnaissance, and it's developers like [@ocervell](https://github.com/ocervell) and a [ton of other developers and hackers from our community](https://github.com/yogeshojha/rengine/graphs/contributors) who inspire and drive us forward. Thank you, [@ocervell](https://github.com/ocervell), for your outstanding work and unwavering commitment to reNgine. Checkout our contributers here: [Contributers](https://github.com/yogeshojha/rengine/graphs/contributors)  ### Documentation You can find detailed documentation at [https://rengine.wiki](https://rengine.wiki)  ### Table of Contents * [About reNgine](#about-rengine) * [Workflow](#workflow) * [Features](#features) * [Scan Engine](#scan-engine) * [Quick Installation](#quick-installation) * [What's new in reNgine 2.0](#changelog) * [Screenshots](#screenshots) * [Contributing](#contributing) * [reNgine Support](#rengine-support) * [Support and Sponsoring](#support-and-sponsoring) * [reNgine Bug Bounty Program](#rengine-bug-bounty-program) * [License](#license)  ### About reNgine reNgine is not an ordinary reconnaissance suite; it's a game-changer! We've turbocharged the traditional workflow with groundbreaking features that is sure to ease your reconnaissance game. reNgine redefines the art of reconnaissance with highly configurable scan engines, recon data correlation, continuous monitoring, GPT powered Vulnerability Report, Project Management and role based access control etc. 🦾&nbsp;&nbsp; reNgine has advanced reconnaissance capabilities, harnessing a range of open-source tools to deliver a comprehensive web application reconnaissance experience. With it's intuitive User Interface, it excels in subdomain discovery, pinpointing IP addresses and open ports, collecting endpoints, conducting directory and file fuzzing, capturing screenshots, and performing vulnerability scans. To summarize, it does end-to-end reconnaissance. With WHOIS identification and WAF detection, it offers deep insights into target domains. Additionally, reNgine also identifies misconfigured S3 buckets and find interesting subdomains and URLS, based on specific keywords to helps you identify your next target, making it an go to tool for efficient reconnaissance. 🗃️&nbsp; &nbsp; Say goodbye to recon data chaos! reNgine seamlessly integrates with a database, providing you with unmatched data correlation and organization. Forgot the hassle of grepping through json, txt or csv files. Plus, our custom query language lets you filter reconnaissance data effortlessly using natural language like operators such as filtering all alive subdomains with `http_status=200` and also filter all subdomains that are alive and has admin in name `http_status=200&name=admin` 🔧&nbsp;&nbsp; reNgine offers unparalleled flexibility through its highly configurable scan engines, based on a YAML-based configuration. It offers the freedom to create and customize recon scan engines based on any kind of requirement, users can tailor them to their specific objectives and preferences, from thread management to timeout settings and rate-limit configurations, everything is customizable. Additionally, reNgine offers a range of pre-configured scan engines right out of the box, including Full Scan, Passive Scan, Screenshot Gathering, and the OSINT Scan Engine. These ready-to-use engines eliminate the need for extensive manual setup, aligning perfectly with reNgine's core mission of simplifying the reconnaissance process and enabling users to effortlessly access the right reconnaissance data with minimal effort. 💎&nbsp;&nbsp;Subscans: Subscan is a game-changing feature in reNgine, setting it apart as the only open-source tool of its kind to offer this capability. With Subscan, waiting for the entire pipeline to complete is a thing of the past. Now, users can swiftly respond to newfound discoveries during reconnaissance. Whether you've stumbled upon an intriguing subdomain and wish to conduct a focused port scan or want to delve deeper with a vulnerability assessment, reNgine has you covered. 📃&nbsp;&nbsp; PDF Reports: In addition to its robust reconnaissance capabilities, reNgine goes the extra mile by simplifying the report generation process, recognizing the crucial role that PDF reports play in the realm of end-to-end reconnaissance. Users can effortlessly generate and customize PDF reports to suit their exact needs. Whether it's a Full Scan Report, Vulnerability Report, or a concise reconnaissance report, reNgine provides the flexibility to choose the report type that best communicates your findings. Moreover, the level of customization is unparalleled, allowing users to select report colors, fine-tune executive summaries, and even add personalized touches like company names and footers. With GPT integration, your reports aren't just a report, with remediation steps, and impacts, you get 360-degree view of the vulnerabilities you've uncovered. 🔖&nbsp; &nbsp; Say Hello to Projects! reNgine 2.0 introduces a powerful addition that enables you to efficiently organize your web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. Each projects will have separate dashboard and all the scan results will be separated from each projects, while scan engines and configuration will be shared across all the projects. ⚙️&nbsp; &nbsp; Roles and Permissions! Begining reNgine 2.0, we've taken your web application reconnaissance to a whole new level of control and security. Now, you can assign distinct roles to your team members—Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - 🔐 Sys Admin: Sys Admin is a super user that has permission to modify system and scan related configurations, scan engines, create new users, add new tools etc. Super user can initiate scans and subscans effortlessly. - 🔍 Penetration Tester: Penetration Tester will be allowed to modify and initiate scans and subscans, add or update targets, etc. A penetration tester will not be allowed to modify system configurations. - 📊 Auditor: Auditor can only view and download the report. An auditor can not change any system or scan related configurations nor can initiate any scans or subscans. 🚀&nbsp;&nbsp; GPT Vulnerability Report Generation: Get ready for the future of penetration testing reports with reNgine's groundbreaking feature: "GPT-Powered Report Generation"! With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments that read like they were written by a human security expert! **But that's not all!** Our GPT-driven reports go the extra mile by scouring the web for related news articles, blogs, and references, so you have a 360-degree view of the vulnerabilities you've uncovered. With reNgine 2.0 revolutionize your penetration testing game and impress your clients with reports that are not just informative but engaging and comprehensive with detailed analysis on impact assessment and remediation strategies. 🥷&nbsp;&nbsp; GPT-Powered Attack Surface Generation: With reNgine 2.0, reNgine seamlessly integrates with GPT to identify the attacks that you can likely perform on a subdomain. By making use of reconnaissance data such as page title, open ports, subdomain name etc, reNgine can advice you the attacks you could perform on a target. reNgine will also provide you the rationale on why the specific attack is likely to be successful. 🧭&nbsp;&nbsp;Continuous monitoring: Continuous monitoring is at the core of reNgine's mission, and it's robust continuous monitoring feature ensures that their targets are under constant scrutiny. With the flexibility to schedule scans at regular intervals, penetration testers can effortlessly stay informed about their targets. What sets reNgine apart is its seamless integration with popular notification channels such as Discord, Slack, and Telegram, delivering real-time alerts for newly discovered subdomains, vulnerabilities, or any changes in reconnaissance data.  ### Workflow <img src="https://github.com/yogeshojha/rengine/assets/17223002/10c475b8-b4a8-440d-9126-77fe2038a386">  ### Features * Reconnaissance: * Subdomain Discovery * IP and Open Ports Identification * Endpoints Discovery * Directory/Files fuzzing * Screenshot Gathering * Vulnerability Scan * Nuclei * Dalfox XSS Scanner * CRLFuzzer * Misconfigured S3 Scanner * WHOIS Identification * WAF Detection * OSINT Capabilities * Meta info Gathering * Employees Gathering * Email Address gathering * Google Dorking for sensitive info and urls * Projects, create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. * Perform Advanced Query lookup using natural language alike and, or, not operations * Highly configurable YAML-based Scan Engines * Support for Parallel Scans * Support for Subscans * Recon Data visualization * GPT Vulnerability Description, Impact and Remediation generation * GPT Attack Surface Generator * Multiple Roles and Permissions to cater a team's need * Customizable Alerts/Notifications on Slack, Discord, and Telegram * Automatically report Vulnerabilities to HackerOne * Recon Notes and Todos * Clocked Scans (Run reconnaissance exactly at X Hours and Y minutes) and Periodic Scans (Runs reconnaissance every X minutes/- hours/days/week) * Proxy Support * Screenshot Gallery with Filters * Powerful recon data filtering with autosuggestions * Recon Data changes, find new/removed subdomains/endpoints * Tag targets into the Organization * Smart Duplicate endpoint removal based on page title and content length to cleanup the reconnaissance data * Identify Interesting Subdomains * Custom GF patterns and custom Nuclei Templates * Edit tool-related configuration files (Nuclei, Subfinder, Naabu, amass) * Add external tools from Github/Go * Interoperable with other tools, Import/Export Subdomains/Endpoints * Import Targets via IP and/or CIDRs * Report Generation * Toolbox: Comes bundled with most commonly used tools during penetration testing such as whois lookup, CMS detector, CVE lookup, etc. * Identification of related domains and related TLDs for targets * Find actionable insights such as Most Common Vulnerability, Most Common CVE ID, Most Vulnerable Target/Subdomain, etc.  ### Scan Engine ```yaml subdomain_discovery: { 'uses_tools': [ 'subfinder', 'ctfr', 'sublist3r', 'tlsx', 'oneforall', 'netlas' ], 'enable_http_crawl': true, 'threads': 30, 'timeout': 5, } http_crawl: {} port_scan: { 'enable_http_crawl': true, 'timeout': 5, # 'exclude_ports': [], # 'exclude_subdomains': true, 'ports': ['top-100'], 'rate_limit': 150, 'threads': 30, 'passive': false, # 'use_naabu_config': false, # 'enable_nmap': true, # 'nmap_cmd': '', # 'nmap_script': '', # 'nmap_script_args': '' } osint: { 'discover': [ 'emails', 'metainfo', 'employees' ], 'dorks': [ 'login_pages', 'admin_panels', 'dashboard_pages', 'stackoverflow', 'social_media', 'project_management', 'code_sharing', 'config_files', 'jenkins', 'wordpress_files', 'php_error', 'exposed_documents', 'db_files', 'git_exposed' ], 'custom_dorks': [ { 'lookup_site': 'google.com', 'lookup_keywords': '/home/' }, { 'lookup_site': '_target_', 'lookup_extensions': 'jpg,png' } ], 'intensity': 'normal', 'documents_limit': 50 } dir_file_fuzz: { 'auto_calibration': true, 'enable_http_crawl': true, 'rate_limit': 150, 'extensions': ['html', 'php','git','yaml','conf','cnf','config','gz','env','log','db','mysql','bak','asp','aspx','txt','conf','sql','json','yml','pdf'], 'follow_redirect': false, 'max_time': 0, 'match_http_status': [200, 204], 'recursive_level': 2, 'stop_on_error': false, 'timeout': 5, 'threads': 30, 'wordlist_name': 'dicc' } fetch_url: { 'uses_tools': [ 'gospider', 'hakrawler', 'waybackurls', 'gospider', 'katana' ], 'remove_duplicate_endpoints': true, 'duplicate_fields': [ 'content_length', 'page_title' ], 'enable_http_crawl': true, 'gf_patterns': ['debug_logic', 'idor', 'interestingEXT', 'interestingparams', 'interestingsubs', 'lfi', 'rce', 'redirect', 'sqli', 'ssrf', 'ssti', 'xss'], 'ignore_file_extensions': ['png', 'jpg', 'jpeg', 'gif', 'mp4', 'mpeg', 'mp3'] # 'exclude_subdomains': true } vulnerability_scan: { 'run_nuclei': false, 'run_dalfox': false, 'run_crlfuzz': false, 'run_s3scanner': true, 'enable_http_crawl': true, 'concurrency': 50, 'intensity': 'normal', 'rate_limit': 150, 'retries': 1, 'timeout': 5, 'fetch_gpt_report': true, 'nuclei': { 'use_conf': false, 'severities': [ 'unknown', 'info', 'low', 'medium', 'high', 'critical' ], # 'tags': [], # 'templates': [], # 'custom_templates': [], }, 's3scanner': { 'threads': 100, 'providers': [ 'aws', 'gcp', 'digitalocean', 'dreamhost', 'linode' ] } } waf_detection: {} screenshot: { 'enable_http_crawl': true, 'intensity': 'normal', 'timeout': 10, 'threads': 40 } # custom_header: "Cookie: Test" ```  ### Quick Installation **Note:** Only Ubuntu/VPS 1. Clone this repo ```bash git clone https://github.com/yogeshojha/rengine && cd rengine ``` 1. Edit the `.env` file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```bash nano .env ``` 1. **Optional, only for non-interactive install**: In the `.env` file, **please make sure to change the super admin values!** ```bash DJANGO_SUPERUSER_USERNAME=yourUsername DJANGO_SUPERUSER_EMAIL=YourMail@example.com DJANGO_SUPERUSER_PASSWORD=yourStrongPassword ``` If you need to carry out a non-interactive installation, you can setup the login, email and password of the web interface admin directly from the .env file (instead of manually setting them from prompts during the installation process). This option can be interesting for automated installation (via ansible, vagrant, etc.). `DJANGO_SUPERUSER_USERNAME`: web interface admin username (used to login to the web interface). `DJANGO_SUPERUSER_EMAIL`: web interface admin email. `DJANGO_SUPERUSER_PASSWORD`: web interface admin password (used to login to the web interface). 1. In the dotenv file, you may also modify the Scaling Configurations ```bash MAX_CONCURRENCY=80 MIN_CONCURRENCY=10 ``` `MAX_CONCURRENCY`: This parameter specifies the maximum number of reNgine's concurrent Celery worker processes that can be spawned. In this case, it's set to 80, meaning that the application can utilize up to 80 concurrent worker processes to execute tasks concurrently. This is useful for handling a high volume of scans or when you want to scale up processing power during periods of high demand. If you have more CPU cores, you will need to increase this for maximised performance. `MIN_CONCURRENCY`: On the other hand, MIN_CONCURRENCY specifies the minimum number of concurrent worker processes that should be maintained, even during periods of lower demand. In this example, it's set to 10, which means that even when there are fewer tasks to process, at least 10 worker processes will be kept running. This helps ensure that the application can respond promptly to incoming tasks without the overhead of repeatedly starting and stopping worker processes. These settings allow for dynamic scaling of Celery workers, ensuring that the application efficiently manages its workload by adjusting the number of concurrent workers based on the workload's size and complexity 1. Run the installation script, Please keep an eye for any prompt, you will also be asked for username and password for reNgine. ```bash sudo ./install.sh ``` Or for a non-interactive installation, use `-n` argument (make sure you've modified the `.env` file before launching the installation). ```bash sudo ./install.sh -n ``` If `install.sh` does not have install permission, please change it, `chmod +x install.sh` **reNgine can now be accessed from <https://127.0.0.1> or if you're on the VPS <https://your_vps_ip_address>** **Unless you are on development branch, please do not access reNgine via any ports** ### Installation (Mac/Windows/Other) Installation instructions can be found at [https://reNgine.wiki/install/detailed/](https://reNgine.wiki/2.0/install/detailed/) ### Updating 1. Updating is as simple as running the following command: ```bash cd rengine && sudo ./update.sh ``` If `update.sh` does not have execution permissions, please change it, `sudo chmod +x update.sh` **NOTE:** if you're updating from 1.3.6 and you're getting a 'password authentication failed' error, consider uninstalling 1.3.6 first, then install 2.x.x as you'd normally do. ### Changelog [Please find the latest release notes and changelog here.](https://rengine.wiki/changelog/)  ### Screenshots #### Scan Results  #### General Usage <img src="https://user-images.githubusercontent.com/17223002/164993781-b6012995-522b-480a-a8bf-911193d35894.gif"> #### Initiating Subscan <img src="https://user-images.githubusercontent.com/17223002/164993749-1ad343d6-8ce7-43d6-aee7-b3add0321da7.gif"> #### Recon Data filtering <img src="https://user-images.githubusercontent.com/17223002/164993687-b63f3de8-e033-4ac0-808e-a2aa377d3cf8.gif"> #### Report Generation <img src="https://user-images.githubusercontent.com/17223002/164993689-c796c6cd-eb61-43f4-800d-08aba9740088.gif"> #### Toolbox <img src="https://user-images.githubusercontent.com/17223002/164993751-d687e88a-eb79-440f-9dc0-0ad006901620.gif"> #### Adding Custom tool in Tools Arsenal <img src="https://user-images.githubusercontent.com/17223002/164993670-466f6459-9499-498b-a9bd-526476d735a7.gif">  ### Contributing Contributions are what make the open-source community such an amazing place to learn, inspire and create. Every contributions you make is **greatly appreciated**. Your contributions can be as simple as fixing the indentation or UI, or as complex as adding new modules and features. See the [Contributing Guide](.github/CONTRIBUTING.md) to get started. You can also [join our Discord channel #development](https://discord.gg/JuhHdHTtwd) for any development related questions.  ### First-time Open Source contributors Please note that reNgine is beginner friendly. If you have never done open-source before, we encourage you to do so. **We will be happy and proud of your first PR ever.** You can start by resolving any [open issues](https://github.com/yogeshojha/rengine/issues).  ### reNgine Support Please do not use GitHub for support requests. Instead, [join our Discord channel #support](https://discord.gg/azv6fzhNCE).  ### Support and Sponsoring Over the past few years, I have been working hard on reNgine to add new features with the sole aim of making it the de facto standard for reconnaissance. I spend most of my free time and weekends working on reNgine. I do this in addition to my day job. I am happy to have received such overwhelming support from the community. But to keep this project alive, I am looking for financial support. | Paypal | Bitcoin | Ethereum | | :-------------------------------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | |[https://www.paypal.com/paypalme/yogeshojha11](https://www.paypal.com/paypalme/yogeshojha11) | `35AiKyNswNZ4TZUSdriHopSCjNMPi63BCX` | `0xe7A337Da6ff98A28513C26A7Fec8C9b42A63d346` OR * Add a [GitHub Star](https://github.com/yogeshojha/rengine) to the project. * Tweet about this project, or maybe blogs? * Maybe nominate me for [GitHub Stars?](https://stars.github.com/nominate/) * Join DigitalOcean using my [referral link](https://m.do.co/c/e353502d19fc) your profit is **$100** and I get $25 DO credit. This will help me test reNgine on VPS before I release any major features. It takes a considerable amount of time to add new features and make sure everything works. Donating is your way of saying: **reNgine is awesome**. Any support is greatly appreciated! Thank you!  ### reNgine Bug Bounty Program [](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine) Security researchers, welcome aboard! I'm excited to announce the reNgine bug bounty programme in collaboration with [huntr.dev](https://huntr.dev), which means that you will be rewarded for any vulnerabilities you find in reNgine. Thank you for your interest in reporting reNgine vulnerabilities! If you are aware of any potential security vulnerabilities in reNgine, we encourage you to report them immediately via [huntr.dev](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine). **Please do not disclose vulnerabilities via Github issues/blogs/tweets after/before reporting to huntr.dev as this is explicitly against the disclosure policy of huntr.dev and reNgine and will not be considered for monetary rewards.** Please note that the reNgine maintainer does not set the bounty amount. The bounty reward is determined by an industry-first equation developed by huntr.dev to understand the popularity, impact and value of repositories to the open-source community. **What do I expect from security researchers?** * Patience: Please note that I am currently the only maintainer in reNgine and it will take some time to validate your report. I ask for your patience during this process. * Respect for privacy and security reports: Please do not publicly disclose any vulnerabilities (including GitHub issues) before or after reporting them on huntr.dev! This is against the disclosure policy and will not be rewarded. * Respect the rules **What do you get in return?** * Thanks from the maintainer * Monetary rewards * CVE ID(s) Please find the [FAQ](https://www.huntr.dev/faq) and [Responsible disclosure policy](https://www.huntr.dev/policy/) from huntr.dev.  ### License Distributed under the GNU GPL v3 License. See [LICENSE](LICENSE) for more information.  <p align="right">(ChatGPT was used to write some or most part of this README section.)</p>

C0wnuts

3dd700357a4bd5701b07ede4511f66042655be00

64b7f291240b3b8853e3cec7ee6230827c97b907

```suggestion `DJANGO_SUPERUSER_USERNAME`: web interface admin username (used to login to the web interface). ```

AnonymousWP

yogeshojha/rengine

Add non-interactive installation parameter

Add a non-interactive installation method via a new parameter to be passed to the install.sh script. Essential for automated/industrialized systems (e.g. via Ansible or another automated environment creation system).

2023-10-12 01:09:15+00:00

2023-11-21 12:49:22+00:00

README.md

<p align="center"> <a href="https://rengine.wiki"><img src=".github/screenshots/banner.gif" alt=""/></a> </p> <p align="center"><a href="https://github.com/yogeshojha/rengine/releases" target="_blank"><img src="https://img.shields.io/badge/version-v2.0.0-informational?&logo=none" alt="reNgine Latest Version" /></a>&nbsp;<a href="https://www.gnu.org/licenses/gpl-3.0" target="_blank"><img src="https://img.shields.io/badge/License-GPLv3-red.svg?&logo=none" alt="License" /></a>&nbsp;<a href="#" target="_blank"><img src="https://img.shields.io/badge/first--timers--only-friendly-blue.svg?&logo=none" alt="" /></a>&nbsp;<a href="https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine" target="_blank"><img src="https://cdn.huntr.dev/huntr_security_badge_mono.svg" alt="" /></a>&nbsp;</p> <p align="center"> <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Asia-2023-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/Open--Source--Summit-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://cyberweek.ae/2021/hitb-armory/" target="_blank"><img src="https://img.shields.io/badge/HITB--Armory-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=7uvP6MaQOX0" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://drive.google.com/file/d/1Bh8lbf-Dztt5ViHJVACyrXMiglyICPQ2/view?usp=sharing" target="_blank"><img src="https://img.shields.io/badge/Defcon--Demolabs--29-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=A1oNOIc0h5A" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Europe-2020-blue.svg?&logo=none" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml/badge.svg" alt="" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/actions/workflows/build.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/build.yml/badge.svg" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://discord.gg/H6WzebwX3H" target="_blank"><img src="https://img.shields.io/discord/880363103689277461" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://opensourcesecurityindex.io/" target="_blank" rel="noopener"> <img style="width: 282px; height: 56px" src="https://opensourcesecurityindex.io/badge.svg" alt="Open Source Security Index - Fastest Growing Open Source Security Projects" width="282" height="56" /> </a> </p> <h3>reNgine 2.0-jasper<br>Redefining the future of reconnaissance!</h3> <h4>What is reNgine?</h4> <p align="left">reNgine is your go-to web application reconnaissance suite that's designed to simplify and streamline the reconnaissance process for security professionals, penetration testers, and bug bounty hunters. With its highly configurable engines, data correlation capabilities, continuous monitoring, database-backed reconnaissance data, and an intuitive user interface, reNgine redefines how you gather critical information about your target web applications. Traditional reconnaissance tools often fall short in terms of configurability and efficiency. reNgine addresses these shortcomings and emerges as a excellent alternative to existing commercial tools. reNgine was created to address the limitations of traditional reconnaissance tools and provide a better alternative, even surpassing some commercial offerings. Whether you're a bug bounty hunter, a penetration tester, or a corporate security team, reNgine is your go-to solution for automating and enhancing your information-gathering efforts. </p> reNgine 2.0-jasper is out now, you can [watch reNgine 2.0-jasper release trailer here!](https://youtu.be/VwkOWqiWW5g) reNgine 2.0-Jasper would not have been possible without [@ocervell](https://github.com/ocervell) valuable contributions. [@ocervell](https://github.com/ocervell) did majority of the refactoring if not all and also added a ton of features. Together, we wish to shape the future of web application reconnaissance, and it's developers like [@ocervell](https://github.com/ocervell) and a [ton of other developers and hackers from our community](https://github.com/yogeshojha/rengine/graphs/contributors) who inspire and drive us forward. Thank you, [@ocervell](https://github.com/ocervell), for your outstanding work and unwavering commitment to reNgine. Checkout our contributers here: [Contributers](https://github.com/yogeshojha/rengine/graphs/contributors)  ### Documentation You can find detailed documentation at [https://rengine.wiki](https://rengine.wiki)  ### Table of Contents * [About reNgine](#about-rengine) * [Workflow](#workflow) * [Features](#features) * [Scan Engine](#scan-engine) * [Quick Installation](#quick-installation) * [What's new in reNgine 2.0](#changelog) * [Screenshots](#screenshots) * [Contributing](#contributing) * [reNgine Support](#rengine-support) * [Support and Sponsoring](#support-and-sponsoring) * [reNgine Bug Bounty Program](#rengine-bug-bounty-program) * [License](#license)  ### About reNgine reNgine is not an ordinary reconnaissance suite; it's a game-changer! We've turbocharged the traditional workflow with groundbreaking features that is sure to ease your reconnaissance game. reNgine redefines the art of reconnaissance with highly configurable scan engines, recon data correlation, continuous monitoring, GPT powered Vulnerability Report, Project Management and role based access control etc. 🦾&nbsp;&nbsp; reNgine has advanced reconnaissance capabilities, harnessing a range of open-source tools to deliver a comprehensive web application reconnaissance experience. With it's intuitive User Interface, it excels in subdomain discovery, pinpointing IP addresses and open ports, collecting endpoints, conducting directory and file fuzzing, capturing screenshots, and performing vulnerability scans. To summarize, it does end-to-end reconnaissance. With WHOIS identification and WAF detection, it offers deep insights into target domains. Additionally, reNgine also identifies misconfigured S3 buckets and find interesting subdomains and URLS, based on specific keywords to helps you identify your next target, making it an go to tool for efficient reconnaissance. 🗃️&nbsp; &nbsp; Say goodbye to recon data chaos! reNgine seamlessly integrates with a database, providing you with unmatched data correlation and organization. Forgot the hassle of grepping through json, txt or csv files. Plus, our custom query language lets you filter reconnaissance data effortlessly using natural language like operators such as filtering all alive subdomains with `http_status=200` and also filter all subdomains that are alive and has admin in name `http_status=200&name=admin` 🔧&nbsp;&nbsp; reNgine offers unparalleled flexibility through its highly configurable scan engines, based on a YAML-based configuration. It offers the freedom to create and customize recon scan engines based on any kind of requirement, users can tailor them to their specific objectives and preferences, from thread management to timeout settings and rate-limit configurations, everything is customizable. Additionally, reNgine offers a range of pre-configured scan engines right out of the box, including Full Scan, Passive Scan, Screenshot Gathering, and the OSINT Scan Engine. These ready-to-use engines eliminate the need for extensive manual setup, aligning perfectly with reNgine's core mission of simplifying the reconnaissance process and enabling users to effortlessly access the right reconnaissance data with minimal effort. 💎&nbsp;&nbsp;Subscans: Subscan is a game-changing feature in reNgine, setting it apart as the only open-source tool of its kind to offer this capability. With Subscan, waiting for the entire pipeline to complete is a thing of the past. Now, users can swiftly respond to newfound discoveries during reconnaissance. Whether you've stumbled upon an intriguing subdomain and wish to conduct a focused port scan or want to delve deeper with a vulnerability assessment, reNgine has you covered. 📃&nbsp;&nbsp; PDF Reports: In addition to its robust reconnaissance capabilities, reNgine goes the extra mile by simplifying the report generation process, recognizing the crucial role that PDF reports play in the realm of end-to-end reconnaissance. Users can effortlessly generate and customize PDF reports to suit their exact needs. Whether it's a Full Scan Report, Vulnerability Report, or a concise reconnaissance report, reNgine provides the flexibility to choose the report type that best communicates your findings. Moreover, the level of customization is unparalleled, allowing users to select report colors, fine-tune executive summaries, and even add personalized touches like company names and footers. With GPT integration, your reports aren't just a report, with remediation steps, and impacts, you get 360-degree view of the vulnerabilities you've uncovered. 🔖&nbsp; &nbsp; Say Hello to Projects! reNgine 2.0 introduces a powerful addition that enables you to efficiently organize your web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. Each projects will have separate dashboard and all the scan results will be separated from each projects, while scan engines and configuration will be shared across all the projects. ⚙️&nbsp; &nbsp; Roles and Permissions! Begining reNgine 2.0, we've taken your web application reconnaissance to a whole new level of control and security. Now, you can assign distinct roles to your team members—Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - 🔐 Sys Admin: Sys Admin is a super user that has permission to modify system and scan related configurations, scan engines, create new users, add new tools etc. Super user can initiate scans and subscans effortlessly. - 🔍 Penetration Tester: Penetration Tester will be allowed to modify and initiate scans and subscans, add or update targets, etc. A penetration tester will not be allowed to modify system configurations. - 📊 Auditor: Auditor can only view and download the report. An auditor can not change any system or scan related configurations nor can initiate any scans or subscans. 🚀&nbsp;&nbsp; GPT Vulnerability Report Generation: Get ready for the future of penetration testing reports with reNgine's groundbreaking feature: "GPT-Powered Report Generation"! With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments that read like they were written by a human security expert! **But that's not all!** Our GPT-driven reports go the extra mile by scouring the web for related news articles, blogs, and references, so you have a 360-degree view of the vulnerabilities you've uncovered. With reNgine 2.0 revolutionize your penetration testing game and impress your clients with reports that are not just informative but engaging and comprehensive with detailed analysis on impact assessment and remediation strategies. 🥷&nbsp;&nbsp; GPT-Powered Attack Surface Generation: With reNgine 2.0, reNgine seamlessly integrates with GPT to identify the attacks that you can likely perform on a subdomain. By making use of reconnaissance data such as page title, open ports, subdomain name etc, reNgine can advice you the attacks you could perform on a target. reNgine will also provide you the rationale on why the specific attack is likely to be successful. 🧭&nbsp;&nbsp;Continuous monitoring: Continuous monitoring is at the core of reNgine's mission, and it's robust continuous monitoring feature ensures that their targets are under constant scrutiny. With the flexibility to schedule scans at regular intervals, penetration testers can effortlessly stay informed about their targets. What sets reNgine apart is its seamless integration with popular notification channels such as Discord, Slack, and Telegram, delivering real-time alerts for newly discovered subdomains, vulnerabilities, or any changes in reconnaissance data.  ### Workflow <img src="https://github.com/yogeshojha/rengine/assets/17223002/10c475b8-b4a8-440d-9126-77fe2038a386">  ### Features * Reconnaissance: * Subdomain Discovery * IP and Open Ports Identification * Endpoints Discovery * Directory/Files fuzzing * Screenshot Gathering * Vulnerability Scan * Nuclei * Dalfox XSS Scanner * CRLFuzzer * Misconfigured S3 Scanner * WHOIS Identification * WAF Detection * OSINT Capabilities * Meta info Gathering * Employees Gathering * Email Address gathering * Google Dorking for sensitive info and urls * Projects, create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. * Perform Advanced Query lookup using natural language alike and, or, not operations * Highly configurable YAML-based Scan Engines * Support for Parallel Scans * Support for Subscans * Recon Data visualization * GPT Vulnerability Description, Impact and Remediation generation * GPT Attack Surface Generator * Multiple Roles and Permissions to cater a team's need * Customizable Alerts/Notifications on Slack, Discord, and Telegram * Automatically report Vulnerabilities to HackerOne * Recon Notes and Todos * Clocked Scans (Run reconnaissance exactly at X Hours and Y minutes) and Periodic Scans (Runs reconnaissance every X minutes/- hours/days/week) * Proxy Support * Screenshot Gallery with Filters * Powerful recon data filtering with autosuggestions * Recon Data changes, find new/removed subdomains/endpoints * Tag targets into the Organization * Smart Duplicate endpoint removal based on page title and content length to cleanup the reconnaissance data * Identify Interesting Subdomains * Custom GF patterns and custom Nuclei Templates * Edit tool-related configuration files (Nuclei, Subfinder, Naabu, amass) * Add external tools from Github/Go * Interoperable with other tools, Import/Export Subdomains/Endpoints * Import Targets via IP and/or CIDRs * Report Generation * Toolbox: Comes bundled with most commonly used tools during penetration testing such as whois lookup, CMS detector, CVE lookup, etc. * Identification of related domains and related TLDs for targets * Find actionable insights such as Most Common Vulnerability, Most Common CVE ID, Most Vulnerable Target/Subdomain, etc.  ### Scan Engine ```yaml subdomain_discovery: { 'uses_tools': [ 'subfinder', 'ctfr', 'sublist3r', 'tlsx', 'oneforall', 'netlas' ], 'enable_http_crawl': true, 'threads': 30, 'timeout': 5, } http_crawl: {} port_scan: { 'enable_http_crawl': true, 'timeout': 5, # 'exclude_ports': [], # 'exclude_subdomains': true, 'ports': ['top-100'], 'rate_limit': 150, 'threads': 30, 'passive': false, # 'use_naabu_config': false, # 'enable_nmap': true, # 'nmap_cmd': '', # 'nmap_script': '', # 'nmap_script_args': '' } osint: { 'discover': [ 'emails', 'metainfo', 'employees' ], 'dorks': [ 'login_pages', 'admin_panels', 'dashboard_pages', 'stackoverflow', 'social_media', 'project_management', 'code_sharing', 'config_files', 'jenkins', 'wordpress_files', 'php_error', 'exposed_documents', 'db_files', 'git_exposed' ], 'custom_dorks': [ { 'lookup_site': 'google.com', 'lookup_keywords': '/home/' }, { 'lookup_site': '_target_', 'lookup_extensions': 'jpg,png' } ], 'intensity': 'normal', 'documents_limit': 50 } dir_file_fuzz: { 'auto_calibration': true, 'enable_http_crawl': true, 'rate_limit': 150, 'extensions': ['html', 'php','git','yaml','conf','cnf','config','gz','env','log','db','mysql','bak','asp','aspx','txt','conf','sql','json','yml','pdf'], 'follow_redirect': false, 'max_time': 0, 'match_http_status': [200, 204], 'recursive_level': 2, 'stop_on_error': false, 'timeout': 5, 'threads': 30, 'wordlist_name': 'dicc' } fetch_url: { 'uses_tools': [ 'gospider', 'hakrawler', 'waybackurls', 'gospider', 'katana' ], 'remove_duplicate_endpoints': true, 'duplicate_fields': [ 'content_length', 'page_title' ], 'enable_http_crawl': true, 'gf_patterns': ['debug_logic', 'idor', 'interestingEXT', 'interestingparams', 'interestingsubs', 'lfi', 'rce', 'redirect', 'sqli', 'ssrf', 'ssti', 'xss'], 'ignore_file_extensions': ['png', 'jpg', 'jpeg', 'gif', 'mp4', 'mpeg', 'mp3'] # 'exclude_subdomains': true } vulnerability_scan: { 'run_nuclei': false, 'run_dalfox': false, 'run_crlfuzz': false, 'run_s3scanner': true, 'enable_http_crawl': true, 'concurrency': 50, 'intensity': 'normal', 'rate_limit': 150, 'retries': 1, 'timeout': 5, 'fetch_gpt_report': true, 'nuclei': { 'use_conf': false, 'severities': [ 'unknown', 'info', 'low', 'medium', 'high', 'critical' ], # 'tags': [], # 'templates': [], # 'custom_templates': [], }, 's3scanner': { 'threads': 100, 'providers': [ 'aws', 'gcp', 'digitalocean', 'dreamhost', 'linode' ] } } waf_detection: {} screenshot: { 'enable_http_crawl': true, 'intensity': 'normal', 'timeout': 10, 'threads': 40 } # custom_header: "Cookie: Test" ```  ### Quick Installation **Note:** Only Ubuntu/VPS 1. Clone this repo ```bash git clone https://github.com/yogeshojha/rengine && cd rengine ``` 1. Edit the dotenv file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```bash nano .env ``` 1. In the dotenv file, you may also modify the Scaling Configurations ```bash MAX_CONCURRENCY=80 MIN_CONCURRENCY=10 ``` MAX_CONCURRENCY: This parameter specifies the maximum number of reNgine's concurrent Celery worker processes that can be spawned. In this case, it's set to 80, meaning that the application can utilize up to 80 concurrent worker processes to execute tasks concurrently. This is useful for handling a high volume of scans or when you want to scale up processing power during periods of high demand. If you have more CPU cores, you will need to increase this for maximised performance. MIN_CONCURRENCY: On the other hand, MIN_CONCURRENCY specifies the minimum number of concurrent worker processes that should be maintained, even during periods of lower demand. In this example, it's set to 10, which means that even when there are fewer tasks to process, at least 10 worker processes will be kept running. This helps ensure that the application can respond promptly to incoming tasks without the overhead of repeatedly starting and stopping worker processes. These settings allow for dynamic scaling of Celery workers, ensuring that the application efficiently manages its workload by adjusting the number of concurrent workers based on the workload's size and complexity 1. Run the installation script, Please keep an eye for any prompt, you will also be asked for username and password for reNgine. ```bash sudo ./install.sh ``` If `install.sh` does not have install permission, please change it, `chmod +x install.sh` **reNgine can now be accessed from <https://127.0.0.1> or if you're on the VPS <https://your_vps_ip_address>** **Unless you are on development branch, please do not access reNgine via any ports** ### Installation (Mac/Windows/Other) Installation instructions can be found at [https://reNgine.wiki/install/detailed/](https://reNgine.wiki/2.0/install/detailed/) ### Updating 1. Updating is as simple as running the following command: ```bash cd rengine && sudo ./update.sh ``` If `update.sh` does not have execution permissions, please change it, `sudo chmod +x update.sh` **NOTE:** if you're updating from 1.3.6 and you're getting a 'password authentication failed' error, consider uninstalling 1.3.6 first, then install 2.x.x as you'd normally do. ### Changelog [Please find the latest release notes and changelog here.](https://rengine.wiki/changelog/)  ### Screenshots #### Scan Results  #### General Usage <img src="https://user-images.githubusercontent.com/17223002/164993781-b6012995-522b-480a-a8bf-911193d35894.gif"> #### Initiating Subscan <img src="https://user-images.githubusercontent.com/17223002/164993749-1ad343d6-8ce7-43d6-aee7-b3add0321da7.gif"> #### Recon Data filtering <img src="https://user-images.githubusercontent.com/17223002/164993687-b63f3de8-e033-4ac0-808e-a2aa377d3cf8.gif"> #### Report Generation <img src="https://user-images.githubusercontent.com/17223002/164993689-c796c6cd-eb61-43f4-800d-08aba9740088.gif"> #### Toolbox <img src="https://user-images.githubusercontent.com/17223002/164993751-d687e88a-eb79-440f-9dc0-0ad006901620.gif"> #### Adding Custom tool in Tools Arsenal <img src="https://user-images.githubusercontent.com/17223002/164993670-466f6459-9499-498b-a9bd-526476d735a7.gif">  ### Contributing Contributions are what make the open-source community such an amazing place to learn, inspire and create. Every contributions you make is **greatly appreciated**. Your contributions can be as simple as fixing the indentation or UI, or as complex as adding new modules and features. See the [Contributing Guide](.github/CONTRIBUTING.md) to get started. You can also [join our Discord channel #development](https://discord.gg/JuhHdHTtwd) for any development related questions.  ### First-time Open Source contributors Please note that reNgine is beginner friendly. If you have never done open-source before, we encourage you to do so. **We will be happy and proud of your first PR ever.** You can start by resolving any [open issues](https://github.com/yogeshojha/rengine/issues).  ### reNgine Support Please do not use GitHub for support requests. Instead, [join our Discord channel #support](https://discord.gg/azv6fzhNCE).  ### Support and Sponsoring Over the past few years, I have been working hard on reNgine to add new features with the sole aim of making it the de facto standard for reconnaissance. I spend most of my free time and weekends working on reNgine. I do this in addition to my day job. I am happy to have received such overwhelming support from the community. But to keep this project alive, I am looking for financial support. | Paypal | Bitcoin | Ethereum | | :-------------------------------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | |[https://www.paypal.com/paypalme/yogeshojha11](https://www.paypal.com/paypalme/yogeshojha11) | `35AiKyNswNZ4TZUSdriHopSCjNMPi63BCX` | `0xe7A337Da6ff98A28513C26A7Fec8C9b42A63d346` OR * Add a [GitHub Star](https://github.com/yogeshojha/rengine) to the project. * Tweet about this project, or maybe blogs? * Maybe nominate me for [GitHub Stars?](https://stars.github.com/nominate/) * Join DigitalOcean using my [referral link](https://m.do.co/c/e353502d19fc) your profit is **$100** and I get $25 DO credit. This will help me test reNgine on VPS before I release any major features. It takes a considerable amount of time to add new features and make sure everything works. Donating is your way of saying: **reNgine is awesome**. Any support is greatly appreciated! Thank you!  ### reNgine Bug Bounty Program [](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine) Security researchers, welcome aboard! I'm excited to announce the reNgine bug bounty programme in collaboration with [huntr.dev](https://huntr.dev), which means that you will be rewarded for any vulnerabilities you find in reNgine. Thank you for your interest in reporting reNgine vulnerabilities! If you are aware of any potential security vulnerabilities in reNgine, we encourage you to report them immediately via [huntr.dev](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine). **Please do not disclose vulnerabilities via Github issues/blogs/tweets after/before reporting to huntr.dev as this is explicitly against the disclosure policy of huntr.dev and reNgine and will not be considered for monetary rewards.** Please note that the reNgine maintainer does not set the bounty amount. The bounty reward is determined by an industry-first equation developed by huntr.dev to understand the popularity, impact and value of repositories to the open-source community. **What do I expect from security researchers?** * Patience: Please note that I am currently the only maintainer in reNgine and it will take some time to validate your report. I ask for your patience during this process. * Respect for privacy and security reports: Please do not publicly disclose any vulnerabilities (including GitHub issues) before or after reporting them on huntr.dev! This is against the disclosure policy and will not be rewarded. * Respect the rules **What do you get in return?** * Thanks from the maintainer * Monetary rewards * CVE ID(s) Please find the [FAQ](https://www.huntr.dev/faq) and [Responsible disclosure policy](https://www.huntr.dev/policy/) from huntr.dev.  ### License Distributed under the GNU GPL v3 License. See [LICENSE](LICENSE) for more information.  <p align="right">(ChatGPT was used to write some or most part of this README section.)</p>

<p align="center"> <a href="https://rengine.wiki"><img src=".github/screenshots/banner.gif" alt=""/></a> </p> <p align="center"><a href="https://github.com/yogeshojha/rengine/releases" target="_blank"><img src="https://img.shields.io/badge/version-v2.0.0-informational?&logo=none" alt="reNgine Latest Version" /></a>&nbsp;<a href="https://www.gnu.org/licenses/gpl-3.0" target="_blank"><img src="https://img.shields.io/badge/License-GPLv3-red.svg?&logo=none" alt="License" /></a>&nbsp;<a href="#" target="_blank"><img src="https://img.shields.io/badge/first--timers--only-friendly-blue.svg?&logo=none" alt="" /></a>&nbsp;<a href="https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine" target="_blank"><img src="https://cdn.huntr.dev/huntr_security_badge_mono.svg" alt="" /></a>&nbsp;</p> <p align="center"> <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Asia-2023-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/Open--Source--Summit-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://cyberweek.ae/2021/hitb-armory/" target="_blank"><img src="https://img.shields.io/badge/HITB--Armory-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=7uvP6MaQOX0" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://drive.google.com/file/d/1Bh8lbf-Dztt5ViHJVACyrXMiglyICPQ2/view?usp=sharing" target="_blank"><img src="https://img.shields.io/badge/Defcon--Demolabs--29-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=A1oNOIc0h5A" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Europe-2020-blue.svg?&logo=none" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml/badge.svg" alt="" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/actions/workflows/build.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/build.yml/badge.svg" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://discord.gg/H6WzebwX3H" target="_blank"><img src="https://img.shields.io/discord/880363103689277461" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://opensourcesecurityindex.io/" target="_blank" rel="noopener"> <img style="width: 282px; height: 56px" src="https://opensourcesecurityindex.io/badge.svg" alt="Open Source Security Index - Fastest Growing Open Source Security Projects" width="282" height="56" /> </a> </p> <h3>reNgine 2.0-jasper<br>Redefining the future of reconnaissance!</h3> <h4>What is reNgine?</h4> <p align="left">reNgine is your go-to web application reconnaissance suite that's designed to simplify and streamline the reconnaissance process for security professionals, penetration testers, and bug bounty hunters. With its highly configurable engines, data correlation capabilities, continuous monitoring, database-backed reconnaissance data, and an intuitive user interface, reNgine redefines how you gather critical information about your target web applications. Traditional reconnaissance tools often fall short in terms of configurability and efficiency. reNgine addresses these shortcomings and emerges as a excellent alternative to existing commercial tools. reNgine was created to address the limitations of traditional reconnaissance tools and provide a better alternative, even surpassing some commercial offerings. Whether you're a bug bounty hunter, a penetration tester, or a corporate security team, reNgine is your go-to solution for automating and enhancing your information-gathering efforts. </p> reNgine 2.0-jasper is out now, you can [watch reNgine 2.0-jasper release trailer here!](https://youtu.be/VwkOWqiWW5g) reNgine 2.0-Jasper would not have been possible without [@ocervell](https://github.com/ocervell) valuable contributions. [@ocervell](https://github.com/ocervell) did majority of the refactoring if not all and also added a ton of features. Together, we wish to shape the future of web application reconnaissance, and it's developers like [@ocervell](https://github.com/ocervell) and a [ton of other developers and hackers from our community](https://github.com/yogeshojha/rengine/graphs/contributors) who inspire and drive us forward. Thank you, [@ocervell](https://github.com/ocervell), for your outstanding work and unwavering commitment to reNgine. Checkout our contributers here: [Contributers](https://github.com/yogeshojha/rengine/graphs/contributors)  ### Documentation You can find detailed documentation at [https://rengine.wiki](https://rengine.wiki)  ### Table of Contents * [About reNgine](#about-rengine) * [Workflow](#workflow) * [Features](#features) * [Scan Engine](#scan-engine) * [Quick Installation](#quick-installation) * [What's new in reNgine 2.0](#changelog) * [Screenshots](#screenshots) * [Contributing](#contributing) * [reNgine Support](#rengine-support) * [Support and Sponsoring](#support-and-sponsoring) * [reNgine Bug Bounty Program](#rengine-bug-bounty-program) * [License](#license)  ### About reNgine reNgine is not an ordinary reconnaissance suite; it's a game-changer! We've turbocharged the traditional workflow with groundbreaking features that is sure to ease your reconnaissance game. reNgine redefines the art of reconnaissance with highly configurable scan engines, recon data correlation, continuous monitoring, GPT powered Vulnerability Report, Project Management and role based access control etc. 🦾&nbsp;&nbsp; reNgine has advanced reconnaissance capabilities, harnessing a range of open-source tools to deliver a comprehensive web application reconnaissance experience. With it's intuitive User Interface, it excels in subdomain discovery, pinpointing IP addresses and open ports, collecting endpoints, conducting directory and file fuzzing, capturing screenshots, and performing vulnerability scans. To summarize, it does end-to-end reconnaissance. With WHOIS identification and WAF detection, it offers deep insights into target domains. Additionally, reNgine also identifies misconfigured S3 buckets and find interesting subdomains and URLS, based on specific keywords to helps you identify your next target, making it an go to tool for efficient reconnaissance. 🗃️&nbsp; &nbsp; Say goodbye to recon data chaos! reNgine seamlessly integrates with a database, providing you with unmatched data correlation and organization. Forgot the hassle of grepping through json, txt or csv files. Plus, our custom query language lets you filter reconnaissance data effortlessly using natural language like operators such as filtering all alive subdomains with `http_status=200` and also filter all subdomains that are alive and has admin in name `http_status=200&name=admin` 🔧&nbsp;&nbsp; reNgine offers unparalleled flexibility through its highly configurable scan engines, based on a YAML-based configuration. It offers the freedom to create and customize recon scan engines based on any kind of requirement, users can tailor them to their specific objectives and preferences, from thread management to timeout settings and rate-limit configurations, everything is customizable. Additionally, reNgine offers a range of pre-configured scan engines right out of the box, including Full Scan, Passive Scan, Screenshot Gathering, and the OSINT Scan Engine. These ready-to-use engines eliminate the need for extensive manual setup, aligning perfectly with reNgine's core mission of simplifying the reconnaissance process and enabling users to effortlessly access the right reconnaissance data with minimal effort. 💎&nbsp;&nbsp;Subscans: Subscan is a game-changing feature in reNgine, setting it apart as the only open-source tool of its kind to offer this capability. With Subscan, waiting for the entire pipeline to complete is a thing of the past. Now, users can swiftly respond to newfound discoveries during reconnaissance. Whether you've stumbled upon an intriguing subdomain and wish to conduct a focused port scan or want to delve deeper with a vulnerability assessment, reNgine has you covered. 📃&nbsp;&nbsp; PDF Reports: In addition to its robust reconnaissance capabilities, reNgine goes the extra mile by simplifying the report generation process, recognizing the crucial role that PDF reports play in the realm of end-to-end reconnaissance. Users can effortlessly generate and customize PDF reports to suit their exact needs. Whether it's a Full Scan Report, Vulnerability Report, or a concise reconnaissance report, reNgine provides the flexibility to choose the report type that best communicates your findings. Moreover, the level of customization is unparalleled, allowing users to select report colors, fine-tune executive summaries, and even add personalized touches like company names and footers. With GPT integration, your reports aren't just a report, with remediation steps, and impacts, you get 360-degree view of the vulnerabilities you've uncovered. 🔖&nbsp; &nbsp; Say Hello to Projects! reNgine 2.0 introduces a powerful addition that enables you to efficiently organize your web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. Each projects will have separate dashboard and all the scan results will be separated from each projects, while scan engines and configuration will be shared across all the projects. ⚙️&nbsp; &nbsp; Roles and Permissions! Begining reNgine 2.0, we've taken your web application reconnaissance to a whole new level of control and security. Now, you can assign distinct roles to your team members—Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - 🔐 Sys Admin: Sys Admin is a super user that has permission to modify system and scan related configurations, scan engines, create new users, add new tools etc. Super user can initiate scans and subscans effortlessly. - 🔍 Penetration Tester: Penetration Tester will be allowed to modify and initiate scans and subscans, add or update targets, etc. A penetration tester will not be allowed to modify system configurations. - 📊 Auditor: Auditor can only view and download the report. An auditor can not change any system or scan related configurations nor can initiate any scans or subscans. 🚀&nbsp;&nbsp; GPT Vulnerability Report Generation: Get ready for the future of penetration testing reports with reNgine's groundbreaking feature: "GPT-Powered Report Generation"! With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments that read like they were written by a human security expert! **But that's not all!** Our GPT-driven reports go the extra mile by scouring the web for related news articles, blogs, and references, so you have a 360-degree view of the vulnerabilities you've uncovered. With reNgine 2.0 revolutionize your penetration testing game and impress your clients with reports that are not just informative but engaging and comprehensive with detailed analysis on impact assessment and remediation strategies. 🥷&nbsp;&nbsp; GPT-Powered Attack Surface Generation: With reNgine 2.0, reNgine seamlessly integrates with GPT to identify the attacks that you can likely perform on a subdomain. By making use of reconnaissance data such as page title, open ports, subdomain name etc, reNgine can advice you the attacks you could perform on a target. reNgine will also provide you the rationale on why the specific attack is likely to be successful. 🧭&nbsp;&nbsp;Continuous monitoring: Continuous monitoring is at the core of reNgine's mission, and it's robust continuous monitoring feature ensures that their targets are under constant scrutiny. With the flexibility to schedule scans at regular intervals, penetration testers can effortlessly stay informed about their targets. What sets reNgine apart is its seamless integration with popular notification channels such as Discord, Slack, and Telegram, delivering real-time alerts for newly discovered subdomains, vulnerabilities, or any changes in reconnaissance data.  ### Workflow <img src="https://github.com/yogeshojha/rengine/assets/17223002/10c475b8-b4a8-440d-9126-77fe2038a386">  ### Features * Reconnaissance: * Subdomain Discovery * IP and Open Ports Identification * Endpoints Discovery * Directory/Files fuzzing * Screenshot Gathering * Vulnerability Scan * Nuclei * Dalfox XSS Scanner * CRLFuzzer * Misconfigured S3 Scanner * WHOIS Identification * WAF Detection * OSINT Capabilities * Meta info Gathering * Employees Gathering * Email Address gathering * Google Dorking for sensitive info and urls * Projects, create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. * Perform Advanced Query lookup using natural language alike and, or, not operations * Highly configurable YAML-based Scan Engines * Support for Parallel Scans * Support for Subscans * Recon Data visualization * GPT Vulnerability Description, Impact and Remediation generation * GPT Attack Surface Generator * Multiple Roles and Permissions to cater a team's need * Customizable Alerts/Notifications on Slack, Discord, and Telegram * Automatically report Vulnerabilities to HackerOne * Recon Notes and Todos * Clocked Scans (Run reconnaissance exactly at X Hours and Y minutes) and Periodic Scans (Runs reconnaissance every X minutes/- hours/days/week) * Proxy Support * Screenshot Gallery with Filters * Powerful recon data filtering with autosuggestions * Recon Data changes, find new/removed subdomains/endpoints * Tag targets into the Organization * Smart Duplicate endpoint removal based on page title and content length to cleanup the reconnaissance data * Identify Interesting Subdomains * Custom GF patterns and custom Nuclei Templates * Edit tool-related configuration files (Nuclei, Subfinder, Naabu, amass) * Add external tools from Github/Go * Interoperable with other tools, Import/Export Subdomains/Endpoints * Import Targets via IP and/or CIDRs * Report Generation * Toolbox: Comes bundled with most commonly used tools during penetration testing such as whois lookup, CMS detector, CVE lookup, etc. * Identification of related domains and related TLDs for targets * Find actionable insights such as Most Common Vulnerability, Most Common CVE ID, Most Vulnerable Target/Subdomain, etc.  ### Scan Engine ```yaml subdomain_discovery: { 'uses_tools': [ 'subfinder', 'ctfr', 'sublist3r', 'tlsx', 'oneforall', 'netlas' ], 'enable_http_crawl': true, 'threads': 30, 'timeout': 5, } http_crawl: {} port_scan: { 'enable_http_crawl': true, 'timeout': 5, # 'exclude_ports': [], # 'exclude_subdomains': true, 'ports': ['top-100'], 'rate_limit': 150, 'threads': 30, 'passive': false, # 'use_naabu_config': false, # 'enable_nmap': true, # 'nmap_cmd': '', # 'nmap_script': '', # 'nmap_script_args': '' } osint: { 'discover': [ 'emails', 'metainfo', 'employees' ], 'dorks': [ 'login_pages', 'admin_panels', 'dashboard_pages', 'stackoverflow', 'social_media', 'project_management', 'code_sharing', 'config_files', 'jenkins', 'wordpress_files', 'php_error', 'exposed_documents', 'db_files', 'git_exposed' ], 'custom_dorks': [ { 'lookup_site': 'google.com', 'lookup_keywords': '/home/' }, { 'lookup_site': '_target_', 'lookup_extensions': 'jpg,png' } ], 'intensity': 'normal', 'documents_limit': 50 } dir_file_fuzz: { 'auto_calibration': true, 'enable_http_crawl': true, 'rate_limit': 150, 'extensions': ['html', 'php','git','yaml','conf','cnf','config','gz','env','log','db','mysql','bak','asp','aspx','txt','conf','sql','json','yml','pdf'], 'follow_redirect': false, 'max_time': 0, 'match_http_status': [200, 204], 'recursive_level': 2, 'stop_on_error': false, 'timeout': 5, 'threads': 30, 'wordlist_name': 'dicc' } fetch_url: { 'uses_tools': [ 'gospider', 'hakrawler', 'waybackurls', 'gospider', 'katana' ], 'remove_duplicate_endpoints': true, 'duplicate_fields': [ 'content_length', 'page_title' ], 'enable_http_crawl': true, 'gf_patterns': ['debug_logic', 'idor', 'interestingEXT', 'interestingparams', 'interestingsubs', 'lfi', 'rce', 'redirect', 'sqli', 'ssrf', 'ssti', 'xss'], 'ignore_file_extensions': ['png', 'jpg', 'jpeg', 'gif', 'mp4', 'mpeg', 'mp3'] # 'exclude_subdomains': true } vulnerability_scan: { 'run_nuclei': false, 'run_dalfox': false, 'run_crlfuzz': false, 'run_s3scanner': true, 'enable_http_crawl': true, 'concurrency': 50, 'intensity': 'normal', 'rate_limit': 150, 'retries': 1, 'timeout': 5, 'fetch_gpt_report': true, 'nuclei': { 'use_conf': false, 'severities': [ 'unknown', 'info', 'low', 'medium', 'high', 'critical' ], # 'tags': [], # 'templates': [], # 'custom_templates': [], }, 's3scanner': { 'threads': 100, 'providers': [ 'aws', 'gcp', 'digitalocean', 'dreamhost', 'linode' ] } } waf_detection: {} screenshot: { 'enable_http_crawl': true, 'intensity': 'normal', 'timeout': 10, 'threads': 40 } # custom_header: "Cookie: Test" ```  ### Quick Installation **Note:** Only Ubuntu/VPS 1. Clone this repo ```bash git clone https://github.com/yogeshojha/rengine && cd rengine ``` 1. Edit the `.env` file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```bash nano .env ``` 1. **Optional, only for non-interactive install**: In the `.env` file, **please make sure to change the super admin values!** ```bash DJANGO_SUPERUSER_USERNAME=yourUsername DJANGO_SUPERUSER_EMAIL=YourMail@example.com DJANGO_SUPERUSER_PASSWORD=yourStrongPassword ``` If you need to carry out a non-interactive installation, you can setup the login, email and password of the web interface admin directly from the .env file (instead of manually setting them from prompts during the installation process). This option can be interesting for automated installation (via ansible, vagrant, etc.). `DJANGO_SUPERUSER_USERNAME`: web interface admin username (used to login to the web interface). `DJANGO_SUPERUSER_EMAIL`: web interface admin email. `DJANGO_SUPERUSER_PASSWORD`: web interface admin password (used to login to the web interface). 1. In the dotenv file, you may also modify the Scaling Configurations ```bash MAX_CONCURRENCY=80 MIN_CONCURRENCY=10 ``` `MAX_CONCURRENCY`: This parameter specifies the maximum number of reNgine's concurrent Celery worker processes that can be spawned. In this case, it's set to 80, meaning that the application can utilize up to 80 concurrent worker processes to execute tasks concurrently. This is useful for handling a high volume of scans or when you want to scale up processing power during periods of high demand. If you have more CPU cores, you will need to increase this for maximised performance. `MIN_CONCURRENCY`: On the other hand, MIN_CONCURRENCY specifies the minimum number of concurrent worker processes that should be maintained, even during periods of lower demand. In this example, it's set to 10, which means that even when there are fewer tasks to process, at least 10 worker processes will be kept running. This helps ensure that the application can respond promptly to incoming tasks without the overhead of repeatedly starting and stopping worker processes. These settings allow for dynamic scaling of Celery workers, ensuring that the application efficiently manages its workload by adjusting the number of concurrent workers based on the workload's size and complexity 1. Run the installation script, Please keep an eye for any prompt, you will also be asked for username and password for reNgine. ```bash sudo ./install.sh ``` Or for a non-interactive installation, use `-n` argument (make sure you've modified the `.env` file before launching the installation). ```bash sudo ./install.sh -n ``` If `install.sh` does not have install permission, please change it, `chmod +x install.sh` **reNgine can now be accessed from <https://127.0.0.1> or if you're on the VPS <https://your_vps_ip_address>** **Unless you are on development branch, please do not access reNgine via any ports** ### Installation (Mac/Windows/Other) Installation instructions can be found at [https://reNgine.wiki/install/detailed/](https://reNgine.wiki/2.0/install/detailed/) ### Updating 1. Updating is as simple as running the following command: ```bash cd rengine && sudo ./update.sh ``` If `update.sh` does not have execution permissions, please change it, `sudo chmod +x update.sh` **NOTE:** if you're updating from 1.3.6 and you're getting a 'password authentication failed' error, consider uninstalling 1.3.6 first, then install 2.x.x as you'd normally do. ### Changelog [Please find the latest release notes and changelog here.](https://rengine.wiki/changelog/)  ### Screenshots #### Scan Results  #### General Usage <img src="https://user-images.githubusercontent.com/17223002/164993781-b6012995-522b-480a-a8bf-911193d35894.gif"> #### Initiating Subscan <img src="https://user-images.githubusercontent.com/17223002/164993749-1ad343d6-8ce7-43d6-aee7-b3add0321da7.gif"> #### Recon Data filtering <img src="https://user-images.githubusercontent.com/17223002/164993687-b63f3de8-e033-4ac0-808e-a2aa377d3cf8.gif"> #### Report Generation <img src="https://user-images.githubusercontent.com/17223002/164993689-c796c6cd-eb61-43f4-800d-08aba9740088.gif"> #### Toolbox <img src="https://user-images.githubusercontent.com/17223002/164993751-d687e88a-eb79-440f-9dc0-0ad006901620.gif"> #### Adding Custom tool in Tools Arsenal <img src="https://user-images.githubusercontent.com/17223002/164993670-466f6459-9499-498b-a9bd-526476d735a7.gif">  ### Contributing Contributions are what make the open-source community such an amazing place to learn, inspire and create. Every contributions you make is **greatly appreciated**. Your contributions can be as simple as fixing the indentation or UI, or as complex as adding new modules and features. See the [Contributing Guide](.github/CONTRIBUTING.md) to get started. You can also [join our Discord channel #development](https://discord.gg/JuhHdHTtwd) for any development related questions.  ### First-time Open Source contributors Please note that reNgine is beginner friendly. If you have never done open-source before, we encourage you to do so. **We will be happy and proud of your first PR ever.** You can start by resolving any [open issues](https://github.com/yogeshojha/rengine/issues).  ### reNgine Support Please do not use GitHub for support requests. Instead, [join our Discord channel #support](https://discord.gg/azv6fzhNCE).  ### Support and Sponsoring Over the past few years, I have been working hard on reNgine to add new features with the sole aim of making it the de facto standard for reconnaissance. I spend most of my free time and weekends working on reNgine. I do this in addition to my day job. I am happy to have received such overwhelming support from the community. But to keep this project alive, I am looking for financial support. | Paypal | Bitcoin | Ethereum | | :-------------------------------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | |[https://www.paypal.com/paypalme/yogeshojha11](https://www.paypal.com/paypalme/yogeshojha11) | `35AiKyNswNZ4TZUSdriHopSCjNMPi63BCX` | `0xe7A337Da6ff98A28513C26A7Fec8C9b42A63d346` OR * Add a [GitHub Star](https://github.com/yogeshojha/rengine) to the project. * Tweet about this project, or maybe blogs? * Maybe nominate me for [GitHub Stars?](https://stars.github.com/nominate/) * Join DigitalOcean using my [referral link](https://m.do.co/c/e353502d19fc) your profit is **$100** and I get $25 DO credit. This will help me test reNgine on VPS before I release any major features. It takes a considerable amount of time to add new features and make sure everything works. Donating is your way of saying: **reNgine is awesome**. Any support is greatly appreciated! Thank you!  ### reNgine Bug Bounty Program [](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine) Security researchers, welcome aboard! I'm excited to announce the reNgine bug bounty programme in collaboration with [huntr.dev](https://huntr.dev), which means that you will be rewarded for any vulnerabilities you find in reNgine. Thank you for your interest in reporting reNgine vulnerabilities! If you are aware of any potential security vulnerabilities in reNgine, we encourage you to report them immediately via [huntr.dev](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine). **Please do not disclose vulnerabilities via Github issues/blogs/tweets after/before reporting to huntr.dev as this is explicitly against the disclosure policy of huntr.dev and reNgine and will not be considered for monetary rewards.** Please note that the reNgine maintainer does not set the bounty amount. The bounty reward is determined by an industry-first equation developed by huntr.dev to understand the popularity, impact and value of repositories to the open-source community. **What do I expect from security researchers?** * Patience: Please note that I am currently the only maintainer in reNgine and it will take some time to validate your report. I ask for your patience during this process. * Respect for privacy and security reports: Please do not publicly disclose any vulnerabilities (including GitHub issues) before or after reporting them on huntr.dev! This is against the disclosure policy and will not be rewarded. * Respect the rules **What do you get in return?** * Thanks from the maintainer * Monetary rewards * CVE ID(s) Please find the [FAQ](https://www.huntr.dev/faq) and [Responsible disclosure policy](https://www.huntr.dev/policy/) from huntr.dev.  ### License Distributed under the GNU GPL v3 License. See [LICENSE](LICENSE) for more information.  <p align="right">(ChatGPT was used to write some or most part of this README section.)</p>

C0wnuts

3dd700357a4bd5701b07ede4511f66042655be00

64b7f291240b3b8853e3cec7ee6230827c97b907

```suggestion `DJANGO_SUPERUSER_EMAIL`: web interface admin email. ```

AnonymousWP

yogeshojha/rengine

Add non-interactive installation parameter

Add a non-interactive installation method via a new parameter to be passed to the install.sh script. Essential for automated/industrialized systems (e.g. via Ansible or another automated environment creation system).

2023-10-12 01:09:15+00:00

2023-11-21 12:49:22+00:00

README.md

<p align="center"> <a href="https://rengine.wiki"><img src=".github/screenshots/banner.gif" alt=""/></a> </p> <p align="center"><a href="https://github.com/yogeshojha/rengine/releases" target="_blank"><img src="https://img.shields.io/badge/version-v2.0.0-informational?&logo=none" alt="reNgine Latest Version" /></a>&nbsp;<a href="https://www.gnu.org/licenses/gpl-3.0" target="_blank"><img src="https://img.shields.io/badge/License-GPLv3-red.svg?&logo=none" alt="License" /></a>&nbsp;<a href="#" target="_blank"><img src="https://img.shields.io/badge/first--timers--only-friendly-blue.svg?&logo=none" alt="" /></a>&nbsp;<a href="https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine" target="_blank"><img src="https://cdn.huntr.dev/huntr_security_badge_mono.svg" alt="" /></a>&nbsp;</p> <p align="center"> <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Asia-2023-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/Open--Source--Summit-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://cyberweek.ae/2021/hitb-armory/" target="_blank"><img src="https://img.shields.io/badge/HITB--Armory-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=7uvP6MaQOX0" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://drive.google.com/file/d/1Bh8lbf-Dztt5ViHJVACyrXMiglyICPQ2/view?usp=sharing" target="_blank"><img src="https://img.shields.io/badge/Defcon--Demolabs--29-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=A1oNOIc0h5A" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Europe-2020-blue.svg?&logo=none" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml/badge.svg" alt="" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/actions/workflows/build.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/build.yml/badge.svg" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://discord.gg/H6WzebwX3H" target="_blank"><img src="https://img.shields.io/discord/880363103689277461" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://opensourcesecurityindex.io/" target="_blank" rel="noopener"> <img style="width: 282px; height: 56px" src="https://opensourcesecurityindex.io/badge.svg" alt="Open Source Security Index - Fastest Growing Open Source Security Projects" width="282" height="56" /> </a> </p> <h3>reNgine 2.0-jasper<br>Redefining the future of reconnaissance!</h3> <h4>What is reNgine?</h4> <p align="left">reNgine is your go-to web application reconnaissance suite that's designed to simplify and streamline the reconnaissance process for security professionals, penetration testers, and bug bounty hunters. With its highly configurable engines, data correlation capabilities, continuous monitoring, database-backed reconnaissance data, and an intuitive user interface, reNgine redefines how you gather critical information about your target web applications. Traditional reconnaissance tools often fall short in terms of configurability and efficiency. reNgine addresses these shortcomings and emerges as a excellent alternative to existing commercial tools. reNgine was created to address the limitations of traditional reconnaissance tools and provide a better alternative, even surpassing some commercial offerings. Whether you're a bug bounty hunter, a penetration tester, or a corporate security team, reNgine is your go-to solution for automating and enhancing your information-gathering efforts. </p> reNgine 2.0-jasper is out now, you can [watch reNgine 2.0-jasper release trailer here!](https://youtu.be/VwkOWqiWW5g) reNgine 2.0-Jasper would not have been possible without [@ocervell](https://github.com/ocervell) valuable contributions. [@ocervell](https://github.com/ocervell) did majority of the refactoring if not all and also added a ton of features. Together, we wish to shape the future of web application reconnaissance, and it's developers like [@ocervell](https://github.com/ocervell) and a [ton of other developers and hackers from our community](https://github.com/yogeshojha/rengine/graphs/contributors) who inspire and drive us forward. Thank you, [@ocervell](https://github.com/ocervell), for your outstanding work and unwavering commitment to reNgine. Checkout our contributers here: [Contributers](https://github.com/yogeshojha/rengine/graphs/contributors)  ### Documentation You can find detailed documentation at [https://rengine.wiki](https://rengine.wiki)  ### Table of Contents * [About reNgine](#about-rengine) * [Workflow](#workflow) * [Features](#features) * [Scan Engine](#scan-engine) * [Quick Installation](#quick-installation) * [What's new in reNgine 2.0](#changelog) * [Screenshots](#screenshots) * [Contributing](#contributing) * [reNgine Support](#rengine-support) * [Support and Sponsoring](#support-and-sponsoring) * [reNgine Bug Bounty Program](#rengine-bug-bounty-program) * [License](#license)  ### About reNgine reNgine is not an ordinary reconnaissance suite; it's a game-changer! We've turbocharged the traditional workflow with groundbreaking features that is sure to ease your reconnaissance game. reNgine redefines the art of reconnaissance with highly configurable scan engines, recon data correlation, continuous monitoring, GPT powered Vulnerability Report, Project Management and role based access control etc. 🦾&nbsp;&nbsp; reNgine has advanced reconnaissance capabilities, harnessing a range of open-source tools to deliver a comprehensive web application reconnaissance experience. With it's intuitive User Interface, it excels in subdomain discovery, pinpointing IP addresses and open ports, collecting endpoints, conducting directory and file fuzzing, capturing screenshots, and performing vulnerability scans. To summarize, it does end-to-end reconnaissance. With WHOIS identification and WAF detection, it offers deep insights into target domains. Additionally, reNgine also identifies misconfigured S3 buckets and find interesting subdomains and URLS, based on specific keywords to helps you identify your next target, making it an go to tool for efficient reconnaissance. 🗃️&nbsp; &nbsp; Say goodbye to recon data chaos! reNgine seamlessly integrates with a database, providing you with unmatched data correlation and organization. Forgot the hassle of grepping through json, txt or csv files. Plus, our custom query language lets you filter reconnaissance data effortlessly using natural language like operators such as filtering all alive subdomains with `http_status=200` and also filter all subdomains that are alive and has admin in name `http_status=200&name=admin` 🔧&nbsp;&nbsp; reNgine offers unparalleled flexibility through its highly configurable scan engines, based on a YAML-based configuration. It offers the freedom to create and customize recon scan engines based on any kind of requirement, users can tailor them to their specific objectives and preferences, from thread management to timeout settings and rate-limit configurations, everything is customizable. Additionally, reNgine offers a range of pre-configured scan engines right out of the box, including Full Scan, Passive Scan, Screenshot Gathering, and the OSINT Scan Engine. These ready-to-use engines eliminate the need for extensive manual setup, aligning perfectly with reNgine's core mission of simplifying the reconnaissance process and enabling users to effortlessly access the right reconnaissance data with minimal effort. 💎&nbsp;&nbsp;Subscans: Subscan is a game-changing feature in reNgine, setting it apart as the only open-source tool of its kind to offer this capability. With Subscan, waiting for the entire pipeline to complete is a thing of the past. Now, users can swiftly respond to newfound discoveries during reconnaissance. Whether you've stumbled upon an intriguing subdomain and wish to conduct a focused port scan or want to delve deeper with a vulnerability assessment, reNgine has you covered. 📃&nbsp;&nbsp; PDF Reports: In addition to its robust reconnaissance capabilities, reNgine goes the extra mile by simplifying the report generation process, recognizing the crucial role that PDF reports play in the realm of end-to-end reconnaissance. Users can effortlessly generate and customize PDF reports to suit their exact needs. Whether it's a Full Scan Report, Vulnerability Report, or a concise reconnaissance report, reNgine provides the flexibility to choose the report type that best communicates your findings. Moreover, the level of customization is unparalleled, allowing users to select report colors, fine-tune executive summaries, and even add personalized touches like company names and footers. With GPT integration, your reports aren't just a report, with remediation steps, and impacts, you get 360-degree view of the vulnerabilities you've uncovered. 🔖&nbsp; &nbsp; Say Hello to Projects! reNgine 2.0 introduces a powerful addition that enables you to efficiently organize your web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. Each projects will have separate dashboard and all the scan results will be separated from each projects, while scan engines and configuration will be shared across all the projects. ⚙️&nbsp; &nbsp; Roles and Permissions! Begining reNgine 2.0, we've taken your web application reconnaissance to a whole new level of control and security. Now, you can assign distinct roles to your team members—Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - 🔐 Sys Admin: Sys Admin is a super user that has permission to modify system and scan related configurations, scan engines, create new users, add new tools etc. Super user can initiate scans and subscans effortlessly. - 🔍 Penetration Tester: Penetration Tester will be allowed to modify and initiate scans and subscans, add or update targets, etc. A penetration tester will not be allowed to modify system configurations. - 📊 Auditor: Auditor can only view and download the report. An auditor can not change any system or scan related configurations nor can initiate any scans or subscans. 🚀&nbsp;&nbsp; GPT Vulnerability Report Generation: Get ready for the future of penetration testing reports with reNgine's groundbreaking feature: "GPT-Powered Report Generation"! With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments that read like they were written by a human security expert! **But that's not all!** Our GPT-driven reports go the extra mile by scouring the web for related news articles, blogs, and references, so you have a 360-degree view of the vulnerabilities you've uncovered. With reNgine 2.0 revolutionize your penetration testing game and impress your clients with reports that are not just informative but engaging and comprehensive with detailed analysis on impact assessment and remediation strategies. 🥷&nbsp;&nbsp; GPT-Powered Attack Surface Generation: With reNgine 2.0, reNgine seamlessly integrates with GPT to identify the attacks that you can likely perform on a subdomain. By making use of reconnaissance data such as page title, open ports, subdomain name etc, reNgine can advice you the attacks you could perform on a target. reNgine will also provide you the rationale on why the specific attack is likely to be successful. 🧭&nbsp;&nbsp;Continuous monitoring: Continuous monitoring is at the core of reNgine's mission, and it's robust continuous monitoring feature ensures that their targets are under constant scrutiny. With the flexibility to schedule scans at regular intervals, penetration testers can effortlessly stay informed about their targets. What sets reNgine apart is its seamless integration with popular notification channels such as Discord, Slack, and Telegram, delivering real-time alerts for newly discovered subdomains, vulnerabilities, or any changes in reconnaissance data.  ### Workflow <img src="https://github.com/yogeshojha/rengine/assets/17223002/10c475b8-b4a8-440d-9126-77fe2038a386">  ### Features * Reconnaissance: * Subdomain Discovery * IP and Open Ports Identification * Endpoints Discovery * Directory/Files fuzzing * Screenshot Gathering * Vulnerability Scan * Nuclei * Dalfox XSS Scanner * CRLFuzzer * Misconfigured S3 Scanner * WHOIS Identification * WAF Detection * OSINT Capabilities * Meta info Gathering * Employees Gathering * Email Address gathering * Google Dorking for sensitive info and urls * Projects, create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. * Perform Advanced Query lookup using natural language alike and, or, not operations * Highly configurable YAML-based Scan Engines * Support for Parallel Scans * Support for Subscans * Recon Data visualization * GPT Vulnerability Description, Impact and Remediation generation * GPT Attack Surface Generator * Multiple Roles and Permissions to cater a team's need * Customizable Alerts/Notifications on Slack, Discord, and Telegram * Automatically report Vulnerabilities to HackerOne * Recon Notes and Todos * Clocked Scans (Run reconnaissance exactly at X Hours and Y minutes) and Periodic Scans (Runs reconnaissance every X minutes/- hours/days/week) * Proxy Support * Screenshot Gallery with Filters * Powerful recon data filtering with autosuggestions * Recon Data changes, find new/removed subdomains/endpoints * Tag targets into the Organization * Smart Duplicate endpoint removal based on page title and content length to cleanup the reconnaissance data * Identify Interesting Subdomains * Custom GF patterns and custom Nuclei Templates * Edit tool-related configuration files (Nuclei, Subfinder, Naabu, amass) * Add external tools from Github/Go * Interoperable with other tools, Import/Export Subdomains/Endpoints * Import Targets via IP and/or CIDRs * Report Generation * Toolbox: Comes bundled with most commonly used tools during penetration testing such as whois lookup, CMS detector, CVE lookup, etc. * Identification of related domains and related TLDs for targets * Find actionable insights such as Most Common Vulnerability, Most Common CVE ID, Most Vulnerable Target/Subdomain, etc.  ### Scan Engine ```yaml subdomain_discovery: { 'uses_tools': [ 'subfinder', 'ctfr', 'sublist3r', 'tlsx', 'oneforall', 'netlas' ], 'enable_http_crawl': true, 'threads': 30, 'timeout': 5, } http_crawl: {} port_scan: { 'enable_http_crawl': true, 'timeout': 5, # 'exclude_ports': [], # 'exclude_subdomains': true, 'ports': ['top-100'], 'rate_limit': 150, 'threads': 30, 'passive': false, # 'use_naabu_config': false, # 'enable_nmap': true, # 'nmap_cmd': '', # 'nmap_script': '', # 'nmap_script_args': '' } osint: { 'discover': [ 'emails', 'metainfo', 'employees' ], 'dorks': [ 'login_pages', 'admin_panels', 'dashboard_pages', 'stackoverflow', 'social_media', 'project_management', 'code_sharing', 'config_files', 'jenkins', 'wordpress_files', 'php_error', 'exposed_documents', 'db_files', 'git_exposed' ], 'custom_dorks': [ { 'lookup_site': 'google.com', 'lookup_keywords': '/home/' }, { 'lookup_site': '_target_', 'lookup_extensions': 'jpg,png' } ], 'intensity': 'normal', 'documents_limit': 50 } dir_file_fuzz: { 'auto_calibration': true, 'enable_http_crawl': true, 'rate_limit': 150, 'extensions': ['html', 'php','git','yaml','conf','cnf','config','gz','env','log','db','mysql','bak','asp','aspx','txt','conf','sql','json','yml','pdf'], 'follow_redirect': false, 'max_time': 0, 'match_http_status': [200, 204], 'recursive_level': 2, 'stop_on_error': false, 'timeout': 5, 'threads': 30, 'wordlist_name': 'dicc' } fetch_url: { 'uses_tools': [ 'gospider', 'hakrawler', 'waybackurls', 'gospider', 'katana' ], 'remove_duplicate_endpoints': true, 'duplicate_fields': [ 'content_length', 'page_title' ], 'enable_http_crawl': true, 'gf_patterns': ['debug_logic', 'idor', 'interestingEXT', 'interestingparams', 'interestingsubs', 'lfi', 'rce', 'redirect', 'sqli', 'ssrf', 'ssti', 'xss'], 'ignore_file_extensions': ['png', 'jpg', 'jpeg', 'gif', 'mp4', 'mpeg', 'mp3'] # 'exclude_subdomains': true } vulnerability_scan: { 'run_nuclei': false, 'run_dalfox': false, 'run_crlfuzz': false, 'run_s3scanner': true, 'enable_http_crawl': true, 'concurrency': 50, 'intensity': 'normal', 'rate_limit': 150, 'retries': 1, 'timeout': 5, 'fetch_gpt_report': true, 'nuclei': { 'use_conf': false, 'severities': [ 'unknown', 'info', 'low', 'medium', 'high', 'critical' ], # 'tags': [], # 'templates': [], # 'custom_templates': [], }, 's3scanner': { 'threads': 100, 'providers': [ 'aws', 'gcp', 'digitalocean', 'dreamhost', 'linode' ] } } waf_detection: {} screenshot: { 'enable_http_crawl': true, 'intensity': 'normal', 'timeout': 10, 'threads': 40 } # custom_header: "Cookie: Test" ```  ### Quick Installation **Note:** Only Ubuntu/VPS 1. Clone this repo ```bash git clone https://github.com/yogeshojha/rengine && cd rengine ``` 1. Edit the dotenv file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```bash nano .env ``` 1. In the dotenv file, you may also modify the Scaling Configurations ```bash MAX_CONCURRENCY=80 MIN_CONCURRENCY=10 ``` MAX_CONCURRENCY: This parameter specifies the maximum number of reNgine's concurrent Celery worker processes that can be spawned. In this case, it's set to 80, meaning that the application can utilize up to 80 concurrent worker processes to execute tasks concurrently. This is useful for handling a high volume of scans or when you want to scale up processing power during periods of high demand. If you have more CPU cores, you will need to increase this for maximised performance. MIN_CONCURRENCY: On the other hand, MIN_CONCURRENCY specifies the minimum number of concurrent worker processes that should be maintained, even during periods of lower demand. In this example, it's set to 10, which means that even when there are fewer tasks to process, at least 10 worker processes will be kept running. This helps ensure that the application can respond promptly to incoming tasks without the overhead of repeatedly starting and stopping worker processes. These settings allow for dynamic scaling of Celery workers, ensuring that the application efficiently manages its workload by adjusting the number of concurrent workers based on the workload's size and complexity 1. Run the installation script, Please keep an eye for any prompt, you will also be asked for username and password for reNgine. ```bash sudo ./install.sh ``` If `install.sh` does not have install permission, please change it, `chmod +x install.sh` **reNgine can now be accessed from <https://127.0.0.1> or if you're on the VPS <https://your_vps_ip_address>** **Unless you are on development branch, please do not access reNgine via any ports** ### Installation (Mac/Windows/Other) Installation instructions can be found at [https://reNgine.wiki/install/detailed/](https://reNgine.wiki/2.0/install/detailed/) ### Updating 1. Updating is as simple as running the following command: ```bash cd rengine && sudo ./update.sh ``` If `update.sh` does not have execution permissions, please change it, `sudo chmod +x update.sh` **NOTE:** if you're updating from 1.3.6 and you're getting a 'password authentication failed' error, consider uninstalling 1.3.6 first, then install 2.x.x as you'd normally do. ### Changelog [Please find the latest release notes and changelog here.](https://rengine.wiki/changelog/)  ### Screenshots #### Scan Results  #### General Usage <img src="https://user-images.githubusercontent.com/17223002/164993781-b6012995-522b-480a-a8bf-911193d35894.gif"> #### Initiating Subscan <img src="https://user-images.githubusercontent.com/17223002/164993749-1ad343d6-8ce7-43d6-aee7-b3add0321da7.gif"> #### Recon Data filtering <img src="https://user-images.githubusercontent.com/17223002/164993687-b63f3de8-e033-4ac0-808e-a2aa377d3cf8.gif"> #### Report Generation <img src="https://user-images.githubusercontent.com/17223002/164993689-c796c6cd-eb61-43f4-800d-08aba9740088.gif"> #### Toolbox <img src="https://user-images.githubusercontent.com/17223002/164993751-d687e88a-eb79-440f-9dc0-0ad006901620.gif"> #### Adding Custom tool in Tools Arsenal <img src="https://user-images.githubusercontent.com/17223002/164993670-466f6459-9499-498b-a9bd-526476d735a7.gif">  ### Contributing Contributions are what make the open-source community such an amazing place to learn, inspire and create. Every contributions you make is **greatly appreciated**. Your contributions can be as simple as fixing the indentation or UI, or as complex as adding new modules and features. See the [Contributing Guide](.github/CONTRIBUTING.md) to get started. You can also [join our Discord channel #development](https://discord.gg/JuhHdHTtwd) for any development related questions.  ### First-time Open Source contributors Please note that reNgine is beginner friendly. If you have never done open-source before, we encourage you to do so. **We will be happy and proud of your first PR ever.** You can start by resolving any [open issues](https://github.com/yogeshojha/rengine/issues).  ### reNgine Support Please do not use GitHub for support requests. Instead, [join our Discord channel #support](https://discord.gg/azv6fzhNCE).  ### Support and Sponsoring Over the past few years, I have been working hard on reNgine to add new features with the sole aim of making it the de facto standard for reconnaissance. I spend most of my free time and weekends working on reNgine. I do this in addition to my day job. I am happy to have received such overwhelming support from the community. But to keep this project alive, I am looking for financial support. | Paypal | Bitcoin | Ethereum | | :-------------------------------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | |[https://www.paypal.com/paypalme/yogeshojha11](https://www.paypal.com/paypalme/yogeshojha11) | `35AiKyNswNZ4TZUSdriHopSCjNMPi63BCX` | `0xe7A337Da6ff98A28513C26A7Fec8C9b42A63d346` OR * Add a [GitHub Star](https://github.com/yogeshojha/rengine) to the project. * Tweet about this project, or maybe blogs? * Maybe nominate me for [GitHub Stars?](https://stars.github.com/nominate/) * Join DigitalOcean using my [referral link](https://m.do.co/c/e353502d19fc) your profit is **$100** and I get $25 DO credit. This will help me test reNgine on VPS before I release any major features. It takes a considerable amount of time to add new features and make sure everything works. Donating is your way of saying: **reNgine is awesome**. Any support is greatly appreciated! Thank you!  ### reNgine Bug Bounty Program [](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine) Security researchers, welcome aboard! I'm excited to announce the reNgine bug bounty programme in collaboration with [huntr.dev](https://huntr.dev), which means that you will be rewarded for any vulnerabilities you find in reNgine. Thank you for your interest in reporting reNgine vulnerabilities! If you are aware of any potential security vulnerabilities in reNgine, we encourage you to report them immediately via [huntr.dev](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine). **Please do not disclose vulnerabilities via Github issues/blogs/tweets after/before reporting to huntr.dev as this is explicitly against the disclosure policy of huntr.dev and reNgine and will not be considered for monetary rewards.** Please note that the reNgine maintainer does not set the bounty amount. The bounty reward is determined by an industry-first equation developed by huntr.dev to understand the popularity, impact and value of repositories to the open-source community. **What do I expect from security researchers?** * Patience: Please note that I am currently the only maintainer in reNgine and it will take some time to validate your report. I ask for your patience during this process. * Respect for privacy and security reports: Please do not publicly disclose any vulnerabilities (including GitHub issues) before or after reporting them on huntr.dev! This is against the disclosure policy and will not be rewarded. * Respect the rules **What do you get in return?** * Thanks from the maintainer * Monetary rewards * CVE ID(s) Please find the [FAQ](https://www.huntr.dev/faq) and [Responsible disclosure policy](https://www.huntr.dev/policy/) from huntr.dev.  ### License Distributed under the GNU GPL v3 License. See [LICENSE](LICENSE) for more information.  <p align="right">(ChatGPT was used to write some or most part of this README section.)</p>

<p align="center"> <a href="https://rengine.wiki"><img src=".github/screenshots/banner.gif" alt=""/></a> </p> <p align="center"><a href="https://github.com/yogeshojha/rengine/releases" target="_blank"><img src="https://img.shields.io/badge/version-v2.0.0-informational?&logo=none" alt="reNgine Latest Version" /></a>&nbsp;<a href="https://www.gnu.org/licenses/gpl-3.0" target="_blank"><img src="https://img.shields.io/badge/License-GPLv3-red.svg?&logo=none" alt="License" /></a>&nbsp;<a href="#" target="_blank"><img src="https://img.shields.io/badge/first--timers--only-friendly-blue.svg?&logo=none" alt="" /></a>&nbsp;<a href="https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine" target="_blank"><img src="https://cdn.huntr.dev/huntr_security_badge_mono.svg" alt="" /></a>&nbsp;</p> <p align="center"> <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Asia-2023-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/Open--Source--Summit-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://cyberweek.ae/2021/hitb-armory/" target="_blank"><img src="https://img.shields.io/badge/HITB--Armory-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=7uvP6MaQOX0" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://drive.google.com/file/d/1Bh8lbf-Dztt5ViHJVACyrXMiglyICPQ2/view?usp=sharing" target="_blank"><img src="https://img.shields.io/badge/Defcon--Demolabs--29-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=A1oNOIc0h5A" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Europe-2020-blue.svg?&logo=none" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml/badge.svg" alt="" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/actions/workflows/build.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/build.yml/badge.svg" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://discord.gg/H6WzebwX3H" target="_blank"><img src="https://img.shields.io/discord/880363103689277461" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://opensourcesecurityindex.io/" target="_blank" rel="noopener"> <img style="width: 282px; height: 56px" src="https://opensourcesecurityindex.io/badge.svg" alt="Open Source Security Index - Fastest Growing Open Source Security Projects" width="282" height="56" /> </a> </p> <h3>reNgine 2.0-jasper<br>Redefining the future of reconnaissance!</h3> <h4>What is reNgine?</h4> <p align="left">reNgine is your go-to web application reconnaissance suite that's designed to simplify and streamline the reconnaissance process for security professionals, penetration testers, and bug bounty hunters. With its highly configurable engines, data correlation capabilities, continuous monitoring, database-backed reconnaissance data, and an intuitive user interface, reNgine redefines how you gather critical information about your target web applications. Traditional reconnaissance tools often fall short in terms of configurability and efficiency. reNgine addresses these shortcomings and emerges as a excellent alternative to existing commercial tools. reNgine was created to address the limitations of traditional reconnaissance tools and provide a better alternative, even surpassing some commercial offerings. Whether you're a bug bounty hunter, a penetration tester, or a corporate security team, reNgine is your go-to solution for automating and enhancing your information-gathering efforts. </p> reNgine 2.0-jasper is out now, you can [watch reNgine 2.0-jasper release trailer here!](https://youtu.be/VwkOWqiWW5g) reNgine 2.0-Jasper would not have been possible without [@ocervell](https://github.com/ocervell) valuable contributions. [@ocervell](https://github.com/ocervell) did majority of the refactoring if not all and also added a ton of features. Together, we wish to shape the future of web application reconnaissance, and it's developers like [@ocervell](https://github.com/ocervell) and a [ton of other developers and hackers from our community](https://github.com/yogeshojha/rengine/graphs/contributors) who inspire and drive us forward. Thank you, [@ocervell](https://github.com/ocervell), for your outstanding work and unwavering commitment to reNgine. Checkout our contributers here: [Contributers](https://github.com/yogeshojha/rengine/graphs/contributors)  ### Documentation You can find detailed documentation at [https://rengine.wiki](https://rengine.wiki)  ### Table of Contents * [About reNgine](#about-rengine) * [Workflow](#workflow) * [Features](#features) * [Scan Engine](#scan-engine) * [Quick Installation](#quick-installation) * [What's new in reNgine 2.0](#changelog) * [Screenshots](#screenshots) * [Contributing](#contributing) * [reNgine Support](#rengine-support) * [Support and Sponsoring](#support-and-sponsoring) * [reNgine Bug Bounty Program](#rengine-bug-bounty-program) * [License](#license)  ### About reNgine reNgine is not an ordinary reconnaissance suite; it's a game-changer! We've turbocharged the traditional workflow with groundbreaking features that is sure to ease your reconnaissance game. reNgine redefines the art of reconnaissance with highly configurable scan engines, recon data correlation, continuous monitoring, GPT powered Vulnerability Report, Project Management and role based access control etc. 🦾&nbsp;&nbsp; reNgine has advanced reconnaissance capabilities, harnessing a range of open-source tools to deliver a comprehensive web application reconnaissance experience. With it's intuitive User Interface, it excels in subdomain discovery, pinpointing IP addresses and open ports, collecting endpoints, conducting directory and file fuzzing, capturing screenshots, and performing vulnerability scans. To summarize, it does end-to-end reconnaissance. With WHOIS identification and WAF detection, it offers deep insights into target domains. Additionally, reNgine also identifies misconfigured S3 buckets and find interesting subdomains and URLS, based on specific keywords to helps you identify your next target, making it an go to tool for efficient reconnaissance. 🗃️&nbsp; &nbsp; Say goodbye to recon data chaos! reNgine seamlessly integrates with a database, providing you with unmatched data correlation and organization. Forgot the hassle of grepping through json, txt or csv files. Plus, our custom query language lets you filter reconnaissance data effortlessly using natural language like operators such as filtering all alive subdomains with `http_status=200` and also filter all subdomains that are alive and has admin in name `http_status=200&name=admin` 🔧&nbsp;&nbsp; reNgine offers unparalleled flexibility through its highly configurable scan engines, based on a YAML-based configuration. It offers the freedom to create and customize recon scan engines based on any kind of requirement, users can tailor them to their specific objectives and preferences, from thread management to timeout settings and rate-limit configurations, everything is customizable. Additionally, reNgine offers a range of pre-configured scan engines right out of the box, including Full Scan, Passive Scan, Screenshot Gathering, and the OSINT Scan Engine. These ready-to-use engines eliminate the need for extensive manual setup, aligning perfectly with reNgine's core mission of simplifying the reconnaissance process and enabling users to effortlessly access the right reconnaissance data with minimal effort. 💎&nbsp;&nbsp;Subscans: Subscan is a game-changing feature in reNgine, setting it apart as the only open-source tool of its kind to offer this capability. With Subscan, waiting for the entire pipeline to complete is a thing of the past. Now, users can swiftly respond to newfound discoveries during reconnaissance. Whether you've stumbled upon an intriguing subdomain and wish to conduct a focused port scan or want to delve deeper with a vulnerability assessment, reNgine has you covered. 📃&nbsp;&nbsp; PDF Reports: In addition to its robust reconnaissance capabilities, reNgine goes the extra mile by simplifying the report generation process, recognizing the crucial role that PDF reports play in the realm of end-to-end reconnaissance. Users can effortlessly generate and customize PDF reports to suit their exact needs. Whether it's a Full Scan Report, Vulnerability Report, or a concise reconnaissance report, reNgine provides the flexibility to choose the report type that best communicates your findings. Moreover, the level of customization is unparalleled, allowing users to select report colors, fine-tune executive summaries, and even add personalized touches like company names and footers. With GPT integration, your reports aren't just a report, with remediation steps, and impacts, you get 360-degree view of the vulnerabilities you've uncovered. 🔖&nbsp; &nbsp; Say Hello to Projects! reNgine 2.0 introduces a powerful addition that enables you to efficiently organize your web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. Each projects will have separate dashboard and all the scan results will be separated from each projects, while scan engines and configuration will be shared across all the projects. ⚙️&nbsp; &nbsp; Roles and Permissions! Begining reNgine 2.0, we've taken your web application reconnaissance to a whole new level of control and security. Now, you can assign distinct roles to your team members—Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - 🔐 Sys Admin: Sys Admin is a super user that has permission to modify system and scan related configurations, scan engines, create new users, add new tools etc. Super user can initiate scans and subscans effortlessly. - 🔍 Penetration Tester: Penetration Tester will be allowed to modify and initiate scans and subscans, add or update targets, etc. A penetration tester will not be allowed to modify system configurations. - 📊 Auditor: Auditor can only view and download the report. An auditor can not change any system or scan related configurations nor can initiate any scans or subscans. 🚀&nbsp;&nbsp; GPT Vulnerability Report Generation: Get ready for the future of penetration testing reports with reNgine's groundbreaking feature: "GPT-Powered Report Generation"! With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments that read like they were written by a human security expert! **But that's not all!** Our GPT-driven reports go the extra mile by scouring the web for related news articles, blogs, and references, so you have a 360-degree view of the vulnerabilities you've uncovered. With reNgine 2.0 revolutionize your penetration testing game and impress your clients with reports that are not just informative but engaging and comprehensive with detailed analysis on impact assessment and remediation strategies. 🥷&nbsp;&nbsp; GPT-Powered Attack Surface Generation: With reNgine 2.0, reNgine seamlessly integrates with GPT to identify the attacks that you can likely perform on a subdomain. By making use of reconnaissance data such as page title, open ports, subdomain name etc, reNgine can advice you the attacks you could perform on a target. reNgine will also provide you the rationale on why the specific attack is likely to be successful. 🧭&nbsp;&nbsp;Continuous monitoring: Continuous monitoring is at the core of reNgine's mission, and it's robust continuous monitoring feature ensures that their targets are under constant scrutiny. With the flexibility to schedule scans at regular intervals, penetration testers can effortlessly stay informed about their targets. What sets reNgine apart is its seamless integration with popular notification channels such as Discord, Slack, and Telegram, delivering real-time alerts for newly discovered subdomains, vulnerabilities, or any changes in reconnaissance data.  ### Workflow <img src="https://github.com/yogeshojha/rengine/assets/17223002/10c475b8-b4a8-440d-9126-77fe2038a386">  ### Features * Reconnaissance: * Subdomain Discovery * IP and Open Ports Identification * Endpoints Discovery * Directory/Files fuzzing * Screenshot Gathering * Vulnerability Scan * Nuclei * Dalfox XSS Scanner * CRLFuzzer * Misconfigured S3 Scanner * WHOIS Identification * WAF Detection * OSINT Capabilities * Meta info Gathering * Employees Gathering * Email Address gathering * Google Dorking for sensitive info and urls * Projects, create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. * Perform Advanced Query lookup using natural language alike and, or, not operations * Highly configurable YAML-based Scan Engines * Support for Parallel Scans * Support for Subscans * Recon Data visualization * GPT Vulnerability Description, Impact and Remediation generation * GPT Attack Surface Generator * Multiple Roles and Permissions to cater a team's need * Customizable Alerts/Notifications on Slack, Discord, and Telegram * Automatically report Vulnerabilities to HackerOne * Recon Notes and Todos * Clocked Scans (Run reconnaissance exactly at X Hours and Y minutes) and Periodic Scans (Runs reconnaissance every X minutes/- hours/days/week) * Proxy Support * Screenshot Gallery with Filters * Powerful recon data filtering with autosuggestions * Recon Data changes, find new/removed subdomains/endpoints * Tag targets into the Organization * Smart Duplicate endpoint removal based on page title and content length to cleanup the reconnaissance data * Identify Interesting Subdomains * Custom GF patterns and custom Nuclei Templates * Edit tool-related configuration files (Nuclei, Subfinder, Naabu, amass) * Add external tools from Github/Go * Interoperable with other tools, Import/Export Subdomains/Endpoints * Import Targets via IP and/or CIDRs * Report Generation * Toolbox: Comes bundled with most commonly used tools during penetration testing such as whois lookup, CMS detector, CVE lookup, etc. * Identification of related domains and related TLDs for targets * Find actionable insights such as Most Common Vulnerability, Most Common CVE ID, Most Vulnerable Target/Subdomain, etc.  ### Scan Engine ```yaml subdomain_discovery: { 'uses_tools': [ 'subfinder', 'ctfr', 'sublist3r', 'tlsx', 'oneforall', 'netlas' ], 'enable_http_crawl': true, 'threads': 30, 'timeout': 5, } http_crawl: {} port_scan: { 'enable_http_crawl': true, 'timeout': 5, # 'exclude_ports': [], # 'exclude_subdomains': true, 'ports': ['top-100'], 'rate_limit': 150, 'threads': 30, 'passive': false, # 'use_naabu_config': false, # 'enable_nmap': true, # 'nmap_cmd': '', # 'nmap_script': '', # 'nmap_script_args': '' } osint: { 'discover': [ 'emails', 'metainfo', 'employees' ], 'dorks': [ 'login_pages', 'admin_panels', 'dashboard_pages', 'stackoverflow', 'social_media', 'project_management', 'code_sharing', 'config_files', 'jenkins', 'wordpress_files', 'php_error', 'exposed_documents', 'db_files', 'git_exposed' ], 'custom_dorks': [ { 'lookup_site': 'google.com', 'lookup_keywords': '/home/' }, { 'lookup_site': '_target_', 'lookup_extensions': 'jpg,png' } ], 'intensity': 'normal', 'documents_limit': 50 } dir_file_fuzz: { 'auto_calibration': true, 'enable_http_crawl': true, 'rate_limit': 150, 'extensions': ['html', 'php','git','yaml','conf','cnf','config','gz','env','log','db','mysql','bak','asp','aspx','txt','conf','sql','json','yml','pdf'], 'follow_redirect': false, 'max_time': 0, 'match_http_status': [200, 204], 'recursive_level': 2, 'stop_on_error': false, 'timeout': 5, 'threads': 30, 'wordlist_name': 'dicc' } fetch_url: { 'uses_tools': [ 'gospider', 'hakrawler', 'waybackurls', 'gospider', 'katana' ], 'remove_duplicate_endpoints': true, 'duplicate_fields': [ 'content_length', 'page_title' ], 'enable_http_crawl': true, 'gf_patterns': ['debug_logic', 'idor', 'interestingEXT', 'interestingparams', 'interestingsubs', 'lfi', 'rce', 'redirect', 'sqli', 'ssrf', 'ssti', 'xss'], 'ignore_file_extensions': ['png', 'jpg', 'jpeg', 'gif', 'mp4', 'mpeg', 'mp3'] # 'exclude_subdomains': true } vulnerability_scan: { 'run_nuclei': false, 'run_dalfox': false, 'run_crlfuzz': false, 'run_s3scanner': true, 'enable_http_crawl': true, 'concurrency': 50, 'intensity': 'normal', 'rate_limit': 150, 'retries': 1, 'timeout': 5, 'fetch_gpt_report': true, 'nuclei': { 'use_conf': false, 'severities': [ 'unknown', 'info', 'low', 'medium', 'high', 'critical' ], # 'tags': [], # 'templates': [], # 'custom_templates': [], }, 's3scanner': { 'threads': 100, 'providers': [ 'aws', 'gcp', 'digitalocean', 'dreamhost', 'linode' ] } } waf_detection: {} screenshot: { 'enable_http_crawl': true, 'intensity': 'normal', 'timeout': 10, 'threads': 40 } # custom_header: "Cookie: Test" ```  ### Quick Installation **Note:** Only Ubuntu/VPS 1. Clone this repo ```bash git clone https://github.com/yogeshojha/rengine && cd rengine ``` 1. Edit the `.env` file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```bash nano .env ``` 1. **Optional, only for non-interactive install**: In the `.env` file, **please make sure to change the super admin values!** ```bash DJANGO_SUPERUSER_USERNAME=yourUsername DJANGO_SUPERUSER_EMAIL=YourMail@example.com DJANGO_SUPERUSER_PASSWORD=yourStrongPassword ``` If you need to carry out a non-interactive installation, you can setup the login, email and password of the web interface admin directly from the .env file (instead of manually setting them from prompts during the installation process). This option can be interesting for automated installation (via ansible, vagrant, etc.). `DJANGO_SUPERUSER_USERNAME`: web interface admin username (used to login to the web interface). `DJANGO_SUPERUSER_EMAIL`: web interface admin email. `DJANGO_SUPERUSER_PASSWORD`: web interface admin password (used to login to the web interface). 1. In the dotenv file, you may also modify the Scaling Configurations ```bash MAX_CONCURRENCY=80 MIN_CONCURRENCY=10 ``` `MAX_CONCURRENCY`: This parameter specifies the maximum number of reNgine's concurrent Celery worker processes that can be spawned. In this case, it's set to 80, meaning that the application can utilize up to 80 concurrent worker processes to execute tasks concurrently. This is useful for handling a high volume of scans or when you want to scale up processing power during periods of high demand. If you have more CPU cores, you will need to increase this for maximised performance. `MIN_CONCURRENCY`: On the other hand, MIN_CONCURRENCY specifies the minimum number of concurrent worker processes that should be maintained, even during periods of lower demand. In this example, it's set to 10, which means that even when there are fewer tasks to process, at least 10 worker processes will be kept running. This helps ensure that the application can respond promptly to incoming tasks without the overhead of repeatedly starting and stopping worker processes. These settings allow for dynamic scaling of Celery workers, ensuring that the application efficiently manages its workload by adjusting the number of concurrent workers based on the workload's size and complexity 1. Run the installation script, Please keep an eye for any prompt, you will also be asked for username and password for reNgine. ```bash sudo ./install.sh ``` Or for a non-interactive installation, use `-n` argument (make sure you've modified the `.env` file before launching the installation). ```bash sudo ./install.sh -n ``` If `install.sh` does not have install permission, please change it, `chmod +x install.sh` **reNgine can now be accessed from <https://127.0.0.1> or if you're on the VPS <https://your_vps_ip_address>** **Unless you are on development branch, please do not access reNgine via any ports** ### Installation (Mac/Windows/Other) Installation instructions can be found at [https://reNgine.wiki/install/detailed/](https://reNgine.wiki/2.0/install/detailed/) ### Updating 1. Updating is as simple as running the following command: ```bash cd rengine && sudo ./update.sh ``` If `update.sh` does not have execution permissions, please change it, `sudo chmod +x update.sh` **NOTE:** if you're updating from 1.3.6 and you're getting a 'password authentication failed' error, consider uninstalling 1.3.6 first, then install 2.x.x as you'd normally do. ### Changelog [Please find the latest release notes and changelog here.](https://rengine.wiki/changelog/)  ### Screenshots #### Scan Results  #### General Usage <img src="https://user-images.githubusercontent.com/17223002/164993781-b6012995-522b-480a-a8bf-911193d35894.gif"> #### Initiating Subscan <img src="https://user-images.githubusercontent.com/17223002/164993749-1ad343d6-8ce7-43d6-aee7-b3add0321da7.gif"> #### Recon Data filtering <img src="https://user-images.githubusercontent.com/17223002/164993687-b63f3de8-e033-4ac0-808e-a2aa377d3cf8.gif"> #### Report Generation <img src="https://user-images.githubusercontent.com/17223002/164993689-c796c6cd-eb61-43f4-800d-08aba9740088.gif"> #### Toolbox <img src="https://user-images.githubusercontent.com/17223002/164993751-d687e88a-eb79-440f-9dc0-0ad006901620.gif"> #### Adding Custom tool in Tools Arsenal <img src="https://user-images.githubusercontent.com/17223002/164993670-466f6459-9499-498b-a9bd-526476d735a7.gif">  ### Contributing Contributions are what make the open-source community such an amazing place to learn, inspire and create. Every contributions you make is **greatly appreciated**. Your contributions can be as simple as fixing the indentation or UI, or as complex as adding new modules and features. See the [Contributing Guide](.github/CONTRIBUTING.md) to get started. You can also [join our Discord channel #development](https://discord.gg/JuhHdHTtwd) for any development related questions.  ### First-time Open Source contributors Please note that reNgine is beginner friendly. If you have never done open-source before, we encourage you to do so. **We will be happy and proud of your first PR ever.** You can start by resolving any [open issues](https://github.com/yogeshojha/rengine/issues).  ### reNgine Support Please do not use GitHub for support requests. Instead, [join our Discord channel #support](https://discord.gg/azv6fzhNCE).  ### Support and Sponsoring Over the past few years, I have been working hard on reNgine to add new features with the sole aim of making it the de facto standard for reconnaissance. I spend most of my free time and weekends working on reNgine. I do this in addition to my day job. I am happy to have received such overwhelming support from the community. But to keep this project alive, I am looking for financial support. | Paypal | Bitcoin | Ethereum | | :-------------------------------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | |[https://www.paypal.com/paypalme/yogeshojha11](https://www.paypal.com/paypalme/yogeshojha11) | `35AiKyNswNZ4TZUSdriHopSCjNMPi63BCX` | `0xe7A337Da6ff98A28513C26A7Fec8C9b42A63d346` OR * Add a [GitHub Star](https://github.com/yogeshojha/rengine) to the project. * Tweet about this project, or maybe blogs? * Maybe nominate me for [GitHub Stars?](https://stars.github.com/nominate/) * Join DigitalOcean using my [referral link](https://m.do.co/c/e353502d19fc) your profit is **$100** and I get $25 DO credit. This will help me test reNgine on VPS before I release any major features. It takes a considerable amount of time to add new features and make sure everything works. Donating is your way of saying: **reNgine is awesome**. Any support is greatly appreciated! Thank you!  ### reNgine Bug Bounty Program [](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine) Security researchers, welcome aboard! I'm excited to announce the reNgine bug bounty programme in collaboration with [huntr.dev](https://huntr.dev), which means that you will be rewarded for any vulnerabilities you find in reNgine. Thank you for your interest in reporting reNgine vulnerabilities! If you are aware of any potential security vulnerabilities in reNgine, we encourage you to report them immediately via [huntr.dev](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine). **Please do not disclose vulnerabilities via Github issues/blogs/tweets after/before reporting to huntr.dev as this is explicitly against the disclosure policy of huntr.dev and reNgine and will not be considered for monetary rewards.** Please note that the reNgine maintainer does not set the bounty amount. The bounty reward is determined by an industry-first equation developed by huntr.dev to understand the popularity, impact and value of repositories to the open-source community. **What do I expect from security researchers?** * Patience: Please note that I am currently the only maintainer in reNgine and it will take some time to validate your report. I ask for your patience during this process. * Respect for privacy and security reports: Please do not publicly disclose any vulnerabilities (including GitHub issues) before or after reporting them on huntr.dev! This is against the disclosure policy and will not be rewarded. * Respect the rules **What do you get in return?** * Thanks from the maintainer * Monetary rewards * CVE ID(s) Please find the [FAQ](https://www.huntr.dev/faq) and [Responsible disclosure policy](https://www.huntr.dev/policy/) from huntr.dev.  ### License Distributed under the GNU GPL v3 License. See [LICENSE](LICENSE) for more information.  <p align="right">(ChatGPT was used to write some or most part of this README section.)</p>

C0wnuts

3dd700357a4bd5701b07ede4511f66042655be00

64b7f291240b3b8853e3cec7ee6230827c97b907

```suggestion `DJANGO_SUPERUSER_PASSWORD`: web interface admin password (used to login to the web interface). ```

AnonymousWP

yogeshojha/rengine

Add non-interactive installation parameter

Add a non-interactive installation method via a new parameter to be passed to the install.sh script. Essential for automated/industrialized systems (e.g. via Ansible or another automated environment creation system).

2023-10-12 01:09:15+00:00

2023-11-21 12:49:22+00:00

README.md

<p align="center"> <a href="https://rengine.wiki"><img src=".github/screenshots/banner.gif" alt=""/></a> </p> <p align="center"><a href="https://github.com/yogeshojha/rengine/releases" target="_blank"><img src="https://img.shields.io/badge/version-v2.0.0-informational?&logo=none" alt="reNgine Latest Version" /></a>&nbsp;<a href="https://www.gnu.org/licenses/gpl-3.0" target="_blank"><img src="https://img.shields.io/badge/License-GPLv3-red.svg?&logo=none" alt="License" /></a>&nbsp;<a href="#" target="_blank"><img src="https://img.shields.io/badge/first--timers--only-friendly-blue.svg?&logo=none" alt="" /></a>&nbsp;<a href="https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine" target="_blank"><img src="https://cdn.huntr.dev/huntr_security_badge_mono.svg" alt="" /></a>&nbsp;</p> <p align="center"> <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Asia-2023-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/Open--Source--Summit-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://cyberweek.ae/2021/hitb-armory/" target="_blank"><img src="https://img.shields.io/badge/HITB--Armory-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=7uvP6MaQOX0" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://drive.google.com/file/d/1Bh8lbf-Dztt5ViHJVACyrXMiglyICPQ2/view?usp=sharing" target="_blank"><img src="https://img.shields.io/badge/Defcon--Demolabs--29-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=A1oNOIc0h5A" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Europe-2020-blue.svg?&logo=none" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml/badge.svg" alt="" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/actions/workflows/build.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/build.yml/badge.svg" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://discord.gg/H6WzebwX3H" target="_blank"><img src="https://img.shields.io/discord/880363103689277461" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://opensourcesecurityindex.io/" target="_blank" rel="noopener"> <img style="width: 282px; height: 56px" src="https://opensourcesecurityindex.io/badge.svg" alt="Open Source Security Index - Fastest Growing Open Source Security Projects" width="282" height="56" /> </a> </p> <h3>reNgine 2.0-jasper<br>Redefining the future of reconnaissance!</h3> <h4>What is reNgine?</h4> <p align="left">reNgine is your go-to web application reconnaissance suite that's designed to simplify and streamline the reconnaissance process for security professionals, penetration testers, and bug bounty hunters. With its highly configurable engines, data correlation capabilities, continuous monitoring, database-backed reconnaissance data, and an intuitive user interface, reNgine redefines how you gather critical information about your target web applications. Traditional reconnaissance tools often fall short in terms of configurability and efficiency. reNgine addresses these shortcomings and emerges as a excellent alternative to existing commercial tools. reNgine was created to address the limitations of traditional reconnaissance tools and provide a better alternative, even surpassing some commercial offerings. Whether you're a bug bounty hunter, a penetration tester, or a corporate security team, reNgine is your go-to solution for automating and enhancing your information-gathering efforts. </p> reNgine 2.0-jasper is out now, you can [watch reNgine 2.0-jasper release trailer here!](https://youtu.be/VwkOWqiWW5g) reNgine 2.0-Jasper would not have been possible without [@ocervell](https://github.com/ocervell) valuable contributions. [@ocervell](https://github.com/ocervell) did majority of the refactoring if not all and also added a ton of features. Together, we wish to shape the future of web application reconnaissance, and it's developers like [@ocervell](https://github.com/ocervell) and a [ton of other developers and hackers from our community](https://github.com/yogeshojha/rengine/graphs/contributors) who inspire and drive us forward. Thank you, [@ocervell](https://github.com/ocervell), for your outstanding work and unwavering commitment to reNgine. Checkout our contributers here: [Contributers](https://github.com/yogeshojha/rengine/graphs/contributors)  ### Documentation You can find detailed documentation at [https://rengine.wiki](https://rengine.wiki)  ### Table of Contents * [About reNgine](#about-rengine) * [Workflow](#workflow) * [Features](#features) * [Scan Engine](#scan-engine) * [Quick Installation](#quick-installation) * [What's new in reNgine 2.0](#changelog) * [Screenshots](#screenshots) * [Contributing](#contributing) * [reNgine Support](#rengine-support) * [Support and Sponsoring](#support-and-sponsoring) * [reNgine Bug Bounty Program](#rengine-bug-bounty-program) * [License](#license)  ### About reNgine reNgine is not an ordinary reconnaissance suite; it's a game-changer! We've turbocharged the traditional workflow with groundbreaking features that is sure to ease your reconnaissance game. reNgine redefines the art of reconnaissance with highly configurable scan engines, recon data correlation, continuous monitoring, GPT powered Vulnerability Report, Project Management and role based access control etc. 🦾&nbsp;&nbsp; reNgine has advanced reconnaissance capabilities, harnessing a range of open-source tools to deliver a comprehensive web application reconnaissance experience. With it's intuitive User Interface, it excels in subdomain discovery, pinpointing IP addresses and open ports, collecting endpoints, conducting directory and file fuzzing, capturing screenshots, and performing vulnerability scans. To summarize, it does end-to-end reconnaissance. With WHOIS identification and WAF detection, it offers deep insights into target domains. Additionally, reNgine also identifies misconfigured S3 buckets and find interesting subdomains and URLS, based on specific keywords to helps you identify your next target, making it an go to tool for efficient reconnaissance. 🗃️&nbsp; &nbsp; Say goodbye to recon data chaos! reNgine seamlessly integrates with a database, providing you with unmatched data correlation and organization. Forgot the hassle of grepping through json, txt or csv files. Plus, our custom query language lets you filter reconnaissance data effortlessly using natural language like operators such as filtering all alive subdomains with `http_status=200` and also filter all subdomains that are alive and has admin in name `http_status=200&name=admin` 🔧&nbsp;&nbsp; reNgine offers unparalleled flexibility through its highly configurable scan engines, based on a YAML-based configuration. It offers the freedom to create and customize recon scan engines based on any kind of requirement, users can tailor them to their specific objectives and preferences, from thread management to timeout settings and rate-limit configurations, everything is customizable. Additionally, reNgine offers a range of pre-configured scan engines right out of the box, including Full Scan, Passive Scan, Screenshot Gathering, and the OSINT Scan Engine. These ready-to-use engines eliminate the need for extensive manual setup, aligning perfectly with reNgine's core mission of simplifying the reconnaissance process and enabling users to effortlessly access the right reconnaissance data with minimal effort. 💎&nbsp;&nbsp;Subscans: Subscan is a game-changing feature in reNgine, setting it apart as the only open-source tool of its kind to offer this capability. With Subscan, waiting for the entire pipeline to complete is a thing of the past. Now, users can swiftly respond to newfound discoveries during reconnaissance. Whether you've stumbled upon an intriguing subdomain and wish to conduct a focused port scan or want to delve deeper with a vulnerability assessment, reNgine has you covered. 📃&nbsp;&nbsp; PDF Reports: In addition to its robust reconnaissance capabilities, reNgine goes the extra mile by simplifying the report generation process, recognizing the crucial role that PDF reports play in the realm of end-to-end reconnaissance. Users can effortlessly generate and customize PDF reports to suit their exact needs. Whether it's a Full Scan Report, Vulnerability Report, or a concise reconnaissance report, reNgine provides the flexibility to choose the report type that best communicates your findings. Moreover, the level of customization is unparalleled, allowing users to select report colors, fine-tune executive summaries, and even add personalized touches like company names and footers. With GPT integration, your reports aren't just a report, with remediation steps, and impacts, you get 360-degree view of the vulnerabilities you've uncovered. 🔖&nbsp; &nbsp; Say Hello to Projects! reNgine 2.0 introduces a powerful addition that enables you to efficiently organize your web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. Each projects will have separate dashboard and all the scan results will be separated from each projects, while scan engines and configuration will be shared across all the projects. ⚙️&nbsp; &nbsp; Roles and Permissions! Begining reNgine 2.0, we've taken your web application reconnaissance to a whole new level of control and security. Now, you can assign distinct roles to your team members—Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - 🔐 Sys Admin: Sys Admin is a super user that has permission to modify system and scan related configurations, scan engines, create new users, add new tools etc. Super user can initiate scans and subscans effortlessly. - 🔍 Penetration Tester: Penetration Tester will be allowed to modify and initiate scans and subscans, add or update targets, etc. A penetration tester will not be allowed to modify system configurations. - 📊 Auditor: Auditor can only view and download the report. An auditor can not change any system or scan related configurations nor can initiate any scans or subscans. 🚀&nbsp;&nbsp; GPT Vulnerability Report Generation: Get ready for the future of penetration testing reports with reNgine's groundbreaking feature: "GPT-Powered Report Generation"! With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments that read like they were written by a human security expert! **But that's not all!** Our GPT-driven reports go the extra mile by scouring the web for related news articles, blogs, and references, so you have a 360-degree view of the vulnerabilities you've uncovered. With reNgine 2.0 revolutionize your penetration testing game and impress your clients with reports that are not just informative but engaging and comprehensive with detailed analysis on impact assessment and remediation strategies. 🥷&nbsp;&nbsp; GPT-Powered Attack Surface Generation: With reNgine 2.0, reNgine seamlessly integrates with GPT to identify the attacks that you can likely perform on a subdomain. By making use of reconnaissance data such as page title, open ports, subdomain name etc, reNgine can advice you the attacks you could perform on a target. reNgine will also provide you the rationale on why the specific attack is likely to be successful. 🧭&nbsp;&nbsp;Continuous monitoring: Continuous monitoring is at the core of reNgine's mission, and it's robust continuous monitoring feature ensures that their targets are under constant scrutiny. With the flexibility to schedule scans at regular intervals, penetration testers can effortlessly stay informed about their targets. What sets reNgine apart is its seamless integration with popular notification channels such as Discord, Slack, and Telegram, delivering real-time alerts for newly discovered subdomains, vulnerabilities, or any changes in reconnaissance data.  ### Workflow <img src="https://github.com/yogeshojha/rengine/assets/17223002/10c475b8-b4a8-440d-9126-77fe2038a386">  ### Features * Reconnaissance: * Subdomain Discovery * IP and Open Ports Identification * Endpoints Discovery * Directory/Files fuzzing * Screenshot Gathering * Vulnerability Scan * Nuclei * Dalfox XSS Scanner * CRLFuzzer * Misconfigured S3 Scanner * WHOIS Identification * WAF Detection * OSINT Capabilities * Meta info Gathering * Employees Gathering * Email Address gathering * Google Dorking for sensitive info and urls * Projects, create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. * Perform Advanced Query lookup using natural language alike and, or, not operations * Highly configurable YAML-based Scan Engines * Support for Parallel Scans * Support for Subscans * Recon Data visualization * GPT Vulnerability Description, Impact and Remediation generation * GPT Attack Surface Generator * Multiple Roles and Permissions to cater a team's need * Customizable Alerts/Notifications on Slack, Discord, and Telegram * Automatically report Vulnerabilities to HackerOne * Recon Notes and Todos * Clocked Scans (Run reconnaissance exactly at X Hours and Y minutes) and Periodic Scans (Runs reconnaissance every X minutes/- hours/days/week) * Proxy Support * Screenshot Gallery with Filters * Powerful recon data filtering with autosuggestions * Recon Data changes, find new/removed subdomains/endpoints * Tag targets into the Organization * Smart Duplicate endpoint removal based on page title and content length to cleanup the reconnaissance data * Identify Interesting Subdomains * Custom GF patterns and custom Nuclei Templates * Edit tool-related configuration files (Nuclei, Subfinder, Naabu, amass) * Add external tools from Github/Go * Interoperable with other tools, Import/Export Subdomains/Endpoints * Import Targets via IP and/or CIDRs * Report Generation * Toolbox: Comes bundled with most commonly used tools during penetration testing such as whois lookup, CMS detector, CVE lookup, etc. * Identification of related domains and related TLDs for targets * Find actionable insights such as Most Common Vulnerability, Most Common CVE ID, Most Vulnerable Target/Subdomain, etc.  ### Scan Engine ```yaml subdomain_discovery: { 'uses_tools': [ 'subfinder', 'ctfr', 'sublist3r', 'tlsx', 'oneforall', 'netlas' ], 'enable_http_crawl': true, 'threads': 30, 'timeout': 5, } http_crawl: {} port_scan: { 'enable_http_crawl': true, 'timeout': 5, # 'exclude_ports': [], # 'exclude_subdomains': true, 'ports': ['top-100'], 'rate_limit': 150, 'threads': 30, 'passive': false, # 'use_naabu_config': false, # 'enable_nmap': true, # 'nmap_cmd': '', # 'nmap_script': '', # 'nmap_script_args': '' } osint: { 'discover': [ 'emails', 'metainfo', 'employees' ], 'dorks': [ 'login_pages', 'admin_panels', 'dashboard_pages', 'stackoverflow', 'social_media', 'project_management', 'code_sharing', 'config_files', 'jenkins', 'wordpress_files', 'php_error', 'exposed_documents', 'db_files', 'git_exposed' ], 'custom_dorks': [ { 'lookup_site': 'google.com', 'lookup_keywords': '/home/' }, { 'lookup_site': '_target_', 'lookup_extensions': 'jpg,png' } ], 'intensity': 'normal', 'documents_limit': 50 } dir_file_fuzz: { 'auto_calibration': true, 'enable_http_crawl': true, 'rate_limit': 150, 'extensions': ['html', 'php','git','yaml','conf','cnf','config','gz','env','log','db','mysql','bak','asp','aspx','txt','conf','sql','json','yml','pdf'], 'follow_redirect': false, 'max_time': 0, 'match_http_status': [200, 204], 'recursive_level': 2, 'stop_on_error': false, 'timeout': 5, 'threads': 30, 'wordlist_name': 'dicc' } fetch_url: { 'uses_tools': [ 'gospider', 'hakrawler', 'waybackurls', 'gospider', 'katana' ], 'remove_duplicate_endpoints': true, 'duplicate_fields': [ 'content_length', 'page_title' ], 'enable_http_crawl': true, 'gf_patterns': ['debug_logic', 'idor', 'interestingEXT', 'interestingparams', 'interestingsubs', 'lfi', 'rce', 'redirect', 'sqli', 'ssrf', 'ssti', 'xss'], 'ignore_file_extensions': ['png', 'jpg', 'jpeg', 'gif', 'mp4', 'mpeg', 'mp3'] # 'exclude_subdomains': true } vulnerability_scan: { 'run_nuclei': false, 'run_dalfox': false, 'run_crlfuzz': false, 'run_s3scanner': true, 'enable_http_crawl': true, 'concurrency': 50, 'intensity': 'normal', 'rate_limit': 150, 'retries': 1, 'timeout': 5, 'fetch_gpt_report': true, 'nuclei': { 'use_conf': false, 'severities': [ 'unknown', 'info', 'low', 'medium', 'high', 'critical' ], # 'tags': [], # 'templates': [], # 'custom_templates': [], }, 's3scanner': { 'threads': 100, 'providers': [ 'aws', 'gcp', 'digitalocean', 'dreamhost', 'linode' ] } } waf_detection: {} screenshot: { 'enable_http_crawl': true, 'intensity': 'normal', 'timeout': 10, 'threads': 40 } # custom_header: "Cookie: Test" ```  ### Quick Installation **Note:** Only Ubuntu/VPS 1. Clone this repo ```bash git clone https://github.com/yogeshojha/rengine && cd rengine ``` 1. Edit the dotenv file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```bash nano .env ``` 1. In the dotenv file, you may also modify the Scaling Configurations ```bash MAX_CONCURRENCY=80 MIN_CONCURRENCY=10 ``` MAX_CONCURRENCY: This parameter specifies the maximum number of reNgine's concurrent Celery worker processes that can be spawned. In this case, it's set to 80, meaning that the application can utilize up to 80 concurrent worker processes to execute tasks concurrently. This is useful for handling a high volume of scans or when you want to scale up processing power during periods of high demand. If you have more CPU cores, you will need to increase this for maximised performance. MIN_CONCURRENCY: On the other hand, MIN_CONCURRENCY specifies the minimum number of concurrent worker processes that should be maintained, even during periods of lower demand. In this example, it's set to 10, which means that even when there are fewer tasks to process, at least 10 worker processes will be kept running. This helps ensure that the application can respond promptly to incoming tasks without the overhead of repeatedly starting and stopping worker processes. These settings allow for dynamic scaling of Celery workers, ensuring that the application efficiently manages its workload by adjusting the number of concurrent workers based on the workload's size and complexity 1. Run the installation script, Please keep an eye for any prompt, you will also be asked for username and password for reNgine. ```bash sudo ./install.sh ``` If `install.sh` does not have install permission, please change it, `chmod +x install.sh` **reNgine can now be accessed from <https://127.0.0.1> or if you're on the VPS <https://your_vps_ip_address>** **Unless you are on development branch, please do not access reNgine via any ports** ### Installation (Mac/Windows/Other) Installation instructions can be found at [https://reNgine.wiki/install/detailed/](https://reNgine.wiki/2.0/install/detailed/) ### Updating 1. Updating is as simple as running the following command: ```bash cd rengine && sudo ./update.sh ``` If `update.sh` does not have execution permissions, please change it, `sudo chmod +x update.sh` **NOTE:** if you're updating from 1.3.6 and you're getting a 'password authentication failed' error, consider uninstalling 1.3.6 first, then install 2.x.x as you'd normally do. ### Changelog [Please find the latest release notes and changelog here.](https://rengine.wiki/changelog/)  ### Screenshots #### Scan Results  #### General Usage <img src="https://user-images.githubusercontent.com/17223002/164993781-b6012995-522b-480a-a8bf-911193d35894.gif"> #### Initiating Subscan <img src="https://user-images.githubusercontent.com/17223002/164993749-1ad343d6-8ce7-43d6-aee7-b3add0321da7.gif"> #### Recon Data filtering <img src="https://user-images.githubusercontent.com/17223002/164993687-b63f3de8-e033-4ac0-808e-a2aa377d3cf8.gif"> #### Report Generation <img src="https://user-images.githubusercontent.com/17223002/164993689-c796c6cd-eb61-43f4-800d-08aba9740088.gif"> #### Toolbox <img src="https://user-images.githubusercontent.com/17223002/164993751-d687e88a-eb79-440f-9dc0-0ad006901620.gif"> #### Adding Custom tool in Tools Arsenal <img src="https://user-images.githubusercontent.com/17223002/164993670-466f6459-9499-498b-a9bd-526476d735a7.gif">  ### Contributing Contributions are what make the open-source community such an amazing place to learn, inspire and create. Every contributions you make is **greatly appreciated**. Your contributions can be as simple as fixing the indentation or UI, or as complex as adding new modules and features. See the [Contributing Guide](.github/CONTRIBUTING.md) to get started. You can also [join our Discord channel #development](https://discord.gg/JuhHdHTtwd) for any development related questions.  ### First-time Open Source contributors Please note that reNgine is beginner friendly. If you have never done open-source before, we encourage you to do so. **We will be happy and proud of your first PR ever.** You can start by resolving any [open issues](https://github.com/yogeshojha/rengine/issues).  ### reNgine Support Please do not use GitHub for support requests. Instead, [join our Discord channel #support](https://discord.gg/azv6fzhNCE).  ### Support and Sponsoring Over the past few years, I have been working hard on reNgine to add new features with the sole aim of making it the de facto standard for reconnaissance. I spend most of my free time and weekends working on reNgine. I do this in addition to my day job. I am happy to have received such overwhelming support from the community. But to keep this project alive, I am looking for financial support. | Paypal | Bitcoin | Ethereum | | :-------------------------------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | |[https://www.paypal.com/paypalme/yogeshojha11](https://www.paypal.com/paypalme/yogeshojha11) | `35AiKyNswNZ4TZUSdriHopSCjNMPi63BCX` | `0xe7A337Da6ff98A28513C26A7Fec8C9b42A63d346` OR * Add a [GitHub Star](https://github.com/yogeshojha/rengine) to the project. * Tweet about this project, or maybe blogs? * Maybe nominate me for [GitHub Stars?](https://stars.github.com/nominate/) * Join DigitalOcean using my [referral link](https://m.do.co/c/e353502d19fc) your profit is **$100** and I get $25 DO credit. This will help me test reNgine on VPS before I release any major features. It takes a considerable amount of time to add new features and make sure everything works. Donating is your way of saying: **reNgine is awesome**. Any support is greatly appreciated! Thank you!  ### reNgine Bug Bounty Program [](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine) Security researchers, welcome aboard! I'm excited to announce the reNgine bug bounty programme in collaboration with [huntr.dev](https://huntr.dev), which means that you will be rewarded for any vulnerabilities you find in reNgine. Thank you for your interest in reporting reNgine vulnerabilities! If you are aware of any potential security vulnerabilities in reNgine, we encourage you to report them immediately via [huntr.dev](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine). **Please do not disclose vulnerabilities via Github issues/blogs/tweets after/before reporting to huntr.dev as this is explicitly against the disclosure policy of huntr.dev and reNgine and will not be considered for monetary rewards.** Please note that the reNgine maintainer does not set the bounty amount. The bounty reward is determined by an industry-first equation developed by huntr.dev to understand the popularity, impact and value of repositories to the open-source community. **What do I expect from security researchers?** * Patience: Please note that I am currently the only maintainer in reNgine and it will take some time to validate your report. I ask for your patience during this process. * Respect for privacy and security reports: Please do not publicly disclose any vulnerabilities (including GitHub issues) before or after reporting them on huntr.dev! This is against the disclosure policy and will not be rewarded. * Respect the rules **What do you get in return?** * Thanks from the maintainer * Monetary rewards * CVE ID(s) Please find the [FAQ](https://www.huntr.dev/faq) and [Responsible disclosure policy](https://www.huntr.dev/policy/) from huntr.dev.  ### License Distributed under the GNU GPL v3 License. See [LICENSE](LICENSE) for more information.  <p align="right">(ChatGPT was used to write some or most part of this README section.)</p>

<p align="center"> <a href="https://rengine.wiki"><img src=".github/screenshots/banner.gif" alt=""/></a> </p> <p align="center"><a href="https://github.com/yogeshojha/rengine/releases" target="_blank"><img src="https://img.shields.io/badge/version-v2.0.0-informational?&logo=none" alt="reNgine Latest Version" /></a>&nbsp;<a href="https://www.gnu.org/licenses/gpl-3.0" target="_blank"><img src="https://img.shields.io/badge/License-GPLv3-red.svg?&logo=none" alt="License" /></a>&nbsp;<a href="#" target="_blank"><img src="https://img.shields.io/badge/first--timers--only-friendly-blue.svg?&logo=none" alt="" /></a>&nbsp;<a href="https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine" target="_blank"><img src="https://cdn.huntr.dev/huntr_security_badge_mono.svg" alt="" /></a>&nbsp;</p> <p align="center"> <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Asia-2023-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/Open--Source--Summit-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://cyberweek.ae/2021/hitb-armory/" target="_blank"><img src="https://img.shields.io/badge/HITB--Armory-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=7uvP6MaQOX0" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://drive.google.com/file/d/1Bh8lbf-Dztt5ViHJVACyrXMiglyICPQ2/view?usp=sharing" target="_blank"><img src="https://img.shields.io/badge/Defcon--Demolabs--29-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=A1oNOIc0h5A" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Europe-2020-blue.svg?&logo=none" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml/badge.svg" alt="" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/actions/workflows/build.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/build.yml/badge.svg" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://discord.gg/H6WzebwX3H" target="_blank"><img src="https://img.shields.io/discord/880363103689277461" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://opensourcesecurityindex.io/" target="_blank" rel="noopener"> <img style="width: 282px; height: 56px" src="https://opensourcesecurityindex.io/badge.svg" alt="Open Source Security Index - Fastest Growing Open Source Security Projects" width="282" height="56" /> </a> </p> <h3>reNgine 2.0-jasper<br>Redefining the future of reconnaissance!</h3> <h4>What is reNgine?</h4> <p align="left">reNgine is your go-to web application reconnaissance suite that's designed to simplify and streamline the reconnaissance process for security professionals, penetration testers, and bug bounty hunters. With its highly configurable engines, data correlation capabilities, continuous monitoring, database-backed reconnaissance data, and an intuitive user interface, reNgine redefines how you gather critical information about your target web applications. Traditional reconnaissance tools often fall short in terms of configurability and efficiency. reNgine addresses these shortcomings and emerges as a excellent alternative to existing commercial tools. reNgine was created to address the limitations of traditional reconnaissance tools and provide a better alternative, even surpassing some commercial offerings. Whether you're a bug bounty hunter, a penetration tester, or a corporate security team, reNgine is your go-to solution for automating and enhancing your information-gathering efforts. </p> reNgine 2.0-jasper is out now, you can [watch reNgine 2.0-jasper release trailer here!](https://youtu.be/VwkOWqiWW5g) reNgine 2.0-Jasper would not have been possible without [@ocervell](https://github.com/ocervell) valuable contributions. [@ocervell](https://github.com/ocervell) did majority of the refactoring if not all and also added a ton of features. Together, we wish to shape the future of web application reconnaissance, and it's developers like [@ocervell](https://github.com/ocervell) and a [ton of other developers and hackers from our community](https://github.com/yogeshojha/rengine/graphs/contributors) who inspire and drive us forward. Thank you, [@ocervell](https://github.com/ocervell), for your outstanding work and unwavering commitment to reNgine. Checkout our contributers here: [Contributers](https://github.com/yogeshojha/rengine/graphs/contributors)  ### Documentation You can find detailed documentation at [https://rengine.wiki](https://rengine.wiki)  ### Table of Contents * [About reNgine](#about-rengine) * [Workflow](#workflow) * [Features](#features) * [Scan Engine](#scan-engine) * [Quick Installation](#quick-installation) * [What's new in reNgine 2.0](#changelog) * [Screenshots](#screenshots) * [Contributing](#contributing) * [reNgine Support](#rengine-support) * [Support and Sponsoring](#support-and-sponsoring) * [reNgine Bug Bounty Program](#rengine-bug-bounty-program) * [License](#license)  ### About reNgine reNgine is not an ordinary reconnaissance suite; it's a game-changer! We've turbocharged the traditional workflow with groundbreaking features that is sure to ease your reconnaissance game. reNgine redefines the art of reconnaissance with highly configurable scan engines, recon data correlation, continuous monitoring, GPT powered Vulnerability Report, Project Management and role based access control etc. 🦾&nbsp;&nbsp; reNgine has advanced reconnaissance capabilities, harnessing a range of open-source tools to deliver a comprehensive web application reconnaissance experience. With it's intuitive User Interface, it excels in subdomain discovery, pinpointing IP addresses and open ports, collecting endpoints, conducting directory and file fuzzing, capturing screenshots, and performing vulnerability scans. To summarize, it does end-to-end reconnaissance. With WHOIS identification and WAF detection, it offers deep insights into target domains. Additionally, reNgine also identifies misconfigured S3 buckets and find interesting subdomains and URLS, based on specific keywords to helps you identify your next target, making it an go to tool for efficient reconnaissance. 🗃️&nbsp; &nbsp; Say goodbye to recon data chaos! reNgine seamlessly integrates with a database, providing you with unmatched data correlation and organization. Forgot the hassle of grepping through json, txt or csv files. Plus, our custom query language lets you filter reconnaissance data effortlessly using natural language like operators such as filtering all alive subdomains with `http_status=200` and also filter all subdomains that are alive and has admin in name `http_status=200&name=admin` 🔧&nbsp;&nbsp; reNgine offers unparalleled flexibility through its highly configurable scan engines, based on a YAML-based configuration. It offers the freedom to create and customize recon scan engines based on any kind of requirement, users can tailor them to their specific objectives and preferences, from thread management to timeout settings and rate-limit configurations, everything is customizable. Additionally, reNgine offers a range of pre-configured scan engines right out of the box, including Full Scan, Passive Scan, Screenshot Gathering, and the OSINT Scan Engine. These ready-to-use engines eliminate the need for extensive manual setup, aligning perfectly with reNgine's core mission of simplifying the reconnaissance process and enabling users to effortlessly access the right reconnaissance data with minimal effort. 💎&nbsp;&nbsp;Subscans: Subscan is a game-changing feature in reNgine, setting it apart as the only open-source tool of its kind to offer this capability. With Subscan, waiting for the entire pipeline to complete is a thing of the past. Now, users can swiftly respond to newfound discoveries during reconnaissance. Whether you've stumbled upon an intriguing subdomain and wish to conduct a focused port scan or want to delve deeper with a vulnerability assessment, reNgine has you covered. 📃&nbsp;&nbsp; PDF Reports: In addition to its robust reconnaissance capabilities, reNgine goes the extra mile by simplifying the report generation process, recognizing the crucial role that PDF reports play in the realm of end-to-end reconnaissance. Users can effortlessly generate and customize PDF reports to suit their exact needs. Whether it's a Full Scan Report, Vulnerability Report, or a concise reconnaissance report, reNgine provides the flexibility to choose the report type that best communicates your findings. Moreover, the level of customization is unparalleled, allowing users to select report colors, fine-tune executive summaries, and even add personalized touches like company names and footers. With GPT integration, your reports aren't just a report, with remediation steps, and impacts, you get 360-degree view of the vulnerabilities you've uncovered. 🔖&nbsp; &nbsp; Say Hello to Projects! reNgine 2.0 introduces a powerful addition that enables you to efficiently organize your web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. Each projects will have separate dashboard and all the scan results will be separated from each projects, while scan engines and configuration will be shared across all the projects. ⚙️&nbsp; &nbsp; Roles and Permissions! Begining reNgine 2.0, we've taken your web application reconnaissance to a whole new level of control and security. Now, you can assign distinct roles to your team members—Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - 🔐 Sys Admin: Sys Admin is a super user that has permission to modify system and scan related configurations, scan engines, create new users, add new tools etc. Super user can initiate scans and subscans effortlessly. - 🔍 Penetration Tester: Penetration Tester will be allowed to modify and initiate scans and subscans, add or update targets, etc. A penetration tester will not be allowed to modify system configurations. - 📊 Auditor: Auditor can only view and download the report. An auditor can not change any system or scan related configurations nor can initiate any scans or subscans. 🚀&nbsp;&nbsp; GPT Vulnerability Report Generation: Get ready for the future of penetration testing reports with reNgine's groundbreaking feature: "GPT-Powered Report Generation"! With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments that read like they were written by a human security expert! **But that's not all!** Our GPT-driven reports go the extra mile by scouring the web for related news articles, blogs, and references, so you have a 360-degree view of the vulnerabilities you've uncovered. With reNgine 2.0 revolutionize your penetration testing game and impress your clients with reports that are not just informative but engaging and comprehensive with detailed analysis on impact assessment and remediation strategies. 🥷&nbsp;&nbsp; GPT-Powered Attack Surface Generation: With reNgine 2.0, reNgine seamlessly integrates with GPT to identify the attacks that you can likely perform on a subdomain. By making use of reconnaissance data such as page title, open ports, subdomain name etc, reNgine can advice you the attacks you could perform on a target. reNgine will also provide you the rationale on why the specific attack is likely to be successful. 🧭&nbsp;&nbsp;Continuous monitoring: Continuous monitoring is at the core of reNgine's mission, and it's robust continuous monitoring feature ensures that their targets are under constant scrutiny. With the flexibility to schedule scans at regular intervals, penetration testers can effortlessly stay informed about their targets. What sets reNgine apart is its seamless integration with popular notification channels such as Discord, Slack, and Telegram, delivering real-time alerts for newly discovered subdomains, vulnerabilities, or any changes in reconnaissance data.  ### Workflow <img src="https://github.com/yogeshojha/rengine/assets/17223002/10c475b8-b4a8-440d-9126-77fe2038a386">  ### Features * Reconnaissance: * Subdomain Discovery * IP and Open Ports Identification * Endpoints Discovery * Directory/Files fuzzing * Screenshot Gathering * Vulnerability Scan * Nuclei * Dalfox XSS Scanner * CRLFuzzer * Misconfigured S3 Scanner * WHOIS Identification * WAF Detection * OSINT Capabilities * Meta info Gathering * Employees Gathering * Email Address gathering * Google Dorking for sensitive info and urls * Projects, create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. * Perform Advanced Query lookup using natural language alike and, or, not operations * Highly configurable YAML-based Scan Engines * Support for Parallel Scans * Support for Subscans * Recon Data visualization * GPT Vulnerability Description, Impact and Remediation generation * GPT Attack Surface Generator * Multiple Roles and Permissions to cater a team's need * Customizable Alerts/Notifications on Slack, Discord, and Telegram * Automatically report Vulnerabilities to HackerOne * Recon Notes and Todos * Clocked Scans (Run reconnaissance exactly at X Hours and Y minutes) and Periodic Scans (Runs reconnaissance every X minutes/- hours/days/week) * Proxy Support * Screenshot Gallery with Filters * Powerful recon data filtering with autosuggestions * Recon Data changes, find new/removed subdomains/endpoints * Tag targets into the Organization * Smart Duplicate endpoint removal based on page title and content length to cleanup the reconnaissance data * Identify Interesting Subdomains * Custom GF patterns and custom Nuclei Templates * Edit tool-related configuration files (Nuclei, Subfinder, Naabu, amass) * Add external tools from Github/Go * Interoperable with other tools, Import/Export Subdomains/Endpoints * Import Targets via IP and/or CIDRs * Report Generation * Toolbox: Comes bundled with most commonly used tools during penetration testing such as whois lookup, CMS detector, CVE lookup, etc. * Identification of related domains and related TLDs for targets * Find actionable insights such as Most Common Vulnerability, Most Common CVE ID, Most Vulnerable Target/Subdomain, etc.  ### Scan Engine ```yaml subdomain_discovery: { 'uses_tools': [ 'subfinder', 'ctfr', 'sublist3r', 'tlsx', 'oneforall', 'netlas' ], 'enable_http_crawl': true, 'threads': 30, 'timeout': 5, } http_crawl: {} port_scan: { 'enable_http_crawl': true, 'timeout': 5, # 'exclude_ports': [], # 'exclude_subdomains': true, 'ports': ['top-100'], 'rate_limit': 150, 'threads': 30, 'passive': false, # 'use_naabu_config': false, # 'enable_nmap': true, # 'nmap_cmd': '', # 'nmap_script': '', # 'nmap_script_args': '' } osint: { 'discover': [ 'emails', 'metainfo', 'employees' ], 'dorks': [ 'login_pages', 'admin_panels', 'dashboard_pages', 'stackoverflow', 'social_media', 'project_management', 'code_sharing', 'config_files', 'jenkins', 'wordpress_files', 'php_error', 'exposed_documents', 'db_files', 'git_exposed' ], 'custom_dorks': [ { 'lookup_site': 'google.com', 'lookup_keywords': '/home/' }, { 'lookup_site': '_target_', 'lookup_extensions': 'jpg,png' } ], 'intensity': 'normal', 'documents_limit': 50 } dir_file_fuzz: { 'auto_calibration': true, 'enable_http_crawl': true, 'rate_limit': 150, 'extensions': ['html', 'php','git','yaml','conf','cnf','config','gz','env','log','db','mysql','bak','asp','aspx','txt','conf','sql','json','yml','pdf'], 'follow_redirect': false, 'max_time': 0, 'match_http_status': [200, 204], 'recursive_level': 2, 'stop_on_error': false, 'timeout': 5, 'threads': 30, 'wordlist_name': 'dicc' } fetch_url: { 'uses_tools': [ 'gospider', 'hakrawler', 'waybackurls', 'gospider', 'katana' ], 'remove_duplicate_endpoints': true, 'duplicate_fields': [ 'content_length', 'page_title' ], 'enable_http_crawl': true, 'gf_patterns': ['debug_logic', 'idor', 'interestingEXT', 'interestingparams', 'interestingsubs', 'lfi', 'rce', 'redirect', 'sqli', 'ssrf', 'ssti', 'xss'], 'ignore_file_extensions': ['png', 'jpg', 'jpeg', 'gif', 'mp4', 'mpeg', 'mp3'] # 'exclude_subdomains': true } vulnerability_scan: { 'run_nuclei': false, 'run_dalfox': false, 'run_crlfuzz': false, 'run_s3scanner': true, 'enable_http_crawl': true, 'concurrency': 50, 'intensity': 'normal', 'rate_limit': 150, 'retries': 1, 'timeout': 5, 'fetch_gpt_report': true, 'nuclei': { 'use_conf': false, 'severities': [ 'unknown', 'info', 'low', 'medium', 'high', 'critical' ], # 'tags': [], # 'templates': [], # 'custom_templates': [], }, 's3scanner': { 'threads': 100, 'providers': [ 'aws', 'gcp', 'digitalocean', 'dreamhost', 'linode' ] } } waf_detection: {} screenshot: { 'enable_http_crawl': true, 'intensity': 'normal', 'timeout': 10, 'threads': 40 } # custom_header: "Cookie: Test" ```  ### Quick Installation **Note:** Only Ubuntu/VPS 1. Clone this repo ```bash git clone https://github.com/yogeshojha/rengine && cd rengine ``` 1. Edit the `.env` file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```bash nano .env ``` 1. **Optional, only for non-interactive install**: In the `.env` file, **please make sure to change the super admin values!** ```bash DJANGO_SUPERUSER_USERNAME=yourUsername DJANGO_SUPERUSER_EMAIL=YourMail@example.com DJANGO_SUPERUSER_PASSWORD=yourStrongPassword ``` If you need to carry out a non-interactive installation, you can setup the login, email and password of the web interface admin directly from the .env file (instead of manually setting them from prompts during the installation process). This option can be interesting for automated installation (via ansible, vagrant, etc.). `DJANGO_SUPERUSER_USERNAME`: web interface admin username (used to login to the web interface). `DJANGO_SUPERUSER_EMAIL`: web interface admin email. `DJANGO_SUPERUSER_PASSWORD`: web interface admin password (used to login to the web interface). 1. In the dotenv file, you may also modify the Scaling Configurations ```bash MAX_CONCURRENCY=80 MIN_CONCURRENCY=10 ``` `MAX_CONCURRENCY`: This parameter specifies the maximum number of reNgine's concurrent Celery worker processes that can be spawned. In this case, it's set to 80, meaning that the application can utilize up to 80 concurrent worker processes to execute tasks concurrently. This is useful for handling a high volume of scans or when you want to scale up processing power during periods of high demand. If you have more CPU cores, you will need to increase this for maximised performance. `MIN_CONCURRENCY`: On the other hand, MIN_CONCURRENCY specifies the minimum number of concurrent worker processes that should be maintained, even during periods of lower demand. In this example, it's set to 10, which means that even when there are fewer tasks to process, at least 10 worker processes will be kept running. This helps ensure that the application can respond promptly to incoming tasks without the overhead of repeatedly starting and stopping worker processes. These settings allow for dynamic scaling of Celery workers, ensuring that the application efficiently manages its workload by adjusting the number of concurrent workers based on the workload's size and complexity 1. Run the installation script, Please keep an eye for any prompt, you will also be asked for username and password for reNgine. ```bash sudo ./install.sh ``` Or for a non-interactive installation, use `-n` argument (make sure you've modified the `.env` file before launching the installation). ```bash sudo ./install.sh -n ``` If `install.sh` does not have install permission, please change it, `chmod +x install.sh` **reNgine can now be accessed from <https://127.0.0.1> or if you're on the VPS <https://your_vps_ip_address>** **Unless you are on development branch, please do not access reNgine via any ports** ### Installation (Mac/Windows/Other) Installation instructions can be found at [https://reNgine.wiki/install/detailed/](https://reNgine.wiki/2.0/install/detailed/) ### Updating 1. Updating is as simple as running the following command: ```bash cd rengine && sudo ./update.sh ``` If `update.sh` does not have execution permissions, please change it, `sudo chmod +x update.sh` **NOTE:** if you're updating from 1.3.6 and you're getting a 'password authentication failed' error, consider uninstalling 1.3.6 first, then install 2.x.x as you'd normally do. ### Changelog [Please find the latest release notes and changelog here.](https://rengine.wiki/changelog/)  ### Screenshots #### Scan Results  #### General Usage <img src="https://user-images.githubusercontent.com/17223002/164993781-b6012995-522b-480a-a8bf-911193d35894.gif"> #### Initiating Subscan <img src="https://user-images.githubusercontent.com/17223002/164993749-1ad343d6-8ce7-43d6-aee7-b3add0321da7.gif"> #### Recon Data filtering <img src="https://user-images.githubusercontent.com/17223002/164993687-b63f3de8-e033-4ac0-808e-a2aa377d3cf8.gif"> #### Report Generation <img src="https://user-images.githubusercontent.com/17223002/164993689-c796c6cd-eb61-43f4-800d-08aba9740088.gif"> #### Toolbox <img src="https://user-images.githubusercontent.com/17223002/164993751-d687e88a-eb79-440f-9dc0-0ad006901620.gif"> #### Adding Custom tool in Tools Arsenal <img src="https://user-images.githubusercontent.com/17223002/164993670-466f6459-9499-498b-a9bd-526476d735a7.gif">  ### Contributing Contributions are what make the open-source community such an amazing place to learn, inspire and create. Every contributions you make is **greatly appreciated**. Your contributions can be as simple as fixing the indentation or UI, or as complex as adding new modules and features. See the [Contributing Guide](.github/CONTRIBUTING.md) to get started. You can also [join our Discord channel #development](https://discord.gg/JuhHdHTtwd) for any development related questions.  ### First-time Open Source contributors Please note that reNgine is beginner friendly. If you have never done open-source before, we encourage you to do so. **We will be happy and proud of your first PR ever.** You can start by resolving any [open issues](https://github.com/yogeshojha/rengine/issues).  ### reNgine Support Please do not use GitHub for support requests. Instead, [join our Discord channel #support](https://discord.gg/azv6fzhNCE).  ### Support and Sponsoring Over the past few years, I have been working hard on reNgine to add new features with the sole aim of making it the de facto standard for reconnaissance. I spend most of my free time and weekends working on reNgine. I do this in addition to my day job. I am happy to have received such overwhelming support from the community. But to keep this project alive, I am looking for financial support. | Paypal | Bitcoin | Ethereum | | :-------------------------------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | |[https://www.paypal.com/paypalme/yogeshojha11](https://www.paypal.com/paypalme/yogeshojha11) | `35AiKyNswNZ4TZUSdriHopSCjNMPi63BCX` | `0xe7A337Da6ff98A28513C26A7Fec8C9b42A63d346` OR * Add a [GitHub Star](https://github.com/yogeshojha/rengine) to the project. * Tweet about this project, or maybe blogs? * Maybe nominate me for [GitHub Stars?](https://stars.github.com/nominate/) * Join DigitalOcean using my [referral link](https://m.do.co/c/e353502d19fc) your profit is **$100** and I get $25 DO credit. This will help me test reNgine on VPS before I release any major features. It takes a considerable amount of time to add new features and make sure everything works. Donating is your way of saying: **reNgine is awesome**. Any support is greatly appreciated! Thank you!  ### reNgine Bug Bounty Program [](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine) Security researchers, welcome aboard! I'm excited to announce the reNgine bug bounty programme in collaboration with [huntr.dev](https://huntr.dev), which means that you will be rewarded for any vulnerabilities you find in reNgine. Thank you for your interest in reporting reNgine vulnerabilities! If you are aware of any potential security vulnerabilities in reNgine, we encourage you to report them immediately via [huntr.dev](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine). **Please do not disclose vulnerabilities via Github issues/blogs/tweets after/before reporting to huntr.dev as this is explicitly against the disclosure policy of huntr.dev and reNgine and will not be considered for monetary rewards.** Please note that the reNgine maintainer does not set the bounty amount. The bounty reward is determined by an industry-first equation developed by huntr.dev to understand the popularity, impact and value of repositories to the open-source community. **What do I expect from security researchers?** * Patience: Please note that I am currently the only maintainer in reNgine and it will take some time to validate your report. I ask for your patience during this process. * Respect for privacy and security reports: Please do not publicly disclose any vulnerabilities (including GitHub issues) before or after reporting them on huntr.dev! This is against the disclosure policy and will not be rewarded. * Respect the rules **What do you get in return?** * Thanks from the maintainer * Monetary rewards * CVE ID(s) Please find the [FAQ](https://www.huntr.dev/faq) and [Responsible disclosure policy](https://www.huntr.dev/policy/) from huntr.dev.  ### License Distributed under the GNU GPL v3 License. See [LICENSE](LICENSE) for more information.  <p align="right">(ChatGPT was used to write some or most part of this README section.)</p>

C0wnuts

3dd700357a4bd5701b07ede4511f66042655be00

64b7f291240b3b8853e3cec7ee6230827c97b907

```suggestion Or for a non-interactive installation, use `-n` argument (make sure you've modified the `.env` file before launching the installation). ```

AnonymousWP

yogeshojha/rengine

2.0-jasper release

### Added - Projects: Projects allow you to efficiently organize their web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. - Roles and Permissions: assign distinct roles to your team members: Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - GPT-powered Report Generation: With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments. - API Vault: This feature allows you to organize your API keys such as OpenAI or Netlas API keys. - GPT-powered Attack Surface Generation - URL gathering now is much more efficient, removing duplicate endpoints based on similar HTTP Responses, having the same content_lenth, or page_title. Custom duplicate fields can also be set from the scan engine configuration. - URL Path filtering while initiating scan: For instance, if we want to scan only endpoints starting with https://example.com/start/, we can pass the /start as a path filter while starting the scan. @ocervell - Expanding Target Concept: reNgine 2.0 now accepts IPs, URLS, etc as targets. (#678, #658) Excellent work by @ocervell - A ton of refactoring on reNgine's core to improve scan efficiency. Massive kudos to @ocervell - Created a custom celery workflow to be able to run several tasks in parallel that are not dependent on each other, such OSINT task and subdomain discovery will run in parallel, and directory and file fuzzing, vulnerability scan, screenshot gathering etc. will run in parallel after port scan or url fetching is completed. This will increase the efficiency of scans and instead of having one long flow of tasks, they can run independently on their own. @ocervell - Refactored all tasks to run asynchronously @ocervell - Added a stream_command that allows to read the output of a command live: this means the UI is updated with results while the command runs and does not have to wait until the task completes. Excellent work by @ocervell - Pwndb is now replaced by h8mail. @ocervell - Group Scan Results: reNgine 2.0 allows to group of subdomains based on similar page titles and HTTP status, and also vulnerability grouping based on the same vulnerability title and severity. - Added Support for Nmap: reNgine 2.0 allows to run Nmap scripts and vuln scans on ports found by Naabu. @ocervell - Added support for Shared Scan Variables in Scan Engine Configuration: - `enable_http_crawl`: (true/false) You can disable it to be more stealthy or focus on something different than HTTP - `timeout`: set timeout for all tasks - `rate_limit`: set rate limit for all tasks - `retries`: set retries for all tasks - `custom_header`: set the custom header for all tasks - Added Dalfox for XSS Vulnerability Scan - Added CRLFuzz for CRLF Vulnerability Scan - Added S3Scanner for scanning misconfigured S3 buckets - Improve OSINT Dork results, now detects admin panels, login pages and dashboards - Added Custom Dorks - Improved UI for vulnerability results, clicking on each vulnerability will open up a sidebar with vulnerability details. - Added HTTP Request and Response in vulnerability Results - Under Admin Settings, added an option to allow add/remove/deactivate additional users - Added Option to Preview Scan Report instead of forcing to download - Added Katana for crawling and spidering URLs - Added Netlas for Whois and subdomain gathering - Added TLSX for subdomain gathering - Added CTFR for subdomain gathering - Added historical IP in whois section ### Fixes - GF patterns do not run on 404 endpoints (#574 closed) - Fixes for retrieving whois data (#693 closed) - Related/Associated Domains in Whois section is now fixed ### Removed - Removed pwndb and tor related to it. - Removed tor for pwndb

2023-10-02 07:51:35+00:00

2023-10-07 10:37:23+00:00

README.md

<p align="center"> <a href="https://rengine.wiki"><img src=".github/screenshots/banner.gif" alt=""/></a> </p> <p align="center"><a href="https://github.com/yogeshojha/rengine/releases" target="_blank"><img src="https://img.shields.io/badge/version-v1.2.0-informational?&logo=none" alt="reNgine Latest Version" /></a>&nbsp;<a href="https://www.gnu.org/licenses/gpl-3.0" target="_blank"><img src="https://img.shields.io/badge/License-GPLv3-red.svg?&logo=none" alt="License" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/issues" target="_blank"><img src="https://img.shields.io/github/issues/yogeshojha/rengine?color=red&logo=none" alt="reNgine Issues" /></a>&nbsp;<a href="#" target="_blank"><img src="https://img.shields.io/badge/first--timers--only-friendly-blue.svg?&logo=none" alt="" /></a>&nbsp;<a href="https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine" target="_blank"><img src="https://cdn.huntr.dev/huntr_security_badge_mono.svg" alt="" /></a>&nbsp;</p> <p align="center"> <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/Open--Source--Summit-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://cyberweek.ae/2021/hitb-armory/" target="_blank"><img src="https://img.shields.io/badge/HITB--Armory-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=7uvP6MaQOX0" target="_blank"><img src="https://img.shields.io/badge/Black--Hat--Arsenal-USA--2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://drive.google.com/file/d/1Bh8lbf-Dztt5ViHJVACyrXMiglyICPQ2/view?usp=sharing" target="_blank"><img src="https://img.shields.io/badge/Defcon--Demolabs--29-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=A1oNOIc0h5A" target="_blank"><img src="https://img.shields.io/badge/Black--Hat--Arsenal-Europe--2020-blue.svg?&logo=none" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml/badge.svg" alt="" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/actions/workflows/build.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/build.yml/badge.svg" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://discord.gg/H6WzebwX3H" target="_blank"><img src="https://img.shields.io/discord/880363103689277461" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://opensourcesecurityindex.io/" target="_blank" rel="noopener"> <img style="width: 282px; height: 56px" src="https://opensourcesecurityindex.io/badge.svg" alt="Open Source Security Index - Fastest Growing Open Source Security Projects" width="282" height="56" /> </a> </p> <h3>reNgine 1.1<br>More than just recon!</h3> <h4>The only web application recon tool you will ever need!</h4> <p>Quickly discover the attack surface, and identify vulnerabilities using highly customizable and powerful scan engines. Enjoy peace of mind with reNgine's continuous monitoring, deeper reconnaissance, and open-source powered Vulnerability Scanner.</p> <h4>What is reNgine?</h4> <p align="left">reNgine is a web application reconnaissance suite that focuses on a highly configurable streamlined reconnaissance process via engines, reconnaissance data correlation, continuous monitoring, database backed reconnaissance data and a simple yet intuitive user interface. With features such as sub-scan, deeper co-relation, report generation, etc., reNgine aims to fill the gap in traditional reconnaissance tools and is likely to be a better alternative to existing commercial tools. reNgine makes it easy for penetration testers and security auditors to gather reconnaissance data with minimal configuration. </p>  <p align="center"> ⭐<a href="https://rengine.wiki">reNgine Documentation</a> · <a href="https://rengine.wiki/changelog/">What's new</a> · <a href="https://github.com/yogeshojha/rengine/blob/master/.github/CONTRIBUTING.md">Contribute</a> · <a href="https://github.com/yogeshojha/rengine/issues">Report Bug</a> · <a href="https://github.com/yogeshojha/rengine/issues">Request Feature</a>⭐ </p>  ## Table of Contents * [About reNgine](#about-rengine) * [Features](#features) * [Documentation](#documentation) * [Quick Installation](#quick-installation) * [What's new in reNgine](#changelog) * [reNgine Bug Bounty Program](#rengine-bug-bounty-program) * [Screenshots](#screenshots) * [Contributing](#contributing) * [reNgine Support](#rengine-support) * [Related Projects](#related-projects) * [Support and Sponsoring](#support-and-sponsoring) * [License](#license)  ### reNgine 2.0 codenamed Jasper I am currently working on reNgine 2.0, which will probably be announced sometime between May and August 2023. reNgine 2.0 will be the most advanced reNgine ever, a lot of work will be done in how scans are performed, things such as Pause and Resume Scan, Axiom Integration, more deeper correlation, Project Options, Multiple Tenants, etc. Please submit your feature requests via GitHub issues.  ## About reNgine You can watch [reNgine 1.1 release trailer here.](https://www.youtube.com/watch?v=iy_6F7Vq8Lo) (Recommended) <img src="https://user-images.githubusercontent.com/17223002/164993688-50eb95f2-3653-4ef7-bd3b-ef7a096824ea.jpeg">  reNgine is a web application reconnaissance suite with a focus on a highly configurable, streamlined reconnaissance process. reNgine is backed by a database, with data correlation and organisation, the custom query "like" language for filtering reconnaissance data, reNgine aims to address the shortcomings of the traditional reconnaissance workflow. The developers behind reNgine understand that reconnaissance data can be huge and manually searching for records to attack can be tedious, so features such as identifying subdomains of interest help penetration testers focus on attack rather than reconnaissance. reNgine also focuses on continuous monitoring. Penetration testers can choose to schedule the scan at regular intervals and be notified via notification channels such as Discord, Slack and Telegram of any new subdomains or vulnerabilities identified, or any changes to the recon data. Interoperability is something every reconnaissance tool needs, and reNgine is no different. Starting with reNgine 1.0, we have added features such as import and export of subdomains, endpoints, GF pattern matched endpoints, etc. This allows you to use your favourite reconnaissance workflow in conjunction with reNgine. PDF reports are something every individual or team needs. From reNgine 1.1, reNgine also comes with the option to download PDF reports. You can also choose the type of report, a full scan report or just a reconnaissance report. We also understand that PDF reports need to be customisable. Choose the colour of the report you want, customise the executive summary, etc. You choose how your PDF report looks! reNgine features highly configurable scan engines based on YAML, allowing penetration testers to create as many reconnaissance engines of their choice as they like, configure them as they like, and use them against any targets for scanning. These engines allow penetration testers to use the tools of their choice, with the configuration of their choice. Out of the box, reNgine comes with several scan engines such as Full Scan, Passive Scan, Screenshot Gathering, OSINT Engine, etc. Our focus has always been on finding the right reconnaissance data with the least amount of effort. After several discussions with fellow hackers/pentesters, a screenshot gallery was a must, reNgine also comes with a screenshot gallery, and what's more exciting than having a screenshot gallery with filters, filter screenshots with HTTP status, technology, ports and services. We also want our fellow hackers to stay ahead of the game, so reNgine also comes with automatic vulnerability reporting (ATM only Hackerone is supported, other platforms may come soon). This allows hackers to define their vulnerability reporting template and reNgine will do the rest of the work to report the vulnerability as soon as it is identified.  <img src="https://user-images.githubusercontent.com/17223002/164993945-aabdbb4a-2b9d-4951-ba27-5f2f5abd1d8b.gif"> ## Features * Reconnaissance: Subdomain Discovery, IP and Open Ports Identification, Endpoints Discovery, Directory and Files fuzzing, - Screenshot gathering, Vulnerability scan using Nuclei, WHOIS Identification, WAF Detection etc. * Highly configurable YAML-based Scan Engines * Support for Parallel Scans and Subscans * Automatically report Vulnerabilities to HackerOne * Recon Data visualization * OSINT Capabilities (Meta info Gathering, Employees Gathering, Email Address with an option to look password in the leaked database, - dorks, etc.) * Customizable Alerts/Notifications on Slack, Discord, and Telegram * Perform Advanced Query lookup using natural language alike and, or, not operations * Recon Notes and Todos * Clocked Scans (Run reconnaissance exactly at X Hours and Y minutes) and Periodic Scans (Runs reconnaissance every X minutes/- hours/days/week) * Proxy Support * Screenshot Gallery with Filters * Powerful recon data filtering with autosuggestions * Recon Data changes, find new/removed subdomains/endpoints * Tag targets into the Organization * Identify Interesting Subdomains * Custom GF patterns and custom Nuclei Templates * Edit tool-related configuration files (Nuclei, Subfinder, Naabu, amass) * Add external tools from Github/Go * Interoperable with other tools, Import/Export Subdomains/Endpoints * Import Targets via IP and/or CIDRs * Report Generation * Toolbox: Comes bundled with most commonly used tools such as whois lookup, CMS detector, CVE lookup, etc. * Identification of related domains and related TLDs for targets * Find actionable insights such as Most Common Vulnerability, Most Common CVE ID, Most Vulnerable Target/Subdomain, etc.  ## Documentation You can find reNgine documentation at [https://rengine.wiki](https://rengine.wiki)  ## Quick Installation **Note:** Only Ubuntu/VPS 1. Clone this repo ```bash git clone https://github.com/yogeshojha/rengine && cd rengine ``` 1. Edit the dotenv file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```bash nano .env ``` 1. Run the installation script, Please keep an eye for any prompt, you will also be asked for username and password for reNgine. ```bash sudo ./install.sh ``` If `install.sh` does not have install permission, please change it, `chmod +x install.sh` **reNgine can now be accessed from <https://127.0.0.1> or if you're on the VPS <https://your_vps_ip_address>** A detailed installation guide can also be found [here](https://www.rffuste.com/2022/05/23/rengine-a-brief-overview/). Thanks to Rubén! ## Installation (Mac/Windows/Other) Installation instructions can be found at [https://reNgine.wiki/install/detailed/](https://reNgine.wiki/install/detailed/) ## Updating 1. Updating is as simple as running the following command: ```bash sudo ./update.sh ``` If `update.sh` does not have execution permissions, please change it, `sudo chmod +x update.sh` ## Changelog [Please find the latest release notes and changelog here.](https://rengine.wiki/changelog/)  ## reNgine Bug Bounty Program [](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine) Security researchers, welcome aboard! I'm excited to announce the reNgine bug bounty programme in collaboration with [huntr.dev](https://huntr.dev), which means that you will be rewarded for any vulnerabilities you find in reNgine. Thank you for your interest in reporting reNgine vulnerabilities! If you are aware of any potential security vulnerabilities in reNgine, we encourage you to report them immediately via [huntr.dev](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine). **Please do not disclose vulnerabilities via Github issues/blogs/tweets after/before reporting to huntr.dev as this is explicitly against the disclosure policy of huntr.dev and reNgine and will not be considered for monetary rewards.** Please note that the reNgine maintainer does not set the bounty amount. The bounty reward is determined by an industry-first equation developed by huntr.dev to understand the popularity, impact and value of repositories to the open-source community. **What do I expect from security researchers?** * Patience: Please note that I am currently the only maintainer in reNgine and it will take some time to validate your report. I ask for your patience during this process. * Respect for privacy and security reports: Please do not publicly disclose any vulnerabilities (including GitHub issues) before or after reporting them on huntr.dev! This is against the disclosure policy and will not be rewarded. * Respect the rules **What do you get in return?** * Thanks from the maintainer * Monetary rewards * CVE ID(s) Please find the [FAQ](https://www.huntr.dev/faq) and [Responsible disclosure policy](https://www.huntr.dev/policy/) from huntr.dev.  ## Screenshots ### Scan Results  ### General Usage <img src="https://user-images.githubusercontent.com/17223002/164993781-b6012995-522b-480a-a8bf-911193d35894.gif"> ### Initiating Subscan <img src="https://user-images.githubusercontent.com/17223002/164993749-1ad343d6-8ce7-43d6-aee7-b3add0321da7.gif"> ### Recon Data filtering <img src="https://user-images.githubusercontent.com/17223002/164993687-b63f3de8-e033-4ac0-808e-a2aa377d3cf8.gif"> ### Report Generation <img src="https://user-images.githubusercontent.com/17223002/164993689-c796c6cd-eb61-43f4-800d-08aba9740088.gif"> ### Toolbox <img src="https://user-images.githubusercontent.com/17223002/164993751-d687e88a-eb79-440f-9dc0-0ad006901620.gif"> ### Adding Custom tool in Tools Arsenal <img src="https://user-images.githubusercontent.com/17223002/164993670-466f6459-9499-498b-a9bd-526476d735a7.gif">  ## Contributing Contributions are what make the open-source community such an amazing place to learn, inspire and create. Every contributions you make is **greatly appreciated**. Your contributions can be as simple as fixing the indentation or UI, or as complex as adding new modules and features. See the [Contributing Guide](.github/CONTRIBUTING.md) to get started. You can also [join our Discord channel #development](https://discord.gg/JuhHdHTtwd) for any development related questions.  ### First-time Open Source contributors Please note that reNgine is beginner friendly. If you have never done open-source before, we encourage you to do so. **We will be happy and proud of your first PR ever.** You can start by resolving any [open issues](https://github.com/yogeshojha/rengine/issues).  ## reNgine Support Please do not use GitHub for support requests. Instead, [join our Discord channel #support](https://discord.gg/azv6fzhNCE)  ## Related Projects There are many other great reconnaissance frameworks out there, you can use reNgine in conjunction with those tools. But they are great in their own right, and can sometimes produce better results than reNgine. * [ReconFTW](https://github.com/six2dez/reconftw#sample-video) * [Reconmap](https://github.com/reconmap/reconmap)  ## Support and Sponsoring Over the past few years, I have been working hard on reNgine to add new features with the sole aim of making it the de facto standard for reconnaissance. I spend most of my free time and weekends working on reNgine. I do this in addition to my day job. I am happy to have received such overwhelming support from the community. But to keep this project alive, I am looking for financial support. | Paypal | Bitcoin | Ethereum | | :-------------------------------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | |[https://www.paypal.com/paypalme/yogeshojha11](https://www.paypal.com/paypalme/yogeshojha11) | `35AiKyNswNZ4TZUSdriHopSCjNMPi63BCX` | `0xe7A337Da6ff98A28513C26A7Fec8C9b42A63d346` OR * Add a [GitHub Star](https://github.com/yogeshojha/rengine) to the project. * Tweet about this project, or maybe blogs? * Maybe nominate me for [GitHub Stars?](https://stars.github.com/nominate/) * Join DigitalOcean using my [referral link](https://m.do.co/c/e353502d19fc) your profit is **$100** and I get $25 DO credit. This will help me test reNgine on VPS before I release any major features. It takes a considerable amount of time to add new features and make sure everything works. Donating is your way of saying: **reNgine is awesome**. Any support is greatly appreciated! Thank you!  ## License Distributed under the GNU GPL v3 License. See [LICENSE](LICENSE) for more information. 

<p align="center"> <a href="https://rengine.wiki"><img src=".github/screenshots/banner.gif" alt=""/></a> </p> <p align="center"><a href="https://github.com/yogeshojha/rengine/releases" target="_blank"><img src="https://img.shields.io/badge/version-v2.0.0-informational?&logo=none" alt="reNgine Latest Version" /></a>&nbsp;<a href="https://www.gnu.org/licenses/gpl-3.0" target="_blank"><img src="https://img.shields.io/badge/License-GPLv3-red.svg?&logo=none" alt="License" /></a>&nbsp;<a href="#" target="_blank"><img src="https://img.shields.io/badge/first--timers--only-friendly-blue.svg?&logo=none" alt="" /></a>&nbsp;<a href="https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine" target="_blank"><img src="https://cdn.huntr.dev/huntr_security_badge_mono.svg" alt="" /></a>&nbsp;</p> <p align="center"> <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Asia-2023-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/Open--Source--Summit-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://cyberweek.ae/2021/hitb-armory/" target="_blank"><img src="https://img.shields.io/badge/HITB--Armory-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=7uvP6MaQOX0" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://drive.google.com/file/d/1Bh8lbf-Dztt5ViHJVACyrXMiglyICPQ2/view?usp=sharing" target="_blank"><img src="https://img.shields.io/badge/Defcon--Demolabs--29-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=A1oNOIc0h5A" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Europe-2020-blue.svg?&logo=none" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml/badge.svg" alt="" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/actions/workflows/build.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/build.yml/badge.svg" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://discord.gg/H6WzebwX3H" target="_blank"><img src="https://img.shields.io/discord/880363103689277461" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://opensourcesecurityindex.io/" target="_blank" rel="noopener"> <img style="width: 282px; height: 56px" src="https://opensourcesecurityindex.io/badge.svg" alt="Open Source Security Index - Fastest Growing Open Source Security Projects" width="282" height="56" /> </a> </p> <h3>reNgine 2.0-jasper<br>Redefining the future of reconnaissance!</h3> <h4>What is reNgine?</h4> <p align="left">reNgine is your go-to web application reconnaissance suite that's designed to simplify and streamline the reconnaissance process for security professionals, penetration testers, and bug bounty hunters. With its highly configurable engines, data correlation capabilities, continuous monitoring, database-backed reconnaissance data, and an intuitive user interface, reNgine redefines how you gather critical information about your target web applications. Traditional reconnaissance tools often fall short in terms of configurability and efficiency. reNgine addresses these shortcomings and emerges as a excellent alternative to existing commercial tools. reNgine was created to address the limitations of traditional reconnaissance tools and provide a better alternative, even surpassing some commercial offerings. Whether you're a bug bounty hunter, a penetration tester, or a corporate security team, reNgine is your go-to solution for automating and enhancing your information-gathering efforts. </p> reNgine 2.0-jasper is out now, you can [watch reNgine 2.0-jasper release trailer here!](https://youtu.be/VwkOWqiWW5g) reNgine 2.0-Jasper would not have been possible without [@ocervell](https://github.com/ocervell) valuable contributions. [@ocervell](https://github.com/ocervell) did majority of the refactoring if not all and also added a ton of features. Together, we wish to shape the future of web application reconnaissance, and it's developers like [@ocervell](https://github.com/ocervell) and a [ton of other developers and hackers from our community](https://github.com/yogeshojha/rengine/graphs/contributors) who inspire and drive us forward. Thank you, [@ocervell](https://github.com/ocervell), for your outstanding work and unwavering commitment to reNgine. Checkout our contributers here: [Contributers](https://github.com/yogeshojha/rengine/graphs/contributors)  ### Documentation You can find detailed documentation at [https://rengine.wiki](https://rengine.wiki)  ### Table of Contents * [About reNgine](#about-rengine) * [Workflow](#workflow) * [Features](#features) * [Scan Engine](#scan-engine) * [Quick Installation](#quick-installation) * [What's new in reNgine 2.0](#changelog) * [Screenshots](#screenshots) * [Contributing](#contributing) * [reNgine Support](#rengine-support) * [Support and Sponsoring](#support-and-sponsoring) * [reNgine Bug Bounty Program](#rengine-bug-bounty-program) * [License](#license)  ### About reNgine reNgine is not an ordinary reconnaissance suite; it's a game-changer! We've turbocharged the traditional workflow with groundbreaking features that is sure to ease your reconnaissance game. reNgine redefines the art of reconnaissance with highly configurable scan engines, recon data correlation, continuous monitoring, GPT powered Vulnerability Report, Project Management and role based access control etc. 🦾&nbsp;&nbsp; reNgine has advanced reconnaissance capabilities, harnessing a range of open-source tools to deliver a comprehensive web application reconnaissance experience. With it's intuitive User Interface, it excels in subdomain discovery, pinpointing IP addresses and open ports, collecting endpoints, conducting directory and file fuzzing, capturing screenshots, and performing vulnerability scans. To summarize, it does end-to-end reconnaissance. With WHOIS identification and WAF detection, it offers deep insights into target domains. Additionally, reNgine also identifies misconfigured S3 buckets and find interesting subdomains and URLS, based on specific keywords to helps you identify your next target, making it an go to tool for efficient reconnaissance. 🗃️&nbsp; &nbsp; Say goodbye to recon data chaos! reNgine seamlessly integrates with a database, providing you with unmatched data correlation and organization. Forgot the hassle of grepping through json, txt or csv files. Plus, our custom query language lets you filter reconnaissance data effortlessly using natural language like operators such as filtering all alive subdomains with `http_status=200` and also filter all subdomains that are alive and has admin in name `http_status=200&name=admin` 🔧&nbsp;&nbsp; reNgine offers unparalleled flexibility through its highly configurable scan engines, based on a YAML-based configuration. It offers the freedom to create and customize recon scan engines based on any kind of requirement, users can tailor them to their specific objectives and preferences, from thread management to timeout settings and rate-limit configurations, everything is customizable. Additionally, reNgine offers a range of pre-configured scan engines right out of the box, including Full Scan, Passive Scan, Screenshot Gathering, and the OSINT Scan Engine. These ready-to-use engines eliminate the need for extensive manual setup, aligning perfectly with reNgine's core mission of simplifying the reconnaissance process and enabling users to effortlessly access the right reconnaissance data with minimal effort. 💎&nbsp;&nbsp;Subscans: Subscan is a game-changing feature in reNgine, setting it apart as the only open-source tool of its kind to offer this capability. With Subscan, waiting for the entire pipeline to complete is a thing of the past. Now, users can swiftly respond to newfound discoveries during reconnaissance. Whether you've stumbled upon an intriguing subdomain and wish to conduct a focused port scan or want to delve deeper with a vulnerability assessment, reNgine has you covered. 📃&nbsp;&nbsp; PDF Reports: In addition to its robust reconnaissance capabilities, reNgine goes the extra mile by simplifying the report generation process, recognizing the crucial role that PDF reports play in the realm of end-to-end reconnaissance. Users can effortlessly generate and customize PDF reports to suit their exact needs. Whether it's a Full Scan Report, Vulnerability Report, or a concise reconnaissance report, reNgine provides the flexibility to choose the report type that best communicates your findings. Moreover, the level of customization is unparalleled, allowing users to select report colors, fine-tune executive summaries, and even add personalized touches like company names and footers. With GPT integration, your reports aren't just a report, with remediation steps, and impacts, you get 360-degree view of the vulnerabilities you've uncovered. 🔖&nbsp; &nbsp; Say Hello to Projects! reNgine 2.0 introduces a powerful addition that enables you to efficiently organize your web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. Each projects will have separate dashboard and all the scan results will be separated from each projects, while scan engines and configuration will be shared across all the projects. ⚙️&nbsp; &nbsp; Roles and Permissions! Begining reNgine 2.0, we've taken your web application reconnaissance to a whole new level of control and security. Now, you can assign distinct roles to your team members—Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - 🔐 Sys Admin: Sys Admin is a super user that has permission to modify system and scan related configurations, scan engines, create new users, add new tools etc. Super user can initiate scans and subscans effortlessly. - 🔍 Penetration Tester: Penetration Tester will be allowed to modify and initiate scans and subscans, add or update targets, etc. A penetration tester will not be allowed to modify system configurations. - 📊 Auditor: Auditor can only view and download the report. An auditor can not change any system or scan related configurations nor can initiate any scans or subscans. 🚀&nbsp;&nbsp; GPT Vulnerability Report Generation: Get ready for the future of penetration testing reports with reNgine's groundbreaking feature: "GPT-Powered Report Generation"! With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments that read like they were written by a human security expert! **But that's not all!** Our GPT-driven reports go the extra mile by scouring the web for related news articles, blogs, and references, so you have a 360-degree view of the vulnerabilities you've uncovered. With reNgine 2.0 revolutionize your penetration testing game and impress your clients with reports that are not just informative but engaging and comprehensive with detailed analysis on impact assessment and remediation strategies. 🥷&nbsp;&nbsp; GPT-Powered Attack Surface Generation: With reNgine 2.0, reNgine seamlessly integrates with GPT to identify the attacks that you can likely perform on a subdomain. By making use of reconnaissance data such as page title, open ports, subdomain name etc, reNgine can advice you the attacks you could perform on a target. reNgine will also provide you the rationale on why the specific attack is likely to be successful. 🧭&nbsp;&nbsp;Continuous monitoring: Continuous monitoring is at the core of reNgine's mission, and it's robust continuous monitoring feature ensures that their targets are under constant scrutiny. With the flexibility to schedule scans at regular intervals, penetration testers can effortlessly stay informed about their targets. What sets reNgine apart is its seamless integration with popular notification channels such as Discord, Slack, and Telegram, delivering real-time alerts for newly discovered subdomains, vulnerabilities, or any changes in reconnaissance data.  ### Workflow <img src="https://github.com/yogeshojha/rengine/assets/17223002/10c475b8-b4a8-440d-9126-77fe2038a386">  ### Features * Reconnaissance: * Subdomain Discovery * IP and Open Ports Identification * Endpoints Discovery * Directory/Files fuzzing * Screenshot Gathering * Vulnerability Scan * Nuclei * Dalfox XSS Scanner * CRLFuzzer * Misconfigured S3 Scanner * WHOIS Identification * WAF Detection * OSINT Capabilities * Meta info Gathering * Employees Gathering * Email Address gathering * Google Dorking for sensitive info and urls * Projects, create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. * Perform Advanced Query lookup using natural language alike and, or, not operations * Highly configurable YAML-based Scan Engines * Support for Parallel Scans * Support for Subscans * Recon Data visualization * GPT Vulnerability Description, Impact and Remediation generation * GPT Attack Surface Generator * Multiple Roles and Permissions to cater a team's need * Customizable Alerts/Notifications on Slack, Discord, and Telegram * Automatically report Vulnerabilities to HackerOne * Recon Notes and Todos * Clocked Scans (Run reconnaissance exactly at X Hours and Y minutes) and Periodic Scans (Runs reconnaissance every X minutes/- hours/days/week) * Proxy Support * Screenshot Gallery with Filters * Powerful recon data filtering with autosuggestions * Recon Data changes, find new/removed subdomains/endpoints * Tag targets into the Organization * Smart Duplicate endpoint removal based on page title and content length to cleanup the reconnaissance data * Identify Interesting Subdomains * Custom GF patterns and custom Nuclei Templates * Edit tool-related configuration files (Nuclei, Subfinder, Naabu, amass) * Add external tools from Github/Go * Interoperable with other tools, Import/Export Subdomains/Endpoints * Import Targets via IP and/or CIDRs * Report Generation * Toolbox: Comes bundled with most commonly used tools during penetration testing such as whois lookup, CMS detector, CVE lookup, etc. * Identification of related domains and related TLDs for targets * Find actionable insights such as Most Common Vulnerability, Most Common CVE ID, Most Vulnerable Target/Subdomain, etc.  ### Scan Engine ```yaml subdomain_discovery: { 'uses_tools': [ 'subfinder', 'ctfr', 'sublist3r', 'tlsx', 'oneforall', 'netlas' ], 'enable_http_crawl': true, 'threads': 30, 'timeout': 5, } http_crawl: {} port_scan: { 'enable_http_crawl': true, 'timeout': 5, # 'exclude_ports': [], # 'exclude_subdomains': true, 'ports': ['top-100'], 'rate_limit': 150, 'threads': 30, 'passive': false, # 'use_naabu_config': false, # 'enable_nmap': true, # 'nmap_cmd': '', # 'nmap_script': '', # 'nmap_script_args': '' } osint: { 'discover': [ 'emails', 'metainfo', 'employees' ], 'dorks': [ 'login_pages', 'admin_panels', 'dashboard_pages', 'stackoverflow', 'social_media', 'project_management', 'code_sharing', 'config_files', 'jenkins', 'wordpress_files', 'php_error', 'exposed_documents', 'db_files', 'git_exposed' ], 'custom_dorks': [ { 'lookup_site': 'google.com', 'lookup_keywords': '/home/' }, { 'lookup_site': '_target_', 'lookup_extensions': 'jpg,png' } ], 'intensity': 'normal', 'documents_limit': 50 } dir_file_fuzz: { 'auto_calibration': true, 'enable_http_crawl': true, 'rate_limit': 150, 'extensions': ['html', 'php','git','yaml','conf','cnf','config','gz','env','log','db','mysql','bak','asp','aspx','txt','conf','sql','json','yml','pdf'], 'follow_redirect': false, 'max_time': 0, 'match_http_status': [200, 204], 'recursive_level': 2, 'stop_on_error': false, 'timeout': 5, 'threads': 30, 'wordlist_name': 'dicc' } fetch_url: { 'uses_tools': [ 'gospider', 'hakrawler', 'waybackurls', 'gospider', 'katana' ], 'remove_duplicate_endpoints': true, 'duplicate_fields': [ 'content_length', 'page_title' ], 'enable_http_crawl': true, 'gf_patterns': ['debug_logic', 'idor', 'interestingEXT', 'interestingparams', 'interestingsubs', 'lfi', 'rce', 'redirect', 'sqli', 'ssrf', 'ssti', 'xss'], 'ignore_file_extensions': ['png', 'jpg', 'jpeg', 'gif', 'mp4', 'mpeg', 'mp3'] # 'exclude_subdomains': true } vulnerability_scan: { 'run_nuclei': false, 'run_dalfox': false, 'run_crlfuzz': false, 'run_s3scanner': true, 'enable_http_crawl': true, 'concurrency': 50, 'intensity': 'normal', 'rate_limit': 150, 'retries': 1, 'timeout': 5, 'fetch_gpt_report': true, 'nuclei': { 'use_conf': false, 'severities': [ 'unknown', 'info', 'low', 'medium', 'high', 'critical' ], # 'tags': [], # 'templates': [], # 'custom_templates': [], }, 's3scanner': { 'threads': 100, 'providers': [ 'aws', 'gcp', 'digitalocean', 'dreamhost', 'linode' ] } } waf_detection: {} screenshot: { 'enable_http_crawl': true, 'intensity': 'normal', 'timeout': 10, 'threads': 40 } # custom_header: "Cookie: Test" ```  ### Quick Installation **Note:** Only Ubuntu/VPS 1. Clone this repo ```bash git clone https://github.com/yogeshojha/rengine && cd rengine ``` 1. Edit the dotenv file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```bash nano .env ``` 1. In the dotenv file, you may also modify the Scaling Configurations ```bash MAX_CONCURRENCY=80 MIN_CONCURRENCY=10 ``` MAX_CONCURRENCY: This parameter specifies the maximum number of reNgine's concurrent Celery worker processes that can be spawned. In this case, it's set to 80, meaning that the application can utilize up to 80 concurrent worker processes to execute tasks concurrently. This is useful for handling a high volume of scans or when you want to scale up processing power during periods of high demand. If you have more CPU cores, you will need to increase this for maximised performance. MIN_CONCURRENCY: On the other hand, MIN_CONCURRENCY specifies the minimum number of concurrent worker processes that should be maintained, even during periods of lower demand. In this example, it's set to 10, which means that even when there are fewer tasks to process, at least 10 worker processes will be kept running. This helps ensure that the application can respond promptly to incoming tasks without the overhead of repeatedly starting and stopping worker processes. These settings allow for dynamic scaling of Celery workers, ensuring that the application efficiently manages its workload by adjusting the number of concurrent workers based on the workload's size and complexity 1. Run the installation script, Please keep an eye for any prompt, you will also be asked for username and password for reNgine. ```bash sudo ./install.sh ``` If `install.sh` does not have install permission, please change it, `chmod +x install.sh` **reNgine can now be accessed from <https://127.0.0.1> or if you're on the VPS <https://your_vps_ip_address>** **Unless you are on development branch, please do not access reNgine via any ports** ### Installation (Mac/Windows/Other) Installation instructions can be found at [https://reNgine.wiki/install/detailed/](https://reNgine.wiki/2.0/install/detailed/) ### Updating 1. Updating is as simple as running the following command: ```bash cd rengine && sudo ./update.sh ``` If `update.sh` does not have execution permissions, please change it, `sudo chmod +x update.sh` ### Changelog [Please find the latest release notes and changelog here.](https://rengine.wiki/changelog/)  ### Screenshots #### Scan Results  #### General Usage <img src="https://user-images.githubusercontent.com/17223002/164993781-b6012995-522b-480a-a8bf-911193d35894.gif"> #### Initiating Subscan <img src="https://user-images.githubusercontent.com/17223002/164993749-1ad343d6-8ce7-43d6-aee7-b3add0321da7.gif"> #### Recon Data filtering <img src="https://user-images.githubusercontent.com/17223002/164993687-b63f3de8-e033-4ac0-808e-a2aa377d3cf8.gif"> #### Report Generation <img src="https://user-images.githubusercontent.com/17223002/164993689-c796c6cd-eb61-43f4-800d-08aba9740088.gif"> #### Toolbox <img src="https://user-images.githubusercontent.com/17223002/164993751-d687e88a-eb79-440f-9dc0-0ad006901620.gif"> #### Adding Custom tool in Tools Arsenal <img src="https://user-images.githubusercontent.com/17223002/164993670-466f6459-9499-498b-a9bd-526476d735a7.gif">  ### Contributing Contributions are what make the open-source community such an amazing place to learn, inspire and create. Every contributions you make is **greatly appreciated**. Your contributions can be as simple as fixing the indentation or UI, or as complex as adding new modules and features. See the [Contributing Guide](.github/CONTRIBUTING.md) to get started. You can also [join our Discord channel #development](https://discord.gg/JuhHdHTtwd) for any development related questions.  ### First-time Open Source contributors Please note that reNgine is beginner friendly. If you have never done open-source before, we encourage you to do so. **We will be happy and proud of your first PR ever.** You can start by resolving any [open issues](https://github.com/yogeshojha/rengine/issues).  ### reNgine Support Please do not use GitHub for support requests. Instead, [join our Discord channel #support](https://discord.gg/azv6fzhNCE).  ### Support and Sponsoring Over the past few years, I have been working hard on reNgine to add new features with the sole aim of making it the de facto standard for reconnaissance. I spend most of my free time and weekends working on reNgine. I do this in addition to my day job. I am happy to have received such overwhelming support from the community. But to keep this project alive, I am looking for financial support. | Paypal | Bitcoin | Ethereum | | :-------------------------------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | |[https://www.paypal.com/paypalme/yogeshojha11](https://www.paypal.com/paypalme/yogeshojha11) | `35AiKyNswNZ4TZUSdriHopSCjNMPi63BCX` | `0xe7A337Da6ff98A28513C26A7Fec8C9b42A63d346` OR * Add a [GitHub Star](https://github.com/yogeshojha/rengine) to the project. * Tweet about this project, or maybe blogs? * Maybe nominate me for [GitHub Stars?](https://stars.github.com/nominate/) * Join DigitalOcean using my [referral link](https://m.do.co/c/e353502d19fc) your profit is **$100** and I get $25 DO credit. This will help me test reNgine on VPS before I release any major features. It takes a considerable amount of time to add new features and make sure everything works. Donating is your way of saying: **reNgine is awesome**. Any support is greatly appreciated! Thank you!  ### reNgine Bug Bounty Program [](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine) Security researchers, welcome aboard! I'm excited to announce the reNgine bug bounty programme in collaboration with [huntr.dev](https://huntr.dev), which means that you will be rewarded for any vulnerabilities you find in reNgine. Thank you for your interest in reporting reNgine vulnerabilities! If you are aware of any potential security vulnerabilities in reNgine, we encourage you to report them immediately via [huntr.dev](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine). **Please do not disclose vulnerabilities via Github issues/blogs/tweets after/before reporting to huntr.dev as this is explicitly against the disclosure policy of huntr.dev and reNgine and will not be considered for monetary rewards.** Please note that the reNgine maintainer does not set the bounty amount. The bounty reward is determined by an industry-first equation developed by huntr.dev to understand the popularity, impact and value of repositories to the open-source community. **What do I expect from security researchers?** * Patience: Please note that I am currently the only maintainer in reNgine and it will take some time to validate your report. I ask for your patience during this process. * Respect for privacy and security reports: Please do not publicly disclose any vulnerabilities (including GitHub issues) before or after reporting them on huntr.dev! This is against the disclosure policy and will not be rewarded. * Respect the rules **What do you get in return?** * Thanks from the maintainer * Monetary rewards * CVE ID(s) Please find the [FAQ](https://www.huntr.dev/faq) and [Responsible disclosure policy](https://www.huntr.dev/policy/) from huntr.dev.  ### License Distributed under the GNU GPL v3 License. See [LICENSE](LICENSE) for more information.  <p align="right">(ChatGPT was used to write some or most part of this README section.)</p>

yogeshojha

3c60bc1ee495044794d91edee0c96fff73ab46c7

5413708d243799a5271440c47c6f98d0c51154ca

The hyperlinks seem incorrect or I am mentioned by accident. :p

AnonymousWP

yogeshojha/rengine

2.0-jasper release

### Added - Projects: Projects allow you to efficiently organize their web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. - Roles and Permissions: assign distinct roles to your team members: Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - GPT-powered Report Generation: With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments. - API Vault: This feature allows you to organize your API keys such as OpenAI or Netlas API keys. - GPT-powered Attack Surface Generation - URL gathering now is much more efficient, removing duplicate endpoints based on similar HTTP Responses, having the same content_lenth, or page_title. Custom duplicate fields can also be set from the scan engine configuration. - URL Path filtering while initiating scan: For instance, if we want to scan only endpoints starting with https://example.com/start/, we can pass the /start as a path filter while starting the scan. @ocervell - Expanding Target Concept: reNgine 2.0 now accepts IPs, URLS, etc as targets. (#678, #658) Excellent work by @ocervell - A ton of refactoring on reNgine's core to improve scan efficiency. Massive kudos to @ocervell - Created a custom celery workflow to be able to run several tasks in parallel that are not dependent on each other, such OSINT task and subdomain discovery will run in parallel, and directory and file fuzzing, vulnerability scan, screenshot gathering etc. will run in parallel after port scan or url fetching is completed. This will increase the efficiency of scans and instead of having one long flow of tasks, they can run independently on their own. @ocervell - Refactored all tasks to run asynchronously @ocervell - Added a stream_command that allows to read the output of a command live: this means the UI is updated with results while the command runs and does not have to wait until the task completes. Excellent work by @ocervell - Pwndb is now replaced by h8mail. @ocervell - Group Scan Results: reNgine 2.0 allows to group of subdomains based on similar page titles and HTTP status, and also vulnerability grouping based on the same vulnerability title and severity. - Added Support for Nmap: reNgine 2.0 allows to run Nmap scripts and vuln scans on ports found by Naabu. @ocervell - Added support for Shared Scan Variables in Scan Engine Configuration: - `enable_http_crawl`: (true/false) You can disable it to be more stealthy or focus on something different than HTTP - `timeout`: set timeout for all tasks - `rate_limit`: set rate limit for all tasks - `retries`: set retries for all tasks - `custom_header`: set the custom header for all tasks - Added Dalfox for XSS Vulnerability Scan - Added CRLFuzz for CRLF Vulnerability Scan - Added S3Scanner for scanning misconfigured S3 buckets - Improve OSINT Dork results, now detects admin panels, login pages and dashboards - Added Custom Dorks - Improved UI for vulnerability results, clicking on each vulnerability will open up a sidebar with vulnerability details. - Added HTTP Request and Response in vulnerability Results - Under Admin Settings, added an option to allow add/remove/deactivate additional users - Added Option to Preview Scan Report instead of forcing to download - Added Katana for crawling and spidering URLs - Added Netlas for Whois and subdomain gathering - Added TLSX for subdomain gathering - Added CTFR for subdomain gathering - Added historical IP in whois section ### Fixes - GF patterns do not run on 404 endpoints (#574 closed) - Fixes for retrieving whois data (#693 closed) - Related/Associated Domains in Whois section is now fixed ### Removed - Removed pwndb and tor related to it. - Removed tor for pwndb

2023-10-02 07:51:35+00:00

2023-10-07 10:37:23+00:00

README.md

<p align="center"> <a href="https://rengine.wiki"><img src=".github/screenshots/banner.gif" alt=""/></a> </p> <p align="center"><a href="https://github.com/yogeshojha/rengine/releases" target="_blank"><img src="https://img.shields.io/badge/version-v1.2.0-informational?&logo=none" alt="reNgine Latest Version" /></a>&nbsp;<a href="https://www.gnu.org/licenses/gpl-3.0" target="_blank"><img src="https://img.shields.io/badge/License-GPLv3-red.svg?&logo=none" alt="License" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/issues" target="_blank"><img src="https://img.shields.io/github/issues/yogeshojha/rengine?color=red&logo=none" alt="reNgine Issues" /></a>&nbsp;<a href="#" target="_blank"><img src="https://img.shields.io/badge/first--timers--only-friendly-blue.svg?&logo=none" alt="" /></a>&nbsp;<a href="https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine" target="_blank"><img src="https://cdn.huntr.dev/huntr_security_badge_mono.svg" alt="" /></a>&nbsp;</p> <p align="center"> <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/Open--Source--Summit-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://cyberweek.ae/2021/hitb-armory/" target="_blank"><img src="https://img.shields.io/badge/HITB--Armory-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=7uvP6MaQOX0" target="_blank"><img src="https://img.shields.io/badge/Black--Hat--Arsenal-USA--2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://drive.google.com/file/d/1Bh8lbf-Dztt5ViHJVACyrXMiglyICPQ2/view?usp=sharing" target="_blank"><img src="https://img.shields.io/badge/Defcon--Demolabs--29-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=A1oNOIc0h5A" target="_blank"><img src="https://img.shields.io/badge/Black--Hat--Arsenal-Europe--2020-blue.svg?&logo=none" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml/badge.svg" alt="" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/actions/workflows/build.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/build.yml/badge.svg" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://discord.gg/H6WzebwX3H" target="_blank"><img src="https://img.shields.io/discord/880363103689277461" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://opensourcesecurityindex.io/" target="_blank" rel="noopener"> <img style="width: 282px; height: 56px" src="https://opensourcesecurityindex.io/badge.svg" alt="Open Source Security Index - Fastest Growing Open Source Security Projects" width="282" height="56" /> </a> </p> <h3>reNgine 1.1<br>More than just recon!</h3> <h4>The only web application recon tool you will ever need!</h4> <p>Quickly discover the attack surface, and identify vulnerabilities using highly customizable and powerful scan engines. Enjoy peace of mind with reNgine's continuous monitoring, deeper reconnaissance, and open-source powered Vulnerability Scanner.</p> <h4>What is reNgine?</h4> <p align="left">reNgine is a web application reconnaissance suite that focuses on a highly configurable streamlined reconnaissance process via engines, reconnaissance data correlation, continuous monitoring, database backed reconnaissance data and a simple yet intuitive user interface. With features such as sub-scan, deeper co-relation, report generation, etc., reNgine aims to fill the gap in traditional reconnaissance tools and is likely to be a better alternative to existing commercial tools. reNgine makes it easy for penetration testers and security auditors to gather reconnaissance data with minimal configuration. </p>  <p align="center"> ⭐<a href="https://rengine.wiki">reNgine Documentation</a> · <a href="https://rengine.wiki/changelog/">What's new</a> · <a href="https://github.com/yogeshojha/rengine/blob/master/.github/CONTRIBUTING.md">Contribute</a> · <a href="https://github.com/yogeshojha/rengine/issues">Report Bug</a> · <a href="https://github.com/yogeshojha/rengine/issues">Request Feature</a>⭐ </p>  ## Table of Contents * [About reNgine](#about-rengine) * [Features](#features) * [Documentation](#documentation) * [Quick Installation](#quick-installation) * [What's new in reNgine](#changelog) * [reNgine Bug Bounty Program](#rengine-bug-bounty-program) * [Screenshots](#screenshots) * [Contributing](#contributing) * [reNgine Support](#rengine-support) * [Related Projects](#related-projects) * [Support and Sponsoring](#support-and-sponsoring) * [License](#license)  ### reNgine 2.0 codenamed Jasper I am currently working on reNgine 2.0, which will probably be announced sometime between May and August 2023. reNgine 2.0 will be the most advanced reNgine ever, a lot of work will be done in how scans are performed, things such as Pause and Resume Scan, Axiom Integration, more deeper correlation, Project Options, Multiple Tenants, etc. Please submit your feature requests via GitHub issues.  ## About reNgine You can watch [reNgine 1.1 release trailer here.](https://www.youtube.com/watch?v=iy_6F7Vq8Lo) (Recommended) <img src="https://user-images.githubusercontent.com/17223002/164993688-50eb95f2-3653-4ef7-bd3b-ef7a096824ea.jpeg">  reNgine is a web application reconnaissance suite with a focus on a highly configurable, streamlined reconnaissance process. reNgine is backed by a database, with data correlation and organisation, the custom query "like" language for filtering reconnaissance data, reNgine aims to address the shortcomings of the traditional reconnaissance workflow. The developers behind reNgine understand that reconnaissance data can be huge and manually searching for records to attack can be tedious, so features such as identifying subdomains of interest help penetration testers focus on attack rather than reconnaissance. reNgine also focuses on continuous monitoring. Penetration testers can choose to schedule the scan at regular intervals and be notified via notification channels such as Discord, Slack and Telegram of any new subdomains or vulnerabilities identified, or any changes to the recon data. Interoperability is something every reconnaissance tool needs, and reNgine is no different. Starting with reNgine 1.0, we have added features such as import and export of subdomains, endpoints, GF pattern matched endpoints, etc. This allows you to use your favourite reconnaissance workflow in conjunction with reNgine. PDF reports are something every individual or team needs. From reNgine 1.1, reNgine also comes with the option to download PDF reports. You can also choose the type of report, a full scan report or just a reconnaissance report. We also understand that PDF reports need to be customisable. Choose the colour of the report you want, customise the executive summary, etc. You choose how your PDF report looks! reNgine features highly configurable scan engines based on YAML, allowing penetration testers to create as many reconnaissance engines of their choice as they like, configure them as they like, and use them against any targets for scanning. These engines allow penetration testers to use the tools of their choice, with the configuration of their choice. Out of the box, reNgine comes with several scan engines such as Full Scan, Passive Scan, Screenshot Gathering, OSINT Engine, etc. Our focus has always been on finding the right reconnaissance data with the least amount of effort. After several discussions with fellow hackers/pentesters, a screenshot gallery was a must, reNgine also comes with a screenshot gallery, and what's more exciting than having a screenshot gallery with filters, filter screenshots with HTTP status, technology, ports and services. We also want our fellow hackers to stay ahead of the game, so reNgine also comes with automatic vulnerability reporting (ATM only Hackerone is supported, other platforms may come soon). This allows hackers to define their vulnerability reporting template and reNgine will do the rest of the work to report the vulnerability as soon as it is identified.  <img src="https://user-images.githubusercontent.com/17223002/164993945-aabdbb4a-2b9d-4951-ba27-5f2f5abd1d8b.gif"> ## Features * Reconnaissance: Subdomain Discovery, IP and Open Ports Identification, Endpoints Discovery, Directory and Files fuzzing, - Screenshot gathering, Vulnerability scan using Nuclei, WHOIS Identification, WAF Detection etc. * Highly configurable YAML-based Scan Engines * Support for Parallel Scans and Subscans * Automatically report Vulnerabilities to HackerOne * Recon Data visualization * OSINT Capabilities (Meta info Gathering, Employees Gathering, Email Address with an option to look password in the leaked database, - dorks, etc.) * Customizable Alerts/Notifications on Slack, Discord, and Telegram * Perform Advanced Query lookup using natural language alike and, or, not operations * Recon Notes and Todos * Clocked Scans (Run reconnaissance exactly at X Hours and Y minutes) and Periodic Scans (Runs reconnaissance every X minutes/- hours/days/week) * Proxy Support * Screenshot Gallery with Filters * Powerful recon data filtering with autosuggestions * Recon Data changes, find new/removed subdomains/endpoints * Tag targets into the Organization * Identify Interesting Subdomains * Custom GF patterns and custom Nuclei Templates * Edit tool-related configuration files (Nuclei, Subfinder, Naabu, amass) * Add external tools from Github/Go * Interoperable with other tools, Import/Export Subdomains/Endpoints * Import Targets via IP and/or CIDRs * Report Generation * Toolbox: Comes bundled with most commonly used tools such as whois lookup, CMS detector, CVE lookup, etc. * Identification of related domains and related TLDs for targets * Find actionable insights such as Most Common Vulnerability, Most Common CVE ID, Most Vulnerable Target/Subdomain, etc.  ## Documentation You can find reNgine documentation at [https://rengine.wiki](https://rengine.wiki)  ## Quick Installation **Note:** Only Ubuntu/VPS 1. Clone this repo ```bash git clone https://github.com/yogeshojha/rengine && cd rengine ``` 1. Edit the dotenv file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```bash nano .env ``` 1. Run the installation script, Please keep an eye for any prompt, you will also be asked for username and password for reNgine. ```bash sudo ./install.sh ``` If `install.sh` does not have install permission, please change it, `chmod +x install.sh` **reNgine can now be accessed from <https://127.0.0.1> or if you're on the VPS <https://your_vps_ip_address>** A detailed installation guide can also be found [here](https://www.rffuste.com/2022/05/23/rengine-a-brief-overview/). Thanks to Rubén! ## Installation (Mac/Windows/Other) Installation instructions can be found at [https://reNgine.wiki/install/detailed/](https://reNgine.wiki/install/detailed/) ## Updating 1. Updating is as simple as running the following command: ```bash sudo ./update.sh ``` If `update.sh` does not have execution permissions, please change it, `sudo chmod +x update.sh` ## Changelog [Please find the latest release notes and changelog here.](https://rengine.wiki/changelog/)  ## reNgine Bug Bounty Program [](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine) Security researchers, welcome aboard! I'm excited to announce the reNgine bug bounty programme in collaboration with [huntr.dev](https://huntr.dev), which means that you will be rewarded for any vulnerabilities you find in reNgine. Thank you for your interest in reporting reNgine vulnerabilities! If you are aware of any potential security vulnerabilities in reNgine, we encourage you to report them immediately via [huntr.dev](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine). **Please do not disclose vulnerabilities via Github issues/blogs/tweets after/before reporting to huntr.dev as this is explicitly against the disclosure policy of huntr.dev and reNgine and will not be considered for monetary rewards.** Please note that the reNgine maintainer does not set the bounty amount. The bounty reward is determined by an industry-first equation developed by huntr.dev to understand the popularity, impact and value of repositories to the open-source community. **What do I expect from security researchers?** * Patience: Please note that I am currently the only maintainer in reNgine and it will take some time to validate your report. I ask for your patience during this process. * Respect for privacy and security reports: Please do not publicly disclose any vulnerabilities (including GitHub issues) before or after reporting them on huntr.dev! This is against the disclosure policy and will not be rewarded. * Respect the rules **What do you get in return?** * Thanks from the maintainer * Monetary rewards * CVE ID(s) Please find the [FAQ](https://www.huntr.dev/faq) and [Responsible disclosure policy](https://www.huntr.dev/policy/) from huntr.dev.  ## Screenshots ### Scan Results  ### General Usage <img src="https://user-images.githubusercontent.com/17223002/164993781-b6012995-522b-480a-a8bf-911193d35894.gif"> ### Initiating Subscan <img src="https://user-images.githubusercontent.com/17223002/164993749-1ad343d6-8ce7-43d6-aee7-b3add0321da7.gif"> ### Recon Data filtering <img src="https://user-images.githubusercontent.com/17223002/164993687-b63f3de8-e033-4ac0-808e-a2aa377d3cf8.gif"> ### Report Generation <img src="https://user-images.githubusercontent.com/17223002/164993689-c796c6cd-eb61-43f4-800d-08aba9740088.gif"> ### Toolbox <img src="https://user-images.githubusercontent.com/17223002/164993751-d687e88a-eb79-440f-9dc0-0ad006901620.gif"> ### Adding Custom tool in Tools Arsenal <img src="https://user-images.githubusercontent.com/17223002/164993670-466f6459-9499-498b-a9bd-526476d735a7.gif">  ## Contributing Contributions are what make the open-source community such an amazing place to learn, inspire and create. Every contributions you make is **greatly appreciated**. Your contributions can be as simple as fixing the indentation or UI, or as complex as adding new modules and features. See the [Contributing Guide](.github/CONTRIBUTING.md) to get started. You can also [join our Discord channel #development](https://discord.gg/JuhHdHTtwd) for any development related questions.  ### First-time Open Source contributors Please note that reNgine is beginner friendly. If you have never done open-source before, we encourage you to do so. **We will be happy and proud of your first PR ever.** You can start by resolving any [open issues](https://github.com/yogeshojha/rengine/issues).  ## reNgine Support Please do not use GitHub for support requests. Instead, [join our Discord channel #support](https://discord.gg/azv6fzhNCE)  ## Related Projects There are many other great reconnaissance frameworks out there, you can use reNgine in conjunction with those tools. But they are great in their own right, and can sometimes produce better results than reNgine. * [ReconFTW](https://github.com/six2dez/reconftw#sample-video) * [Reconmap](https://github.com/reconmap/reconmap)  ## Support and Sponsoring Over the past few years, I have been working hard on reNgine to add new features with the sole aim of making it the de facto standard for reconnaissance. I spend most of my free time and weekends working on reNgine. I do this in addition to my day job. I am happy to have received such overwhelming support from the community. But to keep this project alive, I am looking for financial support. | Paypal | Bitcoin | Ethereum | | :-------------------------------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | |[https://www.paypal.com/paypalme/yogeshojha11](https://www.paypal.com/paypalme/yogeshojha11) | `35AiKyNswNZ4TZUSdriHopSCjNMPi63BCX` | `0xe7A337Da6ff98A28513C26A7Fec8C9b42A63d346` OR * Add a [GitHub Star](https://github.com/yogeshojha/rengine) to the project. * Tweet about this project, or maybe blogs? * Maybe nominate me for [GitHub Stars?](https://stars.github.com/nominate/) * Join DigitalOcean using my [referral link](https://m.do.co/c/e353502d19fc) your profit is **$100** and I get $25 DO credit. This will help me test reNgine on VPS before I release any major features. It takes a considerable amount of time to add new features and make sure everything works. Donating is your way of saying: **reNgine is awesome**. Any support is greatly appreciated! Thank you!  ## License Distributed under the GNU GPL v3 License. See [LICENSE](LICENSE) for more information. 

<p align="center"> <a href="https://rengine.wiki"><img src=".github/screenshots/banner.gif" alt=""/></a> </p> <p align="center"><a href="https://github.com/yogeshojha/rengine/releases" target="_blank"><img src="https://img.shields.io/badge/version-v2.0.0-informational?&logo=none" alt="reNgine Latest Version" /></a>&nbsp;<a href="https://www.gnu.org/licenses/gpl-3.0" target="_blank"><img src="https://img.shields.io/badge/License-GPLv3-red.svg?&logo=none" alt="License" /></a>&nbsp;<a href="#" target="_blank"><img src="https://img.shields.io/badge/first--timers--only-friendly-blue.svg?&logo=none" alt="" /></a>&nbsp;<a href="https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine" target="_blank"><img src="https://cdn.huntr.dev/huntr_security_badge_mono.svg" alt="" /></a>&nbsp;</p> <p align="center"> <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Asia-2023-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/Open--Source--Summit-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://cyberweek.ae/2021/hitb-armory/" target="_blank"><img src="https://img.shields.io/badge/HITB--Armory-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=7uvP6MaQOX0" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://drive.google.com/file/d/1Bh8lbf-Dztt5ViHJVACyrXMiglyICPQ2/view?usp=sharing" target="_blank"><img src="https://img.shields.io/badge/Defcon--Demolabs--29-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=A1oNOIc0h5A" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Europe-2020-blue.svg?&logo=none" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml/badge.svg" alt="" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/actions/workflows/build.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/build.yml/badge.svg" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://discord.gg/H6WzebwX3H" target="_blank"><img src="https://img.shields.io/discord/880363103689277461" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://opensourcesecurityindex.io/" target="_blank" rel="noopener"> <img style="width: 282px; height: 56px" src="https://opensourcesecurityindex.io/badge.svg" alt="Open Source Security Index - Fastest Growing Open Source Security Projects" width="282" height="56" /> </a> </p> <h3>reNgine 2.0-jasper<br>Redefining the future of reconnaissance!</h3> <h4>What is reNgine?</h4> <p align="left">reNgine is your go-to web application reconnaissance suite that's designed to simplify and streamline the reconnaissance process for security professionals, penetration testers, and bug bounty hunters. With its highly configurable engines, data correlation capabilities, continuous monitoring, database-backed reconnaissance data, and an intuitive user interface, reNgine redefines how you gather critical information about your target web applications. Traditional reconnaissance tools often fall short in terms of configurability and efficiency. reNgine addresses these shortcomings and emerges as a excellent alternative to existing commercial tools. reNgine was created to address the limitations of traditional reconnaissance tools and provide a better alternative, even surpassing some commercial offerings. Whether you're a bug bounty hunter, a penetration tester, or a corporate security team, reNgine is your go-to solution for automating and enhancing your information-gathering efforts. </p> reNgine 2.0-jasper is out now, you can [watch reNgine 2.0-jasper release trailer here!](https://youtu.be/VwkOWqiWW5g) reNgine 2.0-Jasper would not have been possible without [@ocervell](https://github.com/ocervell) valuable contributions. [@ocervell](https://github.com/ocervell) did majority of the refactoring if not all and also added a ton of features. Together, we wish to shape the future of web application reconnaissance, and it's developers like [@ocervell](https://github.com/ocervell) and a [ton of other developers and hackers from our community](https://github.com/yogeshojha/rengine/graphs/contributors) who inspire and drive us forward. Thank you, [@ocervell](https://github.com/ocervell), for your outstanding work and unwavering commitment to reNgine. Checkout our contributers here: [Contributers](https://github.com/yogeshojha/rengine/graphs/contributors)  ### Documentation You can find detailed documentation at [https://rengine.wiki](https://rengine.wiki)  ### Table of Contents * [About reNgine](#about-rengine) * [Workflow](#workflow) * [Features](#features) * [Scan Engine](#scan-engine) * [Quick Installation](#quick-installation) * [What's new in reNgine 2.0](#changelog) * [Screenshots](#screenshots) * [Contributing](#contributing) * [reNgine Support](#rengine-support) * [Support and Sponsoring](#support-and-sponsoring) * [reNgine Bug Bounty Program](#rengine-bug-bounty-program) * [License](#license)  ### About reNgine reNgine is not an ordinary reconnaissance suite; it's a game-changer! We've turbocharged the traditional workflow with groundbreaking features that is sure to ease your reconnaissance game. reNgine redefines the art of reconnaissance with highly configurable scan engines, recon data correlation, continuous monitoring, GPT powered Vulnerability Report, Project Management and role based access control etc. 🦾&nbsp;&nbsp; reNgine has advanced reconnaissance capabilities, harnessing a range of open-source tools to deliver a comprehensive web application reconnaissance experience. With it's intuitive User Interface, it excels in subdomain discovery, pinpointing IP addresses and open ports, collecting endpoints, conducting directory and file fuzzing, capturing screenshots, and performing vulnerability scans. To summarize, it does end-to-end reconnaissance. With WHOIS identification and WAF detection, it offers deep insights into target domains. Additionally, reNgine also identifies misconfigured S3 buckets and find interesting subdomains and URLS, based on specific keywords to helps you identify your next target, making it an go to tool for efficient reconnaissance. 🗃️&nbsp; &nbsp; Say goodbye to recon data chaos! reNgine seamlessly integrates with a database, providing you with unmatched data correlation and organization. Forgot the hassle of grepping through json, txt or csv files. Plus, our custom query language lets you filter reconnaissance data effortlessly using natural language like operators such as filtering all alive subdomains with `http_status=200` and also filter all subdomains that are alive and has admin in name `http_status=200&name=admin` 🔧&nbsp;&nbsp; reNgine offers unparalleled flexibility through its highly configurable scan engines, based on a YAML-based configuration. It offers the freedom to create and customize recon scan engines based on any kind of requirement, users can tailor them to their specific objectives and preferences, from thread management to timeout settings and rate-limit configurations, everything is customizable. Additionally, reNgine offers a range of pre-configured scan engines right out of the box, including Full Scan, Passive Scan, Screenshot Gathering, and the OSINT Scan Engine. These ready-to-use engines eliminate the need for extensive manual setup, aligning perfectly with reNgine's core mission of simplifying the reconnaissance process and enabling users to effortlessly access the right reconnaissance data with minimal effort. 💎&nbsp;&nbsp;Subscans: Subscan is a game-changing feature in reNgine, setting it apart as the only open-source tool of its kind to offer this capability. With Subscan, waiting for the entire pipeline to complete is a thing of the past. Now, users can swiftly respond to newfound discoveries during reconnaissance. Whether you've stumbled upon an intriguing subdomain and wish to conduct a focused port scan or want to delve deeper with a vulnerability assessment, reNgine has you covered. 📃&nbsp;&nbsp; PDF Reports: In addition to its robust reconnaissance capabilities, reNgine goes the extra mile by simplifying the report generation process, recognizing the crucial role that PDF reports play in the realm of end-to-end reconnaissance. Users can effortlessly generate and customize PDF reports to suit their exact needs. Whether it's a Full Scan Report, Vulnerability Report, or a concise reconnaissance report, reNgine provides the flexibility to choose the report type that best communicates your findings. Moreover, the level of customization is unparalleled, allowing users to select report colors, fine-tune executive summaries, and even add personalized touches like company names and footers. With GPT integration, your reports aren't just a report, with remediation steps, and impacts, you get 360-degree view of the vulnerabilities you've uncovered. 🔖&nbsp; &nbsp; Say Hello to Projects! reNgine 2.0 introduces a powerful addition that enables you to efficiently organize your web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. Each projects will have separate dashboard and all the scan results will be separated from each projects, while scan engines and configuration will be shared across all the projects. ⚙️&nbsp; &nbsp; Roles and Permissions! Begining reNgine 2.0, we've taken your web application reconnaissance to a whole new level of control and security. Now, you can assign distinct roles to your team members—Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - 🔐 Sys Admin: Sys Admin is a super user that has permission to modify system and scan related configurations, scan engines, create new users, add new tools etc. Super user can initiate scans and subscans effortlessly. - 🔍 Penetration Tester: Penetration Tester will be allowed to modify and initiate scans and subscans, add or update targets, etc. A penetration tester will not be allowed to modify system configurations. - 📊 Auditor: Auditor can only view and download the report. An auditor can not change any system or scan related configurations nor can initiate any scans or subscans. 🚀&nbsp;&nbsp; GPT Vulnerability Report Generation: Get ready for the future of penetration testing reports with reNgine's groundbreaking feature: "GPT-Powered Report Generation"! With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments that read like they were written by a human security expert! **But that's not all!** Our GPT-driven reports go the extra mile by scouring the web for related news articles, blogs, and references, so you have a 360-degree view of the vulnerabilities you've uncovered. With reNgine 2.0 revolutionize your penetration testing game and impress your clients with reports that are not just informative but engaging and comprehensive with detailed analysis on impact assessment and remediation strategies. 🥷&nbsp;&nbsp; GPT-Powered Attack Surface Generation: With reNgine 2.0, reNgine seamlessly integrates with GPT to identify the attacks that you can likely perform on a subdomain. By making use of reconnaissance data such as page title, open ports, subdomain name etc, reNgine can advice you the attacks you could perform on a target. reNgine will also provide you the rationale on why the specific attack is likely to be successful. 🧭&nbsp;&nbsp;Continuous monitoring: Continuous monitoring is at the core of reNgine's mission, and it's robust continuous monitoring feature ensures that their targets are under constant scrutiny. With the flexibility to schedule scans at regular intervals, penetration testers can effortlessly stay informed about their targets. What sets reNgine apart is its seamless integration with popular notification channels such as Discord, Slack, and Telegram, delivering real-time alerts for newly discovered subdomains, vulnerabilities, or any changes in reconnaissance data.  ### Workflow <img src="https://github.com/yogeshojha/rengine/assets/17223002/10c475b8-b4a8-440d-9126-77fe2038a386">  ### Features * Reconnaissance: * Subdomain Discovery * IP and Open Ports Identification * Endpoints Discovery * Directory/Files fuzzing * Screenshot Gathering * Vulnerability Scan * Nuclei * Dalfox XSS Scanner * CRLFuzzer * Misconfigured S3 Scanner * WHOIS Identification * WAF Detection * OSINT Capabilities * Meta info Gathering * Employees Gathering * Email Address gathering * Google Dorking for sensitive info and urls * Projects, create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. * Perform Advanced Query lookup using natural language alike and, or, not operations * Highly configurable YAML-based Scan Engines * Support for Parallel Scans * Support for Subscans * Recon Data visualization * GPT Vulnerability Description, Impact and Remediation generation * GPT Attack Surface Generator * Multiple Roles and Permissions to cater a team's need * Customizable Alerts/Notifications on Slack, Discord, and Telegram * Automatically report Vulnerabilities to HackerOne * Recon Notes and Todos * Clocked Scans (Run reconnaissance exactly at X Hours and Y minutes) and Periodic Scans (Runs reconnaissance every X minutes/- hours/days/week) * Proxy Support * Screenshot Gallery with Filters * Powerful recon data filtering with autosuggestions * Recon Data changes, find new/removed subdomains/endpoints * Tag targets into the Organization * Smart Duplicate endpoint removal based on page title and content length to cleanup the reconnaissance data * Identify Interesting Subdomains * Custom GF patterns and custom Nuclei Templates * Edit tool-related configuration files (Nuclei, Subfinder, Naabu, amass) * Add external tools from Github/Go * Interoperable with other tools, Import/Export Subdomains/Endpoints * Import Targets via IP and/or CIDRs * Report Generation * Toolbox: Comes bundled with most commonly used tools during penetration testing such as whois lookup, CMS detector, CVE lookup, etc. * Identification of related domains and related TLDs for targets * Find actionable insights such as Most Common Vulnerability, Most Common CVE ID, Most Vulnerable Target/Subdomain, etc.  ### Scan Engine ```yaml subdomain_discovery: { 'uses_tools': [ 'subfinder', 'ctfr', 'sublist3r', 'tlsx', 'oneforall', 'netlas' ], 'enable_http_crawl': true, 'threads': 30, 'timeout': 5, } http_crawl: {} port_scan: { 'enable_http_crawl': true, 'timeout': 5, # 'exclude_ports': [], # 'exclude_subdomains': true, 'ports': ['top-100'], 'rate_limit': 150, 'threads': 30, 'passive': false, # 'use_naabu_config': false, # 'enable_nmap': true, # 'nmap_cmd': '', # 'nmap_script': '', # 'nmap_script_args': '' } osint: { 'discover': [ 'emails', 'metainfo', 'employees' ], 'dorks': [ 'login_pages', 'admin_panels', 'dashboard_pages', 'stackoverflow', 'social_media', 'project_management', 'code_sharing', 'config_files', 'jenkins', 'wordpress_files', 'php_error', 'exposed_documents', 'db_files', 'git_exposed' ], 'custom_dorks': [ { 'lookup_site': 'google.com', 'lookup_keywords': '/home/' }, { 'lookup_site': '_target_', 'lookup_extensions': 'jpg,png' } ], 'intensity': 'normal', 'documents_limit': 50 } dir_file_fuzz: { 'auto_calibration': true, 'enable_http_crawl': true, 'rate_limit': 150, 'extensions': ['html', 'php','git','yaml','conf','cnf','config','gz','env','log','db','mysql','bak','asp','aspx','txt','conf','sql','json','yml','pdf'], 'follow_redirect': false, 'max_time': 0, 'match_http_status': [200, 204], 'recursive_level': 2, 'stop_on_error': false, 'timeout': 5, 'threads': 30, 'wordlist_name': 'dicc' } fetch_url: { 'uses_tools': [ 'gospider', 'hakrawler', 'waybackurls', 'gospider', 'katana' ], 'remove_duplicate_endpoints': true, 'duplicate_fields': [ 'content_length', 'page_title' ], 'enable_http_crawl': true, 'gf_patterns': ['debug_logic', 'idor', 'interestingEXT', 'interestingparams', 'interestingsubs', 'lfi', 'rce', 'redirect', 'sqli', 'ssrf', 'ssti', 'xss'], 'ignore_file_extensions': ['png', 'jpg', 'jpeg', 'gif', 'mp4', 'mpeg', 'mp3'] # 'exclude_subdomains': true } vulnerability_scan: { 'run_nuclei': false, 'run_dalfox': false, 'run_crlfuzz': false, 'run_s3scanner': true, 'enable_http_crawl': true, 'concurrency': 50, 'intensity': 'normal', 'rate_limit': 150, 'retries': 1, 'timeout': 5, 'fetch_gpt_report': true, 'nuclei': { 'use_conf': false, 'severities': [ 'unknown', 'info', 'low', 'medium', 'high', 'critical' ], # 'tags': [], # 'templates': [], # 'custom_templates': [], }, 's3scanner': { 'threads': 100, 'providers': [ 'aws', 'gcp', 'digitalocean', 'dreamhost', 'linode' ] } } waf_detection: {} screenshot: { 'enable_http_crawl': true, 'intensity': 'normal', 'timeout': 10, 'threads': 40 } # custom_header: "Cookie: Test" ```  ### Quick Installation **Note:** Only Ubuntu/VPS 1. Clone this repo ```bash git clone https://github.com/yogeshojha/rengine && cd rengine ``` 1. Edit the dotenv file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```bash nano .env ``` 1. In the dotenv file, you may also modify the Scaling Configurations ```bash MAX_CONCURRENCY=80 MIN_CONCURRENCY=10 ``` MAX_CONCURRENCY: This parameter specifies the maximum number of reNgine's concurrent Celery worker processes that can be spawned. In this case, it's set to 80, meaning that the application can utilize up to 80 concurrent worker processes to execute tasks concurrently. This is useful for handling a high volume of scans or when you want to scale up processing power during periods of high demand. If you have more CPU cores, you will need to increase this for maximised performance. MIN_CONCURRENCY: On the other hand, MIN_CONCURRENCY specifies the minimum number of concurrent worker processes that should be maintained, even during periods of lower demand. In this example, it's set to 10, which means that even when there are fewer tasks to process, at least 10 worker processes will be kept running. This helps ensure that the application can respond promptly to incoming tasks without the overhead of repeatedly starting and stopping worker processes. These settings allow for dynamic scaling of Celery workers, ensuring that the application efficiently manages its workload by adjusting the number of concurrent workers based on the workload's size and complexity 1. Run the installation script, Please keep an eye for any prompt, you will also be asked for username and password for reNgine. ```bash sudo ./install.sh ``` If `install.sh` does not have install permission, please change it, `chmod +x install.sh` **reNgine can now be accessed from <https://127.0.0.1> or if you're on the VPS <https://your_vps_ip_address>** **Unless you are on development branch, please do not access reNgine via any ports** ### Installation (Mac/Windows/Other) Installation instructions can be found at [https://reNgine.wiki/install/detailed/](https://reNgine.wiki/2.0/install/detailed/) ### Updating 1. Updating is as simple as running the following command: ```bash cd rengine && sudo ./update.sh ``` If `update.sh` does not have execution permissions, please change it, `sudo chmod +x update.sh` ### Changelog [Please find the latest release notes and changelog here.](https://rengine.wiki/changelog/)  ### Screenshots #### Scan Results  #### General Usage <img src="https://user-images.githubusercontent.com/17223002/164993781-b6012995-522b-480a-a8bf-911193d35894.gif"> #### Initiating Subscan <img src="https://user-images.githubusercontent.com/17223002/164993749-1ad343d6-8ce7-43d6-aee7-b3add0321da7.gif"> #### Recon Data filtering <img src="https://user-images.githubusercontent.com/17223002/164993687-b63f3de8-e033-4ac0-808e-a2aa377d3cf8.gif"> #### Report Generation <img src="https://user-images.githubusercontent.com/17223002/164993689-c796c6cd-eb61-43f4-800d-08aba9740088.gif"> #### Toolbox <img src="https://user-images.githubusercontent.com/17223002/164993751-d687e88a-eb79-440f-9dc0-0ad006901620.gif"> #### Adding Custom tool in Tools Arsenal <img src="https://user-images.githubusercontent.com/17223002/164993670-466f6459-9499-498b-a9bd-526476d735a7.gif">  ### Contributing Contributions are what make the open-source community such an amazing place to learn, inspire and create. Every contributions you make is **greatly appreciated**. Your contributions can be as simple as fixing the indentation or UI, or as complex as adding new modules and features. See the [Contributing Guide](.github/CONTRIBUTING.md) to get started. You can also [join our Discord channel #development](https://discord.gg/JuhHdHTtwd) for any development related questions.  ### First-time Open Source contributors Please note that reNgine is beginner friendly. If you have never done open-source before, we encourage you to do so. **We will be happy and proud of your first PR ever.** You can start by resolving any [open issues](https://github.com/yogeshojha/rengine/issues).  ### reNgine Support Please do not use GitHub for support requests. Instead, [join our Discord channel #support](https://discord.gg/azv6fzhNCE).  ### Support and Sponsoring Over the past few years, I have been working hard on reNgine to add new features with the sole aim of making it the de facto standard for reconnaissance. I spend most of my free time and weekends working on reNgine. I do this in addition to my day job. I am happy to have received such overwhelming support from the community. But to keep this project alive, I am looking for financial support. | Paypal | Bitcoin | Ethereum | | :-------------------------------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | |[https://www.paypal.com/paypalme/yogeshojha11](https://www.paypal.com/paypalme/yogeshojha11) | `35AiKyNswNZ4TZUSdriHopSCjNMPi63BCX` | `0xe7A337Da6ff98A28513C26A7Fec8C9b42A63d346` OR * Add a [GitHub Star](https://github.com/yogeshojha/rengine) to the project. * Tweet about this project, or maybe blogs? * Maybe nominate me for [GitHub Stars?](https://stars.github.com/nominate/) * Join DigitalOcean using my [referral link](https://m.do.co/c/e353502d19fc) your profit is **$100** and I get $25 DO credit. This will help me test reNgine on VPS before I release any major features. It takes a considerable amount of time to add new features and make sure everything works. Donating is your way of saying: **reNgine is awesome**. Any support is greatly appreciated! Thank you!  ### reNgine Bug Bounty Program [](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine) Security researchers, welcome aboard! I'm excited to announce the reNgine bug bounty programme in collaboration with [huntr.dev](https://huntr.dev), which means that you will be rewarded for any vulnerabilities you find in reNgine. Thank you for your interest in reporting reNgine vulnerabilities! If you are aware of any potential security vulnerabilities in reNgine, we encourage you to report them immediately via [huntr.dev](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine). **Please do not disclose vulnerabilities via Github issues/blogs/tweets after/before reporting to huntr.dev as this is explicitly against the disclosure policy of huntr.dev and reNgine and will not be considered for monetary rewards.** Please note that the reNgine maintainer does not set the bounty amount. The bounty reward is determined by an industry-first equation developed by huntr.dev to understand the popularity, impact and value of repositories to the open-source community. **What do I expect from security researchers?** * Patience: Please note that I am currently the only maintainer in reNgine and it will take some time to validate your report. I ask for your patience during this process. * Respect for privacy and security reports: Please do not publicly disclose any vulnerabilities (including GitHub issues) before or after reporting them on huntr.dev! This is against the disclosure policy and will not be rewarded. * Respect the rules **What do you get in return?** * Thanks from the maintainer * Monetary rewards * CVE ID(s) Please find the [FAQ](https://www.huntr.dev/faq) and [Responsible disclosure policy](https://www.huntr.dev/policy/) from huntr.dev.  ### License Distributed under the GNU GPL v3 License. See [LICENSE](LICENSE) for more information.  <p align="right">(ChatGPT was used to write some or most part of this README section.)</p>

yogeshojha

3c60bc1ee495044794d91edee0c96fff73ab46c7

5413708d243799a5271440c47c6f98d0c51154ca

Ditto.

AnonymousWP

yogeshojha/rengine

2.0-jasper release

### Added - Projects: Projects allow you to efficiently organize their web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. - Roles and Permissions: assign distinct roles to your team members: Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - GPT-powered Report Generation: With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments. - API Vault: This feature allows you to organize your API keys such as OpenAI or Netlas API keys. - GPT-powered Attack Surface Generation - URL gathering now is much more efficient, removing duplicate endpoints based on similar HTTP Responses, having the same content_lenth, or page_title. Custom duplicate fields can also be set from the scan engine configuration. - URL Path filtering while initiating scan: For instance, if we want to scan only endpoints starting with https://example.com/start/, we can pass the /start as a path filter while starting the scan. @ocervell - Expanding Target Concept: reNgine 2.0 now accepts IPs, URLS, etc as targets. (#678, #658) Excellent work by @ocervell - A ton of refactoring on reNgine's core to improve scan efficiency. Massive kudos to @ocervell - Created a custom celery workflow to be able to run several tasks in parallel that are not dependent on each other, such OSINT task and subdomain discovery will run in parallel, and directory and file fuzzing, vulnerability scan, screenshot gathering etc. will run in parallel after port scan or url fetching is completed. This will increase the efficiency of scans and instead of having one long flow of tasks, they can run independently on their own. @ocervell - Refactored all tasks to run asynchronously @ocervell - Added a stream_command that allows to read the output of a command live: this means the UI is updated with results while the command runs and does not have to wait until the task completes. Excellent work by @ocervell - Pwndb is now replaced by h8mail. @ocervell - Group Scan Results: reNgine 2.0 allows to group of subdomains based on similar page titles and HTTP status, and also vulnerability grouping based on the same vulnerability title and severity. - Added Support for Nmap: reNgine 2.0 allows to run Nmap scripts and vuln scans on ports found by Naabu. @ocervell - Added support for Shared Scan Variables in Scan Engine Configuration: - `enable_http_crawl`: (true/false) You can disable it to be more stealthy or focus on something different than HTTP - `timeout`: set timeout for all tasks - `rate_limit`: set rate limit for all tasks - `retries`: set retries for all tasks - `custom_header`: set the custom header for all tasks - Added Dalfox for XSS Vulnerability Scan - Added CRLFuzz for CRLF Vulnerability Scan - Added S3Scanner for scanning misconfigured S3 buckets - Improve OSINT Dork results, now detects admin panels, login pages and dashboards - Added Custom Dorks - Improved UI for vulnerability results, clicking on each vulnerability will open up a sidebar with vulnerability details. - Added HTTP Request and Response in vulnerability Results - Under Admin Settings, added an option to allow add/remove/deactivate additional users - Added Option to Preview Scan Report instead of forcing to download - Added Katana for crawling and spidering URLs - Added Netlas for Whois and subdomain gathering - Added TLSX for subdomain gathering - Added CTFR for subdomain gathering - Added historical IP in whois section ### Fixes - GF patterns do not run on 404 endpoints (#574 closed) - Fixes for retrieving whois data (#693 closed) - Related/Associated Domains in Whois section is now fixed ### Removed - Removed pwndb and tor related to it. - Removed tor for pwndb

2023-10-02 07:51:35+00:00

2023-10-07 10:37:23+00:00

README.md

<p align="center"> <a href="https://rengine.wiki"><img src=".github/screenshots/banner.gif" alt=""/></a> </p> <p align="center"><a href="https://github.com/yogeshojha/rengine/releases" target="_blank"><img src="https://img.shields.io/badge/version-v1.2.0-informational?&logo=none" alt="reNgine Latest Version" /></a>&nbsp;<a href="https://www.gnu.org/licenses/gpl-3.0" target="_blank"><img src="https://img.shields.io/badge/License-GPLv3-red.svg?&logo=none" alt="License" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/issues" target="_blank"><img src="https://img.shields.io/github/issues/yogeshojha/rengine?color=red&logo=none" alt="reNgine Issues" /></a>&nbsp;<a href="#" target="_blank"><img src="https://img.shields.io/badge/first--timers--only-friendly-blue.svg?&logo=none" alt="" /></a>&nbsp;<a href="https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine" target="_blank"><img src="https://cdn.huntr.dev/huntr_security_badge_mono.svg" alt="" /></a>&nbsp;</p> <p align="center"> <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/Open--Source--Summit-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://cyberweek.ae/2021/hitb-armory/" target="_blank"><img src="https://img.shields.io/badge/HITB--Armory-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=7uvP6MaQOX0" target="_blank"><img src="https://img.shields.io/badge/Black--Hat--Arsenal-USA--2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://drive.google.com/file/d/1Bh8lbf-Dztt5ViHJVACyrXMiglyICPQ2/view?usp=sharing" target="_blank"><img src="https://img.shields.io/badge/Defcon--Demolabs--29-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=A1oNOIc0h5A" target="_blank"><img src="https://img.shields.io/badge/Black--Hat--Arsenal-Europe--2020-blue.svg?&logo=none" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml/badge.svg" alt="" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/actions/workflows/build.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/build.yml/badge.svg" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://discord.gg/H6WzebwX3H" target="_blank"><img src="https://img.shields.io/discord/880363103689277461" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://opensourcesecurityindex.io/" target="_blank" rel="noopener"> <img style="width: 282px; height: 56px" src="https://opensourcesecurityindex.io/badge.svg" alt="Open Source Security Index - Fastest Growing Open Source Security Projects" width="282" height="56" /> </a> </p> <h3>reNgine 1.1<br>More than just recon!</h3> <h4>The only web application recon tool you will ever need!</h4> <p>Quickly discover the attack surface, and identify vulnerabilities using highly customizable and powerful scan engines. Enjoy peace of mind with reNgine's continuous monitoring, deeper reconnaissance, and open-source powered Vulnerability Scanner.</p> <h4>What is reNgine?</h4> <p align="left">reNgine is a web application reconnaissance suite that focuses on a highly configurable streamlined reconnaissance process via engines, reconnaissance data correlation, continuous monitoring, database backed reconnaissance data and a simple yet intuitive user interface. With features such as sub-scan, deeper co-relation, report generation, etc., reNgine aims to fill the gap in traditional reconnaissance tools and is likely to be a better alternative to existing commercial tools. reNgine makes it easy for penetration testers and security auditors to gather reconnaissance data with minimal configuration. </p>  <p align="center"> ⭐<a href="https://rengine.wiki">reNgine Documentation</a> · <a href="https://rengine.wiki/changelog/">What's new</a> · <a href="https://github.com/yogeshojha/rengine/blob/master/.github/CONTRIBUTING.md">Contribute</a> · <a href="https://github.com/yogeshojha/rengine/issues">Report Bug</a> · <a href="https://github.com/yogeshojha/rengine/issues">Request Feature</a>⭐ </p>  ## Table of Contents * [About reNgine](#about-rengine) * [Features](#features) * [Documentation](#documentation) * [Quick Installation](#quick-installation) * [What's new in reNgine](#changelog) * [reNgine Bug Bounty Program](#rengine-bug-bounty-program) * [Screenshots](#screenshots) * [Contributing](#contributing) * [reNgine Support](#rengine-support) * [Related Projects](#related-projects) * [Support and Sponsoring](#support-and-sponsoring) * [License](#license)  ### reNgine 2.0 codenamed Jasper I am currently working on reNgine 2.0, which will probably be announced sometime between May and August 2023. reNgine 2.0 will be the most advanced reNgine ever, a lot of work will be done in how scans are performed, things such as Pause and Resume Scan, Axiom Integration, more deeper correlation, Project Options, Multiple Tenants, etc. Please submit your feature requests via GitHub issues.  ## About reNgine You can watch [reNgine 1.1 release trailer here.](https://www.youtube.com/watch?v=iy_6F7Vq8Lo) (Recommended) <img src="https://user-images.githubusercontent.com/17223002/164993688-50eb95f2-3653-4ef7-bd3b-ef7a096824ea.jpeg">  reNgine is a web application reconnaissance suite with a focus on a highly configurable, streamlined reconnaissance process. reNgine is backed by a database, with data correlation and organisation, the custom query "like" language for filtering reconnaissance data, reNgine aims to address the shortcomings of the traditional reconnaissance workflow. The developers behind reNgine understand that reconnaissance data can be huge and manually searching for records to attack can be tedious, so features such as identifying subdomains of interest help penetration testers focus on attack rather than reconnaissance. reNgine also focuses on continuous monitoring. Penetration testers can choose to schedule the scan at regular intervals and be notified via notification channels such as Discord, Slack and Telegram of any new subdomains or vulnerabilities identified, or any changes to the recon data. Interoperability is something every reconnaissance tool needs, and reNgine is no different. Starting with reNgine 1.0, we have added features such as import and export of subdomains, endpoints, GF pattern matched endpoints, etc. This allows you to use your favourite reconnaissance workflow in conjunction with reNgine. PDF reports are something every individual or team needs. From reNgine 1.1, reNgine also comes with the option to download PDF reports. You can also choose the type of report, a full scan report or just a reconnaissance report. We also understand that PDF reports need to be customisable. Choose the colour of the report you want, customise the executive summary, etc. You choose how your PDF report looks! reNgine features highly configurable scan engines based on YAML, allowing penetration testers to create as many reconnaissance engines of their choice as they like, configure them as they like, and use them against any targets for scanning. These engines allow penetration testers to use the tools of their choice, with the configuration of their choice. Out of the box, reNgine comes with several scan engines such as Full Scan, Passive Scan, Screenshot Gathering, OSINT Engine, etc. Our focus has always been on finding the right reconnaissance data with the least amount of effort. After several discussions with fellow hackers/pentesters, a screenshot gallery was a must, reNgine also comes with a screenshot gallery, and what's more exciting than having a screenshot gallery with filters, filter screenshots with HTTP status, technology, ports and services. We also want our fellow hackers to stay ahead of the game, so reNgine also comes with automatic vulnerability reporting (ATM only Hackerone is supported, other platforms may come soon). This allows hackers to define their vulnerability reporting template and reNgine will do the rest of the work to report the vulnerability as soon as it is identified.  <img src="https://user-images.githubusercontent.com/17223002/164993945-aabdbb4a-2b9d-4951-ba27-5f2f5abd1d8b.gif"> ## Features * Reconnaissance: Subdomain Discovery, IP and Open Ports Identification, Endpoints Discovery, Directory and Files fuzzing, - Screenshot gathering, Vulnerability scan using Nuclei, WHOIS Identification, WAF Detection etc. * Highly configurable YAML-based Scan Engines * Support for Parallel Scans and Subscans * Automatically report Vulnerabilities to HackerOne * Recon Data visualization * OSINT Capabilities (Meta info Gathering, Employees Gathering, Email Address with an option to look password in the leaked database, - dorks, etc.) * Customizable Alerts/Notifications on Slack, Discord, and Telegram * Perform Advanced Query lookup using natural language alike and, or, not operations * Recon Notes and Todos * Clocked Scans (Run reconnaissance exactly at X Hours and Y minutes) and Periodic Scans (Runs reconnaissance every X minutes/- hours/days/week) * Proxy Support * Screenshot Gallery with Filters * Powerful recon data filtering with autosuggestions * Recon Data changes, find new/removed subdomains/endpoints * Tag targets into the Organization * Identify Interesting Subdomains * Custom GF patterns and custom Nuclei Templates * Edit tool-related configuration files (Nuclei, Subfinder, Naabu, amass) * Add external tools from Github/Go * Interoperable with other tools, Import/Export Subdomains/Endpoints * Import Targets via IP and/or CIDRs * Report Generation * Toolbox: Comes bundled with most commonly used tools such as whois lookup, CMS detector, CVE lookup, etc. * Identification of related domains and related TLDs for targets * Find actionable insights such as Most Common Vulnerability, Most Common CVE ID, Most Vulnerable Target/Subdomain, etc.  ## Documentation You can find reNgine documentation at [https://rengine.wiki](https://rengine.wiki)  ## Quick Installation **Note:** Only Ubuntu/VPS 1. Clone this repo ```bash git clone https://github.com/yogeshojha/rengine && cd rengine ``` 1. Edit the dotenv file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```bash nano .env ``` 1. Run the installation script, Please keep an eye for any prompt, you will also be asked for username and password for reNgine. ```bash sudo ./install.sh ``` If `install.sh` does not have install permission, please change it, `chmod +x install.sh` **reNgine can now be accessed from <https://127.0.0.1> or if you're on the VPS <https://your_vps_ip_address>** A detailed installation guide can also be found [here](https://www.rffuste.com/2022/05/23/rengine-a-brief-overview/). Thanks to Rubén! ## Installation (Mac/Windows/Other) Installation instructions can be found at [https://reNgine.wiki/install/detailed/](https://reNgine.wiki/install/detailed/) ## Updating 1. Updating is as simple as running the following command: ```bash sudo ./update.sh ``` If `update.sh` does not have execution permissions, please change it, `sudo chmod +x update.sh` ## Changelog [Please find the latest release notes and changelog here.](https://rengine.wiki/changelog/)  ## reNgine Bug Bounty Program [](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine) Security researchers, welcome aboard! I'm excited to announce the reNgine bug bounty programme in collaboration with [huntr.dev](https://huntr.dev), which means that you will be rewarded for any vulnerabilities you find in reNgine. Thank you for your interest in reporting reNgine vulnerabilities! If you are aware of any potential security vulnerabilities in reNgine, we encourage you to report them immediately via [huntr.dev](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine). **Please do not disclose vulnerabilities via Github issues/blogs/tweets after/before reporting to huntr.dev as this is explicitly against the disclosure policy of huntr.dev and reNgine and will not be considered for monetary rewards.** Please note that the reNgine maintainer does not set the bounty amount. The bounty reward is determined by an industry-first equation developed by huntr.dev to understand the popularity, impact and value of repositories to the open-source community. **What do I expect from security researchers?** * Patience: Please note that I am currently the only maintainer in reNgine and it will take some time to validate your report. I ask for your patience during this process. * Respect for privacy and security reports: Please do not publicly disclose any vulnerabilities (including GitHub issues) before or after reporting them on huntr.dev! This is against the disclosure policy and will not be rewarded. * Respect the rules **What do you get in return?** * Thanks from the maintainer * Monetary rewards * CVE ID(s) Please find the [FAQ](https://www.huntr.dev/faq) and [Responsible disclosure policy](https://www.huntr.dev/policy/) from huntr.dev.  ## Screenshots ### Scan Results  ### General Usage <img src="https://user-images.githubusercontent.com/17223002/164993781-b6012995-522b-480a-a8bf-911193d35894.gif"> ### Initiating Subscan <img src="https://user-images.githubusercontent.com/17223002/164993749-1ad343d6-8ce7-43d6-aee7-b3add0321da7.gif"> ### Recon Data filtering <img src="https://user-images.githubusercontent.com/17223002/164993687-b63f3de8-e033-4ac0-808e-a2aa377d3cf8.gif"> ### Report Generation <img src="https://user-images.githubusercontent.com/17223002/164993689-c796c6cd-eb61-43f4-800d-08aba9740088.gif"> ### Toolbox <img src="https://user-images.githubusercontent.com/17223002/164993751-d687e88a-eb79-440f-9dc0-0ad006901620.gif"> ### Adding Custom tool in Tools Arsenal <img src="https://user-images.githubusercontent.com/17223002/164993670-466f6459-9499-498b-a9bd-526476d735a7.gif">  ## Contributing Contributions are what make the open-source community such an amazing place to learn, inspire and create. Every contributions you make is **greatly appreciated**. Your contributions can be as simple as fixing the indentation or UI, or as complex as adding new modules and features. See the [Contributing Guide](.github/CONTRIBUTING.md) to get started. You can also [join our Discord channel #development](https://discord.gg/JuhHdHTtwd) for any development related questions.  ### First-time Open Source contributors Please note that reNgine is beginner friendly. If you have never done open-source before, we encourage you to do so. **We will be happy and proud of your first PR ever.** You can start by resolving any [open issues](https://github.com/yogeshojha/rengine/issues).  ## reNgine Support Please do not use GitHub for support requests. Instead, [join our Discord channel #support](https://discord.gg/azv6fzhNCE)  ## Related Projects There are many other great reconnaissance frameworks out there, you can use reNgine in conjunction with those tools. But they are great in their own right, and can sometimes produce better results than reNgine. * [ReconFTW](https://github.com/six2dez/reconftw#sample-video) * [Reconmap](https://github.com/reconmap/reconmap)  ## Support and Sponsoring Over the past few years, I have been working hard on reNgine to add new features with the sole aim of making it the de facto standard for reconnaissance. I spend most of my free time and weekends working on reNgine. I do this in addition to my day job. I am happy to have received such overwhelming support from the community. But to keep this project alive, I am looking for financial support. | Paypal | Bitcoin | Ethereum | | :-------------------------------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | |[https://www.paypal.com/paypalme/yogeshojha11](https://www.paypal.com/paypalme/yogeshojha11) | `35AiKyNswNZ4TZUSdriHopSCjNMPi63BCX` | `0xe7A337Da6ff98A28513C26A7Fec8C9b42A63d346` OR * Add a [GitHub Star](https://github.com/yogeshojha/rengine) to the project. * Tweet about this project, or maybe blogs? * Maybe nominate me for [GitHub Stars?](https://stars.github.com/nominate/) * Join DigitalOcean using my [referral link](https://m.do.co/c/e353502d19fc) your profit is **$100** and I get $25 DO credit. This will help me test reNgine on VPS before I release any major features. It takes a considerable amount of time to add new features and make sure everything works. Donating is your way of saying: **reNgine is awesome**. Any support is greatly appreciated! Thank you!  ## License Distributed under the GNU GPL v3 License. See [LICENSE](LICENSE) for more information. 

<p align="center"> <a href="https://rengine.wiki"><img src=".github/screenshots/banner.gif" alt=""/></a> </p> <p align="center"><a href="https://github.com/yogeshojha/rengine/releases" target="_blank"><img src="https://img.shields.io/badge/version-v2.0.0-informational?&logo=none" alt="reNgine Latest Version" /></a>&nbsp;<a href="https://www.gnu.org/licenses/gpl-3.0" target="_blank"><img src="https://img.shields.io/badge/License-GPLv3-red.svg?&logo=none" alt="License" /></a>&nbsp;<a href="#" target="_blank"><img src="https://img.shields.io/badge/first--timers--only-friendly-blue.svg?&logo=none" alt="" /></a>&nbsp;<a href="https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine" target="_blank"><img src="https://cdn.huntr.dev/huntr_security_badge_mono.svg" alt="" /></a>&nbsp;</p> <p align="center"> <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Asia-2023-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=Xk_YH83IQgg" target="_blank"><img src="https://img.shields.io/badge/Open--Source--Summit-2022-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://cyberweek.ae/2021/hitb-armory/" target="_blank"><img src="https://img.shields.io/badge/HITB--Armory-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=7uvP6MaQOX0" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--USA-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://drive.google.com/file/d/1Bh8lbf-Dztt5ViHJVACyrXMiglyICPQ2/view?usp=sharing" target="_blank"><img src="https://img.shields.io/badge/Defcon--Demolabs--29-2021-blue.svg?logo=none" alt="" /></a>&nbsp; <a href="https://www.youtube.com/watch?v=A1oNOIc0h5A" target="_blank"><img src="https://img.shields.io/badge/BlackHat--Arsenal--Europe-2020-blue.svg?&logo=none" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/codeql-analysis.yml/badge.svg" alt="" /></a>&nbsp;<a href="https://github.com/yogeshojha/rengine/actions/workflows/build.yml" target="_blank"><img src="https://github.com/yogeshojha/rengine/actions/workflows/build.yml/badge.svg" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://discord.gg/H6WzebwX3H" target="_blank"><img src="https://img.shields.io/discord/880363103689277461" alt="" /></a>&nbsp; </p> <p align="center"> <a href="https://opensourcesecurityindex.io/" target="_blank" rel="noopener"> <img style="width: 282px; height: 56px" src="https://opensourcesecurityindex.io/badge.svg" alt="Open Source Security Index - Fastest Growing Open Source Security Projects" width="282" height="56" /> </a> </p> <h3>reNgine 2.0-jasper<br>Redefining the future of reconnaissance!</h3> <h4>What is reNgine?</h4> <p align="left">reNgine is your go-to web application reconnaissance suite that's designed to simplify and streamline the reconnaissance process for security professionals, penetration testers, and bug bounty hunters. With its highly configurable engines, data correlation capabilities, continuous monitoring, database-backed reconnaissance data, and an intuitive user interface, reNgine redefines how you gather critical information about your target web applications. Traditional reconnaissance tools often fall short in terms of configurability and efficiency. reNgine addresses these shortcomings and emerges as a excellent alternative to existing commercial tools. reNgine was created to address the limitations of traditional reconnaissance tools and provide a better alternative, even surpassing some commercial offerings. Whether you're a bug bounty hunter, a penetration tester, or a corporate security team, reNgine is your go-to solution for automating and enhancing your information-gathering efforts. </p> reNgine 2.0-jasper is out now, you can [watch reNgine 2.0-jasper release trailer here!](https://youtu.be/VwkOWqiWW5g) reNgine 2.0-Jasper would not have been possible without [@ocervell](https://github.com/ocervell) valuable contributions. [@ocervell](https://github.com/ocervell) did majority of the refactoring if not all and also added a ton of features. Together, we wish to shape the future of web application reconnaissance, and it's developers like [@ocervell](https://github.com/ocervell) and a [ton of other developers and hackers from our community](https://github.com/yogeshojha/rengine/graphs/contributors) who inspire and drive us forward. Thank you, [@ocervell](https://github.com/ocervell), for your outstanding work and unwavering commitment to reNgine. Checkout our contributers here: [Contributers](https://github.com/yogeshojha/rengine/graphs/contributors)  ### Documentation You can find detailed documentation at [https://rengine.wiki](https://rengine.wiki)  ### Table of Contents * [About reNgine](#about-rengine) * [Workflow](#workflow) * [Features](#features) * [Scan Engine](#scan-engine) * [Quick Installation](#quick-installation) * [What's new in reNgine 2.0](#changelog) * [Screenshots](#screenshots) * [Contributing](#contributing) * [reNgine Support](#rengine-support) * [Support and Sponsoring](#support-and-sponsoring) * [reNgine Bug Bounty Program](#rengine-bug-bounty-program) * [License](#license)  ### About reNgine reNgine is not an ordinary reconnaissance suite; it's a game-changer! We've turbocharged the traditional workflow with groundbreaking features that is sure to ease your reconnaissance game. reNgine redefines the art of reconnaissance with highly configurable scan engines, recon data correlation, continuous monitoring, GPT powered Vulnerability Report, Project Management and role based access control etc. 🦾&nbsp;&nbsp; reNgine has advanced reconnaissance capabilities, harnessing a range of open-source tools to deliver a comprehensive web application reconnaissance experience. With it's intuitive User Interface, it excels in subdomain discovery, pinpointing IP addresses and open ports, collecting endpoints, conducting directory and file fuzzing, capturing screenshots, and performing vulnerability scans. To summarize, it does end-to-end reconnaissance. With WHOIS identification and WAF detection, it offers deep insights into target domains. Additionally, reNgine also identifies misconfigured S3 buckets and find interesting subdomains and URLS, based on specific keywords to helps you identify your next target, making it an go to tool for efficient reconnaissance. 🗃️&nbsp; &nbsp; Say goodbye to recon data chaos! reNgine seamlessly integrates with a database, providing you with unmatched data correlation and organization. Forgot the hassle of grepping through json, txt or csv files. Plus, our custom query language lets you filter reconnaissance data effortlessly using natural language like operators such as filtering all alive subdomains with `http_status=200` and also filter all subdomains that are alive and has admin in name `http_status=200&name=admin` 🔧&nbsp;&nbsp; reNgine offers unparalleled flexibility through its highly configurable scan engines, based on a YAML-based configuration. It offers the freedom to create and customize recon scan engines based on any kind of requirement, users can tailor them to their specific objectives and preferences, from thread management to timeout settings and rate-limit configurations, everything is customizable. Additionally, reNgine offers a range of pre-configured scan engines right out of the box, including Full Scan, Passive Scan, Screenshot Gathering, and the OSINT Scan Engine. These ready-to-use engines eliminate the need for extensive manual setup, aligning perfectly with reNgine's core mission of simplifying the reconnaissance process and enabling users to effortlessly access the right reconnaissance data with minimal effort. 💎&nbsp;&nbsp;Subscans: Subscan is a game-changing feature in reNgine, setting it apart as the only open-source tool of its kind to offer this capability. With Subscan, waiting for the entire pipeline to complete is a thing of the past. Now, users can swiftly respond to newfound discoveries during reconnaissance. Whether you've stumbled upon an intriguing subdomain and wish to conduct a focused port scan or want to delve deeper with a vulnerability assessment, reNgine has you covered. 📃&nbsp;&nbsp; PDF Reports: In addition to its robust reconnaissance capabilities, reNgine goes the extra mile by simplifying the report generation process, recognizing the crucial role that PDF reports play in the realm of end-to-end reconnaissance. Users can effortlessly generate and customize PDF reports to suit their exact needs. Whether it's a Full Scan Report, Vulnerability Report, or a concise reconnaissance report, reNgine provides the flexibility to choose the report type that best communicates your findings. Moreover, the level of customization is unparalleled, allowing users to select report colors, fine-tune executive summaries, and even add personalized touches like company names and footers. With GPT integration, your reports aren't just a report, with remediation steps, and impacts, you get 360-degree view of the vulnerabilities you've uncovered. 🔖&nbsp; &nbsp; Say Hello to Projects! reNgine 2.0 introduces a powerful addition that enables you to efficiently organize your web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. Each projects will have separate dashboard and all the scan results will be separated from each projects, while scan engines and configuration will be shared across all the projects. ⚙️&nbsp; &nbsp; Roles and Permissions! Begining reNgine 2.0, we've taken your web application reconnaissance to a whole new level of control and security. Now, you can assign distinct roles to your team members—Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - 🔐 Sys Admin: Sys Admin is a super user that has permission to modify system and scan related configurations, scan engines, create new users, add new tools etc. Super user can initiate scans and subscans effortlessly. - 🔍 Penetration Tester: Penetration Tester will be allowed to modify and initiate scans and subscans, add or update targets, etc. A penetration tester will not be allowed to modify system configurations. - 📊 Auditor: Auditor can only view and download the report. An auditor can not change any system or scan related configurations nor can initiate any scans or subscans. 🚀&nbsp;&nbsp; GPT Vulnerability Report Generation: Get ready for the future of penetration testing reports with reNgine's groundbreaking feature: "GPT-Powered Report Generation"! With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments that read like they were written by a human security expert! **But that's not all!** Our GPT-driven reports go the extra mile by scouring the web for related news articles, blogs, and references, so you have a 360-degree view of the vulnerabilities you've uncovered. With reNgine 2.0 revolutionize your penetration testing game and impress your clients with reports that are not just informative but engaging and comprehensive with detailed analysis on impact assessment and remediation strategies. 🥷&nbsp;&nbsp; GPT-Powered Attack Surface Generation: With reNgine 2.0, reNgine seamlessly integrates with GPT to identify the attacks that you can likely perform on a subdomain. By making use of reconnaissance data such as page title, open ports, subdomain name etc, reNgine can advice you the attacks you could perform on a target. reNgine will also provide you the rationale on why the specific attack is likely to be successful. 🧭&nbsp;&nbsp;Continuous monitoring: Continuous monitoring is at the core of reNgine's mission, and it's robust continuous monitoring feature ensures that their targets are under constant scrutiny. With the flexibility to schedule scans at regular intervals, penetration testers can effortlessly stay informed about their targets. What sets reNgine apart is its seamless integration with popular notification channels such as Discord, Slack, and Telegram, delivering real-time alerts for newly discovered subdomains, vulnerabilities, or any changes in reconnaissance data.  ### Workflow <img src="https://github.com/yogeshojha/rengine/assets/17223002/10c475b8-b4a8-440d-9126-77fe2038a386">  ### Features * Reconnaissance: * Subdomain Discovery * IP and Open Ports Identification * Endpoints Discovery * Directory/Files fuzzing * Screenshot Gathering * Vulnerability Scan * Nuclei * Dalfox XSS Scanner * CRLFuzzer * Misconfigured S3 Scanner * WHOIS Identification * WAF Detection * OSINT Capabilities * Meta info Gathering * Employees Gathering * Email Address gathering * Google Dorking for sensitive info and urls * Projects, create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. * Perform Advanced Query lookup using natural language alike and, or, not operations * Highly configurable YAML-based Scan Engines * Support for Parallel Scans * Support for Subscans * Recon Data visualization * GPT Vulnerability Description, Impact and Remediation generation * GPT Attack Surface Generator * Multiple Roles and Permissions to cater a team's need * Customizable Alerts/Notifications on Slack, Discord, and Telegram * Automatically report Vulnerabilities to HackerOne * Recon Notes and Todos * Clocked Scans (Run reconnaissance exactly at X Hours and Y minutes) and Periodic Scans (Runs reconnaissance every X minutes/- hours/days/week) * Proxy Support * Screenshot Gallery with Filters * Powerful recon data filtering with autosuggestions * Recon Data changes, find new/removed subdomains/endpoints * Tag targets into the Organization * Smart Duplicate endpoint removal based on page title and content length to cleanup the reconnaissance data * Identify Interesting Subdomains * Custom GF patterns and custom Nuclei Templates * Edit tool-related configuration files (Nuclei, Subfinder, Naabu, amass) * Add external tools from Github/Go * Interoperable with other tools, Import/Export Subdomains/Endpoints * Import Targets via IP and/or CIDRs * Report Generation * Toolbox: Comes bundled with most commonly used tools during penetration testing such as whois lookup, CMS detector, CVE lookup, etc. * Identification of related domains and related TLDs for targets * Find actionable insights such as Most Common Vulnerability, Most Common CVE ID, Most Vulnerable Target/Subdomain, etc.  ### Scan Engine ```yaml subdomain_discovery: { 'uses_tools': [ 'subfinder', 'ctfr', 'sublist3r', 'tlsx', 'oneforall', 'netlas' ], 'enable_http_crawl': true, 'threads': 30, 'timeout': 5, } http_crawl: {} port_scan: { 'enable_http_crawl': true, 'timeout': 5, # 'exclude_ports': [], # 'exclude_subdomains': true, 'ports': ['top-100'], 'rate_limit': 150, 'threads': 30, 'passive': false, # 'use_naabu_config': false, # 'enable_nmap': true, # 'nmap_cmd': '', # 'nmap_script': '', # 'nmap_script_args': '' } osint: { 'discover': [ 'emails', 'metainfo', 'employees' ], 'dorks': [ 'login_pages', 'admin_panels', 'dashboard_pages', 'stackoverflow', 'social_media', 'project_management', 'code_sharing', 'config_files', 'jenkins', 'wordpress_files', 'php_error', 'exposed_documents', 'db_files', 'git_exposed' ], 'custom_dorks': [ { 'lookup_site': 'google.com', 'lookup_keywords': '/home/' }, { 'lookup_site': '_target_', 'lookup_extensions': 'jpg,png' } ], 'intensity': 'normal', 'documents_limit': 50 } dir_file_fuzz: { 'auto_calibration': true, 'enable_http_crawl': true, 'rate_limit': 150, 'extensions': ['html', 'php','git','yaml','conf','cnf','config','gz','env','log','db','mysql','bak','asp','aspx','txt','conf','sql','json','yml','pdf'], 'follow_redirect': false, 'max_time': 0, 'match_http_status': [200, 204], 'recursive_level': 2, 'stop_on_error': false, 'timeout': 5, 'threads': 30, 'wordlist_name': 'dicc' } fetch_url: { 'uses_tools': [ 'gospider', 'hakrawler', 'waybackurls', 'gospider', 'katana' ], 'remove_duplicate_endpoints': true, 'duplicate_fields': [ 'content_length', 'page_title' ], 'enable_http_crawl': true, 'gf_patterns': ['debug_logic', 'idor', 'interestingEXT', 'interestingparams', 'interestingsubs', 'lfi', 'rce', 'redirect', 'sqli', 'ssrf', 'ssti', 'xss'], 'ignore_file_extensions': ['png', 'jpg', 'jpeg', 'gif', 'mp4', 'mpeg', 'mp3'] # 'exclude_subdomains': true } vulnerability_scan: { 'run_nuclei': false, 'run_dalfox': false, 'run_crlfuzz': false, 'run_s3scanner': true, 'enable_http_crawl': true, 'concurrency': 50, 'intensity': 'normal', 'rate_limit': 150, 'retries': 1, 'timeout': 5, 'fetch_gpt_report': true, 'nuclei': { 'use_conf': false, 'severities': [ 'unknown', 'info', 'low', 'medium', 'high', 'critical' ], # 'tags': [], # 'templates': [], # 'custom_templates': [], }, 's3scanner': { 'threads': 100, 'providers': [ 'aws', 'gcp', 'digitalocean', 'dreamhost', 'linode' ] } } waf_detection: {} screenshot: { 'enable_http_crawl': true, 'intensity': 'normal', 'timeout': 10, 'threads': 40 } # custom_header: "Cookie: Test" ```  ### Quick Installation **Note:** Only Ubuntu/VPS 1. Clone this repo ```bash git clone https://github.com/yogeshojha/rengine && cd rengine ``` 1. Edit the dotenv file, **please make sure to change the password for postgresql `POSTGRES_PASSWORD`!** ```bash nano .env ``` 1. In the dotenv file, you may also modify the Scaling Configurations ```bash MAX_CONCURRENCY=80 MIN_CONCURRENCY=10 ``` MAX_CONCURRENCY: This parameter specifies the maximum number of reNgine's concurrent Celery worker processes that can be spawned. In this case, it's set to 80, meaning that the application can utilize up to 80 concurrent worker processes to execute tasks concurrently. This is useful for handling a high volume of scans or when you want to scale up processing power during periods of high demand. If you have more CPU cores, you will need to increase this for maximised performance. MIN_CONCURRENCY: On the other hand, MIN_CONCURRENCY specifies the minimum number of concurrent worker processes that should be maintained, even during periods of lower demand. In this example, it's set to 10, which means that even when there are fewer tasks to process, at least 10 worker processes will be kept running. This helps ensure that the application can respond promptly to incoming tasks without the overhead of repeatedly starting and stopping worker processes. These settings allow for dynamic scaling of Celery workers, ensuring that the application efficiently manages its workload by adjusting the number of concurrent workers based on the workload's size and complexity 1. Run the installation script, Please keep an eye for any prompt, you will also be asked for username and password for reNgine. ```bash sudo ./install.sh ``` If `install.sh` does not have install permission, please change it, `chmod +x install.sh` **reNgine can now be accessed from <https://127.0.0.1> or if you're on the VPS <https://your_vps_ip_address>** **Unless you are on development branch, please do not access reNgine via any ports** ### Installation (Mac/Windows/Other) Installation instructions can be found at [https://reNgine.wiki/install/detailed/](https://reNgine.wiki/2.0/install/detailed/) ### Updating 1. Updating is as simple as running the following command: ```bash cd rengine && sudo ./update.sh ``` If `update.sh` does not have execution permissions, please change it, `sudo chmod +x update.sh` ### Changelog [Please find the latest release notes and changelog here.](https://rengine.wiki/changelog/)  ### Screenshots #### Scan Results  #### General Usage <img src="https://user-images.githubusercontent.com/17223002/164993781-b6012995-522b-480a-a8bf-911193d35894.gif"> #### Initiating Subscan <img src="https://user-images.githubusercontent.com/17223002/164993749-1ad343d6-8ce7-43d6-aee7-b3add0321da7.gif"> #### Recon Data filtering <img src="https://user-images.githubusercontent.com/17223002/164993687-b63f3de8-e033-4ac0-808e-a2aa377d3cf8.gif"> #### Report Generation <img src="https://user-images.githubusercontent.com/17223002/164993689-c796c6cd-eb61-43f4-800d-08aba9740088.gif"> #### Toolbox <img src="https://user-images.githubusercontent.com/17223002/164993751-d687e88a-eb79-440f-9dc0-0ad006901620.gif"> #### Adding Custom tool in Tools Arsenal <img src="https://user-images.githubusercontent.com/17223002/164993670-466f6459-9499-498b-a9bd-526476d735a7.gif">  ### Contributing Contributions are what make the open-source community such an amazing place to learn, inspire and create. Every contributions you make is **greatly appreciated**. Your contributions can be as simple as fixing the indentation or UI, or as complex as adding new modules and features. See the [Contributing Guide](.github/CONTRIBUTING.md) to get started. You can also [join our Discord channel #development](https://discord.gg/JuhHdHTtwd) for any development related questions.  ### First-time Open Source contributors Please note that reNgine is beginner friendly. If you have never done open-source before, we encourage you to do so. **We will be happy and proud of your first PR ever.** You can start by resolving any [open issues](https://github.com/yogeshojha/rengine/issues).  ### reNgine Support Please do not use GitHub for support requests. Instead, [join our Discord channel #support](https://discord.gg/azv6fzhNCE).  ### Support and Sponsoring Over the past few years, I have been working hard on reNgine to add new features with the sole aim of making it the de facto standard for reconnaissance. I spend most of my free time and weekends working on reNgine. I do this in addition to my day job. I am happy to have received such overwhelming support from the community. But to keep this project alive, I am looking for financial support. | Paypal | Bitcoin | Ethereum | | :-------------------------------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------------------------------------------------------: | |[https://www.paypal.com/paypalme/yogeshojha11](https://www.paypal.com/paypalme/yogeshojha11) | `35AiKyNswNZ4TZUSdriHopSCjNMPi63BCX` | `0xe7A337Da6ff98A28513C26A7Fec8C9b42A63d346` OR * Add a [GitHub Star](https://github.com/yogeshojha/rengine) to the project. * Tweet about this project, or maybe blogs? * Maybe nominate me for [GitHub Stars?](https://stars.github.com/nominate/) * Join DigitalOcean using my [referral link](https://m.do.co/c/e353502d19fc) your profit is **$100** and I get $25 DO credit. This will help me test reNgine on VPS before I release any major features. It takes a considerable amount of time to add new features and make sure everything works. Donating is your way of saying: **reNgine is awesome**. Any support is greatly appreciated! Thank you!  ### reNgine Bug Bounty Program [](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine) Security researchers, welcome aboard! I'm excited to announce the reNgine bug bounty programme in collaboration with [huntr.dev](https://huntr.dev), which means that you will be rewarded for any vulnerabilities you find in reNgine. Thank you for your interest in reporting reNgine vulnerabilities! If you are aware of any potential security vulnerabilities in reNgine, we encourage you to report them immediately via [huntr.dev](https://huntr.dev/bounties/disclose/?target=https%3A%2F%2Fgithub.com%2Fyogeshojha%2Frengine). **Please do not disclose vulnerabilities via Github issues/blogs/tweets after/before reporting to huntr.dev as this is explicitly against the disclosure policy of huntr.dev and reNgine and will not be considered for monetary rewards.** Please note that the reNgine maintainer does not set the bounty amount. The bounty reward is determined by an industry-first equation developed by huntr.dev to understand the popularity, impact and value of repositories to the open-source community. **What do I expect from security researchers?** * Patience: Please note that I am currently the only maintainer in reNgine and it will take some time to validate your report. I ask for your patience during this process. * Respect for privacy and security reports: Please do not publicly disclose any vulnerabilities (including GitHub issues) before or after reporting them on huntr.dev! This is against the disclosure policy and will not be rewarded. * Respect the rules **What do you get in return?** * Thanks from the maintainer * Monetary rewards * CVE ID(s) Please find the [FAQ](https://www.huntr.dev/faq) and [Responsible disclosure policy](https://www.huntr.dev/policy/) from huntr.dev.  ### License Distributed under the GNU GPL v3 License. See [LICENSE](LICENSE) for more information.  <p align="right">(ChatGPT was used to write some or most part of this README section.)</p>

yogeshojha

3c60bc1ee495044794d91edee0c96fff73ab46c7

5413708d243799a5271440c47c6f98d0c51154ca

Thank you for pointing out ;p

yogeshojha

yogeshojha/rengine

2.0-jasper release

### Added - Projects: Projects allow you to efficiently organize their web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. - Roles and Permissions: assign distinct roles to your team members: Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - GPT-powered Report Generation: With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments. - API Vault: This feature allows you to organize your API keys such as OpenAI or Netlas API keys. - GPT-powered Attack Surface Generation - URL gathering now is much more efficient, removing duplicate endpoints based on similar HTTP Responses, having the same content_lenth, or page_title. Custom duplicate fields can also be set from the scan engine configuration. - URL Path filtering while initiating scan: For instance, if we want to scan only endpoints starting with https://example.com/start/, we can pass the /start as a path filter while starting the scan. @ocervell - Expanding Target Concept: reNgine 2.0 now accepts IPs, URLS, etc as targets. (#678, #658) Excellent work by @ocervell - A ton of refactoring on reNgine's core to improve scan efficiency. Massive kudos to @ocervell - Created a custom celery workflow to be able to run several tasks in parallel that are not dependent on each other, such OSINT task and subdomain discovery will run in parallel, and directory and file fuzzing, vulnerability scan, screenshot gathering etc. will run in parallel after port scan or url fetching is completed. This will increase the efficiency of scans and instead of having one long flow of tasks, they can run independently on their own. @ocervell - Refactored all tasks to run asynchronously @ocervell - Added a stream_command that allows to read the output of a command live: this means the UI is updated with results while the command runs and does not have to wait until the task completes. Excellent work by @ocervell - Pwndb is now replaced by h8mail. @ocervell - Group Scan Results: reNgine 2.0 allows to group of subdomains based on similar page titles and HTTP status, and also vulnerability grouping based on the same vulnerability title and severity. - Added Support for Nmap: reNgine 2.0 allows to run Nmap scripts and vuln scans on ports found by Naabu. @ocervell - Added support for Shared Scan Variables in Scan Engine Configuration: - `enable_http_crawl`: (true/false) You can disable it to be more stealthy or focus on something different than HTTP - `timeout`: set timeout for all tasks - `rate_limit`: set rate limit for all tasks - `retries`: set retries for all tasks - `custom_header`: set the custom header for all tasks - Added Dalfox for XSS Vulnerability Scan - Added CRLFuzz for CRLF Vulnerability Scan - Added S3Scanner for scanning misconfigured S3 buckets - Improve OSINT Dork results, now detects admin panels, login pages and dashboards - Added Custom Dorks - Improved UI for vulnerability results, clicking on each vulnerability will open up a sidebar with vulnerability details. - Added HTTP Request and Response in vulnerability Results - Under Admin Settings, added an option to allow add/remove/deactivate additional users - Added Option to Preview Scan Report instead of forcing to download - Added Katana for crawling and spidering URLs - Added Netlas for Whois and subdomain gathering - Added TLSX for subdomain gathering - Added CTFR for subdomain gathering - Added historical IP in whois section ### Fixes - GF patterns do not run on 404 endpoints (#574 closed) - Fixes for retrieving whois data (#693 closed) - Related/Associated Domains in Whois section is now fixed ### Removed - Removed pwndb and tor related to it. - Removed tor for pwndb

2023-10-02 07:51:35+00:00

2023-10-07 10:37:23+00:00

web/dashboard/views.py

from datetime import timedelta from targetApp.models import Domain from startScan.models import * from django.utils import timezone from django.shortcuts import render, redirect from django.http import HttpResponse from django.db.models.functions import TruncDay from django.contrib.auth.decorators import login_required from django.dispatch import receiver from django.contrib.auth.signals import user_logged_out, user_logged_in from django.contrib import messages from django.contrib.auth import update_session_auth_hash from django.contrib.auth.forms import PasswordChangeForm from django.db.models import Count, Value, CharField, Q def index(request): domain_count = Domain.objects.all().count() endpoint_count = EndPoint.objects.all().count() scan_count = ScanHistory.objects.all().count() subdomain_count = Subdomain.objects.all().count() subdomain_with_ip_count = Subdomain.objects.filter(ip_addresses__isnull=False).count() alive_count = \ Subdomain.objects.all().exclude(http_status__exact=0).count() endpoint_alive_count = \ EndPoint.objects.filter(http_status__exact=200).count() vulnerabilities = Vulnerability.objects.all() info_count = vulnerabilities.filter(severity=0).count() low_count = vulnerabilities.filter(severity=1).count() medium_count = vulnerabilities.filter(severity=2).count() high_count = vulnerabilities.filter(severity=3).count() critical_count = vulnerabilities.filter(severity=4).count() unknown_count = vulnerabilities.filter(severity=-1).count() vulnerability_feed = Vulnerability.objects.all().order_by( '-discovered_date')[:20] activity_feed = ScanActivity.objects.all().order_by('-time')[:20] total_vul_count = info_count + low_count + \ medium_count + high_count + critical_count + unknown_count total_vul_ignore_info_count = low_count + \ medium_count + high_count + critical_count most_common_vulnerability = Vulnerability.objects.values("name", "severity").annotate(count=Count('name')).order_by("-count")[:10] last_week = timezone.now() - timedelta(days=7) count_targets_by_date = Domain.objects.filter( insert_date__gte=last_week).annotate( date=TruncDay('insert_date')).values("date").annotate( created_count=Count('id')).order_by("-date") count_subdomains_by_date = Subdomain.objects.filter( discovered_date__gte=last_week).annotate( date=TruncDay('discovered_date')).values("date").annotate( count=Count('id')).order_by("-date") count_vulns_by_date = Vulnerability.objects.filter( discovered_date__gte=last_week).annotate( date=TruncDay('discovered_date')).values("date").annotate( count=Count('id')).order_by("-date") count_scans_by_date = ScanHistory.objects.filter( start_scan_date__gte=last_week).annotate( date=TruncDay('start_scan_date')).values("date").annotate( count=Count('id')).order_by("-date") count_endpoints_by_date = EndPoint.objects.filter( discovered_date__gte=last_week).annotate( date=TruncDay('discovered_date')).values("date").annotate( count=Count('id')).order_by("-date") last_7_dates = [(timezone.now() - timedelta(days=i)).date() for i in range(0, 7)] targets_in_last_week = [] subdomains_in_last_week = [] vulns_in_last_week = [] scans_in_last_week = [] endpoints_in_last_week = [] for date in last_7_dates: _target = count_targets_by_date.filter(date=date) _subdomain = count_subdomains_by_date.filter(date=date) _vuln = count_vulns_by_date.filter(date=date) _scan = count_scans_by_date.filter(date=date) _endpoint = count_endpoints_by_date.filter(date=date) if _target: targets_in_last_week.append(_target[0]['created_count']) else: targets_in_last_week.append(0) if _subdomain: subdomains_in_last_week.append(_subdomain[0]['count']) else: subdomains_in_last_week.append(0) if _vuln: vulns_in_last_week.append(_vuln[0]['count']) else: vulns_in_last_week.append(0) if _scan: scans_in_last_week.append(_scan[0]['count']) else: scans_in_last_week.append(0) if _endpoint: endpoints_in_last_week.append(_endpoint[0]['count']) else: endpoints_in_last_week.append(0) targets_in_last_week.reverse() subdomains_in_last_week.reverse() vulns_in_last_week.reverse() scans_in_last_week.reverse() endpoints_in_last_week.reverse() context = { 'dashboard_data_active': 'active', 'domain_count': domain_count, 'endpoint_count': endpoint_count, 'scan_count': scan_count, 'subdomain_count': subdomain_count, 'subdomain_with_ip_count': subdomain_with_ip_count, 'alive_count': alive_count, 'endpoint_alive_count': endpoint_alive_count, 'info_count': info_count, 'low_count': low_count, 'medium_count': medium_count, 'high_count': high_count, 'critical_count': critical_count, 'unknown_count': unknown_count, 'most_common_vulnerability': most_common_vulnerability, 'total_vul_count': total_vul_count, 'total_vul_ignore_info_count': total_vul_ignore_info_count, 'vulnerability_feed': vulnerability_feed, 'activity_feed': activity_feed, 'targets_in_last_week': targets_in_last_week, 'subdomains_in_last_week': subdomains_in_last_week, 'vulns_in_last_week': vulns_in_last_week, 'scans_in_last_week': scans_in_last_week, 'endpoints_in_last_week': endpoints_in_last_week, 'last_7_dates': last_7_dates, } context['total_ips'] = IpAddress.objects.all().count() context['most_used_port'] = Port.objects.annotate(count=Count('ports')).order_by('-count')[:7] context['most_used_ip'] = IpAddress.objects.annotate(count=Count('ip_addresses')).order_by('-count').exclude(ip_addresses__isnull=True)[:7] context['most_used_tech'] = Technology.objects.annotate(count=Count('technologies')).order_by('-count')[:7] context['most_common_cve'] = CveId.objects.annotate(nused=Count('cve_ids')).order_by('-nused').values('name', 'nused')[:7] context['most_common_cwe'] = CweId.objects.annotate(nused=Count('cwe_ids')).order_by('-nused').values('name', 'nused')[:7] context['most_common_tags'] = VulnerabilityTags.objects.annotate(nused=Count('vuln_tags')).order_by('-nused').values('name', 'nused')[:7] context['asset_countries'] = CountryISO.objects.annotate(count=Count('ipaddress')).order_by('-count') return render(request, 'dashboard/index.html', context) def profile(request): if request.method == 'POST': form = PasswordChangeForm(request.user, request.POST) if form.is_valid(): user = form.save() update_session_auth_hash(request, user) messages.success( request, 'Your password was successfully changed!') return redirect('profile') else: messages.error(request, 'Please correct the error below.') else: form = PasswordChangeForm(request.user) return render(request, 'dashboard/profile.html', { 'form': form }) @receiver(user_logged_out) def on_user_logged_out(sender, request, **kwargs): messages.add_message( request, messages.INFO, 'You have been successfully logged out. Thank you ' + 'for using reNgine.') @receiver(user_logged_in) def on_user_logged_in(sender, request, **kwargs): messages.add_message( request, messages.INFO, 'Hi @' + request.user.username + ' welcome back!') def search(request): return render(request, 'dashboard/search.html')

import json import logging from datetime import timedelta from django.contrib.auth import get_user_model from django.contrib import messages from django.contrib.auth import update_session_auth_hash from django.contrib.auth.forms import PasswordChangeForm from django.contrib.auth.signals import user_logged_in, user_logged_out from django.contrib import messages from django.db.models import Count from django.db.models.functions import TruncDay from django.dispatch import receiver from django.shortcuts import redirect, render, get_object_or_404 from django.utils import timezone from django.http import HttpResponseRedirect, JsonResponse from django.urls import reverse from rolepermissions.roles import assign_role, clear_roles from rolepermissions.decorators import has_permission_decorator from django.template.defaultfilters import slugify from startScan.models import * from targetApp.models import Domain from dashboard.models import * from reNgine.definitions import * logger = logging.getLogger(__name__) def index(request, slug): try: project = Project.objects.get(slug=slug) except Exception as e: # if project not found redirect to 404 return HttpResponseRedirect(reverse('four_oh_four')) domains = Domain.objects.filter(project=project) subdomains = Subdomain.objects.filter(target_domain__project=project) endpoints = EndPoint.objects.filter(target_domain__project=project) scan_histories = ScanHistory.objects.filter(domain__project=project) vulnerabilities = Vulnerability.objects.filter(target_domain__project=project) scan_activities = ScanActivity.objects.filter(scan_of__in=scan_histories) domain_count = domains.count() endpoint_count = endpoints.count() scan_count = scan_histories.count() subdomain_count = subdomains.count() subdomain_with_ip_count = subdomains.filter(ip_addresses__isnull=False).count() alive_count = subdomains.exclude(http_status__exact=0).count() endpoint_alive_count = endpoints.filter(http_status__exact=200).count() info_count = vulnerabilities.filter(severity=0).count() low_count = vulnerabilities.filter(severity=1).count() medium_count = vulnerabilities.filter(severity=2).count() high_count = vulnerabilities.filter(severity=3).count() critical_count = vulnerabilities.filter(severity=4).count() unknown_count = vulnerabilities.filter(severity=-1).count() vulnerability_feed = vulnerabilities.order_by('-discovered_date')[:50] activity_feed = scan_activities.order_by('-time')[:50] total_vul_count = info_count + low_count + \ medium_count + high_count + critical_count + unknown_count total_vul_ignore_info_count = low_count + \ medium_count + high_count + critical_count last_week = timezone.now() - timedelta(days=7) count_targets_by_date = domains.filter( insert_date__gte=last_week).annotate( date=TruncDay('insert_date')).values("date").annotate( created_count=Count('id')).order_by("-date") count_subdomains_by_date = subdomains.filter( discovered_date__gte=last_week).annotate( date=TruncDay('discovered_date')).values("date").annotate( count=Count('id')).order_by("-date") count_vulns_by_date = vulnerabilities.filter( discovered_date__gte=last_week).annotate( date=TruncDay('discovered_date')).values("date").annotate( count=Count('id')).order_by("-date") count_scans_by_date = scan_histories.filter( start_scan_date__gte=last_week).annotate( date=TruncDay('start_scan_date')).values("date").annotate( count=Count('id')).order_by("-date") count_endpoints_by_date = endpoints.filter( discovered_date__gte=last_week).annotate( date=TruncDay('discovered_date')).values("date").annotate( count=Count('id')).order_by("-date") last_7_dates = [(timezone.now() - timedelta(days=i)).date() for i in range(0, 7)] targets_in_last_week = [] subdomains_in_last_week = [] vulns_in_last_week = [] scans_in_last_week = [] endpoints_in_last_week = [] for date in last_7_dates: _target = count_targets_by_date.filter(date=date) _subdomain = count_subdomains_by_date.filter(date=date) _vuln = count_vulns_by_date.filter(date=date) _scan = count_scans_by_date.filter(date=date) _endpoint = count_endpoints_by_date.filter(date=date) if _target: targets_in_last_week.append(_target[0]['created_count']) else: targets_in_last_week.append(0) if _subdomain: subdomains_in_last_week.append(_subdomain[0]['count']) else: subdomains_in_last_week.append(0) if _vuln: vulns_in_last_week.append(_vuln[0]['count']) else: vulns_in_last_week.append(0) if _scan: scans_in_last_week.append(_scan[0]['count']) else: scans_in_last_week.append(0) if _endpoint: endpoints_in_last_week.append(_endpoint[0]['count']) else: endpoints_in_last_week.append(0) targets_in_last_week.reverse() subdomains_in_last_week.reverse() vulns_in_last_week.reverse() scans_in_last_week.reverse() endpoints_in_last_week.reverse() context = { 'dashboard_data_active': 'active', 'domain_count': domain_count, 'endpoint_count': endpoint_count, 'scan_count': scan_count, 'subdomain_count': subdomain_count, 'subdomain_with_ip_count': subdomain_with_ip_count, 'alive_count': alive_count, 'endpoint_alive_count': endpoint_alive_count, 'info_count': info_count, 'low_count': low_count, 'medium_count': medium_count, 'high_count': high_count, 'critical_count': critical_count, 'unknown_count': unknown_count, 'total_vul_count': total_vul_count, 'total_vul_ignore_info_count': total_vul_ignore_info_count, 'vulnerability_feed': vulnerability_feed, 'activity_feed': activity_feed, 'targets_in_last_week': targets_in_last_week, 'subdomains_in_last_week': subdomains_in_last_week, 'vulns_in_last_week': vulns_in_last_week, 'scans_in_last_week': scans_in_last_week, 'endpoints_in_last_week': endpoints_in_last_week, 'last_7_dates': last_7_dates, 'project': project } ip_addresses = IpAddress.objects.filter(ip_addresses__in=subdomains) context['total_ips'] = ip_addresses.count() context['most_used_port'] = Port.objects.filter(ports__in=ip_addresses).annotate(count=Count('ports')).order_by('-count')[:7] context['most_used_ip'] = ip_addresses.annotate(count=Count('ip_addresses')).order_by('-count').exclude(ip_addresses__isnull=True)[:7] context['most_used_tech'] = Technology.objects.filter(technologies__in=subdomains).annotate(count=Count('technologies')).order_by('-count')[:7] context['most_common_cve'] = CveId.objects.filter(cve_ids__in=vulnerabilities).annotate(nused=Count('cve_ids')).order_by('-nused').values('name', 'nused')[:7] context['most_common_cwe'] = CweId.objects.filter(cwe_ids__in=vulnerabilities).annotate(nused=Count('cwe_ids')).order_by('-nused').values('name', 'nused')[:7] context['most_common_tags'] = VulnerabilityTags.objects.filter(vuln_tags__in=vulnerabilities).annotate(nused=Count('vuln_tags')).order_by('-nused').values('name', 'nused')[:7] context['asset_countries'] = CountryISO.objects.filter(ipaddress__in=ip_addresses).annotate(count=Count('ipaddress')).order_by('-count') return render(request, 'dashboard/index.html', context) def profile(request, slug): if request.method == 'POST': form = PasswordChangeForm(request.user, request.POST) if form.is_valid(): user = form.save() update_session_auth_hash(request, user) messages.success( request, 'Your password was successfully changed!') return redirect('profile') else: messages.error(request, 'Please correct the error below.') else: form = PasswordChangeForm(request.user) return render(request, 'dashboard/profile.html', { 'form': form }) @has_permission_decorator(PERM_MODIFY_SYSTEM_CONFIGURATIONS, redirect_url=FOUR_OH_FOUR_URL) def admin_interface(request, slug): UserModel = get_user_model() users = UserModel.objects.all().order_by('date_joined') return render( request, 'dashboard/admin.html', { 'users': users } ) @has_permission_decorator(PERM_MODIFY_SYSTEM_CONFIGURATIONS, redirect_url=FOUR_OH_FOUR_URL) def admin_interface_update(request, slug): mode = request.GET.get('mode') user_id = request.GET.get('user') if user_id: UserModel = get_user_model() user = UserModel.objects.get(id=user_id) if request.method == 'GET': if mode == 'change_status': user.is_active = not user.is_active user.save() elif request.method == 'POST': if mode == 'delete': try: user.delete() messages.add_message( request, messages.INFO, f'User {user.username} successfully deleted.' ) messageData = {'status': True} except Exception as e: logger.error(e) messageData = {'status': False} elif mode == 'update': try: response = json.loads(request.body) role = response.get('role') change_password = response.get('change_password') clear_roles(user) assign_role(user, role) if change_password: user.set_password(change_password) user.save() messageData = {'status': True} except Exception as e: logger.error(e) messageData = {'status': False, 'error': str(e)} elif mode == 'create': try: response = json.loads(request.body) UserModel = get_user_model() user = UserModel.objects.create_user( username=response.get('username'), password=response.get('password') ) assign_role(user, response.get('role')) messageData = {'status': True} except Exception as e: logger.error(e) messageData = {'status': False, 'error': str(e)} return JsonResponse(messageData) return HttpResponseRedirect(reverse('admin_interface', kwargs={'slug': slug})) @receiver(user_logged_out) def on_user_logged_out(sender, request, **kwargs): messages.add_message( request, messages.INFO, 'You have been successfully logged out. Thank you ' + 'for using reNgine.') @receiver(user_logged_in) def on_user_logged_in(sender, request, **kwargs): messages.add_message( request, messages.INFO, 'Hi @' + request.user.username + ' welcome back!') def search(request, slug): return render(request, 'dashboard/search.html') def four_oh_four(request): return render(request, '404.html') def projects(request, slug): context = {} context['projects'] = Project.objects.all() return render(request, 'dashboard/projects.html', context) def delete_project(request, id): obj = get_object_or_404(Project, id=id) if request.method == "POST": obj.delete() responseData = { 'status': 'true' } messages.add_message( request, messages.INFO, 'Project successfully deleted!') else: responseData = {'status': 'false'} messages.add_message( request, messages.ERROR, 'Oops! Project could not be deleted!') return JsonResponse(responseData) def onboarding(request): context = {} error = '' if request.method == "POST": project_name = request.POST.get('project_name') slug = slugify(project_name) create_username = request.POST.get('create_username') create_password = request.POST.get('create_password') create_user_role = request.POST.get('create_user_role') key_openai = request.POST.get('key_openai') key_netlas = request.POST.get('key_netlas') insert_date = timezone.now() try: Project.objects.create( name=project_name, slug=slug, insert_date=insert_date ) except Exception as e: error = ' Could not create project, Error: ' + str(e) try: if create_username and create_password and create_user_role: UserModel = get_user_model() user = UserModel.objects.create_user( username=create_username, password=create_password ) assign_role(user, create_user_role) except Exception as e: error = ' Could not create User, Error: ' + str(e) if key_openai: openai_api_key = OpenAiAPIKey.objects.first() if openai_api_key: openai_api_key.key = key_openai openai_api_key.save() else: OpenAiAPIKey.objects.create(key=key_openai) if key_netlas: netlas_api_key = NetlasAPIKey.objects.first() if netlas_api_key: netlas_api_key.key = key_netlas netlas_api_key.save() else: NetlasAPIKey.objects.create(key=key_netlas) context['error'] = error # check is any projects exists, then redirect to project list else onboarding projects = Project.objects.all() context['openai_key'] = OpenAiAPIKey.objects.first() context['netlas_key'] = NetlasAPIKey.objects.first() if len(projects): slug = projects[0].slug return HttpResponseRedirect(reverse('dashboardIndex', kwargs={'slug': slug})) return render(request, 'dashboard/onboarding.html', context)

yogeshojha

3c60bc1ee495044794d91edee0c96fff73ab46c7

5413708d243799a5271440c47c6f98d0c51154ca

## Information exposure through an exception [Stack trace information](1) flows to this location and may be exposed to an external user. [Stack trace information](2) flows to this location and may be exposed to an external user. [Show more details](https://github.com/yogeshojha/rengine/security/code-scanning/148)

github-advanced-security[bot]

yogeshojha/rengine

2.0-jasper release

### Added - Projects: Projects allow you to efficiently organize their web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. - Roles and Permissions: assign distinct roles to your team members: Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - GPT-powered Report Generation: With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments. - API Vault: This feature allows you to organize your API keys such as OpenAI or Netlas API keys. - GPT-powered Attack Surface Generation - URL gathering now is much more efficient, removing duplicate endpoints based on similar HTTP Responses, having the same content_lenth, or page_title. Custom duplicate fields can also be set from the scan engine configuration. - URL Path filtering while initiating scan: For instance, if we want to scan only endpoints starting with https://example.com/start/, we can pass the /start as a path filter while starting the scan. @ocervell - Expanding Target Concept: reNgine 2.0 now accepts IPs, URLS, etc as targets. (#678, #658) Excellent work by @ocervell - A ton of refactoring on reNgine's core to improve scan efficiency. Massive kudos to @ocervell - Created a custom celery workflow to be able to run several tasks in parallel that are not dependent on each other, such OSINT task and subdomain discovery will run in parallel, and directory and file fuzzing, vulnerability scan, screenshot gathering etc. will run in parallel after port scan or url fetching is completed. This will increase the efficiency of scans and instead of having one long flow of tasks, they can run independently on their own. @ocervell - Refactored all tasks to run asynchronously @ocervell - Added a stream_command that allows to read the output of a command live: this means the UI is updated with results while the command runs and does not have to wait until the task completes. Excellent work by @ocervell - Pwndb is now replaced by h8mail. @ocervell - Group Scan Results: reNgine 2.0 allows to group of subdomains based on similar page titles and HTTP status, and also vulnerability grouping based on the same vulnerability title and severity. - Added Support for Nmap: reNgine 2.0 allows to run Nmap scripts and vuln scans on ports found by Naabu. @ocervell - Added support for Shared Scan Variables in Scan Engine Configuration: - `enable_http_crawl`: (true/false) You can disable it to be more stealthy or focus on something different than HTTP - `timeout`: set timeout for all tasks - `rate_limit`: set rate limit for all tasks - `retries`: set retries for all tasks - `custom_header`: set the custom header for all tasks - Added Dalfox for XSS Vulnerability Scan - Added CRLFuzz for CRLF Vulnerability Scan - Added S3Scanner for scanning misconfigured S3 buckets - Improve OSINT Dork results, now detects admin panels, login pages and dashboards - Added Custom Dorks - Improved UI for vulnerability results, clicking on each vulnerability will open up a sidebar with vulnerability details. - Added HTTP Request and Response in vulnerability Results - Under Admin Settings, added an option to allow add/remove/deactivate additional users - Added Option to Preview Scan Report instead of forcing to download - Added Katana for crawling and spidering URLs - Added Netlas for Whois and subdomain gathering - Added TLSX for subdomain gathering - Added CTFR for subdomain gathering - Added historical IP in whois section ### Fixes - GF patterns do not run on 404 endpoints (#574 closed) - Fixes for retrieving whois data (#693 closed) - Related/Associated Domains in Whois section is now fixed ### Removed - Removed pwndb and tor related to it. - Removed tor for pwndb

2023-10-02 07:51:35+00:00

2023-10-07 10:37:23+00:00

web/startScan/templates/startScan/history.html

{% extends 'base/base.html' %} {% load static %} {% load humanize %} {% block title %} Scan history {% endblock title %} {% block custom_js_css_link %} <link rel="stylesheet" type="text/css" href="{% static 'plugins/datatable/datatables.css' %}"> <link rel="stylesheet" type="text/css" href="{% static 'plugins/datatable/global.css' %}"> <link rel="stylesheet" type="text/css" href="{% static 'plugins/datatable/custom.css' %}"> {% endblock custom_js_css_link %} {% block breadcrumb_title %} <li class="breadcrumb-item active" aria-current="page">Scan History</li> {% endblock breadcrumb_title %} {% block page_title %} Quick Scan History {% endblock page_title %} {% block main_content %} <div class="row"> <div class="col-12"> <div class="card"> <div class="p-2"> <div class="row"> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <button type="button" class="btn btn-primary dropdown-toggle" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false" id="filterMenu"> Filter <i class="fe-filter"></i> </button> <div id="filteringText" class="mt-2"> </div> <div class="dropdown-menu" style="width: 30%"> <div class="px-4 py-3"> <h4 class="headline-title">Filters</h4> <div class=""> <label for="filterByOrganization" class="form-label">Filter by Organization</label> <select class="form-control" id="filterByOrganization"> </select> </div> <div class=""> <label for="filterByTarget" class="form-label">Filter by Targets</label> <select class="form-control" id="filterByTarget"> </select> </div> <div class=""> <label for="filterByScanType" class="form-label">Filter by Scan Type</label> <select class="form-control" id="filterByScanType"> </select> </div> <div class=""> <label for="filterByScanStatus" class="form-label">Filter by Scan Status</label> <select class="form-control" id="filterByScanStatus"> </select> </div> </div> <div class="dropdown-divider"></div> <a href="#" class="dropdown-ite text-primary float-end" id="resetFilters">Reset Filters</a> </div> </div> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <a class="btn btn-soft-danger float-end disabled ms-1" href="#" onclick="deleteMultipleScan()" id="delete_multiple_button">Delete Multiple Scans</a> </div> </div> </div> </div> </div> </div> <div class="row"> <div class="col-12"> <div class="card"> <form method="POST" id="scan_history_form"> {% csrf_token %} <table id="scan_history_table" class="table style-3 table-hover"> <thead> <tr> <th class="checkbox-column text-center">Serial Number</th> <th class="text-center">Serial Number</th> <th class="">Domain Name</th> <th>Summary</th> <th class="">Scan Engine Used</th> <th>Last Scan</th> <th class="text-center">Status</th> <th class="text-center">Progress</th> <th class="text-center no-sorting">Action</th> </tr> </thead> <tbody> {% for scan_history in scan_history.all %} <tr> <td class="checkbox-column"> {{ scan_history.id }} </td> <td class=""> {{ scan_history.id }} </td> <td class=""> {{ scan_history.domain.name }} <br> {% for organization in scan_history.domain.get_organization %} <span class="badge badge-soft-dark mt-1 me-1" data-toggle="tooltip" data-placement="top" title="Domain {{domain.name}} belongs to organization {{organization.name}}">{{ organization.name }}</span> {% endfor %} </td> <td class="text-left"> <span class="badge badge-pills bg-info mt-1" data-toggle="tooltip" data-placement="top" title="Subdomains">{{scan_history.get_subdomain_count}}</span> <span class="badge badge-pills bg-warning mt-1" data-toggle="tooltip" data-placement="top" title="Endpoints">{{scan_history.get_endpoint_count}}</span> <span class="badge badge-pills bg-danger mt-1" data-toggle="tooltip" data-placement="top" title="{{scan_history.get_critical_vulnerability_count}} Critical, {{scan_history.get_high_vulnerability_count}} High, {{scan_history.get_medium_vulnerability_count}} Medium Vulnerabilities">{{scan_history.get_vulnerability_count}}</span> </td> <td class=""> <span class="badge badge-soft-primary">{{ scan_history.scan_type }}</span> </td> <td> <span data-toggle="tooltip" data-placement="top" title="{{scan_history.start_scan_date}}">{{scan_history.start_scan_date|naturaltime}}</span> </td> <td class="text-center"> {% if scan_history.scan_status == -1 %} <span class="badge badge-soft-warning" data-placement="top" data-toggle="tooltip" data-placement="top" title="Waiting for other scans to complete"><span class="spinner-border spinner-border-sm"></span> Pending</span> {% elif scan_history.scan_status == 0 %} <span class="badge badge-soft-danger">Failed</span> {% if scan_history.error_message %}</br><p class="text-danger">Scan Failed due to: {{scan_history.error_message}}</p>{% endif %} {% elif scan_history.scan_status == 1 %} <span class="badge badge-soft-info"><span class="spinner-border spinner-border-sm"></span> In Progress</span> {% elif scan_history.scan_status == 2 %} <span class="badge badge-soft-success">Successful</span> {% elif scan_history.scan_status == 3 %} <span class="badge badge-soft-danger">Aborted</span> {% else %} <span class="badge badge-soft-danger">Unknown</span> {% endif %} </td> <td class="text-center"> {% if scan_history.scan_status == -1 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-warning" role="progressbar" style="width: 75%" aria-valuenow="75" aria-valuemin="0" aria-valuemax="100"></div> </div> {% elif scan_history.scan_status == 0 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% elif scan_history.scan_status == 1 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-primary progress-bar-striped progress-bar-animated" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% elif scan_history.scan_status == 2 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-success" role="progressbar" style="width: 100%" aria-valuemin="0" aria-valuemax="100"></div> </div> {% elif scan_history.scan_status == 3 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger progress-bar-striped" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% else %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger" role="progressbar" style="width: 100%" aria-valuemin="0" aria-valuemax="100"> </div> </div> {% endif %} </td> <td class="text-center"> <div class="btn-group mb-2 dropstart"> <div class="btn-group"> <a href="{% url 'detail_scan' scan_history.id %}" class="btn btn-soft-primary">View Results</a> <div class="btn-group dropstart" role="group"> <button type="button" class="btn btn-soft-primary dropdown-toggle dropdown-toggle-split" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false"> <i class="mdi mdi-chevron-right"></i> </button> <div class="dropdown-menu" style=""> {% if scan_history.scan_status == 0 or scan_history.scan_status == 2 or scan_history.scan_status == 3 %} <a class="dropdown-item text-primary" href="{% url 'start_scan' scan_history.domain.id %}"> <i class="fe-refresh-ccw"></i>&nbsp;Rescan </a> {% endif %} {% if scan_history.scan_status == 1 or scan_history.scan_status == -1%} <a href="#" class="dropdown-item text-danger" onclick="stop_scan(scan_id={{ scan_history.id }}, subscan_id=null, reload_scan_bar=false, reload_location=true)"> <i class="fe-alert-triangle"></i>&nbsp;Stop Scan</a> {% endif %} {% if scan_history.scan_status == 2 or scan_history.scan_status == 3 %} <a href="#" class="dropdown-item text-danger" onclick="delete_scan('{{ scan_history.id }}')"> <i class="fe-trash-2"></i>&nbsp;Delete Scan Results</a> {% endif %} {% if scan.scan_status != -1%} <div class="dropdown-divider"></div> <a href="#" class="dropdown-item text-dark" onclick="initiate_report({{scan_history.id}}, '{{scan_history.subdomain_discovery}}', '{{scan_history.vulnerability_scan}}', '{{ scan_history.domain.name }}')"> <i class="fe-download"></i>&nbsp;Download Report</a> {% endif %} </div> </div> </div> </div> </td> </tr> {% endfor %} </tbody> </table> </form> </div> </div> </div> <div class="modal fade" id="generateReportModal" tabindex="-1" style="display: none;" aria-hidden="true"> <div class="modal-dialog modal-dialog-centered"> <div class="modal-content"> <div class="modal-header"> <h4 class="modal-title" id="myCenterModalLabel">Download Report</h4> <button type="button" class="btn-close" data-bs-dismiss="modal" aria-label="Close"></button> </div> <div class="modal-body"> <div class="alert alert-light-primary border-0 mb-4" role="alert"> <div id='report_alert_message'></div> </div> <div class="form-group mb-4"> <label for="reportTypeForm">Report Type</label> <select class="form-control" id="report_type_select" name="report_type"> </select> </div> <a id='generateReportButton' href="#" class="btn btn-primary float-end">Download Report</a> </div> </div> </div> </div> {% endblock main_content %} {% block page_level_script %} <script src="{% static 'plugins/datatable/datatables.js' %}"></script> <script> $(document).ready(function() { var table = $('#scan_history_table').DataTable({ headerCallback: function(e, a, t, n, s) { e.getElementsByTagName("th")[0].innerHTML='<div class="form-check mb-2 form-check-primary"><input type="checkbox" class="float-start form-check-input chk-parent" id="head_checkbox" onclick=mainCheckBoxSelected(this)>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>\n' }, "columnDefs":[ { 'visible': false, 'targets': [1] }, { "targets":0, "width":"20px", "className":"", "orderable":!1, render:function(e, a, t, n) { return'<div class="form-check mb-2 form-check-primary"><input type="checkbox" name="targets_checkbox['+ e + ']" class="float-start form-check-input targets_checkbox" value="' + e + '" onchange=toggleMultipleTargetButton()>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>' }, }], "order": [[1, 'desc']], "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center mt-sm-0 mt-3'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "stripeClasses": [], "lengthMenu": [5, 10, 20, 30, 40, 50], "pageLength": 20, "initComplete": function(settings, json) { $('[data-toggle="tooltip"]').tooltip(); table = settings.oInstance.api(); var rows = table.rows({ selected: true }).indexes(); // populate filter menu from datatables // populate targets var selectedData = table.cells(rows, 2).data(); var target_array = []; for (var i = 0; i < selectedData.length; i++) { col1_data = selectedData[i]; domain_name = col1_data.match(/([^\n]+)/g)[0]; target_array.push(domain_name); } target_array = Array.from(new Set(target_array)); for (target in target_array) { select = document.getElementById('filterByTarget'); var option = document.createElement('option'); option.value = target_array[target]; option.innerHTML = target_array[target]; select.appendChild(option); } // populate Scan Type var selectedData = table.cells(rows, 4).data(); var scan_type_array = []; for (var i = 0; i < selectedData.length; i++) { col1_data = extractContent(selectedData[i]); scan_type_array.push(col1_data); } scan_type_array = Array.from(new Set(scan_type_array)); for (engine in scan_type_array) { select = document.getElementById('filterByScanType'); var option = document.createElement('option'); option.value = scan_type_array[engine]; option.innerHTML = scan_type_array[engine]; select.appendChild(option); } } }); multiCheck(table); // filter organization populate $.getJSON(`/api/listOrganizations?&format=json`, function(data) { data = data['organizations'] for (organization in data) { name = htmlEncode(data[organization]['name']); select = document.getElementById('filterByOrganization'); var option = document.createElement('option'); option.value = name; option.innerHTML = name; select.appendChild(option); } }).fail(function() {}); // filtering for scan status var status_types = ['Pending', 'Scanning', 'Aborted', 'Successful', 'Failed']; for (status in status_types) { select = document.getElementById('filterByScanStatus'); var option = document.createElement('option'); option.value = status_types[status]; option.innerHTML = status_types[status]; select.appendChild(option); } var org_filter = document.getElementById('filterByOrganization'); org_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Organization: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by organization ${this.value}`, pos: 'top-center' }); }, false); var status_filter = document.getElementById('filterByScanStatus'); status_filter.addEventListener('click', function() { table.search(this.value).draw(); switch (this.value) { case 'Pending': badge_color = 'warning'; break; case 'Scanning': badge_color = 'info'; break; case 'Aborted': badge_color = 'danger'; break; case 'Failed': badge_color = 'danger'; break; case 'Successful': badge_color = 'success'; break; default: badge_color = 'primary' } document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-${badge_color}">Scan Status: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by scan status ${this.value}`, pos: 'top-center' }); }, false); var engine_filter = document.getElementById('filterByScanType'); engine_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Scan Engine: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by Engine ${this.value}`, pos: 'top-center' }); }, false); var target_filter = document.getElementById('filterByTarget'); target_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Target/Domain: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by Engine ${this.value}`, pos: 'top-center' }); }, false); // reset filtering var reset_filter = document.getElementById('resetFilters'); reset_filter.addEventListener('click', function() { resetFilters(table); }, false); }); function resetFilters(table_obj) { table_obj.search("").draw(); Snackbar.show({ text: `Filters Reset`, pos: 'top-center' }); document.getElementById('filteringText').innerHTML = ''; } function checkedCount() { // this function will count the number of boxes checked item = document.getElementsByClassName("targets_checkbox"); count = 0; for (var i = 0; i < item.length; i++) { if (item[i].checked) { count++; } } return count; } function toggleMultipleTargetButton() { if (checkedCount() > 0) { $("#delete_multiple_button").removeClass("disabled"); } else { $("#delete_multiple_button").addClass("disabled"); } } function mainCheckBoxSelected(checkbox) { if (checkbox.checked) { $("#delete_multiple_button").removeClass("disabled"); $(".targets_checkbox").prop('checked', true); } else { $("#delete_multiple_button").addClass("disabled"); $(".targets_checkbox").prop('checked', false); } } function deleteMultipleScan() { if (!checkedCount()) { swal({ title: '', text: "Oops! No targets has been selected!", type: 'error', padding: '2em' }) } else { // atleast one target is selected swal.queue([{ title: 'Are you sure you want to delete ' + checkedCount() + ' Scans?', text: "This action is irreversible.\nThis will delete all the scan data and vulnerabilities related to the scan.", type: 'warning', showCancelButton: true, confirmButtonText: 'Delete', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { deleteForm = document.getElementById("scan_history_form"); deleteForm.action = "../delete/multiple"; deleteForm.submit(); } }]) } } function initiate_report(id, is_subdomain_scan, is_vulnerability_scan, domain_name) { $('#generateReportModal').modal('show'); $('#report_alert_message').empty(); $('#report_type_select').empty(); if (is_subdomain_scan == 'True' && is_vulnerability_scan == 'True') { $('#report_alert_message').append(` <b>Full Scan</b> will include both Reconnaissance and Vulnerability Report.<br> `); $('#report_type_select').append($('<option>', { value: 'full', text: 'Full Scan Report' })); } if (is_subdomain_scan == 'True') { // eligible for reconnaissance report $('#report_alert_message').append(` <b>Reconnaissance Report</b> will only include Assets Discovered Section.<br> `); $('#report_type_select').append($('<option>', { value: 'recon', text: 'Reconnaissance Report' })); } if (is_vulnerability_scan == 'True'){ // eligible for vulnerability report $('#report_alert_message').append(` <b>Vulnerability Report</b> will only include details of Vulnerabilities Identified. `); $('#report_type_select').append($('<option>', { value: 'vulnerability', text: 'Vulnerability Report' })); } $('#generateReportButton').attr('onClick', `generate_report(${id}, '${domain_name}')`); } function generate_report(id, domain_name) { var report_type = $("#report_type_select option:selected").val(); $('#generateReportModal').modal('hide'); swal.queue([{ title: 'Generating Report!', text: `Please wait until we generate a report for you!`, padding: '2em', onOpen: function() { swal.showLoading() return fetch(`/scan/create_report/${id}?download&report_type=${report_type}`, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }) .then(function(response) { return response.blob(); }).then(function(blob) { const file = new Blob([blob], {type: 'application/pdf'}); // process to auto download it const fileURL = URL.createObjectURL(file); const link = document.createElement('a'); link.href = fileURL; link.download = domain_name + ".pdf"; link.click(); swal.close(); }) .catch(function() { swal.insertQueueStep({ type: 'error', title: 'Oops! Unable to generate report!' }) }) } }]); } </script> {% endblock page_level_script %}

{% extends 'base/base.html' %} {% load static %} {% load humanize %} {% load permission_tags %} {% block title %} Scan history {% endblock title %} {% block custom_js_css_link %} {% endblock custom_js_css_link %} {% block breadcrumb_title %} <li class="breadcrumb-item active" aria-current="page">Scan History</li> {% endblock breadcrumb_title %} {% block page_title %} Quick Scan History {% endblock page_title %} {% block main_content %} <div class="row"> <div class="col-12"> <div class="card"> <div class="p-2"> <div class="row"> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <button type="button" class="btn btn-primary dropdown-toggle" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false" id="filterMenu"> Filter <i class="fe-filter"></i> </button> <div id="filteringText" class="mt-2"> </div> <div class="dropdown-menu" style="width: 30%"> <div class="px-4 py-3"> <h4 class="headline-title">Filters</h4> <div class=""> <label for="filterByOrganization" class="form-label">Filter by Organization</label> <select class="form-control" id="filterByOrganization"> </select> </div> <div class=""> <label for="filterByTarget" class="form-label">Filter by Targets</label> <select class="form-control" id="filterByTarget"> </select> </div> <div class=""> <label for="filterByScanType" class="form-label">Filter by Scan Type</label> <select class="form-control" id="filterByScanType"> </select> </div> <div class=""> <label for="filterByScanStatus" class="form-label">Filter by Scan Status</label> <select class="form-control" id="filterByScanStatus"> </select> </div> </div> <div class="dropdown-divider"></div> <a href="#" class="dropdown-ite text-primary float-end" id="resetFilters">Reset Filters</a> </div> </div> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <a class="btn btn-soft-danger float-end disabled ms-1" href="#" onclick="deleteMultipleScan()" id="delete_multiple_button">Delete Multiple Scans</a> </div> </div> </div> </div> </div> </div> <div class="row"> <div class="col-12"> <div class="card"> <form method="POST" id="scan_history_form"> {% csrf_token %} <table id="scan_history_table" class="table dt-responsive w-100"> <thead> <tr> <th class="checkbox-column text-center">Serial Number</th> <th class="text-center">Serial Number</th> <th class="">Domain Name</th> <th>Summary</th> <th class="">Scan Engine Used</th> <th>Last Scan</th> <th class="text-center">Status</th> <th class="text-center">Progress</th> <th class="text-center no-sorting">Action</th> </tr> </thead> <tbody> {% for scan_history in scan_history.all %} <tr> <td class="checkbox-column"> {{ scan_history.id }} </td> <td class=""> {{ scan_history.id }} </td> <td class=""> {{ scan_history.domain.name }} <br> {% for organization in scan_history.domain.get_organization %} <span class="badge badge-soft-dark mt-1 me-1" data-toggle="tooltip" data-placement="top" title="Domain {{domain.name}} belongs to organization {{organization.name}}">{{ organization.name }}</span> {% endfor %} </td> <td class="text-left"> <span class="badge badge-pills bg-info mt-1" data-toggle="tooltip" data-placement="top" title="Subdomains">{{scan_history.get_subdomain_count}}</span> <span class="badge badge-pills bg-warning mt-1" data-toggle="tooltip" data-placement="top" title="Endpoints">{{scan_history.get_endpoint_count}}</span> <span class="badge badge-pills bg-danger mt-1" data-toggle="tooltip" data-placement="top" title="{{scan_history.get_critical_vulnerability_count}} Critical, {{scan_history.get_high_vulnerability_count}} High, {{scan_history.get_medium_vulnerability_count}} Medium Vulnerabilities">{{scan_history.get_vulnerability_count}}</span> </td> <td class=""> <span class="badge badge-soft-primary">{{ scan_history.scan_type }}</span> </td> <td> <span data-toggle="tooltip" data-placement="top" title="{{scan_history.start_scan_date}}">{{scan_history.start_scan_date|naturaltime}}</span> </td> <td class="text-center"> {% if scan_history.scan_status == -1 %} <span class="badge badge-soft-warning" data-placement="top" data-toggle="tooltip" data-placement="top" title="Waiting for other scans to complete"><span class="spinner-border spinner-border-sm"></span> Pending</span> {% elif scan_history.scan_status == 0 %} <span class="badge badge-soft-danger">Failed</span> {% if scan_history.error_message %}</br><p class="text-danger">Scan Failed due to: {{scan_history.error_message}}</p>{% endif %} {% elif scan_history.scan_status == 1 %} <span class="badge badge-soft-info"><span class="spinner-border spinner-border-sm"></span> In Progress</span> {% elif scan_history.scan_status == 2 %} <span class="badge badge-soft-success">Successful</span> {% elif scan_history.scan_status == 3 %} <span class="badge badge-soft-danger">Aborted</span> {% else %} <span class="badge badge-soft-danger">Unknown</span> {% endif %} </td> <td class="text-center"> {% if scan_history.scan_status == -1 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-warning" role="progressbar" style="width: 75%" aria-valuenow="75" aria-valuemin="0" aria-valuemax="100"></div> </div> {% elif scan_history.scan_status == 0 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% elif scan_history.scan_status == 1 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-primary progress-bar-striped progress-bar-animated" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% elif scan_history.scan_status == 2 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-success" role="progressbar" style="width: 100%" aria-valuemin="0" aria-valuemax="100"></div> </div> {% elif scan_history.scan_status == 3 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger progress-bar-striped" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% else %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger" role="progressbar" style="width: 100%" aria-valuemin="0" aria-valuemax="100"> </div> </div> {% endif %} </td> <td class="text-center"> <div class="btn-group mb-2 dropstart"> <div class="btn-group"> <a href="/scan/{{current_project.slug}}/detail/{{scan_history.id}}" class="btn btn-soft-primary">View Results</a> <div class="btn-group dropstart" role="group"> <button type="button" class="btn btn-soft-primary dropdown-toggle dropdown-toggle-split" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false"> <i class="mdi mdi-chevron-right"></i> </button> <div class="dropdown-menu" style=""> {% if user|can:'initiate_scans_subscans' %} {% if scan_history.scan_status == 0 or scan_history.scan_status == 2 or scan_history.scan_status == 3 %} <a class="dropdown-item text-primary" href="/scan/{{current_project.slug}}/start/{{scan_history.domain.id}}"> <i class="fe-refresh-ccw"></i>&nbsp;Rescan </a> {% endif %} {% if scan_history.scan_status == 1 or scan_history.scan_status == -1%} <a href="#" class="dropdown-item text-danger" onclick="stop_scan(scan_id={{ scan_history.id }}, subscan_id=null, reload_scan_bar=false, reload_location=true)"> <i class="fe-alert-triangle"></i>&nbsp;Stop Scan</a> {% endif %} {% endif %} {% if user|can:'modify_scan_results' %} {% if scan_history.scan_status == 2 or scan_history.scan_status == 3 or scan_history.scan_status == 0 %} <a href="#" class="dropdown-item text-danger" onclick="delete_scan('{{ scan_history.id }}')"> <i class="fe-trash-2"></i>&nbsp;Delete Scan Results</a> {% endif %} <div class="dropdown-divider"></div> {% endif %} {% if scan.scan_status != -1%} <a href="#" class="dropdown-item text-dark" onclick="initiate_report({{scan_history.id}}, '{% if 'subdomain_discovery' in scan_history.scan_type.tasks %}True{% endif %}', '{% if 'vulnerability_scan' in scan_history.scan_type.tasks %}True{% endif %}', '{{ scan_history.domain.name }}')"> <i class="fe-download"></i>&nbsp;Scan Report</a> {% endif %} </div> </div> </div> </div> </td> </tr> {% endfor %} </tbody> </table> </form> </div> </div> </div> <div class="modal fade" id="generateReportModal" tabindex="-1" style="display: none;" aria-hidden="true"> <div class="modal-dialog modal-dialog-centered"> <div class="modal-content"> <div class="modal-header"> <h4 class="modal-title" id="myCenterModalLabel">Download Report</h4> <button type="button" class="btn-close" data-bs-dismiss="modal" aria-label="Close"></button> </div> <div class="modal-body"> <div class="alert alert-light-primary border-0 mb-4" role="alert"> <div id='report_alert_message'></div> </div> <div class="form-group mb-4"> <label for="reportTypeForm">Report Type</label> <select class="form-control" id="report_type_select" name="report_type"> </select> </div> <div class="form-group mb-4"> <div class="form-check" id="report_info_vuln_div"> <input type="checkbox" class="form-check-input" id="report_ignore_info_vuln" checked=""> <label class="form-check-label" for="report_ignore_info_vuln">Ignore Informational Vulnerabilities</label> </div> </div> <a id='generateReportButton' href="#" class="btn btn-primary float-end m-2">Download Report</a> <a id='previewReportButton' href="#" class="btn btn-secondary float-end m-2">Preview Report</a> </div> </div> </div> </div> {% endblock main_content %} {% block page_level_script %} <script> $(document).ready(function() { var table = $('#scan_history_table').DataTable({ headerCallback: function(e, a, t, n, s) { e.getElementsByTagName("th")[0].innerHTML='<div class="form-check mb-2 form-check-primary"><input type="checkbox" class="float-start form-check-input chk-parent" id="head_checkbox" onclick=mainCheckBoxSelected(this)>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>\n' }, "columnDefs":[ { 'visible': false, 'targets': [1] }, { "targets":0, "width":"20px", "className":"", "orderable":!1, render:function(e, a, t, n) { return'<div class="form-check mb-2 form-check-primary"><input type="checkbox" name="targets_checkbox['+ e + ']" class="float-start form-check-input targets_checkbox" value="' + e + '" onchange=toggleMultipleTargetButton()>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>' }, } ], "order": [[1, 'desc']], "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center mt-sm-0 mt-3'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "stripeClasses": [], "lengthMenu": [5, 10, 20, 30, 40, 50], "pageLength": 20, "initComplete": function(settings, json) { $('[data-toggle="tooltip"]').tooltip(); table = settings.oInstance.api(); var rows = table.rows({ selected: true }).indexes(); // populate filter menu from datatables // populate targets var selectedData = table.cells(rows, 2).data(); var target_array = []; for (var i = 0; i < selectedData.length; i++) { col1_data = selectedData[i]; domain_name = col1_data.match(/([^\n]+)/g)[0]; target_array.push(domain_name); } target_array = Array.from(new Set(target_array)); for (target in target_array) { select = document.getElementById('filterByTarget'); var option = document.createElement('option'); option.value = target_array[target]; option.innerHTML = target_array[target]; select.appendChild(option); } // populate Scan Type var selectedData = table.cells(rows, 4).data(); var scan_type_array = []; for (var i = 0; i < selectedData.length; i++) { col1_data = extractContent(selectedData[i]); scan_type_array.push(col1_data); } scan_type_array = Array.from(new Set(scan_type_array)); for (engine in scan_type_array) { select = document.getElementById('filterByScanType'); var option = document.createElement('option'); option.value = scan_type_array[engine]; option.innerHTML = scan_type_array[engine]; select.appendChild(option); } } }); multiCheck(table); // filter organization populate $.getJSON(`/api/listOrganizations?&format=json`, function(data) { data = data['organizations'] for (organization in data) { name = htmlEncode(data[organization]['name']); select = document.getElementById('filterByOrganization'); var option = document.createElement('option'); option.value = name; option.innerHTML = name; select.appendChild(option); } }).fail(function() {}); // filtering for scan status var status_types = ['Pending', 'Scanning', 'Aborted', 'Successful', 'Failed']; for (status in status_types) { select = document.getElementById('filterByScanStatus'); var option = document.createElement('option'); option.value = status_types[status]; option.innerHTML = status_types[status]; select.appendChild(option); } var org_filter = document.getElementById('filterByOrganization'); org_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Organization: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by organization ${this.value}`, pos: 'top-center' }); }, false); var status_filter = document.getElementById('filterByScanStatus'); status_filter.addEventListener('click', function() { table.search(this.value).draw(); switch (this.value) { case 'Pending': badge_color = 'warning'; break; case 'Scanning': badge_color = 'info'; break; case 'Aborted': badge_color = 'danger'; break; case 'Failed': badge_color = 'danger'; break; case 'Successful': badge_color = 'success'; break; default: badge_color = 'primary' } document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-${badge_color}">Scan Status: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by scan status ${this.value}`, pos: 'top-center' }); }, false); var engine_filter = document.getElementById('filterByScanType'); engine_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Scan Engine: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by Engine ${this.value}`, pos: 'top-center' }); }, false); var target_filter = document.getElementById('filterByTarget'); target_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Target/Domain: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by Engine ${this.value}`, pos: 'top-center' }); }, false); // reset filtering var reset_filter = document.getElementById('resetFilters'); reset_filter.addEventListener('click', function() { resetFilters(table); }, false); }); function resetFilters(table_obj) { table_obj.search("").draw(); Snackbar.show({ text: `Filters Reset`, pos: 'top-center' }); document.getElementById('filteringText').innerHTML = ''; } function checkedCount() { // this function will count the number of boxes checked item = document.getElementsByClassName("targets_checkbox"); count = 0; for (var i = 0; i < item.length; i++) { if (item[i].checked) { count++; } } return count; } function toggleMultipleTargetButton() { if (checkedCount() > 0) { $("#delete_multiple_button").removeClass("disabled"); } else { $("#delete_multiple_button").addClass("disabled"); } } function mainCheckBoxSelected(checkbox) { if (checkbox.checked) { $("#delete_multiple_button").removeClass("disabled"); $(".targets_checkbox").prop('checked', true); } else { $("#delete_multiple_button").addClass("disabled"); $(".targets_checkbox").prop('checked', false); } } function deleteMultipleScan() { if (!checkedCount()) { swal({ title: '', text: "Oops! No targets has been selected!", type: 'error', padding: '2em' }) } else { // atleast one target is selected swal.queue([{ title: 'Are you sure you want to delete ' + checkedCount() + ' Scans?', text: "This action is irreversible.\nThis will delete all the scan data and vulnerabilities related to the scan.", type: 'warning', showCancelButton: true, confirmButtonText: 'Delete', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { deleteForm = document.getElementById("scan_history_form"); deleteForm.action = "../delete/multiple"; deleteForm.submit(); } }]) } } // select option listener for report_type_select var report_type = document.getElementById("report_type_select"); report_type.addEventListener("change", function() { if(report_type.value == "recon") { $("#report_info_vuln_div").hide(); } else{ $("#report_info_vuln_div").show(); } }); function initiate_report(id, is_subdomain_scan, is_vulnerability_scan, domain_name) { $('#generateReportModal').modal('show'); $('#report_alert_message').empty(); $('#report_type_select').empty(); if (is_subdomain_scan == 'True' && is_vulnerability_scan == 'True') { $('#report_alert_message').append(` <b>Full Scan</b> will include both Reconnaissance and Vulnerability Report.<br> `); $('#report_type_select').append($('<option>', { value: 'full', text: 'Full Scan Report' })); } if (is_subdomain_scan == 'True') { // eligible for reconnaissance report $('#report_alert_message').append(` <b>Reconnaissance Report</b> will only include Assets Discovered Section.<br> `); $('#report_type_select').append($('<option>', { value: 'recon', text: 'Reconnaissance Report' })); } if (is_vulnerability_scan == 'True'){ // eligible for vulnerability report $('#report_alert_message').append(` <b>Vulnerability Report</b> will only include details of Vulnerabilities Identified. `); $('#report_type_select').append($('<option>', { value: 'vulnerability', text: 'Vulnerability Report' })); } $('#generateReportButton').attr('onClick', `generate_report(${id}, '${domain_name}')`); $('#previewReportButton').attr('onClick', `preview_report(${id}, '${domain_name}')`); } function preview_report(id, domain_name){ var report_type = $("#report_type_select option:selected").val(); var url = `/scan/create_report/${id}?report_type=${report_type}`; if ($('#report_ignore_info_vuln').is(":checked")) { url += `&ignore_info_vuln` } $('#generateReportModal').modal('hide'); window.open(url, '_blank').focus(); } function generate_report(id, domain_name) { var report_type = $("#report_type_select option:selected").val(); var url = `/scan/create_report/${id}?report_type=${report_type}&download`; if ($('#report_ignore_info_vuln').is(":checked")) { url += `&ignore_info_vuln` } $('#generateReportModal').modal('hide'); swal.queue([{ title: 'Generating Report!', text: `Please wait until we generate a report for you!`, padding: '2em', onOpen: function() { swal.showLoading() return fetch(url, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }) .then(function(response) { return response.blob(); }).then(function(blob) { const file = new Blob([blob], {type: 'application/pdf'}); // process to auto download it const fileURL = URL.createObjectURL(file); const link = document.createElement('a'); link.href = fileURL; link.download = domain_name + ".pdf"; link.click(); swal.close(); }) .catch(function() { swal.insertQueueStep({ type: 'error', title: 'Oops! Unable to generate report!' }) }) } }]); } </script> {% endblock page_level_script %}

yogeshojha

3c60bc1ee495044794d91edee0c96fff73ab46c7

5413708d243799a5271440c47c6f98d0c51154ca

## DOM text reinterpreted as HTML [DOM text](1) is reinterpreted as HTML without escaping meta-characters. [Show more details](https://github.com/yogeshojha/rengine/security/code-scanning/153)

github-advanced-security[bot]

yogeshojha/rengine

2.0-jasper release

### Added - Projects: Projects allow you to efficiently organize their web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. - Roles and Permissions: assign distinct roles to your team members: Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - GPT-powered Report Generation: With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments. - API Vault: This feature allows you to organize your API keys such as OpenAI or Netlas API keys. - GPT-powered Attack Surface Generation - URL gathering now is much more efficient, removing duplicate endpoints based on similar HTTP Responses, having the same content_lenth, or page_title. Custom duplicate fields can also be set from the scan engine configuration. - URL Path filtering while initiating scan: For instance, if we want to scan only endpoints starting with https://example.com/start/, we can pass the /start as a path filter while starting the scan. @ocervell - Expanding Target Concept: reNgine 2.0 now accepts IPs, URLS, etc as targets. (#678, #658) Excellent work by @ocervell - A ton of refactoring on reNgine's core to improve scan efficiency. Massive kudos to @ocervell - Created a custom celery workflow to be able to run several tasks in parallel that are not dependent on each other, such OSINT task and subdomain discovery will run in parallel, and directory and file fuzzing, vulnerability scan, screenshot gathering etc. will run in parallel after port scan or url fetching is completed. This will increase the efficiency of scans and instead of having one long flow of tasks, they can run independently on their own. @ocervell - Refactored all tasks to run asynchronously @ocervell - Added a stream_command that allows to read the output of a command live: this means the UI is updated with results while the command runs and does not have to wait until the task completes. Excellent work by @ocervell - Pwndb is now replaced by h8mail. @ocervell - Group Scan Results: reNgine 2.0 allows to group of subdomains based on similar page titles and HTTP status, and also vulnerability grouping based on the same vulnerability title and severity. - Added Support for Nmap: reNgine 2.0 allows to run Nmap scripts and vuln scans on ports found by Naabu. @ocervell - Added support for Shared Scan Variables in Scan Engine Configuration: - `enable_http_crawl`: (true/false) You can disable it to be more stealthy or focus on something different than HTTP - `timeout`: set timeout for all tasks - `rate_limit`: set rate limit for all tasks - `retries`: set retries for all tasks - `custom_header`: set the custom header for all tasks - Added Dalfox for XSS Vulnerability Scan - Added CRLFuzz for CRLF Vulnerability Scan - Added S3Scanner for scanning misconfigured S3 buckets - Improve OSINT Dork results, now detects admin panels, login pages and dashboards - Added Custom Dorks - Improved UI for vulnerability results, clicking on each vulnerability will open up a sidebar with vulnerability details. - Added HTTP Request and Response in vulnerability Results - Under Admin Settings, added an option to allow add/remove/deactivate additional users - Added Option to Preview Scan Report instead of forcing to download - Added Katana for crawling and spidering URLs - Added Netlas for Whois and subdomain gathering - Added TLSX for subdomain gathering - Added CTFR for subdomain gathering - Added historical IP in whois section ### Fixes - GF patterns do not run on 404 endpoints (#574 closed) - Fixes for retrieving whois data (#693 closed) - Related/Associated Domains in Whois section is now fixed ### Removed - Removed pwndb and tor related to it. - Removed tor for pwndb

2023-10-02 07:51:35+00:00

2023-10-07 10:37:23+00:00

web/startScan/templates/startScan/history.html

{% extends 'base/base.html' %} {% load static %} {% load humanize %} {% block title %} Scan history {% endblock title %} {% block custom_js_css_link %} <link rel="stylesheet" type="text/css" href="{% static 'plugins/datatable/datatables.css' %}"> <link rel="stylesheet" type="text/css" href="{% static 'plugins/datatable/global.css' %}"> <link rel="stylesheet" type="text/css" href="{% static 'plugins/datatable/custom.css' %}"> {% endblock custom_js_css_link %} {% block breadcrumb_title %} <li class="breadcrumb-item active" aria-current="page">Scan History</li> {% endblock breadcrumb_title %} {% block page_title %} Quick Scan History {% endblock page_title %} {% block main_content %} <div class="row"> <div class="col-12"> <div class="card"> <div class="p-2"> <div class="row"> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <button type="button" class="btn btn-primary dropdown-toggle" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false" id="filterMenu"> Filter <i class="fe-filter"></i> </button> <div id="filteringText" class="mt-2"> </div> <div class="dropdown-menu" style="width: 30%"> <div class="px-4 py-3"> <h4 class="headline-title">Filters</h4> <div class=""> <label for="filterByOrganization" class="form-label">Filter by Organization</label> <select class="form-control" id="filterByOrganization"> </select> </div> <div class=""> <label for="filterByTarget" class="form-label">Filter by Targets</label> <select class="form-control" id="filterByTarget"> </select> </div> <div class=""> <label for="filterByScanType" class="form-label">Filter by Scan Type</label> <select class="form-control" id="filterByScanType"> </select> </div> <div class=""> <label for="filterByScanStatus" class="form-label">Filter by Scan Status</label> <select class="form-control" id="filterByScanStatus"> </select> </div> </div> <div class="dropdown-divider"></div> <a href="#" class="dropdown-ite text-primary float-end" id="resetFilters">Reset Filters</a> </div> </div> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <a class="btn btn-soft-danger float-end disabled ms-1" href="#" onclick="deleteMultipleScan()" id="delete_multiple_button">Delete Multiple Scans</a> </div> </div> </div> </div> </div> </div> <div class="row"> <div class="col-12"> <div class="card"> <form method="POST" id="scan_history_form"> {% csrf_token %} <table id="scan_history_table" class="table style-3 table-hover"> <thead> <tr> <th class="checkbox-column text-center">Serial Number</th> <th class="text-center">Serial Number</th> <th class="">Domain Name</th> <th>Summary</th> <th class="">Scan Engine Used</th> <th>Last Scan</th> <th class="text-center">Status</th> <th class="text-center">Progress</th> <th class="text-center no-sorting">Action</th> </tr> </thead> <tbody> {% for scan_history in scan_history.all %} <tr> <td class="checkbox-column"> {{ scan_history.id }} </td> <td class=""> {{ scan_history.id }} </td> <td class=""> {{ scan_history.domain.name }} <br> {% for organization in scan_history.domain.get_organization %} <span class="badge badge-soft-dark mt-1 me-1" data-toggle="tooltip" data-placement="top" title="Domain {{domain.name}} belongs to organization {{organization.name}}">{{ organization.name }}</span> {% endfor %} </td> <td class="text-left"> <span class="badge badge-pills bg-info mt-1" data-toggle="tooltip" data-placement="top" title="Subdomains">{{scan_history.get_subdomain_count}}</span> <span class="badge badge-pills bg-warning mt-1" data-toggle="tooltip" data-placement="top" title="Endpoints">{{scan_history.get_endpoint_count}}</span> <span class="badge badge-pills bg-danger mt-1" data-toggle="tooltip" data-placement="top" title="{{scan_history.get_critical_vulnerability_count}} Critical, {{scan_history.get_high_vulnerability_count}} High, {{scan_history.get_medium_vulnerability_count}} Medium Vulnerabilities">{{scan_history.get_vulnerability_count}}</span> </td> <td class=""> <span class="badge badge-soft-primary">{{ scan_history.scan_type }}</span> </td> <td> <span data-toggle="tooltip" data-placement="top" title="{{scan_history.start_scan_date}}">{{scan_history.start_scan_date|naturaltime}}</span> </td> <td class="text-center"> {% if scan_history.scan_status == -1 %} <span class="badge badge-soft-warning" data-placement="top" data-toggle="tooltip" data-placement="top" title="Waiting for other scans to complete"><span class="spinner-border spinner-border-sm"></span> Pending</span> {% elif scan_history.scan_status == 0 %} <span class="badge badge-soft-danger">Failed</span> {% if scan_history.error_message %}</br><p class="text-danger">Scan Failed due to: {{scan_history.error_message}}</p>{% endif %} {% elif scan_history.scan_status == 1 %} <span class="badge badge-soft-info"><span class="spinner-border spinner-border-sm"></span> In Progress</span> {% elif scan_history.scan_status == 2 %} <span class="badge badge-soft-success">Successful</span> {% elif scan_history.scan_status == 3 %} <span class="badge badge-soft-danger">Aborted</span> {% else %} <span class="badge badge-soft-danger">Unknown</span> {% endif %} </td> <td class="text-center"> {% if scan_history.scan_status == -1 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-warning" role="progressbar" style="width: 75%" aria-valuenow="75" aria-valuemin="0" aria-valuemax="100"></div> </div> {% elif scan_history.scan_status == 0 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% elif scan_history.scan_status == 1 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-primary progress-bar-striped progress-bar-animated" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% elif scan_history.scan_status == 2 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-success" role="progressbar" style="width: 100%" aria-valuemin="0" aria-valuemax="100"></div> </div> {% elif scan_history.scan_status == 3 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger progress-bar-striped" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% else %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger" role="progressbar" style="width: 100%" aria-valuemin="0" aria-valuemax="100"> </div> </div> {% endif %} </td> <td class="text-center"> <div class="btn-group mb-2 dropstart"> <div class="btn-group"> <a href="{% url 'detail_scan' scan_history.id %}" class="btn btn-soft-primary">View Results</a> <div class="btn-group dropstart" role="group"> <button type="button" class="btn btn-soft-primary dropdown-toggle dropdown-toggle-split" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false"> <i class="mdi mdi-chevron-right"></i> </button> <div class="dropdown-menu" style=""> {% if scan_history.scan_status == 0 or scan_history.scan_status == 2 or scan_history.scan_status == 3 %} <a class="dropdown-item text-primary" href="{% url 'start_scan' scan_history.domain.id %}"> <i class="fe-refresh-ccw"></i>&nbsp;Rescan </a> {% endif %} {% if scan_history.scan_status == 1 or scan_history.scan_status == -1%} <a href="#" class="dropdown-item text-danger" onclick="stop_scan(scan_id={{ scan_history.id }}, subscan_id=null, reload_scan_bar=false, reload_location=true)"> <i class="fe-alert-triangle"></i>&nbsp;Stop Scan</a> {% endif %} {% if scan_history.scan_status == 2 or scan_history.scan_status == 3 %} <a href="#" class="dropdown-item text-danger" onclick="delete_scan('{{ scan_history.id }}')"> <i class="fe-trash-2"></i>&nbsp;Delete Scan Results</a> {% endif %} {% if scan.scan_status != -1%} <div class="dropdown-divider"></div> <a href="#" class="dropdown-item text-dark" onclick="initiate_report({{scan_history.id}}, '{{scan_history.subdomain_discovery}}', '{{scan_history.vulnerability_scan}}', '{{ scan_history.domain.name }}')"> <i class="fe-download"></i>&nbsp;Download Report</a> {% endif %} </div> </div> </div> </div> </td> </tr> {% endfor %} </tbody> </table> </form> </div> </div> </div> <div class="modal fade" id="generateReportModal" tabindex="-1" style="display: none;" aria-hidden="true"> <div class="modal-dialog modal-dialog-centered"> <div class="modal-content"> <div class="modal-header"> <h4 class="modal-title" id="myCenterModalLabel">Download Report</h4> <button type="button" class="btn-close" data-bs-dismiss="modal" aria-label="Close"></button> </div> <div class="modal-body"> <div class="alert alert-light-primary border-0 mb-4" role="alert"> <div id='report_alert_message'></div> </div> <div class="form-group mb-4"> <label for="reportTypeForm">Report Type</label> <select class="form-control" id="report_type_select" name="report_type"> </select> </div> <a id='generateReportButton' href="#" class="btn btn-primary float-end">Download Report</a> </div> </div> </div> </div> {% endblock main_content %} {% block page_level_script %} <script src="{% static 'plugins/datatable/datatables.js' %}"></script> <script> $(document).ready(function() { var table = $('#scan_history_table').DataTable({ headerCallback: function(e, a, t, n, s) { e.getElementsByTagName("th")[0].innerHTML='<div class="form-check mb-2 form-check-primary"><input type="checkbox" class="float-start form-check-input chk-parent" id="head_checkbox" onclick=mainCheckBoxSelected(this)>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>\n' }, "columnDefs":[ { 'visible': false, 'targets': [1] }, { "targets":0, "width":"20px", "className":"", "orderable":!1, render:function(e, a, t, n) { return'<div class="form-check mb-2 form-check-primary"><input type="checkbox" name="targets_checkbox['+ e + ']" class="float-start form-check-input targets_checkbox" value="' + e + '" onchange=toggleMultipleTargetButton()>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>' }, }], "order": [[1, 'desc']], "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center mt-sm-0 mt-3'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "stripeClasses": [], "lengthMenu": [5, 10, 20, 30, 40, 50], "pageLength": 20, "initComplete": function(settings, json) { $('[data-toggle="tooltip"]').tooltip(); table = settings.oInstance.api(); var rows = table.rows({ selected: true }).indexes(); // populate filter menu from datatables // populate targets var selectedData = table.cells(rows, 2).data(); var target_array = []; for (var i = 0; i < selectedData.length; i++) { col1_data = selectedData[i]; domain_name = col1_data.match(/([^\n]+)/g)[0]; target_array.push(domain_name); } target_array = Array.from(new Set(target_array)); for (target in target_array) { select = document.getElementById('filterByTarget'); var option = document.createElement('option'); option.value = target_array[target]; option.innerHTML = target_array[target]; select.appendChild(option); } // populate Scan Type var selectedData = table.cells(rows, 4).data(); var scan_type_array = []; for (var i = 0; i < selectedData.length; i++) { col1_data = extractContent(selectedData[i]); scan_type_array.push(col1_data); } scan_type_array = Array.from(new Set(scan_type_array)); for (engine in scan_type_array) { select = document.getElementById('filterByScanType'); var option = document.createElement('option'); option.value = scan_type_array[engine]; option.innerHTML = scan_type_array[engine]; select.appendChild(option); } } }); multiCheck(table); // filter organization populate $.getJSON(`/api/listOrganizations?&format=json`, function(data) { data = data['organizations'] for (organization in data) { name = htmlEncode(data[organization]['name']); select = document.getElementById('filterByOrganization'); var option = document.createElement('option'); option.value = name; option.innerHTML = name; select.appendChild(option); } }).fail(function() {}); // filtering for scan status var status_types = ['Pending', 'Scanning', 'Aborted', 'Successful', 'Failed']; for (status in status_types) { select = document.getElementById('filterByScanStatus'); var option = document.createElement('option'); option.value = status_types[status]; option.innerHTML = status_types[status]; select.appendChild(option); } var org_filter = document.getElementById('filterByOrganization'); org_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Organization: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by organization ${this.value}`, pos: 'top-center' }); }, false); var status_filter = document.getElementById('filterByScanStatus'); status_filter.addEventListener('click', function() { table.search(this.value).draw(); switch (this.value) { case 'Pending': badge_color = 'warning'; break; case 'Scanning': badge_color = 'info'; break; case 'Aborted': badge_color = 'danger'; break; case 'Failed': badge_color = 'danger'; break; case 'Successful': badge_color = 'success'; break; default: badge_color = 'primary' } document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-${badge_color}">Scan Status: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by scan status ${this.value}`, pos: 'top-center' }); }, false); var engine_filter = document.getElementById('filterByScanType'); engine_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Scan Engine: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by Engine ${this.value}`, pos: 'top-center' }); }, false); var target_filter = document.getElementById('filterByTarget'); target_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Target/Domain: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by Engine ${this.value}`, pos: 'top-center' }); }, false); // reset filtering var reset_filter = document.getElementById('resetFilters'); reset_filter.addEventListener('click', function() { resetFilters(table); }, false); }); function resetFilters(table_obj) { table_obj.search("").draw(); Snackbar.show({ text: `Filters Reset`, pos: 'top-center' }); document.getElementById('filteringText').innerHTML = ''; } function checkedCount() { // this function will count the number of boxes checked item = document.getElementsByClassName("targets_checkbox"); count = 0; for (var i = 0; i < item.length; i++) { if (item[i].checked) { count++; } } return count; } function toggleMultipleTargetButton() { if (checkedCount() > 0) { $("#delete_multiple_button").removeClass("disabled"); } else { $("#delete_multiple_button").addClass("disabled"); } } function mainCheckBoxSelected(checkbox) { if (checkbox.checked) { $("#delete_multiple_button").removeClass("disabled"); $(".targets_checkbox").prop('checked', true); } else { $("#delete_multiple_button").addClass("disabled"); $(".targets_checkbox").prop('checked', false); } } function deleteMultipleScan() { if (!checkedCount()) { swal({ title: '', text: "Oops! No targets has been selected!", type: 'error', padding: '2em' }) } else { // atleast one target is selected swal.queue([{ title: 'Are you sure you want to delete ' + checkedCount() + ' Scans?', text: "This action is irreversible.\nThis will delete all the scan data and vulnerabilities related to the scan.", type: 'warning', showCancelButton: true, confirmButtonText: 'Delete', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { deleteForm = document.getElementById("scan_history_form"); deleteForm.action = "../delete/multiple"; deleteForm.submit(); } }]) } } function initiate_report(id, is_subdomain_scan, is_vulnerability_scan, domain_name) { $('#generateReportModal').modal('show'); $('#report_alert_message').empty(); $('#report_type_select').empty(); if (is_subdomain_scan == 'True' && is_vulnerability_scan == 'True') { $('#report_alert_message').append(` <b>Full Scan</b> will include both Reconnaissance and Vulnerability Report.<br> `); $('#report_type_select').append($('<option>', { value: 'full', text: 'Full Scan Report' })); } if (is_subdomain_scan == 'True') { // eligible for reconnaissance report $('#report_alert_message').append(` <b>Reconnaissance Report</b> will only include Assets Discovered Section.<br> `); $('#report_type_select').append($('<option>', { value: 'recon', text: 'Reconnaissance Report' })); } if (is_vulnerability_scan == 'True'){ // eligible for vulnerability report $('#report_alert_message').append(` <b>Vulnerability Report</b> will only include details of Vulnerabilities Identified. `); $('#report_type_select').append($('<option>', { value: 'vulnerability', text: 'Vulnerability Report' })); } $('#generateReportButton').attr('onClick', `generate_report(${id}, '${domain_name}')`); } function generate_report(id, domain_name) { var report_type = $("#report_type_select option:selected").val(); $('#generateReportModal').modal('hide'); swal.queue([{ title: 'Generating Report!', text: `Please wait until we generate a report for you!`, padding: '2em', onOpen: function() { swal.showLoading() return fetch(`/scan/create_report/${id}?download&report_type=${report_type}`, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }) .then(function(response) { return response.blob(); }).then(function(blob) { const file = new Blob([blob], {type: 'application/pdf'}); // process to auto download it const fileURL = URL.createObjectURL(file); const link = document.createElement('a'); link.href = fileURL; link.download = domain_name + ".pdf"; link.click(); swal.close(); }) .catch(function() { swal.insertQueueStep({ type: 'error', title: 'Oops! Unable to generate report!' }) }) } }]); } </script> {% endblock page_level_script %}

{% extends 'base/base.html' %} {% load static %} {% load humanize %} {% load permission_tags %} {% block title %} Scan history {% endblock title %} {% block custom_js_css_link %} {% endblock custom_js_css_link %} {% block breadcrumb_title %} <li class="breadcrumb-item active" aria-current="page">Scan History</li> {% endblock breadcrumb_title %} {% block page_title %} Quick Scan History {% endblock page_title %} {% block main_content %} <div class="row"> <div class="col-12"> <div class="card"> <div class="p-2"> <div class="row"> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <button type="button" class="btn btn-primary dropdown-toggle" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false" id="filterMenu"> Filter <i class="fe-filter"></i> </button> <div id="filteringText" class="mt-2"> </div> <div class="dropdown-menu" style="width: 30%"> <div class="px-4 py-3"> <h4 class="headline-title">Filters</h4> <div class=""> <label for="filterByOrganization" class="form-label">Filter by Organization</label> <select class="form-control" id="filterByOrganization"> </select> </div> <div class=""> <label for="filterByTarget" class="form-label">Filter by Targets</label> <select class="form-control" id="filterByTarget"> </select> </div> <div class=""> <label for="filterByScanType" class="form-label">Filter by Scan Type</label> <select class="form-control" id="filterByScanType"> </select> </div> <div class=""> <label for="filterByScanStatus" class="form-label">Filter by Scan Status</label> <select class="form-control" id="filterByScanStatus"> </select> </div> </div> <div class="dropdown-divider"></div> <a href="#" class="dropdown-ite text-primary float-end" id="resetFilters">Reset Filters</a> </div> </div> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <a class="btn btn-soft-danger float-end disabled ms-1" href="#" onclick="deleteMultipleScan()" id="delete_multiple_button">Delete Multiple Scans</a> </div> </div> </div> </div> </div> </div> <div class="row"> <div class="col-12"> <div class="card"> <form method="POST" id="scan_history_form"> {% csrf_token %} <table id="scan_history_table" class="table dt-responsive w-100"> <thead> <tr> <th class="checkbox-column text-center">Serial Number</th> <th class="text-center">Serial Number</th> <th class="">Domain Name</th> <th>Summary</th> <th class="">Scan Engine Used</th> <th>Last Scan</th> <th class="text-center">Status</th> <th class="text-center">Progress</th> <th class="text-center no-sorting">Action</th> </tr> </thead> <tbody> {% for scan_history in scan_history.all %} <tr> <td class="checkbox-column"> {{ scan_history.id }} </td> <td class=""> {{ scan_history.id }} </td> <td class=""> {{ scan_history.domain.name }} <br> {% for organization in scan_history.domain.get_organization %} <span class="badge badge-soft-dark mt-1 me-1" data-toggle="tooltip" data-placement="top" title="Domain {{domain.name}} belongs to organization {{organization.name}}">{{ organization.name }}</span> {% endfor %} </td> <td class="text-left"> <span class="badge badge-pills bg-info mt-1" data-toggle="tooltip" data-placement="top" title="Subdomains">{{scan_history.get_subdomain_count}}</span> <span class="badge badge-pills bg-warning mt-1" data-toggle="tooltip" data-placement="top" title="Endpoints">{{scan_history.get_endpoint_count}}</span> <span class="badge badge-pills bg-danger mt-1" data-toggle="tooltip" data-placement="top" title="{{scan_history.get_critical_vulnerability_count}} Critical, {{scan_history.get_high_vulnerability_count}} High, {{scan_history.get_medium_vulnerability_count}} Medium Vulnerabilities">{{scan_history.get_vulnerability_count}}</span> </td> <td class=""> <span class="badge badge-soft-primary">{{ scan_history.scan_type }}</span> </td> <td> <span data-toggle="tooltip" data-placement="top" title="{{scan_history.start_scan_date}}">{{scan_history.start_scan_date|naturaltime}}</span> </td> <td class="text-center"> {% if scan_history.scan_status == -1 %} <span class="badge badge-soft-warning" data-placement="top" data-toggle="tooltip" data-placement="top" title="Waiting for other scans to complete"><span class="spinner-border spinner-border-sm"></span> Pending</span> {% elif scan_history.scan_status == 0 %} <span class="badge badge-soft-danger">Failed</span> {% if scan_history.error_message %}</br><p class="text-danger">Scan Failed due to: {{scan_history.error_message}}</p>{% endif %} {% elif scan_history.scan_status == 1 %} <span class="badge badge-soft-info"><span class="spinner-border spinner-border-sm"></span> In Progress</span> {% elif scan_history.scan_status == 2 %} <span class="badge badge-soft-success">Successful</span> {% elif scan_history.scan_status == 3 %} <span class="badge badge-soft-danger">Aborted</span> {% else %} <span class="badge badge-soft-danger">Unknown</span> {% endif %} </td> <td class="text-center"> {% if scan_history.scan_status == -1 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-warning" role="progressbar" style="width: 75%" aria-valuenow="75" aria-valuemin="0" aria-valuemax="100"></div> </div> {% elif scan_history.scan_status == 0 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% elif scan_history.scan_status == 1 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-primary progress-bar-striped progress-bar-animated" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% elif scan_history.scan_status == 2 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-success" role="progressbar" style="width: 100%" aria-valuemin="0" aria-valuemax="100"></div> </div> {% elif scan_history.scan_status == 3 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger progress-bar-striped" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% else %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger" role="progressbar" style="width: 100%" aria-valuemin="0" aria-valuemax="100"> </div> </div> {% endif %} </td> <td class="text-center"> <div class="btn-group mb-2 dropstart"> <div class="btn-group"> <a href="/scan/{{current_project.slug}}/detail/{{scan_history.id}}" class="btn btn-soft-primary">View Results</a> <div class="btn-group dropstart" role="group"> <button type="button" class="btn btn-soft-primary dropdown-toggle dropdown-toggle-split" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false"> <i class="mdi mdi-chevron-right"></i> </button> <div class="dropdown-menu" style=""> {% if user|can:'initiate_scans_subscans' %} {% if scan_history.scan_status == 0 or scan_history.scan_status == 2 or scan_history.scan_status == 3 %} <a class="dropdown-item text-primary" href="/scan/{{current_project.slug}}/start/{{scan_history.domain.id}}"> <i class="fe-refresh-ccw"></i>&nbsp;Rescan </a> {% endif %} {% if scan_history.scan_status == 1 or scan_history.scan_status == -1%} <a href="#" class="dropdown-item text-danger" onclick="stop_scan(scan_id={{ scan_history.id }}, subscan_id=null, reload_scan_bar=false, reload_location=true)"> <i class="fe-alert-triangle"></i>&nbsp;Stop Scan</a> {% endif %} {% endif %} {% if user|can:'modify_scan_results' %} {% if scan_history.scan_status == 2 or scan_history.scan_status == 3 or scan_history.scan_status == 0 %} <a href="#" class="dropdown-item text-danger" onclick="delete_scan('{{ scan_history.id }}')"> <i class="fe-trash-2"></i>&nbsp;Delete Scan Results</a> {% endif %} <div class="dropdown-divider"></div> {% endif %} {% if scan.scan_status != -1%} <a href="#" class="dropdown-item text-dark" onclick="initiate_report({{scan_history.id}}, '{% if 'subdomain_discovery' in scan_history.scan_type.tasks %}True{% endif %}', '{% if 'vulnerability_scan' in scan_history.scan_type.tasks %}True{% endif %}', '{{ scan_history.domain.name }}')"> <i class="fe-download"></i>&nbsp;Scan Report</a> {% endif %} </div> </div> </div> </div> </td> </tr> {% endfor %} </tbody> </table> </form> </div> </div> </div> <div class="modal fade" id="generateReportModal" tabindex="-1" style="display: none;" aria-hidden="true"> <div class="modal-dialog modal-dialog-centered"> <div class="modal-content"> <div class="modal-header"> <h4 class="modal-title" id="myCenterModalLabel">Download Report</h4> <button type="button" class="btn-close" data-bs-dismiss="modal" aria-label="Close"></button> </div> <div class="modal-body"> <div class="alert alert-light-primary border-0 mb-4" role="alert"> <div id='report_alert_message'></div> </div> <div class="form-group mb-4"> <label for="reportTypeForm">Report Type</label> <select class="form-control" id="report_type_select" name="report_type"> </select> </div> <div class="form-group mb-4"> <div class="form-check" id="report_info_vuln_div"> <input type="checkbox" class="form-check-input" id="report_ignore_info_vuln" checked=""> <label class="form-check-label" for="report_ignore_info_vuln">Ignore Informational Vulnerabilities</label> </div> </div> <a id='generateReportButton' href="#" class="btn btn-primary float-end m-2">Download Report</a> <a id='previewReportButton' href="#" class="btn btn-secondary float-end m-2">Preview Report</a> </div> </div> </div> </div> {% endblock main_content %} {% block page_level_script %} <script> $(document).ready(function() { var table = $('#scan_history_table').DataTable({ headerCallback: function(e, a, t, n, s) { e.getElementsByTagName("th")[0].innerHTML='<div class="form-check mb-2 form-check-primary"><input type="checkbox" class="float-start form-check-input chk-parent" id="head_checkbox" onclick=mainCheckBoxSelected(this)>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>\n' }, "columnDefs":[ { 'visible': false, 'targets': [1] }, { "targets":0, "width":"20px", "className":"", "orderable":!1, render:function(e, a, t, n) { return'<div class="form-check mb-2 form-check-primary"><input type="checkbox" name="targets_checkbox['+ e + ']" class="float-start form-check-input targets_checkbox" value="' + e + '" onchange=toggleMultipleTargetButton()>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>' }, } ], "order": [[1, 'desc']], "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center mt-sm-0 mt-3'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "stripeClasses": [], "lengthMenu": [5, 10, 20, 30, 40, 50], "pageLength": 20, "initComplete": function(settings, json) { $('[data-toggle="tooltip"]').tooltip(); table = settings.oInstance.api(); var rows = table.rows({ selected: true }).indexes(); // populate filter menu from datatables // populate targets var selectedData = table.cells(rows, 2).data(); var target_array = []; for (var i = 0; i < selectedData.length; i++) { col1_data = selectedData[i]; domain_name = col1_data.match(/([^\n]+)/g)[0]; target_array.push(domain_name); } target_array = Array.from(new Set(target_array)); for (target in target_array) { select = document.getElementById('filterByTarget'); var option = document.createElement('option'); option.value = target_array[target]; option.innerHTML = target_array[target]; select.appendChild(option); } // populate Scan Type var selectedData = table.cells(rows, 4).data(); var scan_type_array = []; for (var i = 0; i < selectedData.length; i++) { col1_data = extractContent(selectedData[i]); scan_type_array.push(col1_data); } scan_type_array = Array.from(new Set(scan_type_array)); for (engine in scan_type_array) { select = document.getElementById('filterByScanType'); var option = document.createElement('option'); option.value = scan_type_array[engine]; option.innerHTML = scan_type_array[engine]; select.appendChild(option); } } }); multiCheck(table); // filter organization populate $.getJSON(`/api/listOrganizations?&format=json`, function(data) { data = data['organizations'] for (organization in data) { name = htmlEncode(data[organization]['name']); select = document.getElementById('filterByOrganization'); var option = document.createElement('option'); option.value = name; option.innerHTML = name; select.appendChild(option); } }).fail(function() {}); // filtering for scan status var status_types = ['Pending', 'Scanning', 'Aborted', 'Successful', 'Failed']; for (status in status_types) { select = document.getElementById('filterByScanStatus'); var option = document.createElement('option'); option.value = status_types[status]; option.innerHTML = status_types[status]; select.appendChild(option); } var org_filter = document.getElementById('filterByOrganization'); org_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Organization: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by organization ${this.value}`, pos: 'top-center' }); }, false); var status_filter = document.getElementById('filterByScanStatus'); status_filter.addEventListener('click', function() { table.search(this.value).draw(); switch (this.value) { case 'Pending': badge_color = 'warning'; break; case 'Scanning': badge_color = 'info'; break; case 'Aborted': badge_color = 'danger'; break; case 'Failed': badge_color = 'danger'; break; case 'Successful': badge_color = 'success'; break; default: badge_color = 'primary' } document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-${badge_color}">Scan Status: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by scan status ${this.value}`, pos: 'top-center' }); }, false); var engine_filter = document.getElementById('filterByScanType'); engine_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Scan Engine: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by Engine ${this.value}`, pos: 'top-center' }); }, false); var target_filter = document.getElementById('filterByTarget'); target_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Target/Domain: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by Engine ${this.value}`, pos: 'top-center' }); }, false); // reset filtering var reset_filter = document.getElementById('resetFilters'); reset_filter.addEventListener('click', function() { resetFilters(table); }, false); }); function resetFilters(table_obj) { table_obj.search("").draw(); Snackbar.show({ text: `Filters Reset`, pos: 'top-center' }); document.getElementById('filteringText').innerHTML = ''; } function checkedCount() { // this function will count the number of boxes checked item = document.getElementsByClassName("targets_checkbox"); count = 0; for (var i = 0; i < item.length; i++) { if (item[i].checked) { count++; } } return count; } function toggleMultipleTargetButton() { if (checkedCount() > 0) { $("#delete_multiple_button").removeClass("disabled"); } else { $("#delete_multiple_button").addClass("disabled"); } } function mainCheckBoxSelected(checkbox) { if (checkbox.checked) { $("#delete_multiple_button").removeClass("disabled"); $(".targets_checkbox").prop('checked', true); } else { $("#delete_multiple_button").addClass("disabled"); $(".targets_checkbox").prop('checked', false); } } function deleteMultipleScan() { if (!checkedCount()) { swal({ title: '', text: "Oops! No targets has been selected!", type: 'error', padding: '2em' }) } else { // atleast one target is selected swal.queue([{ title: 'Are you sure you want to delete ' + checkedCount() + ' Scans?', text: "This action is irreversible.\nThis will delete all the scan data and vulnerabilities related to the scan.", type: 'warning', showCancelButton: true, confirmButtonText: 'Delete', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { deleteForm = document.getElementById("scan_history_form"); deleteForm.action = "../delete/multiple"; deleteForm.submit(); } }]) } } // select option listener for report_type_select var report_type = document.getElementById("report_type_select"); report_type.addEventListener("change", function() { if(report_type.value == "recon") { $("#report_info_vuln_div").hide(); } else{ $("#report_info_vuln_div").show(); } }); function initiate_report(id, is_subdomain_scan, is_vulnerability_scan, domain_name) { $('#generateReportModal').modal('show'); $('#report_alert_message').empty(); $('#report_type_select').empty(); if (is_subdomain_scan == 'True' && is_vulnerability_scan == 'True') { $('#report_alert_message').append(` <b>Full Scan</b> will include both Reconnaissance and Vulnerability Report.<br> `); $('#report_type_select').append($('<option>', { value: 'full', text: 'Full Scan Report' })); } if (is_subdomain_scan == 'True') { // eligible for reconnaissance report $('#report_alert_message').append(` <b>Reconnaissance Report</b> will only include Assets Discovered Section.<br> `); $('#report_type_select').append($('<option>', { value: 'recon', text: 'Reconnaissance Report' })); } if (is_vulnerability_scan == 'True'){ // eligible for vulnerability report $('#report_alert_message').append(` <b>Vulnerability Report</b> will only include details of Vulnerabilities Identified. `); $('#report_type_select').append($('<option>', { value: 'vulnerability', text: 'Vulnerability Report' })); } $('#generateReportButton').attr('onClick', `generate_report(${id}, '${domain_name}')`); $('#previewReportButton').attr('onClick', `preview_report(${id}, '${domain_name}')`); } function preview_report(id, domain_name){ var report_type = $("#report_type_select option:selected").val(); var url = `/scan/create_report/${id}?report_type=${report_type}`; if ($('#report_ignore_info_vuln').is(":checked")) { url += `&ignore_info_vuln` } $('#generateReportModal').modal('hide'); window.open(url, '_blank').focus(); } function generate_report(id, domain_name) { var report_type = $("#report_type_select option:selected").val(); var url = `/scan/create_report/${id}?report_type=${report_type}&download`; if ($('#report_ignore_info_vuln').is(":checked")) { url += `&ignore_info_vuln` } $('#generateReportModal').modal('hide'); swal.queue([{ title: 'Generating Report!', text: `Please wait until we generate a report for you!`, padding: '2em', onOpen: function() { swal.showLoading() return fetch(url, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }) .then(function(response) { return response.blob(); }).then(function(blob) { const file = new Blob([blob], {type: 'application/pdf'}); // process to auto download it const fileURL = URL.createObjectURL(file); const link = document.createElement('a'); link.href = fileURL; link.download = domain_name + ".pdf"; link.click(); swal.close(); }) .catch(function() { swal.insertQueueStep({ type: 'error', title: 'Oops! Unable to generate report!' }) }) } }]); } </script> {% endblock page_level_script %}

yogeshojha

3c60bc1ee495044794d91edee0c96fff73ab46c7

5413708d243799a5271440c47c6f98d0c51154ca

## DOM text reinterpreted as HTML [DOM text](1) is reinterpreted as HTML without escaping meta-characters. [Show more details](https://github.com/yogeshojha/rengine/security/code-scanning/154)

github-advanced-security[bot]

yogeshojha/rengine

2.0-jasper release

### Added - Projects: Projects allow you to efficiently organize their web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. - Roles and Permissions: assign distinct roles to your team members: Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - GPT-powered Report Generation: With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments. - API Vault: This feature allows you to organize your API keys such as OpenAI or Netlas API keys. - GPT-powered Attack Surface Generation - URL gathering now is much more efficient, removing duplicate endpoints based on similar HTTP Responses, having the same content_lenth, or page_title. Custom duplicate fields can also be set from the scan engine configuration. - URL Path filtering while initiating scan: For instance, if we want to scan only endpoints starting with https://example.com/start/, we can pass the /start as a path filter while starting the scan. @ocervell - Expanding Target Concept: reNgine 2.0 now accepts IPs, URLS, etc as targets. (#678, #658) Excellent work by @ocervell - A ton of refactoring on reNgine's core to improve scan efficiency. Massive kudos to @ocervell - Created a custom celery workflow to be able to run several tasks in parallel that are not dependent on each other, such OSINT task and subdomain discovery will run in parallel, and directory and file fuzzing, vulnerability scan, screenshot gathering etc. will run in parallel after port scan or url fetching is completed. This will increase the efficiency of scans and instead of having one long flow of tasks, they can run independently on their own. @ocervell - Refactored all tasks to run asynchronously @ocervell - Added a stream_command that allows to read the output of a command live: this means the UI is updated with results while the command runs and does not have to wait until the task completes. Excellent work by @ocervell - Pwndb is now replaced by h8mail. @ocervell - Group Scan Results: reNgine 2.0 allows to group of subdomains based on similar page titles and HTTP status, and also vulnerability grouping based on the same vulnerability title and severity. - Added Support for Nmap: reNgine 2.0 allows to run Nmap scripts and vuln scans on ports found by Naabu. @ocervell - Added support for Shared Scan Variables in Scan Engine Configuration: - `enable_http_crawl`: (true/false) You can disable it to be more stealthy or focus on something different than HTTP - `timeout`: set timeout for all tasks - `rate_limit`: set rate limit for all tasks - `retries`: set retries for all tasks - `custom_header`: set the custom header for all tasks - Added Dalfox for XSS Vulnerability Scan - Added CRLFuzz for CRLF Vulnerability Scan - Added S3Scanner for scanning misconfigured S3 buckets - Improve OSINT Dork results, now detects admin panels, login pages and dashboards - Added Custom Dorks - Improved UI for vulnerability results, clicking on each vulnerability will open up a sidebar with vulnerability details. - Added HTTP Request and Response in vulnerability Results - Under Admin Settings, added an option to allow add/remove/deactivate additional users - Added Option to Preview Scan Report instead of forcing to download - Added Katana for crawling and spidering URLs - Added Netlas for Whois and subdomain gathering - Added TLSX for subdomain gathering - Added CTFR for subdomain gathering - Added historical IP in whois section ### Fixes - GF patterns do not run on 404 endpoints (#574 closed) - Fixes for retrieving whois data (#693 closed) - Related/Associated Domains in Whois section is now fixed ### Removed - Removed pwndb and tor related to it. - Removed tor for pwndb

2023-10-02 07:51:35+00:00

2023-10-07 10:37:23+00:00

web/startScan/templates/startScan/history.html

{% extends 'base/base.html' %} {% load static %} {% load humanize %} {% block title %} Scan history {% endblock title %} {% block custom_js_css_link %} <link rel="stylesheet" type="text/css" href="{% static 'plugins/datatable/datatables.css' %}"> <link rel="stylesheet" type="text/css" href="{% static 'plugins/datatable/global.css' %}"> <link rel="stylesheet" type="text/css" href="{% static 'plugins/datatable/custom.css' %}"> {% endblock custom_js_css_link %} {% block breadcrumb_title %} <li class="breadcrumb-item active" aria-current="page">Scan History</li> {% endblock breadcrumb_title %} {% block page_title %} Quick Scan History {% endblock page_title %} {% block main_content %} <div class="row"> <div class="col-12"> <div class="card"> <div class="p-2"> <div class="row"> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <button type="button" class="btn btn-primary dropdown-toggle" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false" id="filterMenu"> Filter <i class="fe-filter"></i> </button> <div id="filteringText" class="mt-2"> </div> <div class="dropdown-menu" style="width: 30%"> <div class="px-4 py-3"> <h4 class="headline-title">Filters</h4> <div class=""> <label for="filterByOrganization" class="form-label">Filter by Organization</label> <select class="form-control" id="filterByOrganization"> </select> </div> <div class=""> <label for="filterByTarget" class="form-label">Filter by Targets</label> <select class="form-control" id="filterByTarget"> </select> </div> <div class=""> <label for="filterByScanType" class="form-label">Filter by Scan Type</label> <select class="form-control" id="filterByScanType"> </select> </div> <div class=""> <label for="filterByScanStatus" class="form-label">Filter by Scan Status</label> <select class="form-control" id="filterByScanStatus"> </select> </div> </div> <div class="dropdown-divider"></div> <a href="#" class="dropdown-ite text-primary float-end" id="resetFilters">Reset Filters</a> </div> </div> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <a class="btn btn-soft-danger float-end disabled ms-1" href="#" onclick="deleteMultipleScan()" id="delete_multiple_button">Delete Multiple Scans</a> </div> </div> </div> </div> </div> </div> <div class="row"> <div class="col-12"> <div class="card"> <form method="POST" id="scan_history_form"> {% csrf_token %} <table id="scan_history_table" class="table style-3 table-hover"> <thead> <tr> <th class="checkbox-column text-center">Serial Number</th> <th class="text-center">Serial Number</th> <th class="">Domain Name</th> <th>Summary</th> <th class="">Scan Engine Used</th> <th>Last Scan</th> <th class="text-center">Status</th> <th class="text-center">Progress</th> <th class="text-center no-sorting">Action</th> </tr> </thead> <tbody> {% for scan_history in scan_history.all %} <tr> <td class="checkbox-column"> {{ scan_history.id }} </td> <td class=""> {{ scan_history.id }} </td> <td class=""> {{ scan_history.domain.name }} <br> {% for organization in scan_history.domain.get_organization %} <span class="badge badge-soft-dark mt-1 me-1" data-toggle="tooltip" data-placement="top" title="Domain {{domain.name}} belongs to organization {{organization.name}}">{{ organization.name }}</span> {% endfor %} </td> <td class="text-left"> <span class="badge badge-pills bg-info mt-1" data-toggle="tooltip" data-placement="top" title="Subdomains">{{scan_history.get_subdomain_count}}</span> <span class="badge badge-pills bg-warning mt-1" data-toggle="tooltip" data-placement="top" title="Endpoints">{{scan_history.get_endpoint_count}}</span> <span class="badge badge-pills bg-danger mt-1" data-toggle="tooltip" data-placement="top" title="{{scan_history.get_critical_vulnerability_count}} Critical, {{scan_history.get_high_vulnerability_count}} High, {{scan_history.get_medium_vulnerability_count}} Medium Vulnerabilities">{{scan_history.get_vulnerability_count}}</span> </td> <td class=""> <span class="badge badge-soft-primary">{{ scan_history.scan_type }}</span> </td> <td> <span data-toggle="tooltip" data-placement="top" title="{{scan_history.start_scan_date}}">{{scan_history.start_scan_date|naturaltime}}</span> </td> <td class="text-center"> {% if scan_history.scan_status == -1 %} <span class="badge badge-soft-warning" data-placement="top" data-toggle="tooltip" data-placement="top" title="Waiting for other scans to complete"><span class="spinner-border spinner-border-sm"></span> Pending</span> {% elif scan_history.scan_status == 0 %} <span class="badge badge-soft-danger">Failed</span> {% if scan_history.error_message %}</br><p class="text-danger">Scan Failed due to: {{scan_history.error_message}}</p>{% endif %} {% elif scan_history.scan_status == 1 %} <span class="badge badge-soft-info"><span class="spinner-border spinner-border-sm"></span> In Progress</span> {% elif scan_history.scan_status == 2 %} <span class="badge badge-soft-success">Successful</span> {% elif scan_history.scan_status == 3 %} <span class="badge badge-soft-danger">Aborted</span> {% else %} <span class="badge badge-soft-danger">Unknown</span> {% endif %} </td> <td class="text-center"> {% if scan_history.scan_status == -1 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-warning" role="progressbar" style="width: 75%" aria-valuenow="75" aria-valuemin="0" aria-valuemax="100"></div> </div> {% elif scan_history.scan_status == 0 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% elif scan_history.scan_status == 1 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-primary progress-bar-striped progress-bar-animated" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% elif scan_history.scan_status == 2 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-success" role="progressbar" style="width: 100%" aria-valuemin="0" aria-valuemax="100"></div> </div> {% elif scan_history.scan_status == 3 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger progress-bar-striped" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% else %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger" role="progressbar" style="width: 100%" aria-valuemin="0" aria-valuemax="100"> </div> </div> {% endif %} </td> <td class="text-center"> <div class="btn-group mb-2 dropstart"> <div class="btn-group"> <a href="{% url 'detail_scan' scan_history.id %}" class="btn btn-soft-primary">View Results</a> <div class="btn-group dropstart" role="group"> <button type="button" class="btn btn-soft-primary dropdown-toggle dropdown-toggle-split" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false"> <i class="mdi mdi-chevron-right"></i> </button> <div class="dropdown-menu" style=""> {% if scan_history.scan_status == 0 or scan_history.scan_status == 2 or scan_history.scan_status == 3 %} <a class="dropdown-item text-primary" href="{% url 'start_scan' scan_history.domain.id %}"> <i class="fe-refresh-ccw"></i>&nbsp;Rescan </a> {% endif %} {% if scan_history.scan_status == 1 or scan_history.scan_status == -1%} <a href="#" class="dropdown-item text-danger" onclick="stop_scan(scan_id={{ scan_history.id }}, subscan_id=null, reload_scan_bar=false, reload_location=true)"> <i class="fe-alert-triangle"></i>&nbsp;Stop Scan</a> {% endif %} {% if scan_history.scan_status == 2 or scan_history.scan_status == 3 %} <a href="#" class="dropdown-item text-danger" onclick="delete_scan('{{ scan_history.id }}')"> <i class="fe-trash-2"></i>&nbsp;Delete Scan Results</a> {% endif %} {% if scan.scan_status != -1%} <div class="dropdown-divider"></div> <a href="#" class="dropdown-item text-dark" onclick="initiate_report({{scan_history.id}}, '{{scan_history.subdomain_discovery}}', '{{scan_history.vulnerability_scan}}', '{{ scan_history.domain.name }}')"> <i class="fe-download"></i>&nbsp;Download Report</a> {% endif %} </div> </div> </div> </div> </td> </tr> {% endfor %} </tbody> </table> </form> </div> </div> </div> <div class="modal fade" id="generateReportModal" tabindex="-1" style="display: none;" aria-hidden="true"> <div class="modal-dialog modal-dialog-centered"> <div class="modal-content"> <div class="modal-header"> <h4 class="modal-title" id="myCenterModalLabel">Download Report</h4> <button type="button" class="btn-close" data-bs-dismiss="modal" aria-label="Close"></button> </div> <div class="modal-body"> <div class="alert alert-light-primary border-0 mb-4" role="alert"> <div id='report_alert_message'></div> </div> <div class="form-group mb-4"> <label for="reportTypeForm">Report Type</label> <select class="form-control" id="report_type_select" name="report_type"> </select> </div> <a id='generateReportButton' href="#" class="btn btn-primary float-end">Download Report</a> </div> </div> </div> </div> {% endblock main_content %} {% block page_level_script %} <script src="{% static 'plugins/datatable/datatables.js' %}"></script> <script> $(document).ready(function() { var table = $('#scan_history_table').DataTable({ headerCallback: function(e, a, t, n, s) { e.getElementsByTagName("th")[0].innerHTML='<div class="form-check mb-2 form-check-primary"><input type="checkbox" class="float-start form-check-input chk-parent" id="head_checkbox" onclick=mainCheckBoxSelected(this)>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>\n' }, "columnDefs":[ { 'visible': false, 'targets': [1] }, { "targets":0, "width":"20px", "className":"", "orderable":!1, render:function(e, a, t, n) { return'<div class="form-check mb-2 form-check-primary"><input type="checkbox" name="targets_checkbox['+ e + ']" class="float-start form-check-input targets_checkbox" value="' + e + '" onchange=toggleMultipleTargetButton()>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>' }, }], "order": [[1, 'desc']], "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center mt-sm-0 mt-3'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "stripeClasses": [], "lengthMenu": [5, 10, 20, 30, 40, 50], "pageLength": 20, "initComplete": function(settings, json) { $('[data-toggle="tooltip"]').tooltip(); table = settings.oInstance.api(); var rows = table.rows({ selected: true }).indexes(); // populate filter menu from datatables // populate targets var selectedData = table.cells(rows, 2).data(); var target_array = []; for (var i = 0; i < selectedData.length; i++) { col1_data = selectedData[i]; domain_name = col1_data.match(/([^\n]+)/g)[0]; target_array.push(domain_name); } target_array = Array.from(new Set(target_array)); for (target in target_array) { select = document.getElementById('filterByTarget'); var option = document.createElement('option'); option.value = target_array[target]; option.innerHTML = target_array[target]; select.appendChild(option); } // populate Scan Type var selectedData = table.cells(rows, 4).data(); var scan_type_array = []; for (var i = 0; i < selectedData.length; i++) { col1_data = extractContent(selectedData[i]); scan_type_array.push(col1_data); } scan_type_array = Array.from(new Set(scan_type_array)); for (engine in scan_type_array) { select = document.getElementById('filterByScanType'); var option = document.createElement('option'); option.value = scan_type_array[engine]; option.innerHTML = scan_type_array[engine]; select.appendChild(option); } } }); multiCheck(table); // filter organization populate $.getJSON(`/api/listOrganizations?&format=json`, function(data) { data = data['organizations'] for (organization in data) { name = htmlEncode(data[organization]['name']); select = document.getElementById('filterByOrganization'); var option = document.createElement('option'); option.value = name; option.innerHTML = name; select.appendChild(option); } }).fail(function() {}); // filtering for scan status var status_types = ['Pending', 'Scanning', 'Aborted', 'Successful', 'Failed']; for (status in status_types) { select = document.getElementById('filterByScanStatus'); var option = document.createElement('option'); option.value = status_types[status]; option.innerHTML = status_types[status]; select.appendChild(option); } var org_filter = document.getElementById('filterByOrganization'); org_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Organization: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by organization ${this.value}`, pos: 'top-center' }); }, false); var status_filter = document.getElementById('filterByScanStatus'); status_filter.addEventListener('click', function() { table.search(this.value).draw(); switch (this.value) { case 'Pending': badge_color = 'warning'; break; case 'Scanning': badge_color = 'info'; break; case 'Aborted': badge_color = 'danger'; break; case 'Failed': badge_color = 'danger'; break; case 'Successful': badge_color = 'success'; break; default: badge_color = 'primary' } document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-${badge_color}">Scan Status: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by scan status ${this.value}`, pos: 'top-center' }); }, false); var engine_filter = document.getElementById('filterByScanType'); engine_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Scan Engine: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by Engine ${this.value}`, pos: 'top-center' }); }, false); var target_filter = document.getElementById('filterByTarget'); target_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Target/Domain: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by Engine ${this.value}`, pos: 'top-center' }); }, false); // reset filtering var reset_filter = document.getElementById('resetFilters'); reset_filter.addEventListener('click', function() { resetFilters(table); }, false); }); function resetFilters(table_obj) { table_obj.search("").draw(); Snackbar.show({ text: `Filters Reset`, pos: 'top-center' }); document.getElementById('filteringText').innerHTML = ''; } function checkedCount() { // this function will count the number of boxes checked item = document.getElementsByClassName("targets_checkbox"); count = 0; for (var i = 0; i < item.length; i++) { if (item[i].checked) { count++; } } return count; } function toggleMultipleTargetButton() { if (checkedCount() > 0) { $("#delete_multiple_button").removeClass("disabled"); } else { $("#delete_multiple_button").addClass("disabled"); } } function mainCheckBoxSelected(checkbox) { if (checkbox.checked) { $("#delete_multiple_button").removeClass("disabled"); $(".targets_checkbox").prop('checked', true); } else { $("#delete_multiple_button").addClass("disabled"); $(".targets_checkbox").prop('checked', false); } } function deleteMultipleScan() { if (!checkedCount()) { swal({ title: '', text: "Oops! No targets has been selected!", type: 'error', padding: '2em' }) } else { // atleast one target is selected swal.queue([{ title: 'Are you sure you want to delete ' + checkedCount() + ' Scans?', text: "This action is irreversible.\nThis will delete all the scan data and vulnerabilities related to the scan.", type: 'warning', showCancelButton: true, confirmButtonText: 'Delete', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { deleteForm = document.getElementById("scan_history_form"); deleteForm.action = "../delete/multiple"; deleteForm.submit(); } }]) } } function initiate_report(id, is_subdomain_scan, is_vulnerability_scan, domain_name) { $('#generateReportModal').modal('show'); $('#report_alert_message').empty(); $('#report_type_select').empty(); if (is_subdomain_scan == 'True' && is_vulnerability_scan == 'True') { $('#report_alert_message').append(` <b>Full Scan</b> will include both Reconnaissance and Vulnerability Report.<br> `); $('#report_type_select').append($('<option>', { value: 'full', text: 'Full Scan Report' })); } if (is_subdomain_scan == 'True') { // eligible for reconnaissance report $('#report_alert_message').append(` <b>Reconnaissance Report</b> will only include Assets Discovered Section.<br> `); $('#report_type_select').append($('<option>', { value: 'recon', text: 'Reconnaissance Report' })); } if (is_vulnerability_scan == 'True'){ // eligible for vulnerability report $('#report_alert_message').append(` <b>Vulnerability Report</b> will only include details of Vulnerabilities Identified. `); $('#report_type_select').append($('<option>', { value: 'vulnerability', text: 'Vulnerability Report' })); } $('#generateReportButton').attr('onClick', `generate_report(${id}, '${domain_name}')`); } function generate_report(id, domain_name) { var report_type = $("#report_type_select option:selected").val(); $('#generateReportModal').modal('hide'); swal.queue([{ title: 'Generating Report!', text: `Please wait until we generate a report for you!`, padding: '2em', onOpen: function() { swal.showLoading() return fetch(`/scan/create_report/${id}?download&report_type=${report_type}`, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }) .then(function(response) { return response.blob(); }).then(function(blob) { const file = new Blob([blob], {type: 'application/pdf'}); // process to auto download it const fileURL = URL.createObjectURL(file); const link = document.createElement('a'); link.href = fileURL; link.download = domain_name + ".pdf"; link.click(); swal.close(); }) .catch(function() { swal.insertQueueStep({ type: 'error', title: 'Oops! Unable to generate report!' }) }) } }]); } </script> {% endblock page_level_script %}

{% extends 'base/base.html' %} {% load static %} {% load humanize %} {% load permission_tags %} {% block title %} Scan history {% endblock title %} {% block custom_js_css_link %} {% endblock custom_js_css_link %} {% block breadcrumb_title %} <li class="breadcrumb-item active" aria-current="page">Scan History</li> {% endblock breadcrumb_title %} {% block page_title %} Quick Scan History {% endblock page_title %} {% block main_content %} <div class="row"> <div class="col-12"> <div class="card"> <div class="p-2"> <div class="row"> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <button type="button" class="btn btn-primary dropdown-toggle" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false" id="filterMenu"> Filter <i class="fe-filter"></i> </button> <div id="filteringText" class="mt-2"> </div> <div class="dropdown-menu" style="width: 30%"> <div class="px-4 py-3"> <h4 class="headline-title">Filters</h4> <div class=""> <label for="filterByOrganization" class="form-label">Filter by Organization</label> <select class="form-control" id="filterByOrganization"> </select> </div> <div class=""> <label for="filterByTarget" class="form-label">Filter by Targets</label> <select class="form-control" id="filterByTarget"> </select> </div> <div class=""> <label for="filterByScanType" class="form-label">Filter by Scan Type</label> <select class="form-control" id="filterByScanType"> </select> </div> <div class=""> <label for="filterByScanStatus" class="form-label">Filter by Scan Status</label> <select class="form-control" id="filterByScanStatus"> </select> </div> </div> <div class="dropdown-divider"></div> <a href="#" class="dropdown-ite text-primary float-end" id="resetFilters">Reset Filters</a> </div> </div> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <a class="btn btn-soft-danger float-end disabled ms-1" href="#" onclick="deleteMultipleScan()" id="delete_multiple_button">Delete Multiple Scans</a> </div> </div> </div> </div> </div> </div> <div class="row"> <div class="col-12"> <div class="card"> <form method="POST" id="scan_history_form"> {% csrf_token %} <table id="scan_history_table" class="table dt-responsive w-100"> <thead> <tr> <th class="checkbox-column text-center">Serial Number</th> <th class="text-center">Serial Number</th> <th class="">Domain Name</th> <th>Summary</th> <th class="">Scan Engine Used</th> <th>Last Scan</th> <th class="text-center">Status</th> <th class="text-center">Progress</th> <th class="text-center no-sorting">Action</th> </tr> </thead> <tbody> {% for scan_history in scan_history.all %} <tr> <td class="checkbox-column"> {{ scan_history.id }} </td> <td class=""> {{ scan_history.id }} </td> <td class=""> {{ scan_history.domain.name }} <br> {% for organization in scan_history.domain.get_organization %} <span class="badge badge-soft-dark mt-1 me-1" data-toggle="tooltip" data-placement="top" title="Domain {{domain.name}} belongs to organization {{organization.name}}">{{ organization.name }}</span> {% endfor %} </td> <td class="text-left"> <span class="badge badge-pills bg-info mt-1" data-toggle="tooltip" data-placement="top" title="Subdomains">{{scan_history.get_subdomain_count}}</span> <span class="badge badge-pills bg-warning mt-1" data-toggle="tooltip" data-placement="top" title="Endpoints">{{scan_history.get_endpoint_count}}</span> <span class="badge badge-pills bg-danger mt-1" data-toggle="tooltip" data-placement="top" title="{{scan_history.get_critical_vulnerability_count}} Critical, {{scan_history.get_high_vulnerability_count}} High, {{scan_history.get_medium_vulnerability_count}} Medium Vulnerabilities">{{scan_history.get_vulnerability_count}}</span> </td> <td class=""> <span class="badge badge-soft-primary">{{ scan_history.scan_type }}</span> </td> <td> <span data-toggle="tooltip" data-placement="top" title="{{scan_history.start_scan_date}}">{{scan_history.start_scan_date|naturaltime}}</span> </td> <td class="text-center"> {% if scan_history.scan_status == -1 %} <span class="badge badge-soft-warning" data-placement="top" data-toggle="tooltip" data-placement="top" title="Waiting for other scans to complete"><span class="spinner-border spinner-border-sm"></span> Pending</span> {% elif scan_history.scan_status == 0 %} <span class="badge badge-soft-danger">Failed</span> {% if scan_history.error_message %}</br><p class="text-danger">Scan Failed due to: {{scan_history.error_message}}</p>{% endif %} {% elif scan_history.scan_status == 1 %} <span class="badge badge-soft-info"><span class="spinner-border spinner-border-sm"></span> In Progress</span> {% elif scan_history.scan_status == 2 %} <span class="badge badge-soft-success">Successful</span> {% elif scan_history.scan_status == 3 %} <span class="badge badge-soft-danger">Aborted</span> {% else %} <span class="badge badge-soft-danger">Unknown</span> {% endif %} </td> <td class="text-center"> {% if scan_history.scan_status == -1 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-warning" role="progressbar" style="width: 75%" aria-valuenow="75" aria-valuemin="0" aria-valuemax="100"></div> </div> {% elif scan_history.scan_status == 0 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% elif scan_history.scan_status == 1 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-primary progress-bar-striped progress-bar-animated" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% elif scan_history.scan_status == 2 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-success" role="progressbar" style="width: 100%" aria-valuemin="0" aria-valuemax="100"></div> </div> {% elif scan_history.scan_status == 3 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger progress-bar-striped" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% else %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger" role="progressbar" style="width: 100%" aria-valuemin="0" aria-valuemax="100"> </div> </div> {% endif %} </td> <td class="text-center"> <div class="btn-group mb-2 dropstart"> <div class="btn-group"> <a href="/scan/{{current_project.slug}}/detail/{{scan_history.id}}" class="btn btn-soft-primary">View Results</a> <div class="btn-group dropstart" role="group"> <button type="button" class="btn btn-soft-primary dropdown-toggle dropdown-toggle-split" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false"> <i class="mdi mdi-chevron-right"></i> </button> <div class="dropdown-menu" style=""> {% if user|can:'initiate_scans_subscans' %} {% if scan_history.scan_status == 0 or scan_history.scan_status == 2 or scan_history.scan_status == 3 %} <a class="dropdown-item text-primary" href="/scan/{{current_project.slug}}/start/{{scan_history.domain.id}}"> <i class="fe-refresh-ccw"></i>&nbsp;Rescan </a> {% endif %} {% if scan_history.scan_status == 1 or scan_history.scan_status == -1%} <a href="#" class="dropdown-item text-danger" onclick="stop_scan(scan_id={{ scan_history.id }}, subscan_id=null, reload_scan_bar=false, reload_location=true)"> <i class="fe-alert-triangle"></i>&nbsp;Stop Scan</a> {% endif %} {% endif %} {% if user|can:'modify_scan_results' %} {% if scan_history.scan_status == 2 or scan_history.scan_status == 3 or scan_history.scan_status == 0 %} <a href="#" class="dropdown-item text-danger" onclick="delete_scan('{{ scan_history.id }}')"> <i class="fe-trash-2"></i>&nbsp;Delete Scan Results</a> {% endif %} <div class="dropdown-divider"></div> {% endif %} {% if scan.scan_status != -1%} <a href="#" class="dropdown-item text-dark" onclick="initiate_report({{scan_history.id}}, '{% if 'subdomain_discovery' in scan_history.scan_type.tasks %}True{% endif %}', '{% if 'vulnerability_scan' in scan_history.scan_type.tasks %}True{% endif %}', '{{ scan_history.domain.name }}')"> <i class="fe-download"></i>&nbsp;Scan Report</a> {% endif %} </div> </div> </div> </div> </td> </tr> {% endfor %} </tbody> </table> </form> </div> </div> </div> <div class="modal fade" id="generateReportModal" tabindex="-1" style="display: none;" aria-hidden="true"> <div class="modal-dialog modal-dialog-centered"> <div class="modal-content"> <div class="modal-header"> <h4 class="modal-title" id="myCenterModalLabel">Download Report</h4> <button type="button" class="btn-close" data-bs-dismiss="modal" aria-label="Close"></button> </div> <div class="modal-body"> <div class="alert alert-light-primary border-0 mb-4" role="alert"> <div id='report_alert_message'></div> </div> <div class="form-group mb-4"> <label for="reportTypeForm">Report Type</label> <select class="form-control" id="report_type_select" name="report_type"> </select> </div> <div class="form-group mb-4"> <div class="form-check" id="report_info_vuln_div"> <input type="checkbox" class="form-check-input" id="report_ignore_info_vuln" checked=""> <label class="form-check-label" for="report_ignore_info_vuln">Ignore Informational Vulnerabilities</label> </div> </div> <a id='generateReportButton' href="#" class="btn btn-primary float-end m-2">Download Report</a> <a id='previewReportButton' href="#" class="btn btn-secondary float-end m-2">Preview Report</a> </div> </div> </div> </div> {% endblock main_content %} {% block page_level_script %} <script> $(document).ready(function() { var table = $('#scan_history_table').DataTable({ headerCallback: function(e, a, t, n, s) { e.getElementsByTagName("th")[0].innerHTML='<div class="form-check mb-2 form-check-primary"><input type="checkbox" class="float-start form-check-input chk-parent" id="head_checkbox" onclick=mainCheckBoxSelected(this)>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>\n' }, "columnDefs":[ { 'visible': false, 'targets': [1] }, { "targets":0, "width":"20px", "className":"", "orderable":!1, render:function(e, a, t, n) { return'<div class="form-check mb-2 form-check-primary"><input type="checkbox" name="targets_checkbox['+ e + ']" class="float-start form-check-input targets_checkbox" value="' + e + '" onchange=toggleMultipleTargetButton()>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>' }, } ], "order": [[1, 'desc']], "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center mt-sm-0 mt-3'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "stripeClasses": [], "lengthMenu": [5, 10, 20, 30, 40, 50], "pageLength": 20, "initComplete": function(settings, json) { $('[data-toggle="tooltip"]').tooltip(); table = settings.oInstance.api(); var rows = table.rows({ selected: true }).indexes(); // populate filter menu from datatables // populate targets var selectedData = table.cells(rows, 2).data(); var target_array = []; for (var i = 0; i < selectedData.length; i++) { col1_data = selectedData[i]; domain_name = col1_data.match(/([^\n]+)/g)[0]; target_array.push(domain_name); } target_array = Array.from(new Set(target_array)); for (target in target_array) { select = document.getElementById('filterByTarget'); var option = document.createElement('option'); option.value = target_array[target]; option.innerHTML = target_array[target]; select.appendChild(option); } // populate Scan Type var selectedData = table.cells(rows, 4).data(); var scan_type_array = []; for (var i = 0; i < selectedData.length; i++) { col1_data = extractContent(selectedData[i]); scan_type_array.push(col1_data); } scan_type_array = Array.from(new Set(scan_type_array)); for (engine in scan_type_array) { select = document.getElementById('filterByScanType'); var option = document.createElement('option'); option.value = scan_type_array[engine]; option.innerHTML = scan_type_array[engine]; select.appendChild(option); } } }); multiCheck(table); // filter organization populate $.getJSON(`/api/listOrganizations?&format=json`, function(data) { data = data['organizations'] for (organization in data) { name = htmlEncode(data[organization]['name']); select = document.getElementById('filterByOrganization'); var option = document.createElement('option'); option.value = name; option.innerHTML = name; select.appendChild(option); } }).fail(function() {}); // filtering for scan status var status_types = ['Pending', 'Scanning', 'Aborted', 'Successful', 'Failed']; for (status in status_types) { select = document.getElementById('filterByScanStatus'); var option = document.createElement('option'); option.value = status_types[status]; option.innerHTML = status_types[status]; select.appendChild(option); } var org_filter = document.getElementById('filterByOrganization'); org_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Organization: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by organization ${this.value}`, pos: 'top-center' }); }, false); var status_filter = document.getElementById('filterByScanStatus'); status_filter.addEventListener('click', function() { table.search(this.value).draw(); switch (this.value) { case 'Pending': badge_color = 'warning'; break; case 'Scanning': badge_color = 'info'; break; case 'Aborted': badge_color = 'danger'; break; case 'Failed': badge_color = 'danger'; break; case 'Successful': badge_color = 'success'; break; default: badge_color = 'primary' } document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-${badge_color}">Scan Status: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by scan status ${this.value}`, pos: 'top-center' }); }, false); var engine_filter = document.getElementById('filterByScanType'); engine_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Scan Engine: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by Engine ${this.value}`, pos: 'top-center' }); }, false); var target_filter = document.getElementById('filterByTarget'); target_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Target/Domain: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by Engine ${this.value}`, pos: 'top-center' }); }, false); // reset filtering var reset_filter = document.getElementById('resetFilters'); reset_filter.addEventListener('click', function() { resetFilters(table); }, false); }); function resetFilters(table_obj) { table_obj.search("").draw(); Snackbar.show({ text: `Filters Reset`, pos: 'top-center' }); document.getElementById('filteringText').innerHTML = ''; } function checkedCount() { // this function will count the number of boxes checked item = document.getElementsByClassName("targets_checkbox"); count = 0; for (var i = 0; i < item.length; i++) { if (item[i].checked) { count++; } } return count; } function toggleMultipleTargetButton() { if (checkedCount() > 0) { $("#delete_multiple_button").removeClass("disabled"); } else { $("#delete_multiple_button").addClass("disabled"); } } function mainCheckBoxSelected(checkbox) { if (checkbox.checked) { $("#delete_multiple_button").removeClass("disabled"); $(".targets_checkbox").prop('checked', true); } else { $("#delete_multiple_button").addClass("disabled"); $(".targets_checkbox").prop('checked', false); } } function deleteMultipleScan() { if (!checkedCount()) { swal({ title: '', text: "Oops! No targets has been selected!", type: 'error', padding: '2em' }) } else { // atleast one target is selected swal.queue([{ title: 'Are you sure you want to delete ' + checkedCount() + ' Scans?', text: "This action is irreversible.\nThis will delete all the scan data and vulnerabilities related to the scan.", type: 'warning', showCancelButton: true, confirmButtonText: 'Delete', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { deleteForm = document.getElementById("scan_history_form"); deleteForm.action = "../delete/multiple"; deleteForm.submit(); } }]) } } // select option listener for report_type_select var report_type = document.getElementById("report_type_select"); report_type.addEventListener("change", function() { if(report_type.value == "recon") { $("#report_info_vuln_div").hide(); } else{ $("#report_info_vuln_div").show(); } }); function initiate_report(id, is_subdomain_scan, is_vulnerability_scan, domain_name) { $('#generateReportModal').modal('show'); $('#report_alert_message').empty(); $('#report_type_select').empty(); if (is_subdomain_scan == 'True' && is_vulnerability_scan == 'True') { $('#report_alert_message').append(` <b>Full Scan</b> will include both Reconnaissance and Vulnerability Report.<br> `); $('#report_type_select').append($('<option>', { value: 'full', text: 'Full Scan Report' })); } if (is_subdomain_scan == 'True') { // eligible for reconnaissance report $('#report_alert_message').append(` <b>Reconnaissance Report</b> will only include Assets Discovered Section.<br> `); $('#report_type_select').append($('<option>', { value: 'recon', text: 'Reconnaissance Report' })); } if (is_vulnerability_scan == 'True'){ // eligible for vulnerability report $('#report_alert_message').append(` <b>Vulnerability Report</b> will only include details of Vulnerabilities Identified. `); $('#report_type_select').append($('<option>', { value: 'vulnerability', text: 'Vulnerability Report' })); } $('#generateReportButton').attr('onClick', `generate_report(${id}, '${domain_name}')`); $('#previewReportButton').attr('onClick', `preview_report(${id}, '${domain_name}')`); } function preview_report(id, domain_name){ var report_type = $("#report_type_select option:selected").val(); var url = `/scan/create_report/${id}?report_type=${report_type}`; if ($('#report_ignore_info_vuln').is(":checked")) { url += `&ignore_info_vuln` } $('#generateReportModal').modal('hide'); window.open(url, '_blank').focus(); } function generate_report(id, domain_name) { var report_type = $("#report_type_select option:selected").val(); var url = `/scan/create_report/${id}?report_type=${report_type}&download`; if ($('#report_ignore_info_vuln').is(":checked")) { url += `&ignore_info_vuln` } $('#generateReportModal').modal('hide'); swal.queue([{ title: 'Generating Report!', text: `Please wait until we generate a report for you!`, padding: '2em', onOpen: function() { swal.showLoading() return fetch(url, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }) .then(function(response) { return response.blob(); }).then(function(blob) { const file = new Blob([blob], {type: 'application/pdf'}); // process to auto download it const fileURL = URL.createObjectURL(file); const link = document.createElement('a'); link.href = fileURL; link.download = domain_name + ".pdf"; link.click(); swal.close(); }) .catch(function() { swal.insertQueueStep({ type: 'error', title: 'Oops! Unable to generate report!' }) }) } }]); } </script> {% endblock page_level_script %}

yogeshojha

3c60bc1ee495044794d91edee0c96fff73ab46c7

5413708d243799a5271440c47c6f98d0c51154ca

## DOM text reinterpreted as HTML [DOM text](1) is reinterpreted as HTML without escaping meta-characters. [Show more details](https://github.com/yogeshojha/rengine/security/code-scanning/155)

github-advanced-security[bot]

yogeshojha/rengine

2.0-jasper release

### Added - Projects: Projects allow you to efficiently organize their web application reconnaissance efforts. With this feature, you can create distinct project spaces, each tailored to a specific purpose, such as personal bug bounty hunting, client engagements, or any other specialized recon task. - Roles and Permissions: assign distinct roles to your team members: Sys Admin, Penetration Tester, and Auditor—each with precisely defined permissions to tailor their access and actions within the reNgine ecosystem. - GPT-powered Report Generation: With the power of OpenAI's GPT, reNgine now provides you with detailed vulnerability descriptions, remediation strategies, and impact assessments. - API Vault: This feature allows you to organize your API keys such as OpenAI or Netlas API keys. - GPT-powered Attack Surface Generation - URL gathering now is much more efficient, removing duplicate endpoints based on similar HTTP Responses, having the same content_lenth, or page_title. Custom duplicate fields can also be set from the scan engine configuration. - URL Path filtering while initiating scan: For instance, if we want to scan only endpoints starting with https://example.com/start/, we can pass the /start as a path filter while starting the scan. @ocervell - Expanding Target Concept: reNgine 2.0 now accepts IPs, URLS, etc as targets. (#678, #658) Excellent work by @ocervell - A ton of refactoring on reNgine's core to improve scan efficiency. Massive kudos to @ocervell - Created a custom celery workflow to be able to run several tasks in parallel that are not dependent on each other, such OSINT task and subdomain discovery will run in parallel, and directory and file fuzzing, vulnerability scan, screenshot gathering etc. will run in parallel after port scan or url fetching is completed. This will increase the efficiency of scans and instead of having one long flow of tasks, they can run independently on their own. @ocervell - Refactored all tasks to run asynchronously @ocervell - Added a stream_command that allows to read the output of a command live: this means the UI is updated with results while the command runs and does not have to wait until the task completes. Excellent work by @ocervell - Pwndb is now replaced by h8mail. @ocervell - Group Scan Results: reNgine 2.0 allows to group of subdomains based on similar page titles and HTTP status, and also vulnerability grouping based on the same vulnerability title and severity. - Added Support for Nmap: reNgine 2.0 allows to run Nmap scripts and vuln scans on ports found by Naabu. @ocervell - Added support for Shared Scan Variables in Scan Engine Configuration: - `enable_http_crawl`: (true/false) You can disable it to be more stealthy or focus on something different than HTTP - `timeout`: set timeout for all tasks - `rate_limit`: set rate limit for all tasks - `retries`: set retries for all tasks - `custom_header`: set the custom header for all tasks - Added Dalfox for XSS Vulnerability Scan - Added CRLFuzz for CRLF Vulnerability Scan - Added S3Scanner for scanning misconfigured S3 buckets - Improve OSINT Dork results, now detects admin panels, login pages and dashboards - Added Custom Dorks - Improved UI for vulnerability results, clicking on each vulnerability will open up a sidebar with vulnerability details. - Added HTTP Request and Response in vulnerability Results - Under Admin Settings, added an option to allow add/remove/deactivate additional users - Added Option to Preview Scan Report instead of forcing to download - Added Katana for crawling and spidering URLs - Added Netlas for Whois and subdomain gathering - Added TLSX for subdomain gathering - Added CTFR for subdomain gathering - Added historical IP in whois section ### Fixes - GF patterns do not run on 404 endpoints (#574 closed) - Fixes for retrieving whois data (#693 closed) - Related/Associated Domains in Whois section is now fixed ### Removed - Removed pwndb and tor related to it. - Removed tor for pwndb

2023-10-02 07:51:35+00:00

2023-10-07 10:37:23+00:00

web/startScan/templates/startScan/history.html

{% extends 'base/base.html' %} {% load static %} {% load humanize %} {% block title %} Scan history {% endblock title %} {% block custom_js_css_link %} <link rel="stylesheet" type="text/css" href="{% static 'plugins/datatable/datatables.css' %}"> <link rel="stylesheet" type="text/css" href="{% static 'plugins/datatable/global.css' %}"> <link rel="stylesheet" type="text/css" href="{% static 'plugins/datatable/custom.css' %}"> {% endblock custom_js_css_link %} {% block breadcrumb_title %} <li class="breadcrumb-item active" aria-current="page">Scan History</li> {% endblock breadcrumb_title %} {% block page_title %} Quick Scan History {% endblock page_title %} {% block main_content %} <div class="row"> <div class="col-12"> <div class="card"> <div class="p-2"> <div class="row"> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <button type="button" class="btn btn-primary dropdown-toggle" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false" id="filterMenu"> Filter <i class="fe-filter"></i> </button> <div id="filteringText" class="mt-2"> </div> <div class="dropdown-menu" style="width: 30%"> <div class="px-4 py-3"> <h4 class="headline-title">Filters</h4> <div class=""> <label for="filterByOrganization" class="form-label">Filter by Organization</label> <select class="form-control" id="filterByOrganization"> </select> </div> <div class=""> <label for="filterByTarget" class="form-label">Filter by Targets</label> <select class="form-control" id="filterByTarget"> </select> </div> <div class=""> <label for="filterByScanType" class="form-label">Filter by Scan Type</label> <select class="form-control" id="filterByScanType"> </select> </div> <div class=""> <label for="filterByScanStatus" class="form-label">Filter by Scan Status</label> <select class="form-control" id="filterByScanStatus"> </select> </div> </div> <div class="dropdown-divider"></div> <a href="#" class="dropdown-ite text-primary float-end" id="resetFilters">Reset Filters</a> </div> </div> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <a class="btn btn-soft-danger float-end disabled ms-1" href="#" onclick="deleteMultipleScan()" id="delete_multiple_button">Delete Multiple Scans</a> </div> </div> </div> </div> </div> </div> <div class="row"> <div class="col-12"> <div class="card"> <form method="POST" id="scan_history_form"> {% csrf_token %} <table id="scan_history_table" class="table style-3 table-hover"> <thead> <tr> <th class="checkbox-column text-center">Serial Number</th> <th class="text-center">Serial Number</th> <th class="">Domain Name</th> <th>Summary</th> <th class="">Scan Engine Used</th> <th>Last Scan</th> <th class="text-center">Status</th> <th class="text-center">Progress</th> <th class="text-center no-sorting">Action</th> </tr> </thead> <tbody> {% for scan_history in scan_history.all %} <tr> <td class="checkbox-column"> {{ scan_history.id }} </td> <td class=""> {{ scan_history.id }} </td> <td class=""> {{ scan_history.domain.name }} <br> {% for organization in scan_history.domain.get_organization %} <span class="badge badge-soft-dark mt-1 me-1" data-toggle="tooltip" data-placement="top" title="Domain {{domain.name}} belongs to organization {{organization.name}}">{{ organization.name }}</span> {% endfor %} </td> <td class="text-left"> <span class="badge badge-pills bg-info mt-1" data-toggle="tooltip" data-placement="top" title="Subdomains">{{scan_history.get_subdomain_count}}</span> <span class="badge badge-pills bg-warning mt-1" data-toggle="tooltip" data-placement="top" title="Endpoints">{{scan_history.get_endpoint_count}}</span> <span class="badge badge-pills bg-danger mt-1" data-toggle="tooltip" data-placement="top" title="{{scan_history.get_critical_vulnerability_count}} Critical, {{scan_history.get_high_vulnerability_count}} High, {{scan_history.get_medium_vulnerability_count}} Medium Vulnerabilities">{{scan_history.get_vulnerability_count}}</span> </td> <td class=""> <span class="badge badge-soft-primary">{{ scan_history.scan_type }}</span> </td> <td> <span data-toggle="tooltip" data-placement="top" title="{{scan_history.start_scan_date}}">{{scan_history.start_scan_date|naturaltime}}</span> </td> <td class="text-center"> {% if scan_history.scan_status == -1 %} <span class="badge badge-soft-warning" data-placement="top" data-toggle="tooltip" data-placement="top" title="Waiting for other scans to complete"><span class="spinner-border spinner-border-sm"></span> Pending</span> {% elif scan_history.scan_status == 0 %} <span class="badge badge-soft-danger">Failed</span> {% if scan_history.error_message %}</br><p class="text-danger">Scan Failed due to: {{scan_history.error_message}}</p>{% endif %} {% elif scan_history.scan_status == 1 %} <span class="badge badge-soft-info"><span class="spinner-border spinner-border-sm"></span> In Progress</span> {% elif scan_history.scan_status == 2 %} <span class="badge badge-soft-success">Successful</span> {% elif scan_history.scan_status == 3 %} <span class="badge badge-soft-danger">Aborted</span> {% else %} <span class="badge badge-soft-danger">Unknown</span> {% endif %} </td> <td class="text-center"> {% if scan_history.scan_status == -1 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-warning" role="progressbar" style="width: 75%" aria-valuenow="75" aria-valuemin="0" aria-valuemax="100"></div> </div> {% elif scan_history.scan_status == 0 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% elif scan_history.scan_status == 1 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-primary progress-bar-striped progress-bar-animated" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% elif scan_history.scan_status == 2 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-success" role="progressbar" style="width: 100%" aria-valuemin="0" aria-valuemax="100"></div> </div> {% elif scan_history.scan_status == 3 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger progress-bar-striped" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% else %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger" role="progressbar" style="width: 100%" aria-valuemin="0" aria-valuemax="100"> </div> </div> {% endif %} </td> <td class="text-center"> <div class="btn-group mb-2 dropstart"> <div class="btn-group"> <a href="{% url 'detail_scan' scan_history.id %}" class="btn btn-soft-primary">View Results</a> <div class="btn-group dropstart" role="group"> <button type="button" class="btn btn-soft-primary dropdown-toggle dropdown-toggle-split" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false"> <i class="mdi mdi-chevron-right"></i> </button> <div class="dropdown-menu" style=""> {% if scan_history.scan_status == 0 or scan_history.scan_status == 2 or scan_history.scan_status == 3 %} <a class="dropdown-item text-primary" href="{% url 'start_scan' scan_history.domain.id %}"> <i class="fe-refresh-ccw"></i>&nbsp;Rescan </a> {% endif %} {% if scan_history.scan_status == 1 or scan_history.scan_status == -1%} <a href="#" class="dropdown-item text-danger" onclick="stop_scan(scan_id={{ scan_history.id }}, subscan_id=null, reload_scan_bar=false, reload_location=true)"> <i class="fe-alert-triangle"></i>&nbsp;Stop Scan</a> {% endif %} {% if scan_history.scan_status == 2 or scan_history.scan_status == 3 %} <a href="#" class="dropdown-item text-danger" onclick="delete_scan('{{ scan_history.id }}')"> <i class="fe-trash-2"></i>&nbsp;Delete Scan Results</a> {% endif %} {% if scan.scan_status != -1%} <div class="dropdown-divider"></div> <a href="#" class="dropdown-item text-dark" onclick="initiate_report({{scan_history.id}}, '{{scan_history.subdomain_discovery}}', '{{scan_history.vulnerability_scan}}', '{{ scan_history.domain.name }}')"> <i class="fe-download"></i>&nbsp;Download Report</a> {% endif %} </div> </div> </div> </div> </td> </tr> {% endfor %} </tbody> </table> </form> </div> </div> </div> <div class="modal fade" id="generateReportModal" tabindex="-1" style="display: none;" aria-hidden="true"> <div class="modal-dialog modal-dialog-centered"> <div class="modal-content"> <div class="modal-header"> <h4 class="modal-title" id="myCenterModalLabel">Download Report</h4> <button type="button" class="btn-close" data-bs-dismiss="modal" aria-label="Close"></button> </div> <div class="modal-body"> <div class="alert alert-light-primary border-0 mb-4" role="alert"> <div id='report_alert_message'></div> </div> <div class="form-group mb-4"> <label for="reportTypeForm">Report Type</label> <select class="form-control" id="report_type_select" name="report_type"> </select> </div> <a id='generateReportButton' href="#" class="btn btn-primary float-end">Download Report</a> </div> </div> </div> </div> {% endblock main_content %} {% block page_level_script %} <script src="{% static 'plugins/datatable/datatables.js' %}"></script> <script> $(document).ready(function() { var table = $('#scan_history_table').DataTable({ headerCallback: function(e, a, t, n, s) { e.getElementsByTagName("th")[0].innerHTML='<div class="form-check mb-2 form-check-primary"><input type="checkbox" class="float-start form-check-input chk-parent" id="head_checkbox" onclick=mainCheckBoxSelected(this)>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>\n' }, "columnDefs":[ { 'visible': false, 'targets': [1] }, { "targets":0, "width":"20px", "className":"", "orderable":!1, render:function(e, a, t, n) { return'<div class="form-check mb-2 form-check-primary"><input type="checkbox" name="targets_checkbox['+ e + ']" class="float-start form-check-input targets_checkbox" value="' + e + '" onchange=toggleMultipleTargetButton()>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>' }, }], "order": [[1, 'desc']], "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center mt-sm-0 mt-3'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "stripeClasses": [], "lengthMenu": [5, 10, 20, 30, 40, 50], "pageLength": 20, "initComplete": function(settings, json) { $('[data-toggle="tooltip"]').tooltip(); table = settings.oInstance.api(); var rows = table.rows({ selected: true }).indexes(); // populate filter menu from datatables // populate targets var selectedData = table.cells(rows, 2).data(); var target_array = []; for (var i = 0; i < selectedData.length; i++) { col1_data = selectedData[i]; domain_name = col1_data.match(/([^\n]+)/g)[0]; target_array.push(domain_name); } target_array = Array.from(new Set(target_array)); for (target in target_array) { select = document.getElementById('filterByTarget'); var option = document.createElement('option'); option.value = target_array[target]; option.innerHTML = target_array[target]; select.appendChild(option); } // populate Scan Type var selectedData = table.cells(rows, 4).data(); var scan_type_array = []; for (var i = 0; i < selectedData.length; i++) { col1_data = extractContent(selectedData[i]); scan_type_array.push(col1_data); } scan_type_array = Array.from(new Set(scan_type_array)); for (engine in scan_type_array) { select = document.getElementById('filterByScanType'); var option = document.createElement('option'); option.value = scan_type_array[engine]; option.innerHTML = scan_type_array[engine]; select.appendChild(option); } } }); multiCheck(table); // filter organization populate $.getJSON(`/api/listOrganizations?&format=json`, function(data) { data = data['organizations'] for (organization in data) { name = htmlEncode(data[organization]['name']); select = document.getElementById('filterByOrganization'); var option = document.createElement('option'); option.value = name; option.innerHTML = name; select.appendChild(option); } }).fail(function() {}); // filtering for scan status var status_types = ['Pending', 'Scanning', 'Aborted', 'Successful', 'Failed']; for (status in status_types) { select = document.getElementById('filterByScanStatus'); var option = document.createElement('option'); option.value = status_types[status]; option.innerHTML = status_types[status]; select.appendChild(option); } var org_filter = document.getElementById('filterByOrganization'); org_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Organization: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by organization ${this.value}`, pos: 'top-center' }); }, false); var status_filter = document.getElementById('filterByScanStatus'); status_filter.addEventListener('click', function() { table.search(this.value).draw(); switch (this.value) { case 'Pending': badge_color = 'warning'; break; case 'Scanning': badge_color = 'info'; break; case 'Aborted': badge_color = 'danger'; break; case 'Failed': badge_color = 'danger'; break; case 'Successful': badge_color = 'success'; break; default: badge_color = 'primary' } document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-${badge_color}">Scan Status: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by scan status ${this.value}`, pos: 'top-center' }); }, false); var engine_filter = document.getElementById('filterByScanType'); engine_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Scan Engine: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by Engine ${this.value}`, pos: 'top-center' }); }, false); var target_filter = document.getElementById('filterByTarget'); target_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Target/Domain: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by Engine ${this.value}`, pos: 'top-center' }); }, false); // reset filtering var reset_filter = document.getElementById('resetFilters'); reset_filter.addEventListener('click', function() { resetFilters(table); }, false); }); function resetFilters(table_obj) { table_obj.search("").draw(); Snackbar.show({ text: `Filters Reset`, pos: 'top-center' }); document.getElementById('filteringText').innerHTML = ''; } function checkedCount() { // this function will count the number of boxes checked item = document.getElementsByClassName("targets_checkbox"); count = 0; for (var i = 0; i < item.length; i++) { if (item[i].checked) { count++; } } return count; } function toggleMultipleTargetButton() { if (checkedCount() > 0) { $("#delete_multiple_button").removeClass("disabled"); } else { $("#delete_multiple_button").addClass("disabled"); } } function mainCheckBoxSelected(checkbox) { if (checkbox.checked) { $("#delete_multiple_button").removeClass("disabled"); $(".targets_checkbox").prop('checked', true); } else { $("#delete_multiple_button").addClass("disabled"); $(".targets_checkbox").prop('checked', false); } } function deleteMultipleScan() { if (!checkedCount()) { swal({ title: '', text: "Oops! No targets has been selected!", type: 'error', padding: '2em' }) } else { // atleast one target is selected swal.queue([{ title: 'Are you sure you want to delete ' + checkedCount() + ' Scans?', text: "This action is irreversible.\nThis will delete all the scan data and vulnerabilities related to the scan.", type: 'warning', showCancelButton: true, confirmButtonText: 'Delete', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { deleteForm = document.getElementById("scan_history_form"); deleteForm.action = "../delete/multiple"; deleteForm.submit(); } }]) } } function initiate_report(id, is_subdomain_scan, is_vulnerability_scan, domain_name) { $('#generateReportModal').modal('show'); $('#report_alert_message').empty(); $('#report_type_select').empty(); if (is_subdomain_scan == 'True' && is_vulnerability_scan == 'True') { $('#report_alert_message').append(` <b>Full Scan</b> will include both Reconnaissance and Vulnerability Report.<br> `); $('#report_type_select').append($('<option>', { value: 'full', text: 'Full Scan Report' })); } if (is_subdomain_scan == 'True') { // eligible for reconnaissance report $('#report_alert_message').append(` <b>Reconnaissance Report</b> will only include Assets Discovered Section.<br> `); $('#report_type_select').append($('<option>', { value: 'recon', text: 'Reconnaissance Report' })); } if (is_vulnerability_scan == 'True'){ // eligible for vulnerability report $('#report_alert_message').append(` <b>Vulnerability Report</b> will only include details of Vulnerabilities Identified. `); $('#report_type_select').append($('<option>', { value: 'vulnerability', text: 'Vulnerability Report' })); } $('#generateReportButton').attr('onClick', `generate_report(${id}, '${domain_name}')`); } function generate_report(id, domain_name) { var report_type = $("#report_type_select option:selected").val(); $('#generateReportModal').modal('hide'); swal.queue([{ title: 'Generating Report!', text: `Please wait until we generate a report for you!`, padding: '2em', onOpen: function() { swal.showLoading() return fetch(`/scan/create_report/${id}?download&report_type=${report_type}`, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }) .then(function(response) { return response.blob(); }).then(function(blob) { const file = new Blob([blob], {type: 'application/pdf'}); // process to auto download it const fileURL = URL.createObjectURL(file); const link = document.createElement('a'); link.href = fileURL; link.download = domain_name + ".pdf"; link.click(); swal.close(); }) .catch(function() { swal.insertQueueStep({ type: 'error', title: 'Oops! Unable to generate report!' }) }) } }]); } </script> {% endblock page_level_script %}

{% extends 'base/base.html' %} {% load static %} {% load humanize %} {% load permission_tags %} {% block title %} Scan history {% endblock title %} {% block custom_js_css_link %} {% endblock custom_js_css_link %} {% block breadcrumb_title %} <li class="breadcrumb-item active" aria-current="page">Scan History</li> {% endblock breadcrumb_title %} {% block page_title %} Quick Scan History {% endblock page_title %} {% block main_content %} <div class="row"> <div class="col-12"> <div class="card"> <div class="p-2"> <div class="row"> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <button type="button" class="btn btn-primary dropdown-toggle" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false" id="filterMenu"> Filter <i class="fe-filter"></i> </button> <div id="filteringText" class="mt-2"> </div> <div class="dropdown-menu" style="width: 30%"> <div class="px-4 py-3"> <h4 class="headline-title">Filters</h4> <div class=""> <label for="filterByOrganization" class="form-label">Filter by Organization</label> <select class="form-control" id="filterByOrganization"> </select> </div> <div class=""> <label for="filterByTarget" class="form-label">Filter by Targets</label> <select class="form-control" id="filterByTarget"> </select> </div> <div class=""> <label for="filterByScanType" class="form-label">Filter by Scan Type</label> <select class="form-control" id="filterByScanType"> </select> </div> <div class=""> <label for="filterByScanStatus" class="form-label">Filter by Scan Status</label> <select class="form-control" id="filterByScanStatus"> </select> </div> </div> <div class="dropdown-divider"></div> <a href="#" class="dropdown-ite text-primary float-end" id="resetFilters">Reset Filters</a> </div> </div> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <a class="btn btn-soft-danger float-end disabled ms-1" href="#" onclick="deleteMultipleScan()" id="delete_multiple_button">Delete Multiple Scans</a> </div> </div> </div> </div> </div> </div> <div class="row"> <div class="col-12"> <div class="card"> <form method="POST" id="scan_history_form"> {% csrf_token %} <table id="scan_history_table" class="table dt-responsive w-100"> <thead> <tr> <th class="checkbox-column text-center">Serial Number</th> <th class="text-center">Serial Number</th> <th class="">Domain Name</th> <th>Summary</th> <th class="">Scan Engine Used</th> <th>Last Scan</th> <th class="text-center">Status</th> <th class="text-center">Progress</th> <th class="text-center no-sorting">Action</th> </tr> </thead> <tbody> {% for scan_history in scan_history.all %} <tr> <td class="checkbox-column"> {{ scan_history.id }} </td> <td class=""> {{ scan_history.id }} </td> <td class=""> {{ scan_history.domain.name }} <br> {% for organization in scan_history.domain.get_organization %} <span class="badge badge-soft-dark mt-1 me-1" data-toggle="tooltip" data-placement="top" title="Domain {{domain.name}} belongs to organization {{organization.name}}">{{ organization.name }}</span> {% endfor %} </td> <td class="text-left"> <span class="badge badge-pills bg-info mt-1" data-toggle="tooltip" data-placement="top" title="Subdomains">{{scan_history.get_subdomain_count}}</span> <span class="badge badge-pills bg-warning mt-1" data-toggle="tooltip" data-placement="top" title="Endpoints">{{scan_history.get_endpoint_count}}</span> <span class="badge badge-pills bg-danger mt-1" data-toggle="tooltip" data-placement="top" title="{{scan_history.get_critical_vulnerability_count}} Critical, {{scan_history.get_high_vulnerability_count}} High, {{scan_history.get_medium_vulnerability_count}} Medium Vulnerabilities">{{scan_history.get_vulnerability_count}}</span> </td> <td class=""> <span class="badge badge-soft-primary">{{ scan_history.scan_type }}</span> </td> <td> <span data-toggle="tooltip" data-placement="top" title="{{scan_history.start_scan_date}}">{{scan_history.start_scan_date|naturaltime}}</span> </td> <td class="text-center"> {% if scan_history.scan_status == -1 %} <span class="badge badge-soft-warning" data-placement="top" data-toggle="tooltip" data-placement="top" title="Waiting for other scans to complete"><span class="spinner-border spinner-border-sm"></span> Pending</span> {% elif scan_history.scan_status == 0 %} <span class="badge badge-soft-danger">Failed</span> {% if scan_history.error_message %}</br><p class="text-danger">Scan Failed due to: {{scan_history.error_message}}</p>{% endif %} {% elif scan_history.scan_status == 1 %} <span class="badge badge-soft-info"><span class="spinner-border spinner-border-sm"></span> In Progress</span> {% elif scan_history.scan_status == 2 %} <span class="badge badge-soft-success">Successful</span> {% elif scan_history.scan_status == 3 %} <span class="badge badge-soft-danger">Aborted</span> {% else %} <span class="badge badge-soft-danger">Unknown</span> {% endif %} </td> <td class="text-center"> {% if scan_history.scan_status == -1 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-warning" role="progressbar" style="width: 75%" aria-valuenow="75" aria-valuemin="0" aria-valuemax="100"></div> </div> {% elif scan_history.scan_status == 0 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% elif scan_history.scan_status == 1 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-primary progress-bar-striped progress-bar-animated" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% elif scan_history.scan_status == 2 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-success" role="progressbar" style="width: 100%" aria-valuemin="0" aria-valuemax="100"></div> </div> {% elif scan_history.scan_status == 3 %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger progress-bar-striped" role="progressbar" style="width: {% widthratio scan_history.scanactivity_set.all|length scan_history.scan_type.get_number_of_steps|add:4 100 %}%" aria-valuemin="0" aria-valuemax="4"></div> </div> {% else %} <div class="progress progress-md mt-1"> <div class="progress-bar bg-danger" role="progressbar" style="width: 100%" aria-valuemin="0" aria-valuemax="100"> </div> </div> {% endif %} </td> <td class="text-center"> <div class="btn-group mb-2 dropstart"> <div class="btn-group"> <a href="/scan/{{current_project.slug}}/detail/{{scan_history.id}}" class="btn btn-soft-primary">View Results</a> <div class="btn-group dropstart" role="group"> <button type="button" class="btn btn-soft-primary dropdown-toggle dropdown-toggle-split" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false"> <i class="mdi mdi-chevron-right"></i> </button> <div class="dropdown-menu" style=""> {% if user|can:'initiate_scans_subscans' %} {% if scan_history.scan_status == 0 or scan_history.scan_status == 2 or scan_history.scan_status == 3 %} <a class="dropdown-item text-primary" href="/scan/{{current_project.slug}}/start/{{scan_history.domain.id}}"> <i class="fe-refresh-ccw"></i>&nbsp;Rescan </a> {% endif %} {% if scan_history.scan_status == 1 or scan_history.scan_status == -1%} <a href="#" class="dropdown-item text-danger" onclick="stop_scan(scan_id={{ scan_history.id }}, subscan_id=null, reload_scan_bar=false, reload_location=true)"> <i class="fe-alert-triangle"></i>&nbsp;Stop Scan</a> {% endif %} {% endif %} {% if user|can:'modify_scan_results' %} {% if scan_history.scan_status == 2 or scan_history.scan_status == 3 or scan_history.scan_status == 0 %} <a href="#" class="dropdown-item text-danger" onclick="delete_scan('{{ scan_history.id }}')"> <i class="fe-trash-2"></i>&nbsp;Delete Scan Results</a> {% endif %} <div class="dropdown-divider"></div> {% endif %} {% if scan.scan_status != -1%} <a href="#" class="dropdown-item text-dark" onclick="initiate_report({{scan_history.id}}, '{% if 'subdomain_discovery' in scan_history.scan_type.tasks %}True{% endif %}', '{% if 'vulnerability_scan' in scan_history.scan_type.tasks %}True{% endif %}', '{{ scan_history.domain.name }}')"> <i class="fe-download"></i>&nbsp;Scan Report</a> {% endif %} </div> </div> </div> </div> </td> </tr> {% endfor %} </tbody> </table> </form> </div> </div> </div> <div class="modal fade" id="generateReportModal" tabindex="-1" style="display: none;" aria-hidden="true"> <div class="modal-dialog modal-dialog-centered"> <div class="modal-content"> <div class="modal-header"> <h4 class="modal-title" id="myCenterModalLabel">Download Report</h4> <button type="button" class="btn-close" data-bs-dismiss="modal" aria-label="Close"></button> </div> <div class="modal-body"> <div class="alert alert-light-primary border-0 mb-4" role="alert"> <div id='report_alert_message'></div> </div> <div class="form-group mb-4"> <label for="reportTypeForm">Report Type</label> <select class="form-control" id="report_type_select" name="report_type"> </select> </div> <div class="form-group mb-4"> <div class="form-check" id="report_info_vuln_div"> <input type="checkbox" class="form-check-input" id="report_ignore_info_vuln" checked=""> <label class="form-check-label" for="report_ignore_info_vuln">Ignore Informational Vulnerabilities</label> </div> </div> <a id='generateReportButton' href="#" class="btn btn-primary float-end m-2">Download Report</a> <a id='previewReportButton' href="#" class="btn btn-secondary float-end m-2">Preview Report</a> </div> </div> </div> </div> {% endblock main_content %} {% block page_level_script %} <script> $(document).ready(function() { var table = $('#scan_history_table').DataTable({ headerCallback: function(e, a, t, n, s) { e.getElementsByTagName("th")[0].innerHTML='<div class="form-check mb-2 form-check-primary"><input type="checkbox" class="float-start form-check-input chk-parent" id="head_checkbox" onclick=mainCheckBoxSelected(this)>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>\n' }, "columnDefs":[ { 'visible': false, 'targets': [1] }, { "targets":0, "width":"20px", "className":"", "orderable":!1, render:function(e, a, t, n) { return'<div class="form-check mb-2 form-check-primary"><input type="checkbox" name="targets_checkbox['+ e + ']" class="float-start form-check-input targets_checkbox" value="' + e + '" onchange=toggleMultipleTargetButton()>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>' }, } ], "order": [[1, 'desc']], "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center mt-sm-0 mt-3'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "stripeClasses": [], "lengthMenu": [5, 10, 20, 30, 40, 50], "pageLength": 20, "initComplete": function(settings, json) { $('[data-toggle="tooltip"]').tooltip(); table = settings.oInstance.api(); var rows = table.rows({ selected: true }).indexes(); // populate filter menu from datatables // populate targets var selectedData = table.cells(rows, 2).data(); var target_array = []; for (var i = 0; i < selectedData.length; i++) { col1_data = selectedData[i]; domain_name = col1_data.match(/([^\n]+)/g)[0]; target_array.push(domain_name); } target_array = Array.from(new Set(target_array)); for (target in target_array) { select = document.getElementById('filterByTarget'); var option = document.createElement('option'); option.value = target_array[target]; option.innerHTML = target_array[target]; select.appendChild(option); } // populate Scan Type var selectedData = table.cells(rows, 4).data(); var scan_type_array = []; for (var i = 0; i < selectedData.length; i++) { col1_data = extractContent(selectedData[i]); scan_type_array.push(col1_data); } scan_type_array = Array.from(new Set(scan_type_array)); for (engine in scan_type_array) { select = document.getElementById('filterByScanType'); var option = document.createElement('option'); option.value = scan_type_array[engine]; option.innerHTML = scan_type_array[engine]; select.appendChild(option); } } }); multiCheck(table); // filter organization populate $.getJSON(`/api/listOrganizations?&format=json`, function(data) { data = data['organizations'] for (organization in data) { name = htmlEncode(data[organization]['name']); select = document.getElementById('filterByOrganization'); var option = document.createElement('option'); option.value = name; option.innerHTML = name; select.appendChild(option); } }).fail(function() {}); // filtering for scan status var status_types = ['Pending', 'Scanning', 'Aborted', 'Successful', 'Failed']; for (status in status_types) { select = document.getElementById('filterByScanStatus'); var option = document.createElement('option'); option.value = status_types[status]; option.innerHTML = status_types[status]; select.appendChild(option); } var org_filter = document.getElementById('filterByOrganization'); org_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Organization: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by organization ${this.value}`, pos: 'top-center' }); }, false); var status_filter = document.getElementById('filterByScanStatus'); status_filter.addEventListener('click', function() { table.search(this.value).draw(); switch (this.value) { case 'Pending': badge_color = 'warning'; break; case 'Scanning': badge_color = 'info'; break; case 'Aborted': badge_color = 'danger'; break; case 'Failed': badge_color = 'danger'; break; case 'Successful': badge_color = 'success'; break; default: badge_color = 'primary' } document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-${badge_color}">Scan Status: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by scan status ${this.value}`, pos: 'top-center' }); }, false); var engine_filter = document.getElementById('filterByScanType'); engine_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Scan Engine: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by Engine ${this.value}`, pos: 'top-center' }); }, false); var target_filter = document.getElementById('filterByTarget'); target_filter.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-soft-primary">Target/Domain: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by Engine ${this.value}`, pos: 'top-center' }); }, false); // reset filtering var reset_filter = document.getElementById('resetFilters'); reset_filter.addEventListener('click', function() { resetFilters(table); }, false); }); function resetFilters(table_obj) { table_obj.search("").draw(); Snackbar.show({ text: `Filters Reset`, pos: 'top-center' }); document.getElementById('filteringText').innerHTML = ''; } function checkedCount() { // this function will count the number of boxes checked item = document.getElementsByClassName("targets_checkbox"); count = 0; for (var i = 0; i < item.length; i++) { if (item[i].checked) { count++; } } return count; } function toggleMultipleTargetButton() { if (checkedCount() > 0) { $("#delete_multiple_button").removeClass("disabled"); } else { $("#delete_multiple_button").addClass("disabled"); } } function mainCheckBoxSelected(checkbox) { if (checkbox.checked) { $("#delete_multiple_button").removeClass("disabled"); $(".targets_checkbox").prop('checked', true); } else { $("#delete_multiple_button").addClass("disabled"); $(".targets_checkbox").prop('checked', false); } } function deleteMultipleScan() { if (!checkedCount()) { swal({ title: '', text: "Oops! No targets has been selected!", type: 'error', padding: '2em' }) } else { // atleast one target is selected swal.queue([{ title: 'Are you sure you want to delete ' + checkedCount() + ' Scans?', text: "This action is irreversible.\nThis will delete all the scan data and vulnerabilities related to the scan.", type: 'warning', showCancelButton: true, confirmButtonText: 'Delete', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { deleteForm = document.getElementById("scan_history_form"); deleteForm.action = "../delete/multiple"; deleteForm.submit(); } }]) } } // select option listener for report_type_select var report_type = document.getElementById("report_type_select"); report_type.addEventListener("change", function() { if(report_type.value == "recon") { $("#report_info_vuln_div").hide(); } else{ $("#report_info_vuln_div").show(); } }); function initiate_report(id, is_subdomain_scan, is_vulnerability_scan, domain_name) { $('#generateReportModal').modal('show'); $('#report_alert_message').empty(); $('#report_type_select').empty(); if (is_subdomain_scan == 'True' && is_vulnerability_scan == 'True') { $('#report_alert_message').append(` <b>Full Scan</b> will include both Reconnaissance and Vulnerability Report.<br> `); $('#report_type_select').append($('<option>', { value: 'full', text: 'Full Scan Report' })); } if (is_subdomain_scan == 'True') { // eligible for reconnaissance report $('#report_alert_message').append(` <b>Reconnaissance Report</b> will only include Assets Discovered Section.<br> `); $('#report_type_select').append($('<option>', { value: 'recon', text: 'Reconnaissance Report' })); } if (is_vulnerability_scan == 'True'){ // eligible for vulnerability report $('#report_alert_message').append(` <b>Vulnerability Report</b> will only include details of Vulnerabilities Identified. `); $('#report_type_select').append($('<option>', { value: 'vulnerability', text: 'Vulnerability Report' })); } $('#generateReportButton').attr('onClick', `generate_report(${id}, '${domain_name}')`); $('#previewReportButton').attr('onClick', `preview_report(${id}, '${domain_name}')`); } function preview_report(id, domain_name){ var report_type = $("#report_type_select option:selected").val(); var url = `/scan/create_report/${id}?report_type=${report_type}`; if ($('#report_ignore_info_vuln').is(":checked")) { url += `&ignore_info_vuln` } $('#generateReportModal').modal('hide'); window.open(url, '_blank').focus(); } function generate_report(id, domain_name) { var report_type = $("#report_type_select option:selected").val(); var url = `/scan/create_report/${id}?report_type=${report_type}&download`; if ($('#report_ignore_info_vuln').is(":checked")) { url += `&ignore_info_vuln` } $('#generateReportModal').modal('hide'); swal.queue([{ title: 'Generating Report!', text: `Please wait until we generate a report for you!`, padding: '2em', onOpen: function() { swal.showLoading() return fetch(url, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }) .then(function(response) { return response.blob(); }).then(function(blob) { const file = new Blob([blob], {type: 'application/pdf'}); // process to auto download it const fileURL = URL.createObjectURL(file); const link = document.createElement('a'); link.href = fileURL; link.download = domain_name + ".pdf"; link.click(); swal.close(); }) .catch(function() { swal.insertQueueStep({ type: 'error', title: 'Oops! Unable to generate report!' }) }) } }]); } </script> {% endblock page_level_script %}

yogeshojha

3c60bc1ee495044794d91edee0c96fff73ab46c7

5413708d243799a5271440c47c6f98d0c51154ca

## DOM text reinterpreted as HTML [DOM text](1) is reinterpreted as HTML without escaping meta-characters. [Show more details](https://github.com/yogeshojha/rengine/security/code-scanning/156)

github-advanced-security[bot]

yogeshojha/rengine

Fixes required to get install script working

Update celery version to 5.2.7 in requirements.txt Update go version to 1.20 in Dockerfile These changes are required for install.sh to complete on Ubuntu 22.04.1 LTS (GNU/Linux 5.15.0-66-generic x86_64) (digital ocean droplet). Note: Saw some warning/maybe error related to whatisport (or similar) pypi package. Will investigate at some point.

2023-02-12 05:24:03+00:00

2023-03-02 17:25:01+00:00

web/Dockerfile

# Base image FROM ubuntu:20.04 # Labels and Credits LABEL \ name="reNgine" \ author="Yogesh Ojha <yogesh.ojha11@gmail.com>" \ description="reNgine is a automated pipeline of recon process, useful for information gathering during web application penetration testing." # Environment Variables ENV DEBIAN_FRONTEND="noninteractive" \ DATABASE="postgres" ENV PYTHONDONTWRITEBYTECODE 1 ENV PYTHONUNBUFFERED 1 # Install essentials RUN apt update -y && apt install -y --no-install-recommends \ build-essential \ cmake \ firefox \ gcc \ git \ libpq-dev \ libpq-dev \ libpcap-dev \ netcat \ postgresql \ python3 \ python3-dev \ python3-pip \ python3-netaddr \ wget \ x11-utils \ xvfb \ python3-cffi \ python3-brotli \ libpango-1.0-0 \ libpangoft2-1.0-0 \ geoip-bin \ geoip-database # Download and install go 1.18 RUN wget https://golang.org/dl/go1.18.2.linux-amd64.tar.gz RUN tar -xvf go1.18.2.linux-amd64.tar.gz RUN rm go1.18.2.linux-amd64.tar.gz RUN mv go /usr/local # Download geckodriver RUN wget https://github.com/mozilla/geckodriver/releases/download/v0.26.0/geckodriver-v0.26.0-linux64.tar.gz RUN tar -xvf geckodriver-v0.26.0-linux64.tar.gz RUN rm geckodriver-v0.26.0-linux64.tar.gz RUN mv geckodriver /usr/bin # ENV for Go ENV GOROOT="/usr/local/go" ENV PATH="${PATH}:${GOROOT}/bin" ENV PATH="${PATH}:${GOPATH}/bin" ENV GOPATH=$HOME/go ENV PATH="${PATH}:${GOROOT}/bin:${GOPATH}/bin" # Make directory for app WORKDIR /usr/src/app # set environment variables ENV PYTHONDONTWRITEBYTECODE 1 ENV PYTHONUNBUFFERED 1 # Download Go packages RUN go install -v github.com/hakluke/hakrawler@latest RUN GO111MODULE=on go install -v -v github.com/bp0lr/gauplus@latest RUN GO111MODULE=on go install -v github.com/jaeles-project/gospider@latest RUN go install -v github.com/OWASP/Amass/v3/...@latest RUN go install -v github.com/ffuf/ffuf@latest RUN go install -v github.com/tomnomnom/assetfinder@latest RUN GO111MODULE=on go install -v github.com/tomnomnom/gf@latest RUN GO111MODULE=on go install -v github.com/tomnomnom/unfurl@latest RUN GO111MODULE=on go install -v github.com/tomnomnom/waybackurls@latest RUN GO111MODULE=on go install -v github.com/projectdiscovery/httpx/cmd/httpx@latest RUN GO111MODULE=on go install -v github.com/projectdiscovery/subfinder/v2/cmd/subfinder@latest RUN GO111MODULE=on go install -v github.com/projectdiscovery/nuclei/v2/cmd/nuclei@latest RUN GO111MODULE=on go install -v github.com/projectdiscovery/naabu/v2/cmd/naabu@latest # Update Nuclei and Nuclei-Templates RUN nuclei -update RUN nuclei -update-templates # Copy requirements COPY ./requirements.txt /tmp/requirements.txt RUN pip3 install --upgrade setuptools pip && \ pip3 install -r /tmp/requirements.txt # install eyewitness RUN python3 -m pip install fuzzywuzzy \ selenium \ python-Levenshtein \ pyvirtualdisplay \ netaddr # Copy source code COPY . /usr/src/app/ # httpx seems to have issue, use alias instead!!! RUN echo 'alias httpx="/go/bin/httpx"' >> ~/.bashrc

# Base image FROM ubuntu:20.04 # Labels and Credits LABEL \ name="reNgine" \ author="Yogesh Ojha <yogesh.ojha11@gmail.com>" \ description="reNgine is a automated pipeline of recon process, useful for information gathering during web application penetration testing." # Environment Variables ENV DEBIAN_FRONTEND="noninteractive" \ DATABASE="postgres" ENV PYTHONDONTWRITEBYTECODE 1 ENV PYTHONUNBUFFERED 1 # Install essentials RUN apt update -y && apt install -y --no-install-recommends \ build-essential \ cmake \ firefox \ gcc \ git \ libpq-dev \ libpq-dev \ libpcap-dev \ netcat \ postgresql \ python3 \ python3-dev \ python3-pip \ python3-netaddr \ wget \ x11-utils \ xvfb \ python3-cffi \ python3-brotli \ libpango-1.0-0 \ libpangoft2-1.0-0 \ geoip-bin \ geoip-database # Download and install go 1.18 RUN wget https://golang.org/dl/go1.20.linux-amd64.tar.gz RUN tar -xvf go1.20.linux-amd64.tar.gz RUN rm go1.20.linux-amd64.tar.gz RUN mv go /usr/local # Download geckodriver RUN wget https://github.com/mozilla/geckodriver/releases/download/v0.26.0/geckodriver-v0.26.0-linux64.tar.gz RUN tar -xvf geckodriver-v0.26.0-linux64.tar.gz RUN rm geckodriver-v0.26.0-linux64.tar.gz RUN mv geckodriver /usr/bin # ENV for Go ENV GOROOT="/usr/local/go" ENV PATH="${PATH}:${GOROOT}/bin" ENV PATH="${PATH}:${GOPATH}/bin" ENV GOPATH=$HOME/go ENV PATH="${PATH}:${GOROOT}/bin:${GOPATH}/bin" # Make directory for app WORKDIR /usr/src/app # set environment variables ENV PYTHONDONTWRITEBYTECODE 1 ENV PYTHONUNBUFFERED 1 # Download Go packages RUN go install -v github.com/hakluke/hakrawler@latest RUN GO111MODULE=on go install -v -v github.com/bp0lr/gauplus@latest RUN GO111MODULE=on go install -v github.com/jaeles-project/gospider@latest RUN go install -v github.com/OWASP/Amass/v3/...@latest RUN go install -v github.com/ffuf/ffuf@latest RUN go install -v github.com/tomnomnom/assetfinder@latest RUN GO111MODULE=on go install -v github.com/tomnomnom/gf@latest RUN GO111MODULE=on go install -v github.com/tomnomnom/unfurl@latest RUN GO111MODULE=on go install -v github.com/tomnomnom/waybackurls@latest RUN GO111MODULE=on go install -v github.com/projectdiscovery/httpx/cmd/httpx@latest RUN GO111MODULE=on go install -v github.com/projectdiscovery/subfinder/v2/cmd/subfinder@latest RUN GO111MODULE=on go install -v github.com/projectdiscovery/nuclei/v2/cmd/nuclei@latest RUN GO111MODULE=on go install -v github.com/projectdiscovery/naabu/v2/cmd/naabu@latest # Update Nuclei and Nuclei-Templates RUN nuclei -update RUN nuclei -update-templates # Copy requirements COPY ./requirements.txt /tmp/requirements.txt RUN pip3 install --upgrade setuptools pip && \ pip3 install -r /tmp/requirements.txt # install eyewitness RUN python3 -m pip install fuzzywuzzy \ selenium \ python-Levenshtein \ pyvirtualdisplay \ netaddr # Copy source code COPY . /usr/src/app/ # httpx seems to have issue, use alias instead!!! RUN echo 'alias httpx="/go/bin/httpx"' >> ~/.bashrc

m00tiny

5e3c04c336d3798fbff20f362a8091dface53203

a0ce6a270a30e8e3c224b6141555a530b9b6c50e

- RUN wget https://go.dev/dl/go1.20.1.linux-amd64.tar.gz

cybersaki

yogeshojha/rengine

Release/1.3.1

# Fixes - Fix for #643 Downloading issue for Subdomain and Endpoints - Fix for #627 Too many Targets causes issues while loading datatable - Fix version Numbering issue

2022-08-12 12:46:29+00:00

2022-08-12 12:57:24+00:00

web/targetApp/templates/target/list.html

{% extends 'base/base.html' %} {% load static %} {% load humanize %} {% block title %} List all targets {% endblock title %} {% block custom_js_css_link %} <link rel="stylesheet" type="text/css" href="{% static 'plugins/datatable/datatables.css' %}"> <link rel="stylesheet" type="text/css" href="{% static 'plugins/datatable/global.css' %}"> <link rel="stylesheet" type="text/css" href="{% static 'plugins/datatable/custom.css' %}"> {% endblock custom_js_css_link %} {% block breadcrumb_title %} <li class="breadcrumb-item active"><a href="{% url 'list_target' %}">Targets</a></li> <li class="breadcrumb-item"><a href="#">All Targets</a></li> {% endblock breadcrumb_title %} {% block page_title %} Targets {% endblock page_title %} {% block main_content %} <div class="row"> <div class="col-12"> <div class="card"> <div class="p-2"> <div class="row"> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <button type="button" class="btn btn-primary dropdown-toggle mt-1" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false" id="filterMenu"> Filter <i class="fe-filter"></i> </button> <div class="dropdown-menu" style="width: 30%"> <div class="px-4 py-3"> <h4 class="headline-title">Filters</h4> <label for="filterByOrganization" class="form-label">Filter by Organization</label> <select class="form-control" id="filterByOrganization"> </select> </div> <div class="dropdown-divider"></div> <a href="#" class="dropdown-ite text-primary float-end" id="resetFilters">Reset Filters</a> </div> </div> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <a class="btn btn-soft-danger float-end disabled ms-1 mt-1" href="#" onclick="deleteMultipleTargets()" id="delete_multiple_button">Delete Multiple Targets</a> <a class="btn btn-soft-info float-end disabled ms-1 mt-1" href="#" onclick="scanMultipleTargets()" id="scan_multiple_button">Scan Multiple Targets</a> <a class="btn btn-soft-primary float-end ms-1 mt-1" href="{% url 'add_target' %}" data-toggle="tooltip" data-placement="top" title="Add New Targets">Add Targets</a> </div> </div> </div> </div> </div> </div> <div class="row"> <div class="col-12"> <div class="card"> <form method="POST" id="multiple_targets_form" action="../../scan/start/multiple/"> {% csrf_token %} <table id="list_target_table" class="table style-3 table-hover"> <thead> <tr> <th class="checkbox-column text-center">Serial Number</th> <th>Serial Number</th> <th class="text-center">Domain Name</th> <th>Description</th> <th class="text-center">Last Scanned</th> <th class="text-center">Action</th> </tr> </thead> <tbody> {% for domain in domains.all %} <tr> <td class="checkbox-column"> {{ domain.id }} </td> <td class=""> {{ domain.id }} </td> <td> <b>{{ domain.name }}</b>&nbsp;&nbsp;<a href="#" onclick="get_target_whois('{{domain.name}}')">(view whois)</a> <br> <small class="text-muted">Added {{ domain.insert_date|naturaltime }}</small> {% if domain.get_organization %} <br> {% for organization in domain.get_organization %} <span class="badge badge-soft-primary me-1 mb-1" data-toggle="tooltip" data-placement="top" title="Domain {{domain.name}} belongs to organization {{organization.name}}">{{ organization.name }}</span> {% endfor %} {% endif %} {% if domain.get_recent_scan_id %} <br> <a href="{% url 'detail_scan' domain.get_recent_scan_id %}" class="text-info">Recent Scan&nbsp;<svg xmlns="http://www.w3.org/2000/svg" width="15" height="15" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-external-link"><path d="M18 13v6a2 2 0 0 1-2 2H5a2 2 0 0 1-2-2V8a2 2 0 0 1 2-2h6"></path><polyline points="15 3 21 3 21 9"></polyline><line x1="10" y1="14" x2="21" y2="3"></line></svg></a> {% endif %} </td> <td>{% if domain.description %}{{domain.description}}{% endif %}</td> {% if domain.start_scan_date %} <td class="text-center"><span data-toggle="tooltip" data-placement="top" title="{{domain.start_scan_date}}">{{domain.start_scan_date|naturaltime}}</span></td> {% else %} <td class="text-center"><span class="badge badge-soft-warning">Never Scanned Before</span></td> {% endif %} <td class="text-center"> <div class="btn-group mb-2 dropstart"> <div class="btn-group"> <a class="btn btn-soft-primary" href="{% url 'target_summary' domain.id %}"><i class="fe-info"></i>&nbsp;&nbsp;Target Summary</a> <a href="{% url 'start_scan' domain.id %}" class="btn btn-soft-primary"><i class="fe-zap"></i>&nbsp;&nbsp;Initiate Scan</a> <div class="btn-group dropstart" role="group"> <button type="button" class="btn btn-soft-primary dropdown-toggle dropdown-toggle-split" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false"> <i class="mdi mdi-chevron-right"></i> </button> <div class="dropdown-menu" style=""> <a class="dropdown-item" href="{% url 'schedule_scan' domain.id %}"><i class="fe-clock"></i>&nbsp;&nbsp;Schedule Scan</a> <div class="dropdown-divider"></div> <a class="dropdown-item" href="{% url 'update_target' domain.id %}"><i class="fe-edit-2"></i>&nbsp;&nbsp;Edit Target</a> <a class="dropdown-item text-danger" href="#" onclick="delete_target({{ domain.id }}, '{{ domain.name }}')"><i class="fe-trash-2"></i>&nbsp;&nbsp;Delete target</a> </div> </div> </div> </div> </td> </tr> {% endfor %} </tbody> </table> </form> </div> </div> </div> {% endblock main_content %} {% block page_level_script %} <script src="{% static 'custom/custom.js' %}"></script> <script src="{% static 'targetApp/js/custom_domain.js' %}"></script> <script src="{% static 'plugins/datatable/datatables.js' %}"></script> <script> $(document).ready(function(){ var table = $('#list_target_table').DataTable({ headerCallback:function(e, a, t, n, s) { e.getElementsByTagName("th")[0].innerHTML='<div class="form-check mb-2 form-check-primary"><input type="checkbox" class="float-start form-check-input chk-parent" id="head_checkbox" onclick=mainCheckBoxSelected()>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>\n' }, "columnDefs":[ { 'visible': false, 'targets': [1] }, { "targets":0, "width":"20px", "className":"", "orderable":!1, render:function(e, a, t, n) { return'<div class="form-check mb-2 form-check-primary"><input type="checkbox" name="targets_checkbox['+ e + ']" class="float-start form-check-input targets_checkbox" value="' + e + '" onchange=toggleMultipleTargetButton()>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>' }, }], "order": [[1, 'desc']], "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center mt-sm-0 mt-3'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "stripeClasses": [], "lengthMenu": [5, 10, 20, 30, 40, 50, 100, 500, 1000], "pageLength": 20, }); multiCheck(table); // Handle form submission event $('#frm-example').on('submit', function(e){ var form = this; table.$('input[type="checkbox"]').each(function(){ if(!$.contains(document, this)){ if(this.checked){ $(form).append( $('<input>') .attr('type', 'hidden') .attr('name', this.name) .val(this.value) ); } } }); e.preventDefault(); }); // filter organization populate $.getJSON(`/api/listOrganizations?&format=json`, function(data) { data = data['organizations'] for (organization in data) { name = htmlEncode(data[organization]['name']); select = document.getElementById('filterByOrganization'); var option = document.createElement('option'); option.value = name; option.innerHTML = name; select.appendChild(option); } }).fail(function(){ }); var a = document.getElementById('filterByOrganization'); a.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-primary m-2">Organization: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by organization ${this.value}`, pos: 'top-center' }); }, false); // reset filtering var reset_filter = document.getElementById('resetFilters'); reset_filter.addEventListener('click', function() { resetFilters(table); }, false); $('[data-toggle=tooltip]').tooltip(); }); function resetFilters(table_obj) { table_obj.search("").draw(); Snackbar.show({ text: `Filters Reset`, pos: 'top-center' }); document.getElementById('filteringText').innerHTML = ''; } </script> {% endblock page_level_script %}

{% extends 'base/base.html' %} {% load static %} {% load humanize %} {% block title %} List all targets {% endblock title %} {% block custom_js_css_link %} <link rel="stylesheet" type="text/css" href="{% static 'plugins/datatable/datatables.css' %}"> <link rel="stylesheet" type="text/css" href="{% static 'plugins/datatable/global.css' %}"> <link rel="stylesheet" type="text/css" href="{% static 'plugins/datatable/custom.css' %}"> {% endblock custom_js_css_link %} {% block breadcrumb_title %} <li class="breadcrumb-item active"><a href="{% url 'list_target' %}">Targets</a></li> <li class="breadcrumb-item"><a href="#">All Targets</a></li> {% endblock breadcrumb_title %} {% block page_title %} Targets {% endblock page_title %} {% block main_content %} <div class="row"> <div class="col-12"> <div class="card"> <div class="p-2"> <div class="row"> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <button type="button" class="btn btn-primary dropdown-toggle mt-1" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false" id="filterMenu"> Filter <i class="fe-filter"></i> </button> <div class="dropdown-menu" style="width: 30%"> <div class="px-4 py-3"> <h4 class="headline-title">Filters</h4> <label for="filterByOrganization" class="form-label">Filter by Organization</label> <select class="form-control" id="filterByOrganization"> </select> </div> <div class="dropdown-divider"></div> <a href="#" class="dropdown-ite text-primary float-end" id="resetFilters">Reset Filters</a> </div> </div> <div class="col-xl-6 col-lg-6 col-md-6 col-sm-12 col-12"> <a class="btn btn-soft-danger float-end disabled ms-1 mt-1" href="#" onclick="deleteMultipleTargets()" id="delete_multiple_button">Delete Multiple Targets</a> <a class="btn btn-soft-info float-end disabled ms-1 mt-1" href="#" onclick="scanMultipleTargets()" id="scan_multiple_button">Scan Multiple Targets</a> <a class="btn btn-soft-primary float-end ms-1 mt-1" href="{% url 'add_target' %}" data-toggle="tooltip" data-placement="top" title="Add New Targets">Add Targets</a> </div> </div> </div> </div> </div> </div> <div class="row"> <div class="col-12"> <div class="card"> <form method="POST" id="multiple_targets_form" action="../../scan/start/multiple/"> {% csrf_token %} <table id="list_target_table" class="table style-3 table-hover"> <thead> <tr> <th class="checkbox-column text-center">Serial Number</th> <th>Serial Number</th> <th>Target</th> <th>Description</th> <th class="text-center">Added On</th> <th class="text-center">Last Scanned</th> <th class="text-center">Action</th> </tr> </thead> </table> </form> </div> </div> </div> {% endblock main_content %} {% block page_level_script %} <script src="{% static 'custom/custom.js' %}"></script> <script src="{% static 'targetApp/js/custom_domain.js' %}"></script> <script src="{% static 'plugins/datatable/datatables.js' %}"></script> <script> $(document).ready(function(){ var table = $('#list_target_table').DataTable({ "headerCallback": function(e, a, t, n, s) { e.getElementsByTagName("th")[0].innerHTML='<div class="form-check mb-2 form-check-primary"><input type="checkbox" class="float-start form-check-input chk-parent" id="head_checkbox" onclick=mainCheckBoxSelected()>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>\n' }, "destroy": true, "processing": true, "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center mt-sm-0 mt-3'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "fnCreatedRow": function(nRow, aData, iDataIndex) { $(nRow).attr('id', 'target_row_' + aData['id']); }, "stripeClasses": [], "lengthMenu": [5, 10, 20, 30, 40, 50, 100, 500, 1000], "pageLength": 20, 'serverSide': true, "ajax": '/api/listTargets/?format=datatables', "order": [ [1, "desc"] ], "columns": [ { 'data': 'id' }, { 'data': 'id' }, { 'data': 'name' }, { 'data': 'description' }, { 'data': 'id' }, { 'data': 'start_scan_date' }, { 'data': 'id' }, { 'data': 'organization' }, { 'data': 'most_recent_scan' }, { 'data': 'insert_date' }, { 'data': 'insert_date_humanized' }, { 'data': 'start_scan_date_humanized' }, ], "columnDefs":[ { 'orderable': false, 'targets': [0, 3, 6]}, { 'visible': false, 'targets': [1, 7, 8, 9, 10, 11] }, { "targets":0, "width":"20px", "className":"", "orderable":!1, render:function(e, a, t, n) { return'<div class="form-check mb-2 form-check-primary"><input type="checkbox" name="targets_checkbox['+ e + ']" class="float-start form-check-input targets_checkbox" value="' + e + '" onchange=toggleMultipleTargetButton()>\n<span class="new-control-indicator"></span><span style="visibility:hidden">c</span></div>' } }, { "render": function(data, type, row) { var content = ''; content += `<b>${data}</b>&nbsp;&nbsp;<a href="#" onclick="get_target_whois('${data}')">(view whois)</a>`; if (row.organization) { content += '<br>'; for (var org in row.organization) { content += `<span class="badge badge-soft-primary me-1 mb-1" data-toggle="tooltip" data-placement="top" title="Domain ${data} belongs to organization ${row.organization[org]}">${row.organization[org]}</span>`; } } if (row.most_recent_scan) { content += `<br><a href="/scan/detail/${row.most_recent_scan}" class="text-primary">Recent Scan&nbsp;<svg xmlns="http://www.w3.org/2000/svg" width="15" height="15" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-external-link"><path d="M18 13v6a2 2 0 0 1-2 2H5a2 2 0 0 1-2-2V8a2 2 0 0 1 2-2h6"></path><polyline points="15 3 21 3 21 9"></polyline><line x1="10" y1="14" x2="21" y2="3"></line></svg></a>`; } return content; }, "targets": 2, }, { "render": function(data, type, row) { var content = '<div class="text-center">'; content += ` <span class="badge badge-soft-primary">${row.insert_date}, ${row.insert_date_humanized}</span> `; content += '</div>'; return content; }, "targets": 4, }, { "render": function(data, type, row) { var content = '<div class="text-center">'; if (data) { content += `<span class="badge badge-soft-primary">${row.start_scan_date_humanized}</span>`; } else{ content += '<span class="badge badge-soft-warning">Never Scanned Before</span>'; } content += '</div>'; return content; }, "targets": 5, }, { "render": function(data, type, row) { var content = ''; content += ` <div class="btn-group float-end"> <a class="btn btn-soft-primary" href="/target/summary/${row.id}"><i class="fe-info"></i>&nbsp;&nbsp;Target Summary</a> <a href="/scan/start/${row.id}" class="btn btn-soft-primary"><i class="fe-zap"></i>&nbsp;&nbsp;Initiate Scan</a> <div class="btn-group dropstart" role="group"> <button type="button" class="btn btn-soft-primary dropdown-toggle dropdown-toggle-split" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false"> <i class="mdi mdi-chevron-right"></i> </button> <div class="dropdown-menu" style=""> <a class="dropdown-item" href="/scan/schedule/target/${row.id}"><i class="fe-clock"></i>&nbsp;&nbsp;Schedule Scan</a> <div class="dropdown-divider"></div> <a class="dropdown-item" href="/target/update/target/${row.id}"><i class="fe-edit-2"></i>&nbsp;&nbsp;Edit Target</a> <a class="dropdown-item text-danger" href="#" onclick="delete_target(${row.id}, '${row.name}')"><i class="fe-trash-2"></i>&nbsp;&nbsp;Delete target</a> </div> </div> </div> `; return content; }, "targets": 6, }, ], drawCallback: function() { $('.badge').tooltip({ template: '<div class="tooltip status" role="tooltip"><div class="arrow"></div><div class="tooltip-inner"></div></div>' }) $('.bs-tooltip').tooltip(); }, }); multiCheck(table); // Handle form submission event $('#frm-example').on('submit', function(e){ var form = this; table.$('input[type="checkbox"]').each(function(){ if(!$.contains(document, this)){ if(this.checked){ $(form).append( $('<input>') .attr('type', 'hidden') .attr('name', this.name) .val(this.value) ); } } }); e.preventDefault(); }); // filter organization populate $.getJSON(`/api/listOrganizations?&format=json`, function(data) { data = data['organizations'] for (organization in data) { name = htmlEncode(data[organization]['name']); select = document.getElementById('filterByOrganization'); var option = document.createElement('option'); option.value = name; option.innerHTML = name; select.appendChild(option); } }).fail(function(){ }); var a = document.getElementById('filterByOrganization'); a.addEventListener('click', function() { table.search(this.value).draw(); document.getElementById('filteringText').innerHTML = `<span class="badge badge-primary m-2">Organization: ${this.value} <span id="clearFilterChip" class="badge-link" onclick="document.getElementById('resetFilters').click()">X</span> </span>`; Snackbar.show({ text: `Filtering by organization ${this.value}`, pos: 'top-center' }); }, false); // reset filtering var reset_filter = document.getElementById('resetFilters'); reset_filter.addEventListener('click', function() { resetFilters(table); }, false); $('[data-toggle=tooltip]').tooltip(); }); function resetFilters(table_obj) { table_obj.search("").draw(); Snackbar.show({ text: `Filters Reset`, pos: 'top-center' }); document.getElementById('filteringText').innerHTML = ''; } </script> {% endblock page_level_script %}

yogeshojha

758debc4e79b5dc3f1ee29fcabcacb8e15656a94

0caa3a6f04a26f9f3554e1617c7f369a9b10330e

## DOM text reinterpreted as HTML [DOM text](1) is reinterpreted as HTML without escaping meta-characters. [Show more details](https://github.com/yogeshojha/rengine/security/code-scanning/135)

github-advanced-security[bot]

yogeshojha/rengine

Release/1.3.0

## 1.3.0 **Release Date: July 11, 2022** ## Added - Geographic Distribution of Assets Map ## Fixes - WHOIS Provider Changed - Fixed Dark UI Issues - Fix HTTPX Issue

2022-07-10 17:49:40+00:00

2022-07-18 19:30:04+00:00

web/static/custom/custom.js

function checkall(clickchk, relChkbox) { var checker = $('#' + clickchk); var multichk = $('.' + relChkbox); checker.click(function() { multichk.prop('checked', $(this).prop('checked')); }); } function multiCheck(tb_var) { tb_var.on("change", ".chk-parent", function() { var e = $(this).closest("table").find("td:first-child .child-chk"), a = $(this).is(":checked"); $(e).each(function() { a ? ($(this).prop("checked", !0), $(this).closest("tr").addClass("active")) : ($(this).prop("checked", !1), $(this).closest("tr").removeClass("active")) }) }), tb_var.on("change", "tbody tr .new-control", function() { $(this).parents("tr").toggleClass("active") }) } function GetIEVersion() { var sAgent = window.navigator.userAgent; var Idx = sAgent.indexOf("MSIE"); // If IE, return version number. if (Idx > 0) return parseInt(sAgent.substring(Idx + 5, sAgent.indexOf(".", Idx))); // If IE 11 then look for Updated user agent string. else if (!!navigator.userAgent.match(/Trident\/7\./)) return 11; else return 0; //It is not IE } function truncate(str, n) { return (str.length > n) ? str.substr(0, n - 1) + '&hellip;' : str; }; function return_str_if_not_null(val) { return val ? val : ''; } // seperate hostname and url // Referenced from https://stackoverflow.com/questions/736513/how-do-i-parse-a-url-into-hostname-and-path-in-javascript function getParsedURL(url) { var parser = new URL(url); return parser.pathname + parser.search; }; function getCookie(name) { var cookieValue = null; if (document.cookie && document.cookie !== '') { var cookies = document.cookie.split(';'); for (var i = 0; i < cookies.length; i++) { var cookie = jQuery.trim(cookies[i]); // Does this cookie string begin with the name we want? if (cookie.substring(0, name.length + 1) === (name + '=')) { cookieValue = decodeURIComponent(cookie.substring(name.length + 1)); break; } } } return cookieValue; } // Source: https://portswigger.net/web-security/cross-site-scripting/preventing#encode-data-on-output function htmlEncode(str) { return String(str).replace(/[^\w. ]/gi, function(c) { return '&#' + c.charCodeAt(0) + ';'; }); } // Source: https://portswigger.net/web-security/cross-site-scripting/preventing#encode-data-on-output function jsEscape(str) { return String(str).replace(/[^\w. ]/gi, function(c) { return '\\u' + ('0000' + c.charCodeAt(0).toString(16)).slice(-4); }); } function deleteScheduledScan(id) { const delAPI = "../delete/scheduled_task/" + id; swal.queue([{ title: 'Are you sure you want to delete this?', text: "This action can not be undone.", icon: 'warning', showCancelButton: true, confirmButtonText: 'Delete', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { return fetch(delAPI, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }).then(function(response) { return response.json(); }).then(function(data) { // TODO Look for better way return location.reload(); }).catch(function() { swal.insertQueueStep({ icon: 'error', title: 'Oops! Unable to delete the scheduled task!' }) }) } }]) } function change_scheduled_task_status(id, checkbox) { if (checkbox.checked) { text_msg = 'Schedule Scan Started'; } else { text_msg = 'Schedule Scan Stopped'; } Snackbar.show({ text: text_msg, pos: 'top-right', duration: 2500 }); const taskStatusApi = "../toggle/scheduled_task/" + id; return fetch(taskStatusApi, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }) } function change_vuln_status(id) { const vulnStatusApi = "../toggle/vuln_status/" + id; return fetch(vulnStatusApi, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }) } // splits really long strings into multiple lines // Souce: https://stackoverflow.com/a/52395960 function split_into_lines(str, maxWidth) { const newLineStr = "</br>"; done = false; res = ''; do { found = false; // Inserts new line at first whitespace of the line for (i = maxWidth - 1; i >= 0; i--) { if (test_white_space(str.charAt(i))) { res = res + [str.slice(0, i), newLineStr].join(''); str = str.slice(i + 1); found = true; break; } } // Inserts new line at maxWidth position, the word is too long to wrap if (!found) { res += [str.slice(0, maxWidth), newLineStr].join(''); str = str.slice(maxWidth); } if (str.length < maxWidth) done = true; } while (!done); return res + str; } function test_white_space(x) { const white = new RegExp(/^\s$/); return white.test(x.charAt(0)); }; // span values function will seperate the values by comma and put badge around it function parse_comma_values_into_span(data, color, outline = null) { if (data) { var badge = `<span class='badge badge-soft-` + color + ` m-1'>`; var data_with_span = ""; data.split(/\s*,\s*/).forEach(function(split_vals) { data_with_span += badge + split_vals + "</span>"; }); return data_with_span; } return ''; } function get_severity_badge(severity) { switch (severity) { case 'Info': return "<span class='badge badge-soft-primary'>&nbsp;&nbsp;INFO&nbsp;&nbsp;</span>"; break; case 'Low': return "<span class='badge badge-low'>&nbsp;&nbsp;LOW&nbsp;&nbsp;</span>"; break; case 'Medium': return "<span class='badge badge-soft-warning'>&nbsp;&nbsp;MEDIUM&nbsp;&nbsp;</span>"; break; case 'High': return "<span class='badge badge-soft-danger'>&nbsp;&nbsp;HIGH&nbsp;&nbsp;</span>"; break; case 'Critical': return "<span class='badge badge-critical'>&nbsp;&nbsp;CRITICAL&nbsp;&nbsp;</span>"; break; case 'Unknown': return "<span class='badge badge-soft-info'>&nbsp;&nbsp;UNKNOWN&nbsp;&nbsp;</span>"; default: return ""; } } // Source: https://stackoverflow.com/a/54733055 function typingEffect(words, id, i) { let word = words[i].split(""); var loopTyping = function() { if (word.length > 0) { let elem = document.getElementById(id); elem.setAttribute('placeholder', elem.getAttribute('placeholder') + word.shift()); } else { deletingEffect(words, id, i); return false; }; timer = setTimeout(loopTyping, 150); }; loopTyping(); }; function deletingEffect(words, id, i) { let word = words[i].split(""); var loopDeleting = function() { if (word.length > 0) { word.pop(); document.getElementById(id).setAttribute('placeholder', word.join("")); } else { if (words.length > (i + 1)) { i++; } else { i = 0; }; typingEffect(words, id, i); return false; }; timer = setTimeout(loopDeleting, 90); }; loopDeleting(); }; function fullScreenDiv(id, btn) { let fullscreen = document.querySelector(id); let button = document.querySelector(btn); document.fullscreenElement && document.exitFullscreen() || document.querySelector(id).requestFullscreen() fullscreen.setAttribute("style", "overflow:auto"); } function get_randid() { return '_' + Math.random().toString(36).substr(2, 9); } function delete_all_scan_results() { const delAPI = "../scan/delete/scan_results/"; swal.queue([{ title: 'Are you sure you want to delete all scan results?', text: "You won't be able to revert this!", icon: 'warning', showCancelButton: true, confirmButtonText: 'Delete', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { return fetch(delAPI, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }).then(function(response) { return response.json(); }).then(function(data) { // TODO Look for better way return location.reload(); }).catch(function() { swal.insertQueueStep({ icon: 'error', title: 'Oops! Unable to delete Delete scan results!' }) }) } }]) } function delete_all_screenshots() { const delAPI = "../scan/delete/screenshots/"; swal.queue([{ title: 'Are you sure you want to delete all Screenshots?', text: "You won't be able to revert this!", icon: 'warning', showCancelButton: true, confirmButtonText: 'Delete', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { return fetch(delAPI, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }).then(function(response) { return response.json(); }).then(function(data) { // TODO Look for better way return location.reload(); }).catch(function() { swal.insertQueueStep({ icon: 'error', title: 'Oops! Unable to delete Empty Screenshots!' }) }) } }]) } function load_image_from_url(src, append_to_id) { img = document.createElement('img'); img.src = src; img.style.width = '100%'; document.getElementById(append_to_id).appendChild(img); } function setTooltip(btn, message) { hide_all_tooltips(); const instance = tippy(document.querySelector(btn)); instance.setContent(message); instance.show(); setTimeout(function() { instance.hide(); }, 500); } function hide_all_tooltips() { $(".tooltip").tooltip("hide"); } function get_response_time_text(response_time) { if (response_time) { var text_color = 'danger'; if (response_time < 0.5) { text_color = 'success' } else if (response_time >= 0.5 && response_time < 1) { text_color = 'warning' } return `<span class="text-${text_color}">${response_time.toFixed(4)}s</span>`; } return ''; } function parse_technology(data, color, scan_id = null, domain_id=null) { var badge = `<span data-toggle="tooltip" title="Technology" class='badge-link badge badge-soft-` + color + ` mt-1 me-1'`; var data_with_span = ""; for (var key in data) { if (scan_id) { data_with_span += badge + ` onclick="get_tech_details('${data[key]['name']}', ${scan_id}, domain_id=null)">` + data[key]['name'] + "</span>"; } else if (domain_id) { data_with_span += badge + ` onclick="get_tech_details('${data[key]['name']}', scan_id=null, domain_id=domain_id)">` + data[key]['name'] + "</span>"; } } return data_with_span; } // span values function will seperate the values by comma and put badge around it function parse_ip(data, cdn) { if (cdn) { var badge = `<span class='badge badge-soft-warning m-1 bs-tooltip' title="CDN IP Address">`; } else { var badge = `<span class='badge badge-soft-primary m-1'>`; } var data_with_span = ""; data.split(/\s*,\s*/).forEach(function(split_vals) { data_with_span += badge + split_vals + "</span>"; }); return data_with_span; } //to remove the image element if there is no screenshot captured function removeImageElement(element) { element.parentElement.remove(); } // https://stackoverflow.com/a/18197341/9338140 function download(filename, text) { var element = document.createElement('a'); element.setAttribute('href', 'data:text/plain;charset=utf-8,' + encodeURIComponent(text)); element.setAttribute('download', filename); element.style.display = 'none'; document.body.appendChild(element); element.click(); document.body.removeChild(element); } function vuln_status_change(checkbox, id) { if (checkbox.checked) { checkbox.parentNode.parentNode.parentNode.className = "table-success text-strike"; } else { checkbox.parentNode.parentNode.parentNode.classList.remove("table-success"); checkbox.parentNode.parentNode.parentNode.classList.remove("text-strike"); } change_vuln_status(id); } function report_hackerone(vulnerability_id, severity) { message = "" if (severity == 'Info' || severity == 'Low' || severity == 'Medium') { message = "We do not recommended sending this vulnerability report to hackerone due to the severity, do you still want to report this?" } else { message = "This vulnerability report will be sent to Hackerone."; } const vulnerability_report_api = "../../api/vulnerability/report/?vulnerability_id=" + vulnerability_id; swal.queue([{ title: 'Reporting vulnerability to hackerone', text: message, icon: 'warning', showCancelButton: true, confirmButtonText: 'Report', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { return fetch(vulnerability_report_api, { method: 'GET', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }).then(function(response) { return response.json(); }).then(function(data) { console.log(data.status) if (data.status == 111) { swal.insertQueueStep({ icon: 'error', title: 'Target does not has team_handle to send report to.' }) } else if (data.status == 201) { swal.insertQueueStep({ icon: 'success', title: 'Vulnerability report successfully submitted to hackerone.' }) } else if (data.status == 400) { swal.insertQueueStep({ icon: 'error', title: 'Invalid Report.' }) } else if (data.status == 401) { swal.insertQueueStep({ icon: 'error', title: 'Hackerone authentication failed.' }) } else if (data.status == 403) { swal.insertQueueStep({ icon: 'error', title: 'API Key forbidden by Hackerone.' }) } else if (data.status == 423) { swal.insertQueueStep({ icon: 'error', title: 'Too many requests.' }) } }).catch(function() { swal.insertQueueStep({ icon: 'error', title: 'Oops! Unable to send vulnerability report to hackerone, check your target team_handle or hackerone configurarions!' }) }) } }]) } function get_interesting_subdomains(target_id, scan_history_id) { if (target_id) { url = `/api/listInterestingEndpoints/?target_id=${target_id}&format=datatables`; non_orderable_targets = [0, 1, 2, 3]; } else if (scan_history_id) { url = `/api/listInterestingSubdomains/?scan_id=${scan_history_id}&format=datatables`; non_orderable_targets = []; } var interesting_subdomain_table = $('#interesting_subdomains').DataTable({ "drawCallback": function(settings, start, end, max, total, pre) { // if no interesting subdomains are found, hide the datatable and show no interesting subdomains found badge if (this.fnSettings().fnRecordsTotal() == 0) { $('#interesting_subdomain_div').empty(); // $('#interesting_subdomain_div').append(`<div class="card-header bg-primary py-3 text-white"> // <div class="card-widgets"> // <a href="#" data-toggle="remove"><i class="mdi mdi-close"></i></a> // </div> // <h5 class="card-title mb-0 text-white"><i class="mdi mdi-fire-alert me-2"></i>Interesting subdomains could not be identified</h5> // </div> // <div id="cardCollpase4" class="collapse show"> // <div class="card-body"> // reNgine could not identify any interesting subdomains. You can customize interesting subdomain keywords <a href="/scanEngine/interesting/lookup/">from here</a> and this section would be automatically updated. // </div> // </div>`); } else { // show nav bar $('.interesting-tab-show').removeAttr('style'); $('#interesting_subdomain_alert_count').html(`${this.fnSettings().fnRecordsTotal()} Interesting Subdomains`) $('#interesting_subdomain_count_badge').empty(); $('#interesting_subdomain_count_badge').html(`<span class="badge badge-soft-primary me-1">${this.fnSettings().fnRecordsTotal()}</span>`); } }, "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "processing": true, "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", "destroy": true, "bInfo": false, "stripeClasses": [], 'serverSide': true, "ajax": url, "order": [ [3, "desc"] ], "lengthMenu": [5, 10, 20, 50, 100], "pageLength": 10, "columns": [{ 'data': 'name' }, { 'data': 'page_title' }, { 'data': 'http_status' }, { 'data': 'content_length' }, { 'data': 'http_url' }, { 'data': 'technologies' }, ], "columnDefs": [{ "orderable": false, "targets": non_orderable_targets }, { "targets": [4], "visible": false, "searchable": false, }, { "targets": [5], "visible": false, "searchable": true, }, { "className": "text-center", "targets": [2] }, { "render": function(data, type, row) { tech_badge = ''; if (row['technologies']) { // tech_badge = `</br>` + parse_technology(row['technologies'], "primary", outline=true, scan_id=null); } if (row['http_url']) { return `<a href="` + row['http_url'] + `" class="text-primary" target="_blank">` + data + `</a>` + tech_badge; } return `<a href="https://` + data + `" class="text-primary" target="_blank">` + data + `</a>` + tech_badge; }, "targets": 0 }, { "render": function(data, type, row) { // display badge based on http status // green for http status 2XX, orange for 3XX and warning for everything else if (data >= 200 && data < 300) { return "<span class='badge badge-pills badge-soft-success'>" + data + "</span>"; } else if (data >= 300 && data < 400) { return "<span class='badge badge-pills badge-soft-warning'>" + data + "</span>"; } else if (data == 0) { // datatable throws error when no data is returned return ""; } return `<span class='badge badge-pills badge-soft-danger'>` + data + `</span>`; }, "targets": 2, }, ], }); } function get_interesting_endpoint(target_id, scan_history_id) { var non_orderable_targets = []; if (target_id) { url = `/api/listInterestingEndpoints/?target_id=${target_id}&format=datatables`; // non_orderable_targets = [0, 1, 2, 3]; } else if (scan_history_id) { url = `/api/listInterestingEndpoints/?scan_id=${scan_history_id}&format=datatables`; // non_orderable_targets = [0, 1, 2, 3]; } $('#interesting_endpoints').DataTable({ "drawCallback": function(settings, start, end, max, total, pre) { if (this.fnSettings().fnRecordsTotal() == 0) { $('#interesting_endpoint_div').remove(); } else { $('.interesting-tab-show').removeAttr('style'); $('#interesting_endpoint_alert_count').html(`, ${this.fnSettings().fnRecordsTotal()} Interesting Endpoints`) $('#interesting_endpoint_count_badge').empty(); $('#interesting_endpoint_count_badge').html(`<span class="badge badge-soft-primary me-1">${this.fnSettings().fnRecordsTotal()}</span>`); } }, "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "processing": true, "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", 'serverSide': true, "destroy": true, "bInfo": false, "ajax": url, "order": [ [3, "desc"] ], "lengthMenu": [5, 10, 20, 50, 100], "pageLength": 10, "columns": [{ 'data': 'http_url' }, { 'data': 'page_title' }, { 'data': 'http_status' }, { 'data': 'content_length' }, ], "columnDefs": [{ "orderable": false, "targets": non_orderable_targets }, { "className": "text-center", "targets": [2] }, { "render": function(data, type, row) { var url = split_into_lines(data, 70); return "<a href='" + data + "' target='_blank' class='text-primary'>" + url + "</a>"; }, "targets": 0 }, { "render": function(data, type, row) { // display badge based on http status // green for http status 2XX, orange for 3XX and warning for everything else if (data >= 200 && data < 300) { return "<span class='badge badge-pills badge-soft-success'>" + data + "</span>"; } else if (data >= 300 && data < 400) { return "<span class='badge badge-pills badge-soft-warning'>" + data + "</span>"; } else if (data == 0) { // datatable throws error when no data is returned return ""; } return `<span class='badge badge-pills badge-soft-danger'>` + data + `</span>`; }, "targets": 2, }, ], }); } function get_important_subdomains(target_id, scan_history_id) { var url = `/api/querySubdomains/?only_important&no_lookup_interesting&format=json`; if (target_id) { url += `&target_id=${target_id}`; } else if (scan_history_id) { url += `&scan_id=${scan_history_id}`; } $.getJSON(url, function(data) { $('#important-count').empty(); $('#important-subdomains-list').empty(); if (data['subdomains'].length > 0) { $('#important-count').html(`<span class="badge badge-soft-primary ms-1 me-1">${data['subdomains'].length}</span>`); for (var val in data['subdomains']) { subdomain = data['subdomains'][val]; div_id = 'important_' + subdomain['id']; $("#important-subdomains-list").append(` <div id="${div_id}"> <p> <span id="subdomain_${subdomain['id']}"> ${subdomain['name']} <span class=""> <a href="javascript:;" data-clipboard-action="copy" class="m-1 float-end badge-link text-info copyable text-primary" data-toggle="tooltip" data-placement="top" title="Copy Subdomain!" data-clipboard-target="#subdomain_${subdomain['id']}"> <svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-copy"><rect x="9" y="9" width="13" height="13" rx="2" ry="2"></rect><path d="M5 15H4a2 2 0 0 1-2-2V4a2 2 0 0 1 2-2h9a2 2 0 0 1 2 2v1"></path></svg></span> </a> </span> </p> </div> <hr /> `); } } else { $('#important-count').html(`<span class="badge badge-soft-primary ms-1 me-1">0</span>`); $('#important-subdomains-list').append(`<p>No subdomains markerd as important!</p>`); } $('.bs-tooltip').tooltip(); }); } function mark_important_subdomain(row, subdomain_id) { if (row) { parentNode = row.parentNode.parentNode.parentNode.parentNode; if (parentNode.classList.contains('table-danger')) { parentNode.classList.remove('table-danger'); } else { parentNode.className = "table-danger"; } } var data = {'subdomain_id': subdomain_id} const subdomainImpApi = "/api/toggle/subdomain/important/"; if ($("#important_subdomain_" + subdomain_id).length == 0) { $("#subdomain-" + subdomain_id).prepend(`<span id="important_subdomain_${subdomain_id}"></span>`); setTooltip("#subdomain-" + subdomain_id, 'Marked Important!'); } else { $("#important_subdomain_" + subdomain_id).remove(); setTooltip("#subdomain-" + subdomain_id, 'Marked Un-Important!'); } return fetch(subdomainImpApi, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken"), 'Content-Type': 'application/json' }, body: JSON.stringify(data) }); } function delete_scan(id) { const delAPI = "../delete/scan/" + id; swal.queue([{ title: 'Are you sure you want to delete this scan history?', text: "You won't be able to revert this!", icon: 'warning', showCancelButton: true, confirmButtonText: 'Delete', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { return fetch(delAPI, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }).then(function(response) { return response.json(); }).then(function(data) { // TODO Look for better way return location.reload(); }).catch(function() { swal.insertQueueStep({ icon: 'error', title: 'Oops! Unable to delete the scan history!' }) }) } }]); } function stop_scan(scan_id=null, subscan_id=null, reload_scan_bar=true, reload_location=false) { const stopAPI = "/api/action/stop/scan/"; if (scan_id) { var data = {'scan_id': scan_id} } else if (subscan_id) { var data = {'subscan_id': subscan_id} } swal.queue([{ title: 'Are you sure you want to stop this scan?', text: "You won't be able to revert this!", icon: 'warning', showCancelButton: true, confirmButtonText: 'Stop', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { return fetch(stopAPI, { method: 'POST', credentials: "same-origin", body: JSON.stringify(data), headers: { "X-CSRFToken": getCookie("csrftoken"), "Content-Type": 'application/json', } }).then(function(response) { return response.json(); }).then(function(data) { // TODO Look for better way if (data.status) { Snackbar.show({ text: 'Scan Successfully Aborted.', pos: 'top-right', duration: 1500 }); if (reload_scan_bar) { getScanStatusSidebar(); } if (reload_location) { window.location.reload(); } } else { Snackbar.show({ text: 'Oops! Could not abort the scan. ' + data.message, pos: 'top-right', duration: 1500 }); } }).catch(function() { swal.insertQueueStep({ icon: 'error', title: 'Oops! Unable to stop the scan' }) }) } }]) } function extractContent(s) { var span = document.createElement('span'); span.innerHTML = s; return span.textContent || span.innerText; }; function delete_datatable_rows(table_id, rows_id, show_snackbar = true, snackbar_title) { // this function will delete the datatables rows after actions such as delete // table_id => datatable_id with # // rows_ids: list/array => list of all numerical ids to delete, to maintain consistency // rows id will always follow this pattern: datatable_id_row_n // show_snackbar = bool => whether to show snackbar or not! // snackbar_title: str => snackbar title if show_snackbar = True var table = $(table_id).DataTable(); for (var row in rows_id) { table.row(table_id + '_row_' + rows_id[row]).remove().draw(); } Snackbar.show({ text: snackbar_title, pos: 'top-right', duration: 1500, actionTextColor: '#fff', backgroundColor: '#e7515a', }); } function delete_subscan(subscan_id) { // This function will delete the sunscans using rest api // Supported method: POST const delAPI = "/api/action/rows/delete/"; var data = { 'type': 'subscan', 'rows': [subscan_id] } swal.queue([{ title: 'Are you sure you want to delete this subscan?', text: "You won't be able to revert this!", icon: 'warning', showCancelButton: true, confirmButtonText: 'Delete', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { return fetch(delAPI, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken"), "Content-Type": "application/json" }, body: JSON.stringify(data) }).then(function(response) { return response.json(); }).then(function(response) { if (response['status']) { delete_datatable_rows('#subscan_history_table', [subscan_id], show_snackbar = true, '1 Subscan Deleted!') } }).catch(function() { swal.insertQueueStep({ icon: 'error', title: 'Oops! Unable to delete the scan history!' }) }) } }]) } function show_subscan_results(subscan_id) { // This function will popup a modal and show the subscan results // modal being used is from base var api_url = '/api/fetch/results/subscan/?format=json'; var data = { 'subscan_id': subscan_id }; Swal.fire({ title: 'Fetching Results...' }); swal.showLoading(); fetch(api_url, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken"), 'Content-Type': 'application/json' }, body: JSON.stringify(data) }).then(response => response.json()).then(function(response) { console.log(response); swal.close(); if (response['subscan']['status'] == -1) { swal.fire("Error!", "Scan has not yet started! Please wait for other scans to complete...", "warning", { button: "Okay", }); return; } else if (response['subscan']['status'] == 1) { swal.fire("Error!", "Scan is in progress! Please come back later...", "warning", { button: "Okay", }); return; } $('#xl-modal-title').empty(); $('#xl-modal-content').empty(); $('#xl-modal-footer').empty(); var task_name = ''; if (response['subscan']['task'] == 'port_scan') { task_name = 'Port Scan'; } else if (response['subscan']['task'] == 'vulnerability_scan') { task_name = 'Vulnerability Scan'; } else if (response['subscan']['task'] == 'fetch_url') { task_name = 'EndPoint Gathering'; } else if (response['subscan']['task'] == 'dir_file_fuzz') { task_name = 'Directory and Files Fuzzing'; } $('#xl-modal_title').html(`${task_name} Results on ${response['subscan']['subdomain_name']}`); var scan_status = ''; var badge_color = 'danger'; if (response['subscan']['status'] == 0) { scan_status = 'Failed'; } else if (response['subscan']['status'] == 2) { scan_status = 'Successful'; badge_color = 'success'; } else if (response['subscan']['status'] == 3) { scan_status = 'Aborted'; } else { scan_status = 'Unknown'; } $('#xl-modal-content').append(`<div>Scan Status: <span class="badge bg-${badge_color}">${scan_status}</span></div>`); console.log(response); $('#xl-modal-content').append(`<div class="mt-1">Engine Used: <span class="badge bg-primary">${htmlEncode(response['subscan']['engine'])}</span></div>`); if (response['result'].length > 0) { if (response['subscan']['task'] == 'port_scan') { $('#xl-modal-content').append(`<div id="port_results_li"></div>`); for (var ip in response['result']) { var ip_addr = response['result'][ip]['address']; var id_name = `ip_${ip_addr}`; $('#port_results_li').append(`<h5>IP Address: ${ip_addr}</br></br>${response['result'][ip]['ports'].length} Ports Open</h5>`); $('#port_results_li').append(`<ul id="${id_name}"></ul>`); for (var port_obj in response['result'][ip]['ports']) { var port = response['result'][ip]['ports'][port_obj]; var port_color = 'primary'; if (port["is_uncommon"]) { port_color = 'danger'; } $('#port_results_li ul').append(`<li><span class="ms-1 mt-1 me-1 badge badge-soft-${port_color}">${port['number']}</span>/<span class="ms-1 mt-1 me-1 badge badge-soft-${port_color}">${port['service_name']}</span>/<span class="ms-1 mt-1 me-1 badge badge-soft-${port_color}">${port['description']}</span></li>`); } } $('#xl-modal-footer').append(`<span class="text-danger">* Uncommon Ports</span>`); } else if (response['subscan']['task'] == 'vulnerability_scan') { render_vulnerability_in_xl_modal(vuln_count = response['result'].length, subdomain_name = response['subscan']['subdomain_name'], result = response['result']); } else if (response['subscan']['task'] == 'fetch_url') { render_endpoint_in_xlmodal(endpoint_count = response['result'].length, subdomain_name = response['subscan']['subdomain_name'], result = response['result']); } else if (response['subscan']['task'] == 'dir_file_fuzz') { if (response['result'][0]['directory_files'].length == 0) { $('#xl-modal-content').append(` <div class="alert alert-info mt-2" role="alert"> <i class="mdi mdi-alert-circle-outline me-2"></i> ${task_name} could not fetch any results. </div> `); } else { render_directories_in_xl_modal(response['result'][0]['directory_files'].length, response['subscan']['subdomain_name'], response['result'][0]['directory_files']); } } } else { $('#xl-modal-content').append(` <div class="alert alert-info mt-2" role="alert"> <i class="mdi mdi-alert-circle-outline me-2"></i> ${task_name} could not fetch any results. </div> `); } $('#modal_xl_scroll_dialog').modal('show'); $("body").tooltip({ selector: '[data-toggle=tooltip]' }); }); } function get_http_status_badge(data) { if (data >= 200 && data < 300) { return "<span class='badge badge-soft-success'>" + data + "</span>"; } else if (data >= 300 && data < 400) { return "<span class='badge badge-soft-warning'>" + data + "</span>"; } else if (data == 0) { // datatable throws error when no data is returned return ""; } return "<span class='badge badge-soft-danger'>" + data + "</span>"; } function render_endpoint_in_xlmodal(endpoint_count, subdomain_name, result) { // This function renders endpoints datatable in xl modal // Used in Subscan results and subdomain to endpoints modal $('#xl-modal-content').append(`<h5> ${endpoint_count} Endpoints Discovered on subdomain ${subdomain_name}</h5>`); $('#xl-modal-content').append(` <div class=""> <table id="endpoint-modal-datatable" class="table dt-responsive nowrap w-100"> <thead> <tr> <th>HTTP URL</th> <th>Status</th> <th>Page Title</th> <th>Tags</th> <th>Content Type</th> <th>Content Length</th> <th>Response Time</th> </tr> </thead> <tbody id="endpoint_tbody"> </tbody> </table> </div> `); $('#endpoint_tbody').empty(); for (var endpoint_obj in result) { var endpoint = result[endpoint_obj]; var tech_badge = ''; var web_server = ''; if (endpoint['technologies']) { tech_badge = '<div>' + parse_technology(endpoint['technologies'], "primary", outline = true); } if (endpoint['webserver']) { web_server = `<span class='m-1 badge badge-soft-info' data-toggle="tooltip" data-placement="top" title="Web Server">${endpoint['webserver']}</span>`; } var url = split_into_lines(endpoint['http_url'], 70); var rand_id = get_randid(); tech_badge += web_server + '</div>'; var http_url_td = "<a href='" + endpoint['http_url'] + `' target='_blank' class='text-primary'>` + url + "</a>" + tech_badge; $('#endpoint_tbody').append(` <tr> <td>${http_url_td}</td> <td>${get_http_status_badge(endpoint['http_status'])}</td> <td>${return_str_if_not_null(endpoint['page_title'])}</td> <td>${parse_comma_values_into_span(endpoint['matched_gf_patterns'], "danger", outline=true)}</td> <td>${return_str_if_not_null(endpoint['content_type'])}</td> <td>${return_str_if_not_null(endpoint['content_length'])}</td> <td>${get_response_time_text(endpoint['response_time'])}</td> </tr> `); } $("#endpoint-modal-datatable").DataTable({ "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", "order": [ [5, "desc"] ], drawCallback: function() { $(".dataTables_paginate > .pagination").addClass("pagination-rounded") } }); } function render_vulnerability_in_xl_modal(vuln_count, subdomain_name, result) { // This function will render the vulnerability datatable in xl modal $('#xl-modal-content').append(`<h5> ${vuln_count} Vulnerabilities Discovered on subdomain ${subdomain_name}</h5>`); $('#xl-modal-content').append(`<ol id="vuln_results_ol" class="list-group list-group-numbered"></ol>`); $('#xl-modal-content').append(` <div class=""> <table id="vulnerability-modal-datatable" class="table dt-responsive nowrap w-100"> <thead> <tr> <th>Type</th> <th>Title</th> <th class="text-center">Severity</th> <th>CVSS Score</th> <th>CVE/CWE</th> <th>Vulnerable URL</th> <th>Description</th> <th class="text-center dt-no-sorting">Action</th> </tr> </thead> <tbody id="vuln_tbody"> </tbody> </table> </div> `); $('#vuln_tbody').empty(); for (var vuln in result) { var vuln_obj = result[vuln]; var vuln_type = vuln_obj['type'] ? `<span class="badge badge-soft-primary">&nbsp;&nbsp;${vuln_obj['type'].toUpperCase()}&nbsp;&nbsp;</span>` : ''; var tags = ''; var cvss_metrics_badge = ''; switch (vuln_obj['severity']) { case 'Info': color = 'primary' badge_color = 'soft-primary' break; case 'Low': color = 'low' badge_color = 'soft-warning' break; case 'Medium': color = 'warning' badge_color = 'soft-warning' break; case 'High': color = 'danger' badge_color = 'soft-danger' break; case 'Critical': color = 'critical' badge_color = 'critical' break; default: } if (vuln_obj['tags']) { tags = '<div>'; vuln_obj['tags'].forEach(tag => { tags += `<span class="badge badge-${badge_color} me-1 mb-1" data-toggle="tooltip" data-placement="top" title="Tags">${tag.name}</span>`; }); tags += '</div>'; } if (vuln_obj['cvss_metrics']) { cvss_metrics_badge = `<div><span class="badge badge-outline-primary my-1" data-toggle="tooltip" data-placement="top" title="CVSS Metrics">${vuln_obj['cvss_metrics']}</span></div>`; } var vuln_title = `<b class="text-${color}">` + vuln_obj['name'] + `</b>` + cvss_metrics_badge + tags; var badge = 'danger'; var cvss_score = ''; if (vuln_obj['cvss_score']) { if (vuln_obj['cvss_score'] > 0.1 && vuln_obj['cvss_score'] <= 3.9) { badge = 'info'; } else if (vuln_obj['cvss_score'] > 3.9 && vuln_obj['cvss_score'] <= 6.9) { badge = 'warning'; } else if (vuln_obj['cvss_score'] > 6.9 && vuln_obj['cvss_score'] <= 8.9) { badge = 'danger'; } cvss_score = `<span class="badge badge-outline-${badge}" data-toggle="tooltip" data-placement="top" title="CVSS Score">${vuln_obj['cvss_score']}</span>`; } var cve_cwe_badge = '<div>'; if (vuln_obj['cve_ids']) { vuln_obj['cve_ids'].forEach(cve => { cve_cwe_badge += `<a href="https://google.com/search?q=${cve.name.toUpperCase()}" target="_blank" class="badge badge-outline-primary me-1 mt-1" data-toggle="tooltip" data-placement="top" title="CVE ID">${cve.name.toUpperCase()}</a>`; }); } if (vuln_obj['cwe_ids']) { vuln_obj['cwe_ids'].forEach(cwe => { cve_cwe_badge += `<a href="https://google.com/search?q=${cwe.name.toUpperCase()}" target="_blank" class="badge badge-outline-primary me-1 mt-1" data-toggle="tooltip" data-placement="top" title="CWE ID">${cwe.name.toUpperCase()}</a>`; }); } cve_cwe_badge += '</div>'; var http_url = vuln_obj['http_url'].includes('http') ? "<a href='" + htmlEncode(vuln_obj['http_url']) + "' target='_blank' class='text-danger'>" + htmlEncode(vuln_obj['http_url']) + "</a>" : vuln_obj['http_url']; var description = vuln_obj['description'] ? `<div>${split_into_lines(vuln_obj['description'], 30)}</div>` : ''; // show extracted results, and show matcher names, matcher names can be in badges if (vuln_obj['matcher_name']) { description += `<span class="badge badge-soft-primary" data-toggle="tooltip" data-placement="top" title="Matcher Name">${vuln_obj['matcher_name']}</span>`; } if (vuln_obj['extracted_results'] && vuln_obj['extracted_results'].length > 0) { description += `<br><a class="mt-2" data-bs-toggle="collapse" href="#results_${vuln_obj['id']}" aria-expanded="false" aria-controls="results_${vuln_obj['id']}">Extracted Results <i class="fe-chevron-down"></i></a>`; description += `<div class="collapse" id="results_${vuln_obj['id']}"><ul>`; vuln_obj['extracted_results'].forEach(results => { description += `<li>${results}</li>`; }); description += '</ul></div>'; } if (vuln_obj['references'] && vuln_obj['references'].length > 0) { description += `<br><a class="mt-2" data-bs-toggle="collapse" href="#references_${vuln_obj['id']}" aria-expanded="false" aria-controls="references_${vuln_obj['id']}">References <i class="fe-chevron-down"></i></a>`; description += `<div class="collapse" id="references_${vuln_obj['id']}"><ul>`; vuln_obj['references'].forEach(reference => { description += `<li><a href="${reference.url}" target="_blank">${reference.url}</a></li>`; }); description += '</ul></div>'; } if (vuln_obj['curl_command']) { description += `<br><a class="mt-2" data-bs-toggle="collapse" href="#curl_command_${vuln_obj['id']}" aria-expanded="false" aria-controls="curl_command_${vuln_obj['id']}">CURL command <i class="fe-terminal"></i></a>`; description += `<div class="collapse" id="curl_command_${vuln_obj['id']}"><ul>`; description += `<li><code>${split_into_lines(htmlEncode(vuln_obj['curl_command']), 30)}</code></li>`; description += '</ul></div>'; } var action_icon = vuln_obj['hackerone_report_id'] ? '' : ` <div class="btn-group mb-2 dropstart"> <a href="#" class="text-dark dropdown-toggle float-end" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false"> <svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="feather feather-more-horizontal"><circle cx="12" cy="12" r="1"></circle><circle cx="19" cy="12" r="1"></circle><circle cx="5" cy="12" r="1"></circle></svg> </a> <div class="dropdown-menu" style=""> <a class="dropdown-item" href="javascript:report_hackerone(${vuln_obj['id']}, '${vuln_obj['severity']}');">Report to Hackerone</a> </div> </div>`; $('#vuln_tbody').append(` <tr> <td>${vuln_type}</td> <td>${vuln_title}</td> <td class="text-center">${get_severity_badge(vuln_obj['severity'])}</td> <td class="text-center">${cvss_score}</td> <td>${cve_cwe_badge}</td> <td>${http_url}</td> <td>${description}</td> <td>${action_icon}</td> </tr> `); } $("#vulnerability-modal-datatable").DataTable({ "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", "order": [ [5, "desc"] ], drawCallback: function() { $(".dataTables_paginate > .pagination").addClass("pagination-rounded") } }); } function render_directories_in_xl_modal(directory_count, subdomain_name, result) { $('#xl-modal-content').append(`<h5> ${directory_count} Directories Discovered on subdomain ${subdomain_name}</h5>`); $('#xl-modal-content').append(` <div class=""> <table id="directory-modal-datatable" class="table dt-responsive nowrap w-100"> <thead> <tr> <th>Directory</th> <th class="text-center">HTTP Status</th> <th>Content Length</th> <th>Lines</th> <th>Words</th> </tr> </thead> <tbody id="directory_tbody"> </tbody> </table> </div> `); $('#directory_tbody').empty(); for (var dir_obj in result) { var dir = result[dir_obj]; $('#directory_tbody').append(` <tr> <td><a href="${dir.url}" target="_blank">${dir.name}</a></td> <td class="text-center">${get_http_status_badge(dir.http_status)}</td> <td>${dir.length}</td> <td>${dir.lines}</td> <td>${dir.words}</td> </tr> `); } var interesting_keywords_array = []; var dir_modal_table = $("#directory-modal-datatable").DataTable({ "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", "order": [ [2, "desc"] ], drawCallback: function() { $(".dataTables_paginate > .pagination").addClass("pagination-rounded"); } }); // TODO: Find interetsing dirs // fetch("/api/listInterestingKeywords") // .then(response => { // return response.json(); // }) // .then(data => { // interesting_keywords_array = data; // dir_modal_table.rows().every(function(){ // console.log(this.data()); // }); // }); } function get_and_render_subscan_history(subdomain_id, subdomain_name) { // This function displays the subscan history in a modal for any particular subdomain var data = { 'subdomain_id': subdomain_id }; fetch('/api/listSubScans/?format=json', { method: 'POST', credentials: "same-origin", body: JSON.stringify(data), headers: { "X-CSRFToken": getCookie("csrftoken"), "Content-Type": 'application/json', } }).then(function(response) { return response.json(); }).then(function(data) { console.log(data); if (data['status']) { $('#modal_title').html('Subscan History for subdomain ' + subdomain_name); $('#modal-content').empty(); $('#modal-content').append(`<div id="subscan_history_table"></div>`); $('#subscan_history_table').empty(); for (var result in data['results']) { var result_obj = data['results'][result]; var error_message = ''; var task_name = get_task_name(result_obj); if (result_obj.status == 0) { color = 'danger'; bg_color = 'bg-soft-danger'; status_badge = '<span class="float-end badge bg-danger">Failed</span>'; error_message = `</br><span class="text-danger">Error: ${result_obj.error_message}`; } else if (result_obj.status == 3) { color = 'danger'; bg_color = 'bg-soft-danger'; status_badge = '<span class="float-end badge bg-danger">Aborted</span>'; } else if (result_obj.status == 2) { color = 'success'; bg_color = 'bg-soft-success'; status_badge = '<span class="float-end badge bg-success">Task Completed</span>'; } $('#subscan_history_table').append(` <div class="card border-${color} border mini-card"> <a href="#" class="text-reset item-hovered" onclick="show_subscan_results(${result_obj['id']})"> <div class="card-header ${bg_color} text-${color} mini-card-header"> ${task_name} on <b>${result_obj.subdomain_name}</b> using engine <b>${htmlEncode(result_obj.engine)}</b> </div> <div class="card-body mini-card-body"> <p class="card-text"> ${status_badge} <span class=""> Task Completed ${result_obj.completed_ago} ago </span> Took ${result_obj.time_taken} ${error_message} </p> </div> </a> </div> `); } $('#modal_dialog').modal('show'); } }); } function fetch_whois(domain_name, save_db) { // this function will fetch WHOIS record for any subdomain and also display // snackbar once whois is fetched var url = `/api/tools/whois/?format=json&ip_domain=${domain_name}`; if (save_db) { url += '&save_db'; } $('[data-toggle="tooltip"]').tooltip('hide'); Snackbar.show({ text: 'Fetching WHOIS...', pos: 'top-right', duration: 1500, }); $("#whois_not_fetched_alert").hide(); $("#whois_fetching_alert").show(); fetch(url, {}).then(res => res.json()) .then(function(response) { $("#whois_fetching_alert").hide(); document.getElementById('domain_age').innerHTML = response['domain']['domain_age'] + ' ' + response['domain']['date_created']; document.getElementById('ip_address').innerHTML = response['domain']['ip_address']; document.getElementById('ip_geolocation').innerHTML = response['domain']['geolocation']; document.getElementById('registrant_name').innerHTML = response['registrant']['name']; console.log(response['registrant']['organization']) document.getElementById('registrant_organization').innerHTML = response['registrant']['organization'] ? response['registrant']['organization'] : ' '; document.getElementById('registrant_address').innerHTML = response['registrant']['address'] + ' ' + response['registrant']['city'] + ' ' + response['registrant']['state'] + ' ' + response['registrant']['country']; document.getElementById('registrant_phone_numbers').innerHTML = response['registrant']['tel']; document.getElementById('registrant_fax').innerHTML = response['registrant']['fax']; Snackbar.show({ text: 'Whois Fetched...', pos: 'top-right', duration: 3000 }); $("#whois_fetched_alert").show(); $("#whois_fetched_alert").fadeTo(2000, 500).slideUp(1500, function() { $("#whois_fetched_alert").slideUp(500); }); }).catch(function(error) { console.log(error); }); } function get_target_whois(domain_name) { // this function will fetch whois from db, if not fetched, will make a fresh // query and will display whois on a modal var url = `/api/tools/whois/?format=json&ip_domain=${domain_name}&fetch_from_db` Swal.fire({ title: `Fetching WHOIS details for ${domain_name}...` }); swal.showLoading(); fetch(url, { method: 'GET', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken"), 'Content-Type': 'application/json' }, }).then(response => response.json()).then(function(response) { console.log(response); if (response.status) { swal.close(); display_whois_on_modal(response); } else { fetch(`/api/tools/whois/?format=json&ip_domain=${domain_name}&save_db`, { method: 'GET', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken"), 'Content-Type': 'application/json' }, }).then(response => response.json()).then(function(response) { console.log(response); if (response.status) { swal.close(); display_whois_on_modal(response); } else { Swal.fire({ title: 'Oops!', text: `reNgine could not fetch WHOIS records for ${domain_name}!`, icon: 'error' }); } }); } }); } function get_domain_whois(domain_name, show_add_target_btn=false) { // this function will get whois for domains that are not targets, this will // not store whois into db nor create target var url = `/api/tools/whois/?format=json&ip_domain=${domain_name}` Swal.fire({ title: `Fetching WHOIS details for ${domain_name}...` }); $('.modal').modal('hide'); swal.showLoading(); fetch(url, { method: 'GET', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken"), 'Content-Type': 'application/json' }, }).then(response => response.json()).then(function(response) { swal.close(); if (response.status) { display_whois_on_modal(response, show_add_target_btn=show_add_target_btn); } else { Swal.fire({ title: 'Oops!', text: `reNgine could not fetch WHOIS records for ${domain_name}! ${response['message']}`, icon: 'error' }); } }); } function display_whois_on_modal(response, show_add_target_btn=false) { // this function will display whois data on modal, should be followed after get_domain_whois() $('#modal_dialog').modal('show'); $('#modal-content').empty(); $("#modal-footer").empty(); content = `<div class="row mt-3"> <div class="col-sm-3"> <div class="nav flex-column nav-pills nav-pills-tab" id="v-pills-tab" role="tablist" aria-orientation="vertical"> <a class="nav-link active show mb-1" id="v-pills-domain-tab" data-bs-toggle="pill" href="#v-pills-domain" role="tab" aria-controls="v-pills-domain-tab" aria-selected="true">Domain info</a> <a class="nav-link mb-1" id="v-pills-whois-tab" data-bs-toggle="pill" href="#v-pills-whois" role="tab" aria-controls="v-pills-whois" aria-selected="false">Whois</a> <a class="nav-link mb-1" id="v-pills-nameserver-tab" data-bs-toggle="pill" href="#v-pills-nameserver" role="tab" aria-controls="v-pills-nameserver"aria-selected="false">Nameservers</a> <a class="nav-link mb-1" id="v-pills-history-tab" data-bs-toggle="pill" href="#v-pills-history" role="tab" aria-controls="v-pills-history"aria-selected="false">NS History</a> <a class="nav-link mb-1" id="v-pills-related-tab" data-bs-toggle="pill" href="#v-pills-related" role="tab" aria-controls="v-pills-related"aria-selected="false">Related Domains`; if (response['related_domains'].length) { content += `<span class="badge badge-soft-info float-end">${response['related_domains'].length}</span>` } content += `</a>`; content += `<a class="nav-link mb-1" id="v-pills-related-tld-tab" data-bs-toggle="pill" href="#v-pills-related-tld" role="tab" aria-controls="v-pills-related-tld"aria-selected="false">Related TLDs`; if (response['related_tlds'].length) { content += `<span class="badge badge-soft-info float-end">${response['related_tlds'].length}</span>` } content += `</span></a>`; content += `</div></div> <div class="col-sm-9"> <div class="tab-content pt-0"> <div class="tab-pane fade active show" id="v-pills-domain" role="tabpanel" aria-labelledby="v-pills-domain-tab" data-simplebar style="max-height: 300px; min-height: 300px;"> <h4 class="header-title">Domain Information</h4> <table class="domain_details_table table table-hover table-borderless"> <tr style="display: none"> <th>&nbsp;</th> <th>&nbsp;</th> </tr> <tr> <td>Domain Name</td> <td>${response['ip_domain'] ? response['ip_domain']: "-"}</td> </tr> <tr> <td>Domain age</td> <td>${response['domain']['domain_age'] ? response['domain']['domain_age']: "-"}</td> </tr> <tr> <td>IP Address</td> <td>${response['domain']['ip_address'] ? response['domain']['ip_address']: "-" }</td> </tr> <tr> <td>IP Geolocation</td> <td> ${response['domain']['geolocation_iso'] ? `<img src="https://domainbigdata.com/img/flags-iso/flat/24/${response['domain']['geolocation_iso']}.png" alt="${response['domain']['geolocation_iso']}">` : ""} &nbsp;&nbsp;${response['domain']['geolocation'] ? response['domain']['geolocation'] : "-"}</td> </tr> </table> <h4 class="header-title mt-3">Registrant Information</h4> <table class="domain_details_table table table-hover table-borderless"> <tr style="display: none"> <th>&nbsp;</th> <th>&nbsp;</th> </tr> <tr> <td>Name</td> <td>${response['registrant']['name'] ? response['registrant']['name']: "-"}</td> </tr> <tr> <td>Email</td> <td>${response['registrant']['email'] ? response['registrant']['email']: "-"}</td> </tr> <tr> <td>Organization</td> <td>${response['registrant']['organization'] ? response['registrant']['organization']: "-"}</td> </tr> <tr> <td>Address</td> <td>${response['registrant']['address'] ? response['registrant']['address']: "-"}</td> </tr> <tr> <td>Phone Numbers</td> <td>${response['registrant']['tel'] ? response['registrant']['tel']: "-"}</td> </tr> <tr> <td>Fax</td> <td>${response['registrant']['fax'] ? response['registrant']['fax']: "-"}</td> </tr> </table> </div> <div class="tab-pane fade" id="v-pills-whois" role="tabpanel" aria-labelledby="v-pills-whois-tab"> <pre data-simplebar style="max-height: 310px; min-height: 310px;">${response['whois'] ? response['whois'] : "No Whois Data found!"}</pre> </div> <div class="tab-pane fade" id="v-pills-history" role="tabpanel" aria-labelledby="v-pills-history-tab" data-simplebar style="max-height: 300px; min-height: 300px;">`; if (response['nameserver']['history'].length) { content += `<table class="table table-striped mb-0"> <thead class="table-dark"> <td>Date</td> <td>Action</td> <td>NameServer</td> </thead> <tbody>`; for (var history in response['nameserver']['history']) { var obj = response['nameserver']['history'][history]; content += ` <tr> <td>${obj['date']? obj['date'] : '-'}</td> <td>${obj['action']? obj['action'] : '-'}</td> <td>${obj['server']? obj['server'] : '-'}</td> </tr> `; } content += `</tbody></table>` } else { content += 'No DNS history records found.'; } content += ` </div> <div class="tab-pane fade" id="v-pills-nameserver" role="tabpanel" aria-labelledby="v-pills-nameserver-tab" data-simplebar style="max-height: 300px; min-height: 300px;">`; if (response['nameserver']['records'].length) { content += `<table class="table table-striped mb-0"> <thead class="table-dark"> <td>Type</td> <td>Hostname</td> <td>Address</td> <td>TTL</td> <td>Class</td> <td>Preference</td> </thead> <tbody>`; for (var record in response['nameserver']['records']) { var obj = response['nameserver']['records'][record]; content += ` <tr> <td><span class="badge badge-soft-primary me-1 ms-1">${obj['type']? obj['type'] : '-'}</span</td> <td>${obj['hostname']? obj['hostname'] : '-'}</td> <td>${obj['address']? obj['address'] : '-'}</td> <td>${obj['ttl']? obj['ttl'] : '-'}</td> <td>${obj['ns_class']? obj['ns_class'] : '-'}</td> <td>${obj['preference']? obj['preference'] : '-'}</td> </tr>`; } content += `</tbody></table>`; } else { content += `No DNS history records found.`; } content += ` </div> <div class="tab-pane fade" id="v-pills-related" role="tabpanel" aria-labelledby="v-pills-related-tab" data-simplebar style="max-height: 300px; min-height: 300px;"> `; if (!response['related_domains'].length) { content += `<div class="alert alert-warning" role="alert"> <i class="mdi mdi-alert-outline me-2"></i> Oops! Could not find any related domains. </div>`; } for (var domain in response['related_domains']) { var domain_obj = response['related_domains'][domain]; content += `<span class="badge badge-soft-primary badge-link waves-effect waves-light me-1" data-toggle="tooltip" title="Add ${domain_obj} as target." onclick="add_target('${domain_obj}')">${domain_obj}</span>` } content += ` </div> <div class="tab-pane fade" id="v-pills-related-tld" role="tabpanel" aria-labelledby="v-pills-related-tld-tab" data-simplebar style="max-height: 300px; min-height: 300px;"> `; if (!response['related_tlds'].length) { content += `<div class="alert alert-warning" role="alert"> <i class="mdi mdi-alert-outline me-2"></i> Oops! Could not find any related TLds. </div>`; } for (var domain in response['related_tlds']) { var domain_obj = response['related_tlds'][domain]; content += `<span class="badge badge-soft-primary badge-link waves-effect waves-light me-1" data-toggle="tooltip" title="Add ${domain_obj} as target." onclick="add_target('${domain_obj}')">${domain_obj}</span>` } content += ` </div> </div> </div> </div>`; if (show_add_target_btn) { content += `<div class="text-center"> <button class="btn btn-primary float-end" type="submit" id="search_whois_toolbox_btn" onclick="add_target('${response['ip_domain']}')">Add ${response['ip_domain']} as target</button> </div>` } $('#modal-content').append(content); $('[data-toggle="tooltip"]').tooltip(); } function show_quick_add_target_modal() { // this function will display the modal to add target $('#modal_title').html('Add target'); $('#modal-content').empty(); $('#modal-content').append(` If you would like to add IP/CIDRs, multiple domain, Please <a href="/target/add/target">click here.</a> <div class="mb-3"> <label for="target_name_modal" class="form-label">Target Name</label> <input class="form-control" type="text" id="target_name_modal" required="" placeholder="yourdomain.com"> </div> <div class="mb-3"> <label for="target_description_modal" class="form-label">Description (Optional)</label> <input class="form-control" type="text" id="target_description_modal" required="" placeholder="Target Description"> </div> <div class="mb-3"> <label for="h1_handle_modal" class="form-label">Hackerone Target Team Handle (Optional)</label> <input class="form-control" type="text" id="h1_handle_modal" placeholder="hackerone.com/team_handle, Only enter team_handle after /"> </div> <div class="mb-3 text-center"> <button class="btn btn-primary float-end" type="submit" id="add_target_modal" onclick="add_quick_target()">Add Target</button> </div> `); $('#modal_dialog').modal('show'); } function add_quick_target() { // this function will be a onclick for add target button on add_target modal $('#modal_dialog').modal('hide'); var domain_name = $('#target_name_modal').val(); var description = $('#target_description_modal').val(); var h1_handle = $('#h1_handle_modal').val(); const data = { 'domain_name': domain_name, 'h1_team_handle': h1_handle, 'description': description }; add_target(domain_name, h1_handle = h1_handle, description = description); } function add_target(domain_name, h1_handle = null, description = null) { // this function will add domain_name as target const add_api = '/api/add/target/?format=json'; const data = { 'domain_name': domain_name, 'h1_team_handle': h1_handle, 'description': description }; swal.queue([{ title: 'Add Target', text: `Would you like to add ${domain_name} as target?`, icon: 'info', showCancelButton: true, confirmButtonText: 'Add Target', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { return fetch(add_api, { method: 'POST', credentials: "same-origin", headers: { 'X-CSRFToken': getCookie("csrftoken"), 'Content-Type': 'application/json' }, body: JSON.stringify(data) }).then(function(response) { return response.json(); }).then(function(data) { if (data.status) { swal.queue([{ title: 'Target Successfully added!', text: `Do you wish to initiate the scan on new target?`, icon: 'success', showCancelButton: true, confirmButtonText: 'Initiate Scan', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { window.location = `/scan/start/${data.domain_id}`; } }]); } else { swal.insertQueueStep({ icon: 'error', title: data.message }); } }).catch(function() { swal.insertQueueStep({ icon: 'error', title: 'Oops! Unable to delete the scan history!' }); }) } }]); } function loadSubscanHistoryWidget(scan_history_id = null, domain_id = null) { // This function will load the subscan history widget if (scan_history_id) { var data = { 'scan_history_id': scan_history_id } } if (domain_id) { var data = { 'domain_id': domain_id } } fetch('/api/listSubScans/?format=json', { method: 'POST', credentials: "same-origin", body: JSON.stringify(data), headers: { "X-CSRFToken": getCookie("csrftoken"), "Content-Type": 'application/json', } }).then(function(response) { return response.json(); }).then(function(data) { console.log(data); $('#subscan_history_widget').empty(); if (data['status']) { $('#sub_scan_history_count').append(` <span class="badge badge-soft-primary me-1">${data['results'].length}</span> `) for (var result in data['results']) { var error_message = ''; var result_obj = data['results'][result]; var task_name = get_task_name(result_obj); if (result_obj.status == 0) { color = 'danger'; bg_color = 'bg-soft-danger'; status_badge = '<span class="float-end badge bg-danger">Failed</span>'; error_message = `</br><span class="text-danger">Error: ${result_obj.error_message}`; } else if (result_obj.status == 3) { color = 'danger'; bg_color = 'bg-soft-danger'; status_badge = '<span class="float-end badge bg-danger">Aborted</span>'; } else if (result_obj.status == 2) { color = 'success'; bg_color = 'bg-soft-success'; status_badge = '<span class="float-end badge bg-success">Task Completed</span>'; } else if (result_obj.status == 1) { color = 'primary'; bg_color = 'bg-soft-primary'; status_badge = '<span class="float-end badge bg-primary">Running</span>'; } $('#subscan_history_widget').append(` <div class="card border-${color} border mini-card"> <a href="#" class="text-reset item-hovered" onclick="show_subscan_results(${result_obj['id']})"> <div class="card-header ${bg_color} text-${color} mini-card-header"> ${task_name} on <b>${result_obj.subdomain_name}</b> </div> <div class="card-body mini-card-body"> <p class="card-text"> ${status_badge} <span class=""> Task Completed ${result_obj.completed_ago} ago </span> Took ${result_obj.time_taken} ${error_message} </p> </div> </a> </div> `); } } else { $('#sub_scan_history_count').append(` <span class="badge badge-soft-primary me-1">0</span> `) $('#subscan_history_widget').append(` <div class="alert alert-warning alert-dismissible fade show mt-2" role="alert"> <button type="button" class="btn-close" data-bs-dismiss="alert" aria-label="Close"></button> No Subscans has been initiated for any subdomains. You can select individual subdomains and initiate subscans like Directory Fuzzing, Vulnerability Scan etc. </div> `); } }); } function get_ips(scan_id=null, domain_id=null){ // this function will fetch and render ips in widget var url = '/api/queryIps/?'; if (scan_id) { url += `scan_id=${scan_id}`; } if (domain_id) { url += `target_id=${domain_id}`; } url += `&format=json`; $.getJSON(url, function(data) { $('#ip-address-count').empty(); for (var val in data['ips']){ ip = data['ips'][val] badge_color = ip['is_cdn'] ? 'warning' : 'primary'; if (scan_id) { $("#ip-address").append(`<span class='badge badge-soft-${badge_color} m-1 badge-link' data-toggle="tooltip" title="${ip['ports'].length} Ports Open." onclick="get_ip_details('${ip['address']}', scan_id=${scan_id}, domain_id=null)">${ip['address']}</span>`); } else if (domain_id) { $("#ip-address").append(`<span class='badge badge-soft-${badge_color} m-1 badge-link' data-toggle="tooltip" title="${ip['ports'].length} Ports Open." onclick="get_ip_details('${ip['address']}', scan_id=null, domain_id=${domain_id})">${ip['address']}</span>`); } // $("#ip-address").append(`<span class='badge badge-soft-${badge_color} m-1' data-toggle="modal" data-target="#tabsModal">${ip['address']}</span>`); } $('#ip-address-count').html(`<span class="badge badge-soft-primary me-1">${data['ips'].length}</span>`); $("body").tooltip({ selector: '[data-toggle=tooltip]' }); }); } function get_technologies(scan_id=null, domain_id=null){ // this function will fetch and render tech in widget var url = '/api/queryTechnologies/?'; if (scan_id) { url += `scan_id=${scan_id}`; } if (domain_id) { url += `target_id=${domain_id}`; } url += `&format=json`; $.getJSON(url, function(data) { $('#technologies-count').empty(); for (var val in data['technologies']){ tech = data['technologies'][val] if (scan_id) { $("#technologies").append(`<span class='badge badge-soft-primary m-1 badge-link' data-toggle="tooltip" title="${tech['count']} Subdomains use this technology." onclick="get_tech_details('${tech['name']}', scan_id=${scan_id}, domain_id=null)">${tech['name']}</span>`); } else if (domain_id) { $("#technologies").append(`<span class='badge badge-soft-primary m-1 badge-link' data-toggle="tooltip" title="${tech['count']} Subdomains use this technology." onclick="get_tech_details('${tech['name']}', scan_id=null, domain_id=${domain_id})">${tech['name']}</span>`); } } $('#technologies-count').html(`<span class="badge badge-soft-primary me-1">${data['technologies'].length}</span>`); $("body").tooltip({ selector: '[data-toggle=tooltip]' }); }); } function get_ports(scan_id=null, domain_id=null){ // this function will fetch and render ports in widget var url = '/api/queryPorts/?'; if (scan_id) { url += `scan_id=${scan_id}`; } if (domain_id) { url += `target_id=${domain_id}`; } url += `&format=json`; $.getJSON(url, function(data) { $('#ports-count').empty(); for (var val in data['ports']){ port = data['ports'][val] badge_color = port['is_uncommon'] ? 'danger' : 'primary'; if (scan_id) { $("#ports").append(`<span class='badge badge-soft-${badge_color} m-1 badge-link' data-toggle="tooltip" title="${port['description']}" onclick="get_port_details('${port['number']}', scan_id=${scan_id}, domain_id=null)">${port['number']}/${port['service_name']}</span>`); } else if (domain_id){ $("#ports").append(`<span class='badge badge-soft-${badge_color} m-1 badge-link' data-toggle="tooltip" title="${port['description']}" onclick="get_port_details('${port['number']}', scan_id=null, domain_id=${domain_id})">${port['number']}/${port['service_name']}</span>`); } } $('#ports-count').html(`<span class="badge badge-soft-primary me-1">${data['ports'].length}</span>`); $("body").tooltip({ selector: '[data-toggle=tooltip]' }); }); } function get_ip_details(ip_address, scan_id=null, domain_id=null){ var port_url = `/api/queryPorts/?ip_address=${ip_address}`; var subdomain_url = `/api/querySubdomains/?ip_address=${ip_address}`; if (scan_id) { port_url += `&scan_id=${scan_id}`; subdomain_url += `&scan_id=${scan_id}`; } else if(domain_id){ port_url += `&target_id=${domain_id}`; subdomain_url += `&target_id=${domain_id}`; } port_url += `&format=json`; subdomain_url += `&format=json`; var interesting_badge = `<span class="m-1 badge badge-soft-danger bs-tooltip" title="Interesting Subdomain">Interesting</span>`; var port_loader = `<span class="inner-div spinner-border text-primary align-self-center loader-sm" id="port-modal-loader"></span>`; var subdomain_loader = `<span class="inner-div spinner-border text-primary align-self-center loader-sm" id="subdomain-modal-loader"></span>`; // add tab modal title $('#modal_title').html('Details for IP: <b>' + ip_address + '</b>'); $('#modal-content').empty(); $('#modal-tabs').empty(); $('#modal-content').append(`<ul class='nav nav-tabs nav-bordered' id="modal_tab_nav"></ul><div id="modal_tab_content" class="tab-content"></div>`); $('#modal_tab_nav').append(`<li class="nav-item"><a class="nav-link active" data-bs-toggle="tab" href="#modal_content_port" aria-expanded="true"><span id="modal-open-ports-count"></span>Open Ports &nbsp;${port_loader}</a></li>`); $('#modal_tab_nav').append(`<li class="nav-item"><a class="nav-link" data-bs-toggle="tab" href="#modal_content_subdomain" aria-expanded="false"><span id="modal-subdomain-count"></span>Subdomains &nbsp;${subdomain_loader}</a></li>`) // add content area $('#modal_tab_content').empty(); $('#modal_tab_content').append(`<div class="tab-pane show active" id="modal_content_port"></div><div class="tab-pane" id="modal_content_subdomain"></div>`); $('#modal-open-ports').append(`<div class="modal-text" id="modal-text-open-port"></div>`); $('#modal-text-open-port').append(`<ul id="modal-open-port-text"></ul>`); $('#modal_content_port').append(`<ul id="modal_port_ul"></ul>`); $('#modal_content_subdomain').append(`<ul id="modal_subdomain_ul"></ul>`); $.getJSON(port_url, function(data) { $('#modal_content_port').empty(); $('#modal_content_port').append(`<p> IP Addresses ${ip_address} has ${data['ports'].length} Open Ports`); $('#modal-open-ports-count').html(`<b>${data['ports'].length}</b>&nbsp;&nbsp;`); for (port in data['ports']){ port_obj = data['ports'][port]; badge_color = port_obj['is_uncommon'] ? 'danger' : 'info'; $("#modal_content_port").append(`<li class="mt-1">${port_obj['description']} <b class="text-${badge_color}">(${port_obj['number']}/${port_obj['service_name']})</b></li>`) } $("#port-modal-loader").remove(); }); $('#modal_dialog').modal('show'); // query subdomains $.getJSON(subdomain_url, function(data) { $('#modal_content_subdomain').empty(); $('#modal_content_subdomain').append(`<p>${data['subdomains'].length} Subdomains are associated with IP ${ip_address}`); $('#modal-subdomain-count').html(`<b>${data['subdomains'].length}</b>&nbsp;&nbsp;`); for (subdomain in data['subdomains']){ subdomain_obj = data['subdomains'][subdomain]; badge_color = subdomain_obj['http_status'] >= 400 ? 'danger' : ''; li_id = get_randid(); if (subdomain_obj['http_url']) { $("#modal_content_subdomain").append(`<li class="mt-1" id="${li_id}"><a href='${subdomain_obj['http_url']}' target="_blank" class="text-${badge_color}">${subdomain_obj['name']}</a></li>`) } else { $("#modal_content_subdomain").append(`<li class="mt-1 text-${badge_color}" id="${li_id}">${subdomain_obj['name']}</li>`); } if (subdomain_obj['http_status']) { $("#"+li_id).append(get_http_badge(subdomain_obj['http_status'])); $('.bs-tooltip').tooltip(); } if (subdomain_obj['is_interesting']) { $("#"+li_id).append(interesting_badge) } } $("#modal-text-subdomain").append(`<span class="float-end text-danger">*Subdomains highlighted are 40X HTTP Status</span>`); $("#subdomain-modal-loader").remove(); }); } function get_port_details(port, scan_id=null, domain_id=null){ var ip_url = `/api/queryIps/?port=${port}`; var subdomain_url = `/api/querySubdomains/?port=${port}`; if (scan_id) { ip_url += `&scan_id=${scan_id}`; subdomain_url += `&scan_id=${scan_id}`; } else if(domain_id){ ip_url += `&target_id=${domain_id}`; subdomain_url += `&target_id=${domain_id}`; } ip_url += `&format=json`; subdomain_url += `&format=json`; var interesting_badge = `<span class="m-1 badge badge-soft-danger bs-tooltip" title="Interesting Subdomain">Interesting</span>`; var ip_spinner = `<span class="spinner-border spinner-border-sm me-1" id="ip-modal-loader"></span>`; var subdomain_spinner = `<span class="spinner-border spinner-border-sm me-1" id="subdomain-modal-loader"></span>`; $('#modal_title').html('Details for Port: <b>' + port + '</b>'); $('#modal-content').empty(); $('#modal-tabs').empty(); $('#modal-content').append(`<ul class='nav nav-tabs nav-bordered' id="modal_tab_nav"></ul><div id="modal_tab_content" class="tab-content"></div>`); $('#modal_tab_nav').append(`<li class="nav-item"><a class="nav-link active" data-bs-toggle="tab" href="#modal_content_ip" aria-expanded="true"><span id="modal-ip-count"></span>IP Address&nbsp;${ip_spinner}</a></li>`); $('#modal_tab_nav').append(`<li class="nav-item"><a class="nav-link" data-bs-toggle="tab" href="#modal_content_subdomain" aria-expanded="false"><span id="modal-subdomain-count"></span>Subdomains&nbsp;${subdomain_spinner}</a></li>`) // add content area $('#modal_tab_content').append(`<div class="tab-pane show active" id="modal_content_ip"></div><div class="tab-pane" id="modal_content_subdomain"></div>`); $('#modal_content_ip').append(`<ul id="modal_ip_ul"></ul>`); $('#modal_content_subdomain').append(`<ul id="modal_subdomain_ul"></ul>`); $('#modal_dialog').modal('show'); $.getJSON(ip_url, function(data) { $('#modal_ip_ul').empty(); $('#modal_ip_ul').append(`<p>${data['ips'].length} IP Addresses have Port ${port} Open`); $('#modal-ip-count').html(`<b>${data['ips'].length}</b>&nbsp;&nbsp;`); for (ip in data['ips']){ ip_obj = data['ips'][ip]; text_color = ip_obj['is_cdn'] ? 'warning' : ''; $("#modal_ip_ul").append(`<li class='mt-1 text-${text_color}'>${ip_obj['address']}</li>`) } $('#modal_ip_ul').append(`<span class="float-end text-warning">*IP Address highlighted are CDN IP Address</span>`); $("#ip-modal-loader").remove(); }); // query subdomains $.getJSON(subdomain_url, function(data) { $('#modal_subdomain_ul').empty(); $('#modal_subdomain_ul').append(`<p>${data['subdomains'].length} Subdomains have Port ${port} Open`); $('#modal-subdomain-count').html(`<b>${data['subdomains'].length}</b>&nbsp;&nbsp;`); for (subdomain in data['subdomains']){ subdomain_obj = data['subdomains'][subdomain]; badge_color = subdomain_obj['http_status'] >= 400 ? 'danger' : ''; li_id = get_randid(); if (subdomain_obj['http_url']) { $("#modal_subdomain_ul").append(`<li id="${li_id}" class="mt-1"><a href='${subdomain_obj['http_url']}' target="_blank" class="text-${badge_color}">${subdomain_obj['name']}</a></li>`) } else { $("#modal_subdomain_ul").append(`<li class="mt-1 text-${badge_color}" id="${li_id}">${subdomain_obj['name']}</li>`); } if (subdomain_obj['http_status']) { $("#"+li_id).append(get_http_badge(subdomain_obj['http_status'])); $('.bs-tooltip').tooltip(); } if (subdomain_obj['is_interesting']) { $("#"+li_id).append(interesting_badge) } } $("#modal_subdomain_ul").append(`<span class="float-end text-danger">*Subdomains highlighted are 40X HTTP Status</span>`); $("#subdomain-modal-loader").remove(); }); } function get_tech_details(tech, scan_id=null, domain_id=null){ var url = `/api/querySubdomains/?tech=${tech}`; if (scan_id) { url += `&scan_id=${scan_id}`; } else if(domain_id){ url += `&target_id=${domain_id}`; } url += `&format=json`; var interesting_badge = `<span class="m-1 badge badge-soft-danger bs-tooltip" title="Interesting Subdomain">Interesting</span>`; // render tab modal $('.modal-title').html('Details for Technology: <b>' + tech + '</b>'); $('#modal_dialog').modal('show'); $('.modal-text').empty(); $('#modal-footer').empty(); $('.modal-text').append(`<div class='outer-div' id="modal-loader"><span class="inner-div spinner-border text-primary align-self-center loader-sm"></span></div>`); // query subdomains $.getJSON(url, function(data) { $('#modal-loader').empty(); $('#modal-content').empty(); $('#modal-content').append(`${data['subdomains'].length} Subdomains are using ${tech}`); for (subdomain in data['subdomains']){ subdomain_obj = data['subdomains'][subdomain]; badge_color = subdomain_obj['http_status'] >= 400 ? 'danger' : ''; li_id = get_randid(); if (subdomain_obj['http_url']) { $("#modal-content").append(`<li id="${li_id}"><a href='${subdomain_obj['http_url']}' target="_blank" class="text-${badge_color}">${subdomain_obj['name']}</a></li>`) } else { $("#modal-content").append(`<li class="text-${badge_color}" id="${li_id}">${subdomain_obj['name']}</li>`); } if (subdomain_obj['http_status']) { $("#"+li_id).append(get_http_badge(subdomain_obj['http_status'])); $('.bs-tooltip').tooltip(); } if (subdomain_obj['is_interesting']) { $("#"+li_id).append(interesting_badge) } } $("#modal-content").append(`<span class="float-end text-danger">*Subdomains highlighted are 40X HTTP Status</span>`); $("#subdomain-modal-loader").remove(); }).fail(function(){ $('#modal-loader').empty(); }); } function get_http_badge(http_status){ switch (true) { case (http_status >= 400): badge_color = 'danger' break; case (http_status >= 300): badge_color = 'warning' break; case (http_status >= 200): badge_color = 'success' break; default: badge_color = 'danger' } if (http_status) { badge = `<span class="badge badge-soft-${badge_color} me-1 ms-1 bs-tooltip" data-placement="top" title="HTTP Status">${http_status}</span>`; return badge } } function get_and_render_cve_details(cve_id){ var api_url = `/api/tools/cve_details/?cve_id=${cve_id}&format=json`; Swal.fire({ title: 'Fetching CVE Details...' }); swal.showLoading(); fetch(api_url, { method: 'GET', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken"), "Content-Type": "application/json" }, }).then(response => response.json()).then(function(response) { console.log(response); swal.close(); if (response.status) { $('#xl-modal-title').empty(); $('#xl-modal-content').empty(); $('#xl-modal-footer').empty(); $('#xl-modal_title').html(`CVE Details of ${cve_id}`); var cvss_score_badge = 'danger'; if (response.result.cvss > 0.1 && response.result.cvss <= 3.9) { cvss_score_badge = 'info'; } else if (response.result.cvss > 3.9 && response.result.cvss <= 6.9) { cvss_score_badge = 'warning'; } content = `<div class="row mt-3"> <div class="col-sm-3"> <div class="nav flex-column nav-pills nav-pills-tab" id="v-pills-tab" role="tablist" aria-orientation="vertical"> <a class="nav-link active show mb-1" id="v-pills-cve-details-tab" data-bs-toggle="pill" href="#v-pills-cve-details" role="tab" aria-controls="v-pills-cve-details-tab" aria-selected="true">CVE Details</a> <a class="nav-link mb-1" id="v-pills-affected-products-tab" data-bs-toggle="pill" href="#v-pills-affected-products" role="tab" aria-controls="v-pills-affected-products-tab" aria-selected="true">Affected Products</a> <a class="nav-link mb-1" id="v-pills-affected-versions-tab" data-bs-toggle="pill" href="#v-pills-affected-versions" role="tab" aria-controls="v-pills-affected-versions-tab" aria-selected="true">Affected Versions</a> <a class="nav-link mb-1" id="v-pills-cve-references-tab" data-bs-toggle="pill" href="#v-pills-cve-references" role="tab" aria-controls="v-pills-cve-references-tab" aria-selected="true">References</a> </div> </div> <div class="col-sm-9"> <div class="tab-content pt-0">`; content += ` <div class="tab-pane fade active show" id="v-pills-cve-details" role="tabpanel" aria-labelledby="v-pills-cve-details-tab" data-simplebar style="max-height: 600px; min-height: 600px;"> <h4 class="header-title">${cve_id}</h4> <div class="alert alert-warning" role="alert"> ${response.result.summary} </div> <span class="badge badge-soft-primary">Assigner: ${response.result.assigner}</span> <span class="badge badge-outline-primary">CVSS Vector: ${response.result['cvss-vector']}</span> <table class="domain_details_table table table-hover table-borderless"> <tr style="display: none"> <th>&nbsp;</th> <th>&nbsp;</th> </tr> <tr> <td>CVSS Score</td> <td><span class="badge badge-soft-${cvss_score_badge}">${response.result.cvss ? response.result.cvss: "-"}</span></td> </tr> <tr> <td>Confidentiality Impact</td> <td>${response.result.impact.confidentiality ? response.result.impact.confidentiality: "N/A"}</td> </tr> <tr> <td>Integrity Impact</td> <td>${response.result.impact.integrity ? response.result.impact.integrity: "N/A"}</td> </tr> <tr> <td>Availability Impact</td> <td>${response.result.impact.availability ? response.result.impact.availability: "N/A"}</td> </tr> <tr> <td>Access Complexity</td> <td>${response.result.access.complexity ? response.result.access.complexity: "N/A"}</td> </tr> <tr> <td>Authentication</td> <td>${response.result.access.authentication ? response.result.access.authentication: "N/A"}</td> </tr> <tr> <td>CWE ID</td> <td><span class="badge badge-outline-danger">${response.result.cwe ? response.result.cwe: "N/A"}</span></td> </tr> </table> </div> `; content += `<div class="tab-pane fade" id="v-pills-cve-references" role="tabpanel" aria-labelledby="v-pills-cve-references-tab" data-simplebar style="max-height: 600px; min-height: 600px;"> <ul>`; for (var reference in response.result.references) { content += `<li><a href="${response.result.references[reference]}" target="_blank">${response.result.references[reference]}</a></li>`; } content += `</ul></div>`; content += `<div class="tab-pane fade" id="v-pills-affected-products" role="tabpanel" aria-labelledby="v-pills-affected-products-tab" data-simplebar style="max-height: 600px; min-height: 600px;"> <ul>`; for (var prod in response.result.vulnerable_product) { content += `<li>${response.result.vulnerable_product[prod]}</li>`; } content += `</ul></div>`; content += `<div class="tab-pane fade" id="v-pills-affected-versions" role="tabpanel" aria-labelledby="v-pills-affected-versions-tab" data-simplebar style="max-height: 600px; min-height: 600px;"> <ul>`; for (var conf in response.result.vulnerable_configuration) { content += `<li>${response.result.vulnerable_configuration[conf]['id']}</li>`; } content += `</ul></div>`; content += `</div></div></div>`; $('#xl-modal-content').append(content); $('#modal_xl_scroll_dialog').modal('show'); $("body").tooltip({ selector: '[data-toggle=tooltip]' }); } else{ swal.fire("Error!", response.message, "error", { button: "Okay", }); } }); } function get_most_vulnerable_target(scan_id=null, target_id=null, ignore_info=false, limit=50){ $('#most_vulnerable_target_div').empty(); $('#most_vulnerable_spinner').append(`<div class="spinner-border text-primary m-2" role="status"></div>`); var data = {}; if (scan_id) { data['scan_history_id'] = scan_id; } else if (target_id) { data['target_id'] = target_id; } data['ignore_info'] = ignore_info; data['limit'] = limit; fetch('/api/fetch/most_vulnerable/?format=json', { method: 'POST', credentials: "same-origin", body: JSON.stringify(data), headers: { "X-CSRFToken": getCookie("csrftoken"), "Content-Type": 'application/json', } }).then(function(response) { return response.json(); }).then(function(response) { $('#most_vulnerable_spinner').empty(); if (response.status) { $('#most_vulnerable_target_div').append(` <table class="table table-borderless table-nowrap table-hover table-centered m-0"> <thead> <tr> <th style="width: 60%">Target</th> <th style="width: 30%">Vulnerabilities Count</th> </tr> </thead> <tbody id="most_vulnerable_target_tbody"> </tbody> </table> `); for (var res in response.result) { var targ_obj = response.result[res]; var tr = `<tr onclick="window.location='/scan/detail/vuln?domain=${targ_obj.name}';" style="cursor: pointer;">`; if (scan_id || target_id) { tr = `<tr onclick="window.location='/scan/detail/vuln?subdomain=${targ_obj.name}';" style="cursor: pointer;">`; } $('#most_vulnerable_target_tbody').append(` ${tr} <td> <h5 class="m-0 fw-normal">${targ_obj.name}</h5> </td> <td> <span class="badge badge-outline-danger">${targ_obj.vuln_count} Vulnerabilities</span> </td> </tr> `); } } else{ $('#most_vulnerable_target_div').append(` <div class="mt-4 alert alert-warning"> Could not find most vulnerable targets. </br> Once the vulnerability scan is performed, reNgine will identify the most vulnerable targets.</div> `); } }); } function get_most_common_vulnerability(scan_id=null, target_id=null, ignore_info=false, limit=50){ $('#most_common_vuln_div').empty(); $('#most_common_vuln_spinner').append(`<div class="spinner-border text-primary m-2" role="status"></div>`); var data = {}; if (scan_id) { data['scan_history_id'] = scan_id; } else if (target_id) { data['target_id'] = target_id; } data['ignore_info'] = ignore_info; data['limit'] = limit; fetch('/api/fetch/most_common_vulnerability/?format=json', { method: 'POST', credentials: "same-origin", body: JSON.stringify(data), headers: { "X-CSRFToken": getCookie("csrftoken"), "Content-Type": 'application/json', } }).then(function(response) { return response.json(); }).then(function(response) { $('#most_common_vuln_spinner').empty(); if (response.status) { $('#most_common_vuln_div').append(` <table class="table table-borderless table-nowrap table-hover table-centered m-0"> <thead> <tr> <th style="width: 60%">Vulnerability Name</th> <th style="width: 20%">Count</th> <th style="width: 20%">Severity</th> </tr> </thead> <tbody id="most_common_vuln_tbody"> </tbody> </table> `); for (var res in response.result) { var vuln_obj = response.result[res]; var vuln_badge = ''; switch (vuln_obj.severity) { case -1: vuln_badge = get_severity_badge('Unknown'); break; case 0: vuln_badge = get_severity_badge('Info'); break; case 1: vuln_badge = get_severity_badge('Low'); break; case 2: vuln_badge = get_severity_badge('Medium'); break; case 3: vuln_badge = get_severity_badge('High'); break; case 4: vuln_badge = get_severity_badge('Critical'); break; default: vuln_badge = get_severity_badge('Unknown'); } $('#most_common_vuln_tbody').append(` <tr onclick="window.location='/scan/detail/vuln?vulnerability_name=${vuln_obj.name}';" style="cursor: pointer;"> <td> <h5 class="m-0 fw-normal">${vuln_obj.name}</h5> </td> <td> <span class="badge badge-outline-danger">${vuln_obj.count}</span> </td> <td> ${vuln_badge} </td> </tr> `); } } else{ $('#most_common_vuln_div').append(` <div class="mt-4 alert alert-warning"> Could not find Most Common Vulnerabilities. </br> Once the vulnerability scan is performed, reNgine will identify the Most Common Vulnerabilities.</div> `); } }); } function highlight_search(search_keyword, content){ // this function will send the highlighted text from search keyword var reg = new RegExp('('+search_keyword+')', 'gi'); return content.replace(reg, '<mark>$1</mark>'); } function validURL(str) { // checks for valid http url var pattern = new RegExp('^(https?:\\/\\/)?'+ // protocol '((([a-z\\d]([a-z\\d-]*[a-z\\d])*)\\.)+[a-z]{2,}|'+ // domain name '((\\d{1,3}\\.){3}\\d{1,3}))'+ // OR ip (v4) address '(\\:\\d+)?(\\/[-a-z\\d%_.~+]*)*'+ // port and path '(\\?[;&a-z\\d%_.~+=-]*)?'+ // query string '(\\#[-a-z\\d_]*)?$','i'); // fragment locator return !!pattern.test(str); }

function checkall(clickchk, relChkbox) { var checker = $('#' + clickchk); var multichk = $('.' + relChkbox); checker.click(function() { multichk.prop('checked', $(this).prop('checked')); }); } function multiCheck(tb_var) { tb_var.on("change", ".chk-parent", function() { var e = $(this).closest("table").find("td:first-child .child-chk"), a = $(this).is(":checked"); $(e).each(function() { a ? ($(this).prop("checked", !0), $(this).closest("tr").addClass("active")) : ($(this).prop("checked", !1), $(this).closest("tr").removeClass("active")) }) }), tb_var.on("change", "tbody tr .new-control", function() { $(this).parents("tr").toggleClass("active") }) } function GetIEVersion() { var sAgent = window.navigator.userAgent; var Idx = sAgent.indexOf("MSIE"); // If IE, return version number. if (Idx > 0) return parseInt(sAgent.substring(Idx + 5, sAgent.indexOf(".", Idx))); // If IE 11 then look for Updated user agent string. else if (!!navigator.userAgent.match(/Trident\/7\./)) return 11; else return 0; //It is not IE } function truncate(str, n) { return (str.length > n) ? str.substr(0, n - 1) + '&hellip;' : str; }; function return_str_if_not_null(val) { return val ? val : ''; } // seperate hostname and url // Referenced from https://stackoverflow.com/questions/736513/how-do-i-parse-a-url-into-hostname-and-path-in-javascript function getParsedURL(url) { var parser = new URL(url); return parser.pathname + parser.search; }; function getCookie(name) { var cookieValue = null; if (document.cookie && document.cookie !== '') { var cookies = document.cookie.split(';'); for (var i = 0; i < cookies.length; i++) { var cookie = jQuery.trim(cookies[i]); // Does this cookie string begin with the name we want? if (cookie.substring(0, name.length + 1) === (name + '=')) { cookieValue = decodeURIComponent(cookie.substring(name.length + 1)); break; } } } return cookieValue; } // Source: https://portswigger.net/web-security/cross-site-scripting/preventing#encode-data-on-output function htmlEncode(str) { return String(str).replace(/[^\w. ]/gi, function(c) { return '&#' + c.charCodeAt(0) + ';'; }); } // Source: https://portswigger.net/web-security/cross-site-scripting/preventing#encode-data-on-output function jsEscape(str) { return String(str).replace(/[^\w. ]/gi, function(c) { return '\\u' + ('0000' + c.charCodeAt(0).toString(16)).slice(-4); }); } function deleteScheduledScan(id) { const delAPI = "../delete/scheduled_task/" + id; swal.queue([{ title: 'Are you sure you want to delete this?', text: "This action can not be undone.", icon: 'warning', showCancelButton: true, confirmButtonText: 'Delete', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { return fetch(delAPI, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }).then(function(response) { return response.json(); }).then(function(data) { // TODO Look for better way return location.reload(); }).catch(function() { swal.insertQueueStep({ icon: 'error', title: 'Oops! Unable to delete the scheduled task!' }) }) } }]) } function change_scheduled_task_status(id, checkbox) { if (checkbox.checked) { text_msg = 'Schedule Scan Started'; } else { text_msg = 'Schedule Scan Stopped'; } Snackbar.show({ text: text_msg, pos: 'top-right', duration: 2500 }); const taskStatusApi = "../toggle/scheduled_task/" + id; return fetch(taskStatusApi, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }) } function change_vuln_status(id) { const vulnStatusApi = "../toggle/vuln_status/" + id; return fetch(vulnStatusApi, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }) } // splits really long strings into multiple lines // Souce: https://stackoverflow.com/a/52395960 function split_into_lines(str, maxWidth) { const newLineStr = "</br>"; done = false; res = ''; do { found = false; // Inserts new line at first whitespace of the line for (i = maxWidth - 1; i >= 0; i--) { if (test_white_space(str.charAt(i))) { res = res + [str.slice(0, i), newLineStr].join(''); str = str.slice(i + 1); found = true; break; } } // Inserts new line at maxWidth position, the word is too long to wrap if (!found) { res += [str.slice(0, maxWidth), newLineStr].join(''); str = str.slice(maxWidth); } if (str.length < maxWidth) done = true; } while (!done); return res + str; } function test_white_space(x) { const white = new RegExp(/^\s$/); return white.test(x.charAt(0)); }; // span values function will seperate the values by comma and put badge around it function parse_comma_values_into_span(data, color, outline = null) { if (data) { var badge = `<span class='badge badge-soft-` + color + ` m-1'>`; var data_with_span = ""; data.split(/\s*,\s*/).forEach(function(split_vals) { data_with_span += badge + split_vals + "</span>"; }); return data_with_span; } return ''; } function get_severity_badge(severity) { switch (severity) { case 'Info': return "<span class='badge badge-soft-primary'>&nbsp;&nbsp;INFO&nbsp;&nbsp;</span>"; break; case 'Low': return "<span class='badge badge-low'>&nbsp;&nbsp;LOW&nbsp;&nbsp;</span>"; break; case 'Medium': return "<span class='badge badge-soft-warning'>&nbsp;&nbsp;MEDIUM&nbsp;&nbsp;</span>"; break; case 'High': return "<span class='badge badge-soft-danger'>&nbsp;&nbsp;HIGH&nbsp;&nbsp;</span>"; break; case 'Critical': return "<span class='badge badge-critical'>&nbsp;&nbsp;CRITICAL&nbsp;&nbsp;</span>"; break; case 'Unknown': return "<span class='badge badge-soft-info'>&nbsp;&nbsp;UNKNOWN&nbsp;&nbsp;</span>"; default: return ""; } } // Source: https://stackoverflow.com/a/54733055 function typingEffect(words, id, i) { let word = words[i].split(""); var loopTyping = function() { if (word.length > 0) { let elem = document.getElementById(id); elem.setAttribute('placeholder', elem.getAttribute('placeholder') + word.shift()); } else { deletingEffect(words, id, i); return false; }; timer = setTimeout(loopTyping, 150); }; loopTyping(); }; function deletingEffect(words, id, i) { let word = words[i].split(""); var loopDeleting = function() { if (word.length > 0) { word.pop(); document.getElementById(id).setAttribute('placeholder', word.join("")); } else { if (words.length > (i + 1)) { i++; } else { i = 0; }; typingEffect(words, id, i); return false; }; timer = setTimeout(loopDeleting, 90); }; loopDeleting(); }; function fullScreenDiv(id, btn) { let fullscreen = document.querySelector(id); let button = document.querySelector(btn); document.fullscreenElement && document.exitFullscreen() || document.querySelector(id).requestFullscreen() fullscreen.setAttribute("style", "overflow:auto"); } function get_randid() { return '_' + Math.random().toString(36).substr(2, 9); } function delete_all_scan_results() { const delAPI = "../scan/delete/scan_results/"; swal.queue([{ title: 'Are you sure you want to delete all scan results?', text: "You won't be able to revert this!", icon: 'warning', showCancelButton: true, confirmButtonText: 'Delete', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { return fetch(delAPI, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }).then(function(response) { return response.json(); }).then(function(data) { // TODO Look for better way return location.reload(); }).catch(function() { swal.insertQueueStep({ icon: 'error', title: 'Oops! Unable to delete Delete scan results!' }) }) } }]) } function delete_all_screenshots() { const delAPI = "../scan/delete/screenshots/"; swal.queue([{ title: 'Are you sure you want to delete all Screenshots?', text: "You won't be able to revert this!", icon: 'warning', showCancelButton: true, confirmButtonText: 'Delete', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { return fetch(delAPI, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }).then(function(response) { return response.json(); }).then(function(data) { // TODO Look for better way return location.reload(); }).catch(function() { swal.insertQueueStep({ icon: 'error', title: 'Oops! Unable to delete Empty Screenshots!' }) }) } }]) } function load_image_from_url(src, append_to_id) { img = document.createElement('img'); img.src = src; img.style.width = '100%'; document.getElementById(append_to_id).appendChild(img); } function setTooltip(btn, message) { hide_all_tooltips(); const instance = tippy(document.querySelector(btn)); instance.setContent(message); instance.show(); setTimeout(function() { instance.hide(); }, 500); } function hide_all_tooltips() { $(".tooltip").tooltip("hide"); } function get_response_time_text(response_time) { if (response_time) { var text_color = 'danger'; if (response_time < 0.5) { text_color = 'success' } else if (response_time >= 0.5 && response_time < 1) { text_color = 'warning' } return `<span class="text-${text_color}">${response_time.toFixed(4)}s</span>`; } return ''; } function parse_technology(data, color, scan_id = null, domain_id=null) { var badge = `<span data-toggle="tooltip" title="Technology" class='badge-link badge badge-soft-` + color + ` mt-1 me-1'`; var data_with_span = ""; for (var key in data) { if (scan_id) { data_with_span += badge + ` onclick="get_tech_details('${data[key]['name']}', ${scan_id}, domain_id=null)">` + data[key]['name'] + "</span>"; } else if (domain_id) { data_with_span += badge + ` onclick="get_tech_details('${data[key]['name']}', scan_id=null, domain_id=domain_id)">` + data[key]['name'] + "</span>"; } } return data_with_span; } // span values function will seperate the values by comma and put badge around it function parse_ip(data, cdn) { if (cdn) { var badge = `<span class='badge badge-soft-warning m-1 bs-tooltip' title="CDN IP Address">`; } else { var badge = `<span class='badge badge-soft-primary m-1'>`; } var data_with_span = ""; data.split(/\s*,\s*/).forEach(function(split_vals) { data_with_span += badge + split_vals + "</span>"; }); return data_with_span; } //to remove the image element if there is no screenshot captured function removeImageElement(element) { element.parentElement.remove(); } // https://stackoverflow.com/a/18197341/9338140 function download(filename, text) { var element = document.createElement('a'); element.setAttribute('href', 'data:text/plain;charset=utf-8,' + encodeURIComponent(text)); element.setAttribute('download', filename); element.style.display = 'none'; document.body.appendChild(element); element.click(); document.body.removeChild(element); } function vuln_status_change(checkbox, id) { if (checkbox.checked) { checkbox.parentNode.parentNode.parentNode.className = "table-success text-strike"; } else { checkbox.parentNode.parentNode.parentNode.classList.remove("table-success"); checkbox.parentNode.parentNode.parentNode.classList.remove("text-strike"); } change_vuln_status(id); } function report_hackerone(vulnerability_id, severity) { message = "" if (severity == 'Info' || severity == 'Low' || severity == 'Medium') { message = "We do not recommended sending this vulnerability report to hackerone due to the severity, do you still want to report this?" } else { message = "This vulnerability report will be sent to Hackerone."; } const vulnerability_report_api = "../../api/vulnerability/report/?vulnerability_id=" + vulnerability_id; swal.queue([{ title: 'Reporting vulnerability to hackerone', text: message, icon: 'warning', showCancelButton: true, confirmButtonText: 'Report', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { return fetch(vulnerability_report_api, { method: 'GET', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }).then(function(response) { return response.json(); }).then(function(data) { console.log(data.status) if (data.status == 111) { swal.insertQueueStep({ icon: 'error', title: 'Target does not has team_handle to send report to.' }) } else if (data.status == 201) { swal.insertQueueStep({ icon: 'success', title: 'Vulnerability report successfully submitted to hackerone.' }) } else if (data.status == 400) { swal.insertQueueStep({ icon: 'error', title: 'Invalid Report.' }) } else if (data.status == 401) { swal.insertQueueStep({ icon: 'error', title: 'Hackerone authentication failed.' }) } else if (data.status == 403) { swal.insertQueueStep({ icon: 'error', title: 'API Key forbidden by Hackerone.' }) } else if (data.status == 423) { swal.insertQueueStep({ icon: 'error', title: 'Too many requests.' }) } }).catch(function() { swal.insertQueueStep({ icon: 'error', title: 'Oops! Unable to send vulnerability report to hackerone, check your target team_handle or hackerone configurarions!' }) }) } }]) } function get_interesting_subdomains(target_id, scan_history_id) { if (target_id) { url = `/api/listInterestingEndpoints/?target_id=${target_id}&format=datatables`; non_orderable_targets = [0, 1, 2, 3]; } else if (scan_history_id) { url = `/api/listInterestingSubdomains/?scan_id=${scan_history_id}&format=datatables`; non_orderable_targets = []; } var interesting_subdomain_table = $('#interesting_subdomains').DataTable({ "drawCallback": function(settings, start, end, max, total, pre) { // if no interesting subdomains are found, hide the datatable and show no interesting subdomains found badge if (this.fnSettings().fnRecordsTotal() == 0) { $('#interesting_subdomain_div').empty(); // $('#interesting_subdomain_div').append(`<div class="card-header bg-primary py-3 text-white"> // <div class="card-widgets"> // <a href="#" data-toggle="remove"><i class="mdi mdi-close"></i></a> // </div> // <h5 class="card-title mb-0 text-white"><i class="mdi mdi-fire-alert me-2"></i>Interesting subdomains could not be identified</h5> // </div> // <div id="cardCollpase4" class="collapse show"> // <div class="card-body"> // reNgine could not identify any interesting subdomains. You can customize interesting subdomain keywords <a href="/scanEngine/interesting/lookup/">from here</a> and this section would be automatically updated. // </div> // </div>`); } else { // show nav bar $('.interesting-tab-show').removeAttr('style'); $('#interesting_subdomain_alert_count').html(`${this.fnSettings().fnRecordsTotal()} Interesting Subdomains`) $('#interesting_subdomain_count_badge').empty(); $('#interesting_subdomain_count_badge').html(`<span class="badge badge-soft-primary me-1">${this.fnSettings().fnRecordsTotal()}</span>`); } }, "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "processing": true, "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", "destroy": true, "bInfo": false, "stripeClasses": [], 'serverSide': true, "ajax": url, "order": [ [3, "desc"] ], "lengthMenu": [5, 10, 20, 50, 100], "pageLength": 10, "columns": [{ 'data': 'name' }, { 'data': 'page_title' }, { 'data': 'http_status' }, { 'data': 'content_length' }, { 'data': 'http_url' }, { 'data': 'technologies' }, ], "columnDefs": [{ "orderable": false, "targets": non_orderable_targets }, { "targets": [4], "visible": false, "searchable": false, }, { "targets": [5], "visible": false, "searchable": true, }, { "className": "text-center", "targets": [2] }, { "render": function(data, type, row) { tech_badge = ''; if (row['technologies']) { // tech_badge = `</br>` + parse_technology(row['technologies'], "primary", outline=true, scan_id=null); } if (row['http_url']) { return `<a href="` + row['http_url'] + `" class="text-primary" target="_blank">` + data + `</a>` + tech_badge; } return `<a href="https://` + data + `" class="text-primary" target="_blank">` + data + `</a>` + tech_badge; }, "targets": 0 }, { "render": function(data, type, row) { // display badge based on http status // green for http status 2XX, orange for 3XX and warning for everything else if (data >= 200 && data < 300) { return "<span class='badge badge-pills badge-soft-success'>" + data + "</span>"; } else if (data >= 300 && data < 400) { return "<span class='badge badge-pills badge-soft-warning'>" + data + "</span>"; } else if (data == 0) { // datatable throws error when no data is returned return ""; } return `<span class='badge badge-pills badge-soft-danger'>` + data + `</span>`; }, "targets": 2, }, ], }); } function get_interesting_endpoint(target_id, scan_history_id) { var non_orderable_targets = []; if (target_id) { url = `/api/listInterestingEndpoints/?target_id=${target_id}&format=datatables`; // non_orderable_targets = [0, 1, 2, 3]; } else if (scan_history_id) { url = `/api/listInterestingEndpoints/?scan_id=${scan_history_id}&format=datatables`; // non_orderable_targets = [0, 1, 2, 3]; } $('#interesting_endpoints').DataTable({ "drawCallback": function(settings, start, end, max, total, pre) { if (this.fnSettings().fnRecordsTotal() == 0) { $('#interesting_endpoint_div').remove(); } else { $('.interesting-tab-show').removeAttr('style'); $('#interesting_endpoint_alert_count').html(`, ${this.fnSettings().fnRecordsTotal()} Interesting Endpoints`) $('#interesting_endpoint_count_badge').empty(); $('#interesting_endpoint_count_badge').html(`<span class="badge badge-soft-primary me-1">${this.fnSettings().fnRecordsTotal()}</span>`); } }, "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "processing": true, "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", 'serverSide': true, "destroy": true, "bInfo": false, "ajax": url, "order": [ [3, "desc"] ], "lengthMenu": [5, 10, 20, 50, 100], "pageLength": 10, "columns": [{ 'data': 'http_url' }, { 'data': 'page_title' }, { 'data': 'http_status' }, { 'data': 'content_length' }, ], "columnDefs": [{ "orderable": false, "targets": non_orderable_targets }, { "className": "text-center", "targets": [2] }, { "render": function(data, type, row) { var url = split_into_lines(data, 70); return "<a href='" + data + "' target='_blank' class='text-primary'>" + url + "</a>"; }, "targets": 0 }, { "render": function(data, type, row) { // display badge based on http status // green for http status 2XX, orange for 3XX and warning for everything else if (data >= 200 && data < 300) { return "<span class='badge badge-pills badge-soft-success'>" + data + "</span>"; } else if (data >= 300 && data < 400) { return "<span class='badge badge-pills badge-soft-warning'>" + data + "</span>"; } else if (data == 0) { // datatable throws error when no data is returned return ""; } return `<span class='badge badge-pills badge-soft-danger'>` + data + `</span>`; }, "targets": 2, }, ], }); } function get_important_subdomains(target_id, scan_history_id) { var url = `/api/querySubdomains/?only_important&no_lookup_interesting&format=json`; if (target_id) { url += `&target_id=${target_id}`; } else if (scan_history_id) { url += `&scan_id=${scan_history_id}`; } $.getJSON(url, function(data) { $('#important-count').empty(); $('#important-subdomains-list').empty(); if (data['subdomains'].length > 0) { $('#important-count').html(`<span class="badge badge-soft-primary ms-1 me-1">${data['subdomains'].length}</span>`); for (var val in data['subdomains']) { subdomain = data['subdomains'][val]; div_id = 'important_' + subdomain['id']; $("#important-subdomains-list").append(` <div id="${div_id}"> <p> <span id="subdomain_${subdomain['id']}"> ${subdomain['name']} <span class=""> <a href="javascript:;" data-clipboard-action="copy" class="m-1 float-end badge-link text-info copyable text-primary" data-toggle="tooltip" data-placement="top" title="Copy Subdomain!" data-clipboard-target="#subdomain_${subdomain['id']}"> <svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-copy"><rect x="9" y="9" width="13" height="13" rx="2" ry="2"></rect><path d="M5 15H4a2 2 0 0 1-2-2V4a2 2 0 0 1 2-2h9a2 2 0 0 1 2 2v1"></path></svg></span> </a> </span> </p> </div> <hr /> `); } } else { $('#important-count').html(`<span class="badge badge-soft-primary ms-1 me-1">0</span>`); $('#important-subdomains-list').append(`<p>No subdomains markerd as important!</p>`); } $('.bs-tooltip').tooltip(); }); } function mark_important_subdomain(row, subdomain_id) { if (row) { parentNode = row.parentNode.parentNode.parentNode.parentNode; if (parentNode.classList.contains('table-danger')) { parentNode.classList.remove('table-danger'); } else { parentNode.className = "table-danger"; } } var data = {'subdomain_id': subdomain_id} const subdomainImpApi = "/api/toggle/subdomain/important/"; if ($("#important_subdomain_" + subdomain_id).length == 0) { $("#subdomain-" + subdomain_id).prepend(`<span id="important_subdomain_${subdomain_id}"></span>`); setTooltip("#subdomain-" + subdomain_id, 'Marked Important!'); } else { $("#important_subdomain_" + subdomain_id).remove(); setTooltip("#subdomain-" + subdomain_id, 'Marked Un-Important!'); } return fetch(subdomainImpApi, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken"), 'Content-Type': 'application/json' }, body: JSON.stringify(data) }); } function delete_scan(id) { const delAPI = "../delete/scan/" + id; swal.queue([{ title: 'Are you sure you want to delete this scan history?', text: "You won't be able to revert this!", icon: 'warning', showCancelButton: true, confirmButtonText: 'Delete', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { return fetch(delAPI, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken") } }).then(function(response) { return response.json(); }).then(function(data) { // TODO Look for better way return location.reload(); }).catch(function() { swal.insertQueueStep({ icon: 'error', title: 'Oops! Unable to delete the scan history!' }) }) } }]); } function stop_scan(scan_id=null, subscan_id=null, reload_scan_bar=true, reload_location=false) { const stopAPI = "/api/action/stop/scan/"; if (scan_id) { var data = {'scan_id': scan_id} } else if (subscan_id) { var data = {'subscan_id': subscan_id} } swal.queue([{ title: 'Are you sure you want to stop this scan?', text: "You won't be able to revert this!", icon: 'warning', showCancelButton: true, confirmButtonText: 'Stop', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { return fetch(stopAPI, { method: 'POST', credentials: "same-origin", body: JSON.stringify(data), headers: { "X-CSRFToken": getCookie("csrftoken"), "Content-Type": 'application/json', } }).then(function(response) { return response.json(); }).then(function(data) { // TODO Look for better way if (data.status) { Snackbar.show({ text: 'Scan Successfully Aborted.', pos: 'top-right', duration: 1500 }); if (reload_scan_bar) { getScanStatusSidebar(); } if (reload_location) { window.location.reload(); } } else { Snackbar.show({ text: 'Oops! Could not abort the scan. ' + data.message, pos: 'top-right', duration: 1500 }); } }).catch(function() { swal.insertQueueStep({ icon: 'error', title: 'Oops! Unable to stop the scan' }) }) } }]) } function extractContent(s) { var span = document.createElement('span'); span.innerHTML = s; return span.textContent || span.innerText; }; function delete_datatable_rows(table_id, rows_id, show_snackbar = true, snackbar_title) { // this function will delete the datatables rows after actions such as delete // table_id => datatable_id with # // rows_ids: list/array => list of all numerical ids to delete, to maintain consistency // rows id will always follow this pattern: datatable_id_row_n // show_snackbar = bool => whether to show snackbar or not! // snackbar_title: str => snackbar title if show_snackbar = True var table = $(table_id).DataTable(); for (var row in rows_id) { table.row(table_id + '_row_' + rows_id[row]).remove().draw(); } Snackbar.show({ text: snackbar_title, pos: 'top-right', duration: 1500, actionTextColor: '#fff', backgroundColor: '#e7515a', }); } function delete_subscan(subscan_id) { // This function will delete the sunscans using rest api // Supported method: POST const delAPI = "/api/action/rows/delete/"; var data = { 'type': 'subscan', 'rows': [subscan_id] } swal.queue([{ title: 'Are you sure you want to delete this subscan?', text: "You won't be able to revert this!", icon: 'warning', showCancelButton: true, confirmButtonText: 'Delete', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { return fetch(delAPI, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken"), "Content-Type": "application/json" }, body: JSON.stringify(data) }).then(function(response) { return response.json(); }).then(function(response) { if (response['status']) { delete_datatable_rows('#subscan_history_table', [subscan_id], show_snackbar = true, '1 Subscan Deleted!') } }).catch(function() { swal.insertQueueStep({ icon: 'error', title: 'Oops! Unable to delete the scan history!' }) }) } }]) } function show_subscan_results(subscan_id) { // This function will popup a modal and show the subscan results // modal being used is from base var api_url = '/api/fetch/results/subscan/?format=json'; var data = { 'subscan_id': subscan_id }; Swal.fire({ title: 'Fetching Results...' }); swal.showLoading(); fetch(api_url, { method: 'POST', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken"), 'Content-Type': 'application/json' }, body: JSON.stringify(data) }).then(response => response.json()).then(function(response) { console.log(response); swal.close(); if (response['subscan']['status'] == -1) { swal.fire("Error!", "Scan has not yet started! Please wait for other scans to complete...", "warning", { button: "Okay", }); return; } else if (response['subscan']['status'] == 1) { swal.fire("Error!", "Scan is in progress! Please come back later...", "warning", { button: "Okay", }); return; } $('#xl-modal-title').empty(); $('#xl-modal-content').empty(); $('#xl-modal-footer').empty(); var task_name = ''; if (response['subscan']['task'] == 'port_scan') { task_name = 'Port Scan'; } else if (response['subscan']['task'] == 'vulnerability_scan') { task_name = 'Vulnerability Scan'; } else if (response['subscan']['task'] == 'fetch_url') { task_name = 'EndPoint Gathering'; } else if (response['subscan']['task'] == 'dir_file_fuzz') { task_name = 'Directory and Files Fuzzing'; } $('#xl-modal_title').html(`${task_name} Results on ${response['subscan']['subdomain_name']}`); var scan_status = ''; var badge_color = 'danger'; if (response['subscan']['status'] == 0) { scan_status = 'Failed'; } else if (response['subscan']['status'] == 2) { scan_status = 'Successful'; badge_color = 'success'; } else if (response['subscan']['status'] == 3) { scan_status = 'Aborted'; } else { scan_status = 'Unknown'; } $('#xl-modal-content').append(`<div>Scan Status: <span class="badge bg-${badge_color}">${scan_status}</span></div>`); console.log(response); $('#xl-modal-content').append(`<div class="mt-1">Engine Used: <span class="badge bg-primary">${htmlEncode(response['subscan']['engine'])}</span></div>`); if (response['result'].length > 0) { if (response['subscan']['task'] == 'port_scan') { $('#xl-modal-content').append(`<div id="port_results_li"></div>`); for (var ip in response['result']) { var ip_addr = response['result'][ip]['address']; var id_name = `ip_${ip_addr}`; $('#port_results_li').append(`<h5>IP Address: ${ip_addr}</br></br>${response['result'][ip]['ports'].length} Ports Open</h5>`); $('#port_results_li').append(`<ul id="${id_name}"></ul>`); for (var port_obj in response['result'][ip]['ports']) { var port = response['result'][ip]['ports'][port_obj]; var port_color = 'primary'; if (port["is_uncommon"]) { port_color = 'danger'; } $('#port_results_li ul').append(`<li><span class="ms-1 mt-1 me-1 badge badge-soft-${port_color}">${port['number']}</span>/<span class="ms-1 mt-1 me-1 badge badge-soft-${port_color}">${port['service_name']}</span>/<span class="ms-1 mt-1 me-1 badge badge-soft-${port_color}">${port['description']}</span></li>`); } } $('#xl-modal-footer').append(`<span class="text-danger">* Uncommon Ports</span>`); } else if (response['subscan']['task'] == 'vulnerability_scan') { render_vulnerability_in_xl_modal(vuln_count = response['result'].length, subdomain_name = response['subscan']['subdomain_name'], result = response['result']); } else if (response['subscan']['task'] == 'fetch_url') { render_endpoint_in_xlmodal(endpoint_count = response['result'].length, subdomain_name = response['subscan']['subdomain_name'], result = response['result']); } else if (response['subscan']['task'] == 'dir_file_fuzz') { if (response['result'][0]['directory_files'].length == 0) { $('#xl-modal-content').append(` <div class="alert alert-info mt-2" role="alert"> <i class="mdi mdi-alert-circle-outline me-2"></i> ${task_name} could not fetch any results. </div> `); } else { render_directories_in_xl_modal(response['result'][0]['directory_files'].length, response['subscan']['subdomain_name'], response['result'][0]['directory_files']); } } } else { $('#xl-modal-content').append(` <div class="alert alert-info mt-2" role="alert"> <i class="mdi mdi-alert-circle-outline me-2"></i> ${task_name} could not fetch any results. </div> `); } $('#modal_xl_scroll_dialog').modal('show'); $("body").tooltip({ selector: '[data-toggle=tooltip]' }); }); } function get_http_status_badge(data) { if (data >= 200 && data < 300) { return "<span class='badge badge-soft-success'>" + data + "</span>"; } else if (data >= 300 && data < 400) { return "<span class='badge badge-soft-warning'>" + data + "</span>"; } else if (data == 0) { // datatable throws error when no data is returned return ""; } return "<span class='badge badge-soft-danger'>" + data + "</span>"; } function render_endpoint_in_xlmodal(endpoint_count, subdomain_name, result) { // This function renders endpoints datatable in xl modal // Used in Subscan results and subdomain to endpoints modal $('#xl-modal-content').append(`<h5> ${endpoint_count} Endpoints Discovered on subdomain ${subdomain_name}</h5>`); $('#xl-modal-content').append(` <div class=""> <table id="endpoint-modal-datatable" class="table dt-responsive nowrap w-100"> <thead> <tr> <th>HTTP URL</th> <th>Status</th> <th>Page Title</th> <th>Tags</th> <th>Content Type</th> <th>Content Length</th> <th>Response Time</th> </tr> </thead> <tbody id="endpoint_tbody"> </tbody> </table> </div> `); $('#endpoint_tbody').empty(); for (var endpoint_obj in result) { var endpoint = result[endpoint_obj]; var tech_badge = ''; var web_server = ''; if (endpoint['technologies']) { tech_badge = '<div>' + parse_technology(endpoint['technologies'], "primary", outline = true); } if (endpoint['webserver']) { web_server = `<span class='m-1 badge badge-soft-info' data-toggle="tooltip" data-placement="top" title="Web Server">${endpoint['webserver']}</span>`; } var url = split_into_lines(endpoint['http_url'], 70); var rand_id = get_randid(); tech_badge += web_server + '</div>'; var http_url_td = "<a href='" + endpoint['http_url'] + `' target='_blank' class='text-primary'>` + url + "</a>" + tech_badge; $('#endpoint_tbody').append(` <tr> <td>${http_url_td}</td> <td>${get_http_status_badge(endpoint['http_status'])}</td> <td>${return_str_if_not_null(endpoint['page_title'])}</td> <td>${parse_comma_values_into_span(endpoint['matched_gf_patterns'], "danger", outline=true)}</td> <td>${return_str_if_not_null(endpoint['content_type'])}</td> <td>${return_str_if_not_null(endpoint['content_length'])}</td> <td>${get_response_time_text(endpoint['response_time'])}</td> </tr> `); } $("#endpoint-modal-datatable").DataTable({ "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", "order": [ [5, "desc"] ], drawCallback: function() { $(".dataTables_paginate > .pagination").addClass("pagination-rounded") } }); } function render_vulnerability_in_xl_modal(vuln_count, subdomain_name, result) { // This function will render the vulnerability datatable in xl modal $('#xl-modal-content').append(`<h5> ${vuln_count} Vulnerabilities Discovered on subdomain ${subdomain_name}</h5>`); $('#xl-modal-content').append(`<ol id="vuln_results_ol" class="list-group list-group-numbered"></ol>`); $('#xl-modal-content').append(` <div class=""> <table id="vulnerability-modal-datatable" class="table dt-responsive nowrap w-100"> <thead> <tr> <th>Type</th> <th>Title</th> <th class="text-center">Severity</th> <th>CVSS Score</th> <th>CVE/CWE</th> <th>Vulnerable URL</th> <th>Description</th> <th class="text-center dt-no-sorting">Action</th> </tr> </thead> <tbody id="vuln_tbody"> </tbody> </table> </div> `); $('#vuln_tbody').empty(); for (var vuln in result) { var vuln_obj = result[vuln]; var vuln_type = vuln_obj['type'] ? `<span class="badge badge-soft-primary">&nbsp;&nbsp;${vuln_obj['type'].toUpperCase()}&nbsp;&nbsp;</span>` : ''; var tags = ''; var cvss_metrics_badge = ''; switch (vuln_obj['severity']) { case 'Info': color = 'primary' badge_color = 'soft-primary' break; case 'Low': color = 'low' badge_color = 'soft-warning' break; case 'Medium': color = 'warning' badge_color = 'soft-warning' break; case 'High': color = 'danger' badge_color = 'soft-danger' break; case 'Critical': color = 'critical' badge_color = 'critical' break; default: } if (vuln_obj['tags']) { tags = '<div>'; vuln_obj['tags'].forEach(tag => { tags += `<span class="badge badge-${badge_color} me-1 mb-1" data-toggle="tooltip" data-placement="top" title="Tags">${tag.name}</span>`; }); tags += '</div>'; } if (vuln_obj['cvss_metrics']) { cvss_metrics_badge = `<div><span class="badge badge-outline-primary my-1" data-toggle="tooltip" data-placement="top" title="CVSS Metrics">${vuln_obj['cvss_metrics']}</span></div>`; } var vuln_title = `<b class="text-${color}">` + vuln_obj['name'] + `</b>` + cvss_metrics_badge + tags; var badge = 'danger'; var cvss_score = ''; if (vuln_obj['cvss_score']) { if (vuln_obj['cvss_score'] > 0.1 && vuln_obj['cvss_score'] <= 3.9) { badge = 'info'; } else if (vuln_obj['cvss_score'] > 3.9 && vuln_obj['cvss_score'] <= 6.9) { badge = 'warning'; } else if (vuln_obj['cvss_score'] > 6.9 && vuln_obj['cvss_score'] <= 8.9) { badge = 'danger'; } cvss_score = `<span class="badge badge-outline-${badge}" data-toggle="tooltip" data-placement="top" title="CVSS Score">${vuln_obj['cvss_score']}</span>`; } var cve_cwe_badge = '<div>'; if (vuln_obj['cve_ids']) { vuln_obj['cve_ids'].forEach(cve => { cve_cwe_badge += `<a href="https://google.com/search?q=${cve.name.toUpperCase()}" target="_blank" class="badge badge-outline-primary me-1 mt-1" data-toggle="tooltip" data-placement="top" title="CVE ID">${cve.name.toUpperCase()}</a>`; }); } if (vuln_obj['cwe_ids']) { vuln_obj['cwe_ids'].forEach(cwe => { cve_cwe_badge += `<a href="https://google.com/search?q=${cwe.name.toUpperCase()}" target="_blank" class="badge badge-outline-primary me-1 mt-1" data-toggle="tooltip" data-placement="top" title="CWE ID">${cwe.name.toUpperCase()}</a>`; }); } cve_cwe_badge += '</div>'; var http_url = vuln_obj['http_url'].includes('http') ? "<a href='" + htmlEncode(vuln_obj['http_url']) + "' target='_blank' class='text-danger'>" + htmlEncode(vuln_obj['http_url']) + "</a>" : vuln_obj['http_url']; var description = vuln_obj['description'] ? `<div>${split_into_lines(vuln_obj['description'], 30)}</div>` : ''; // show extracted results, and show matcher names, matcher names can be in badges if (vuln_obj['matcher_name']) { description += `<span class="badge badge-soft-primary" data-toggle="tooltip" data-placement="top" title="Matcher Name">${vuln_obj['matcher_name']}</span>`; } if (vuln_obj['extracted_results'] && vuln_obj['extracted_results'].length > 0) { description += `<br><a class="mt-2" data-bs-toggle="collapse" href="#results_${vuln_obj['id']}" aria-expanded="false" aria-controls="results_${vuln_obj['id']}">Extracted Results <i class="fe-chevron-down"></i></a>`; description += `<div class="collapse" id="results_${vuln_obj['id']}"><ul>`; vuln_obj['extracted_results'].forEach(results => { description += `<li>${results}</li>`; }); description += '</ul></div>'; } if (vuln_obj['references'] && vuln_obj['references'].length > 0) { description += `<br><a class="mt-2" data-bs-toggle="collapse" href="#references_${vuln_obj['id']}" aria-expanded="false" aria-controls="references_${vuln_obj['id']}">References <i class="fe-chevron-down"></i></a>`; description += `<div class="collapse" id="references_${vuln_obj['id']}"><ul>`; vuln_obj['references'].forEach(reference => { description += `<li><a href="${reference.url}" target="_blank">${reference.url}</a></li>`; }); description += '</ul></div>'; } if (vuln_obj['curl_command']) { description += `<br><a class="mt-2" data-bs-toggle="collapse" href="#curl_command_${vuln_obj['id']}" aria-expanded="false" aria-controls="curl_command_${vuln_obj['id']}">CURL command <i class="fe-terminal"></i></a>`; description += `<div class="collapse" id="curl_command_${vuln_obj['id']}"><ul>`; description += `<li><code>${split_into_lines(htmlEncode(vuln_obj['curl_command']), 30)}</code></li>`; description += '</ul></div>'; } var action_icon = vuln_obj['hackerone_report_id'] ? '' : ` <div class="btn-group mb-2 dropstart"> <a href="#" class="text-dark dropdown-toggle float-end" data-bs-toggle="dropdown" aria-haspopup="true" aria-expanded="false"> <svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="feather feather-more-horizontal"><circle cx="12" cy="12" r="1"></circle><circle cx="19" cy="12" r="1"></circle><circle cx="5" cy="12" r="1"></circle></svg> </a> <div class="dropdown-menu" style=""> <a class="dropdown-item" href="javascript:report_hackerone(${vuln_obj['id']}, '${vuln_obj['severity']}');">Report to Hackerone</a> </div> </div>`; $('#vuln_tbody').append(` <tr> <td>${vuln_type}</td> <td>${vuln_title}</td> <td class="text-center">${get_severity_badge(vuln_obj['severity'])}</td> <td class="text-center">${cvss_score}</td> <td>${cve_cwe_badge}</td> <td>${http_url}</td> <td>${description}</td> <td>${action_icon}</td> </tr> `); } $("#vulnerability-modal-datatable").DataTable({ "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", "order": [ [5, "desc"] ], drawCallback: function() { $(".dataTables_paginate > .pagination").addClass("pagination-rounded") } }); } function render_directories_in_xl_modal(directory_count, subdomain_name, result) { $('#xl-modal-content').append(`<h5> ${directory_count} Directories Discovered on subdomain ${subdomain_name}</h5>`); $('#xl-modal-content').append(` <div class=""> <table id="directory-modal-datatable" class="table dt-responsive nowrap w-100"> <thead> <tr> <th>Directory</th> <th class="text-center">HTTP Status</th> <th>Content Length</th> <th>Lines</th> <th>Words</th> </tr> </thead> <tbody id="directory_tbody"> </tbody> </table> </div> `); $('#directory_tbody').empty(); for (var dir_obj in result) { var dir = result[dir_obj]; $('#directory_tbody').append(` <tr> <td><a href="${dir.url}" target="_blank">${dir.name}</a></td> <td class="text-center">${get_http_status_badge(dir.http_status)}</td> <td>${dir.length}</td> <td>${dir.lines}</td> <td>${dir.words}</td> </tr> `); } var interesting_keywords_array = []; var dir_modal_table = $("#directory-modal-datatable").DataTable({ "oLanguage": { "oPaginate": { "sPrevious": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-left"><line x1="19" y1="12" x2="5" y2="12"></line><polyline points="12 19 5 12 12 5"></polyline></svg>', "sNext": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-arrow-right"><line x1="5" y1="12" x2="19" y2="12"></line><polyline points="12 5 19 12 12 19"></polyline></svg>' }, "sInfo": "Showing page _PAGE_ of _PAGES_", "sSearch": '<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" class="feather feather-search"><circle cx="11" cy="11" r="8"></circle><line x1="21" y1="21" x2="16.65" y2="16.65"></line></svg>', "sSearchPlaceholder": "Search...", "sLengthMenu": "Results : _MENU_", }, "dom": "<'dt--top-section'<'row'<'col-12 col-sm-6 d-flex justify-content-sm-start justify-content-center'f><'col-12 col-sm-6 d-flex justify-content-sm-end justify-content-center'l>>>" + "<'table-responsive'tr>" + "<'dt--bottom-section d-sm-flex justify-content-sm-between text-center'<'dt--pages-count mb-sm-0 mb-3'i><'dt--pagination'p>>", "order": [ [2, "desc"] ], drawCallback: function() { $(".dataTables_paginate > .pagination").addClass("pagination-rounded"); } }); // TODO: Find interetsing dirs // fetch("/api/listInterestingKeywords") // .then(response => { // return response.json(); // }) // .then(data => { // interesting_keywords_array = data; // dir_modal_table.rows().every(function(){ // console.log(this.data()); // }); // }); } function get_and_render_subscan_history(subdomain_id, subdomain_name) { // This function displays the subscan history in a modal for any particular subdomain var data = { 'subdomain_id': subdomain_id }; fetch('/api/listSubScans/?format=json', { method: 'POST', credentials: "same-origin", body: JSON.stringify(data), headers: { "X-CSRFToken": getCookie("csrftoken"), "Content-Type": 'application/json', } }).then(function(response) { return response.json(); }).then(function(data) { console.log(data); if (data['status']) { $('#modal_title').html('Subscan History for subdomain ' + subdomain_name); $('#modal-content').empty(); $('#modal-content').append(`<div id="subscan_history_table"></div>`); $('#subscan_history_table').empty(); for (var result in data['results']) { var result_obj = data['results'][result]; var error_message = ''; var task_name = get_task_name(result_obj); if (result_obj.status == 0) { color = 'danger'; bg_color = 'bg-soft-danger'; status_badge = '<span class="float-end badge bg-danger">Failed</span>'; error_message = `</br><span class="text-danger">Error: ${result_obj.error_message}`; } else if (result_obj.status == 3) { color = 'danger'; bg_color = 'bg-soft-danger'; status_badge = '<span class="float-end badge bg-danger">Aborted</span>'; } else if (result_obj.status == 2) { color = 'success'; bg_color = 'bg-soft-success'; status_badge = '<span class="float-end badge bg-success">Task Completed</span>'; } $('#subscan_history_table').append(` <div class="card border-${color} border mini-card"> <a href="#" class="text-reset item-hovered" onclick="show_subscan_results(${result_obj['id']})"> <div class="card-header ${bg_color} text-${color} mini-card-header"> ${task_name} on <b>${result_obj.subdomain_name}</b> using engine <b>${htmlEncode(result_obj.engine)}</b> </div> <div class="card-body mini-card-body"> <p class="card-text"> ${status_badge} <span class=""> Task Completed ${result_obj.completed_ago} ago </span> Took ${result_obj.time_taken} ${error_message} </p> </div> </a> </div> `); } $('#modal_dialog').modal('show'); } }); } function fetch_whois(domain_name, save_db) { // this function will fetch WHOIS record for any subdomain and also display // snackbar once whois is fetched var url = `/api/tools/whois/?format=json&ip_domain=${domain_name}`; if (save_db) { url += '&save_db'; } $('[data-toggle="tooltip"]').tooltip('hide'); Snackbar.show({ text: 'Fetching WHOIS...', pos: 'top-right', duration: 1500, }); $("#whois_not_fetched_alert").hide(); $("#whois_fetching_alert").show(); fetch(url, {}).then(res => res.json()) .then(function(response) { $("#whois_fetching_alert").hide(); document.getElementById('domain_age').innerHTML = response['domain']['domain_age'] + ' ' + response['domain']['date_created']; document.getElementById('ip_address').innerHTML = response['domain']['ip_address']; document.getElementById('ip_geolocation').innerHTML = response['domain']['geolocation']; document.getElementById('registrant_name').innerHTML = response['registrant']['name']; console.log(response['registrant']['organization']) document.getElementById('registrant_organization').innerHTML = response['registrant']['organization'] ? response['registrant']['organization'] : ' '; document.getElementById('registrant_address').innerHTML = response['registrant']['address'] + ' ' + response['registrant']['city'] + ' ' + response['registrant']['state'] + ' ' + response['registrant']['country']; document.getElementById('registrant_phone_numbers').innerHTML = response['registrant']['tel']; document.getElementById('registrant_fax').innerHTML = response['registrant']['fax']; Snackbar.show({ text: 'Whois Fetched...', pos: 'top-right', duration: 3000 }); $("#whois_fetched_alert").show(); $("#whois_fetched_alert").fadeTo(2000, 500).slideUp(1500, function() { $("#whois_fetched_alert").slideUp(500); }); }).catch(function(error) { console.log(error); }); } function get_target_whois(domain_name) { // this function will fetch whois from db, if not fetched, will make a fresh // query and will display whois on a modal var url = `/api/tools/whois/?format=json&ip_domain=${domain_name}&fetch_from_db` Swal.fire({ title: `Fetching WHOIS details for ${domain_name}...` }); swal.showLoading(); fetch(url, { method: 'GET', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken"), 'Content-Type': 'application/json' }, }).then(response => response.json()).then(function(response) { console.log(response); if (response.status) { swal.close(); display_whois_on_modal(response); } else { fetch(`/api/tools/whois/?format=json&ip_domain=${domain_name}&save_db`, { method: 'GET', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken"), 'Content-Type': 'application/json' }, }).then(response => response.json()).then(function(response) { console.log(response); if (response.status) { swal.close(); display_whois_on_modal(response); } else { Swal.fire({ title: 'Oops!', text: `reNgine could not fetch WHOIS records for ${domain_name}!`, icon: 'error' }); } }); } }); } function get_domain_whois(domain_name, show_add_target_btn=false) { // this function will get whois for domains that are not targets, this will // not store whois into db nor create target var url = `/api/tools/whois/?format=json&ip_domain=${domain_name}` Swal.fire({ title: `Fetching WHOIS details for ${domain_name}...` }); $('.modal').modal('hide'); swal.showLoading(); fetch(url, { method: 'GET', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken"), 'Content-Type': 'application/json' }, }).then(response => response.json()).then(function(response) { swal.close(); if (response.status) { display_whois_on_modal(response, show_add_target_btn=show_add_target_btn); } else { Swal.fire({ title: 'Oops!', text: `reNgine could not fetch WHOIS records for ${domain_name}! ${response['message']}`, icon: 'error' }); } }); } function display_whois_on_modal(response, show_add_target_btn=false) { console.log(response); // this function will display whois data on modal, should be followed after get_domain_whois() $('#modal_dialog').modal('show'); $('#modal-content').empty(); $("#modal-footer").empty(); content = ` <div class="row mt-3"> <div class="col-sm-3"> <div class="nav flex-column nav-pills nav-pills-tab" id="v-pills-tab" role="tablist" aria-orientation="vertical"> <a class="nav-link active show mb-1" id="v-pills-domain-tab" data-bs-toggle="pill" href="#v-pills-domain" role="tab" aria-controls="v-pills-domain-tab" aria-selected="true">Domain info</a> <a class="nav-link mb-1" id="v-pills-whois-tab" data-bs-toggle="pill" href="#v-pills-whois" role="tab" aria-controls="v-pills-whois" aria-selected="false">Whois</a> <a class="nav-link mb-1" id="v-pills-nameserver-tab" data-bs-toggle="pill" href="#v-pills-nameserver" role="tab" aria-controls="v-pills-nameserver" aria-selected="false">Nameservers</a> <a class="nav-link mb-1" id="v-pills-history-tab" data-bs-toggle="pill" href="#v-pills-history" role="tab" aria-controls="v-pills-history" aria-selected="false">NS History</a> </div> </div> <!-- end col--> <div class="col-sm-9"> <div class="tab-content pt-0"> <div class="tab-pane fade active show" id="v-pills-domain" role="tabpanel" aria-labelledby="v-pills-domain-tab" data-simplebar style="max-height: 300px; min-height: 300px;"> <h4 class="header-title text-primary"><span class="fe-info"></span>&nbsp;Contact Information</h4> <ul class="nav nav-tabs nav-bordered nav-justified"> <li class="nav-item"> <a href="#registrant-tab" data-bs-toggle="tab" aria-expanded="false" class="nav-link active"> Registrant </a> </li> <li class="nav-item"> <a href="#admin-tab" data-bs-toggle="tab" aria-expanded="true" class="nav-link"> Admin </a> </li> <li class="nav-item"> <a href="#technical-tab" data-bs-toggle="tab" aria-expanded="false" class="nav-link"> Technical </a> </li> </ul> <div class="tab-content"> <div class="tab-pane active" id="registrant-tab"> <div class="table-responsive"> <table class="table mb-0"> <tbody> <tr class=""> <td><b>Name</b></td> <td><span class="fe-user"></span>&nbsp;${response.registrant.name}</td> </tr> <tr class="table-primary"> <td><b>Organization</b></td> <td><span class="fe-briefcase"></span>&nbsp;${response.registrant.organization}</td> </tr> <tr class=""> <td><b>Email</b></td> <td><span class="fe-mail"></span>&nbsp;${response.registrant.email}</td> </tr> <tr class="table-info"> <td><b>Phone/Fax</b></td> <td> <span class="fe-phone"></span>&nbsp;${response.registrant.phone} <span class="fe-printer"></span>&nbsp;${response.registrant.fax} </td> </tr> <tr class=""> <td><b>Address</b></td> <td><span class="fe-home"></span>&nbsp;${response.registrant.address}</td> </tr> <tr class="table-danger"> <td><b>Address</b></td> <td><b>City: </b>${response.registrant.city} <b>State: </b>${response.registrant.state} <b>Zip Code: </b>${response.registrant.zipcode} <b>Country: </b>${response.registrant.country} </td> </tr> </tbody> </table> </div> </div> <div class="tab-pane" id="admin-tab"> <div class="table-responsive"> <table class="table mb-0"> <tbody> <tr class="table-primary"> <td><b>Name</b></td> <td><span class="fe-user"></span>&nbsp;${response.admin.name}</td> </tr> <tr class=""> <td><b>Organization</b></td> <td><span class="fe-briefcase"></span>&nbsp;${response.admin.organization}</td> </tr> <tr class="table-info"> <td><b>Admin ID</b></td> <td><span class="fe-user"></span>&nbsp;${response.admin.id}</td> </tr> <tr class=""> <td><b>Email</b></td> <td><span class="fe-mail"></span>&nbsp;${response.admin.email}</td> </tr> <tr class="table-success"> <td><b>Phone/Fax</b></td> <td> <span class="fe-phone"></span>&nbsp;${response.admin.phone} <span class="fe-printer"></span>&nbsp;${response.admin.fax} </td> </tr> <tr class=""> <td><b>Address</b></td> <td><span class="fe-home"></span>&nbsp;${response.admin.address}</td> </tr> <tr class="table-danger"> <td><b>Address</b></td> <td><b>City: </b>${response.admin.city} <b>State: </b>${response.admin.state} <b>Zip Code: </b>${response.admin.zipcode} <b>Country: </b>${response.admin.country} </td> </tr> </tbody> </table> </div> </div> <div class="tab-pane" id="technical-tab"> <div class="table-responsive"> <table class="table mb-0"> <tbody> <tr class="table-info"> <td><b>Name</b></td> <td><span class="fe-user"></span>&nbsp;${response.technical_contact.name}</td> </tr> <tr class=""> <td><b>Organization</b></td> <td><span class="fe-briefcase"></span>&nbsp;${response.technical_contact.organization}</td> </tr> <tr class="table-primary"> <td><b>Tech ID</b></td> <td><span class="fe-user"></span>&nbsp;${response.technical_contact.id}</td> </tr> <tr class=""> <td><b>Email</b></td> <td><span class="fe-mail"></span>&nbsp;${response.technical_contact.email}</td> </tr> <tr class="table-success"> <td><b>Phone/Fax</b></td> <td> <span class="fe-phone"></span>&nbsp;${response.technical_contact.phone} <span class="fe-printer"></span>&nbsp;${response.technical_contact.fax} </td> </tr> <tr> <td><b>Address</b></td> <td><span class="fe-home"></span>&nbsp;${response.technical_contact.address}</td> </tr> <tr class="table-danger"> <td><b>Address</b></td> <td><b>City: </b>${response.technical_contact.city} <b>State: </b>${response.technical_contact.state} <b>Zip Code: </b>${response.technical_contact.zipcode} <b>Country: </b>${response.technical_contact.country} </td> </tr> </tbody> </table> </div> </div> </div> </div> <div class="tab-pane fade" id="v-pills-whois" role="tabpanel" aria-labelledby="v-pills-whois-tab"> <pre data-simplebar style="max-height: 310px; min-height: 310px;">${response.raw_text}</pre> </div> <div class="tab-pane fade" id="v-pills-history" role="tabpanel" aria-labelledby="v-pills-history-tab" data-simplebar style="max-height: 300px; min-height: 300px;"> </div> <div class="tab-pane fade" id="v-pills-nameserver" role="tabpanel" aria-labelledby="v-pills-nameserver-tab" data-simplebar style="max-height: 300px; min-height: 300px;"> `; for (var ns in response.nameservers) { var ns_object = response.nameservers[ns]; content += `<span class="badge badge-soft-primary me-1 ms-1">${ns_object}</span>`; } content += ` </div> <div class="tab-pane fade" id="v-pills-related" role="tabpanel" aria-labelledby="v-pills-related-tab" data-simplebar style="max-height: 300px; min-height: 300px;"> <!--<span class="badge badge-soft-primary badge-link waves-effect waves-light me-1" data-toggle="tooltip" title="Add {{domain}} as target." onclick="add_target('{{domain}}')">{{domain}}</span>--> </div> <div class="tab-pane fade" id="v-pills-related-tld" role="tabpanel" aria-labelledby="v-pills-related-tld-tab" data-simplebar style="max-height: 300px; min-height: 300px;"> <!--<span class="badge badge-soft-primary badge-link waves-effect waves-light me-1" data-toggle="tooltip" title="Add {{domain}} as target." onclick="add_target('{{domain}}')">{{domain}}</span>--> </div> </div> </div> </div> `; if (show_add_target_btn) { content += `<div class="text-center"> <button class="btn btn-primary float-end mt-4" type="submit" id="search_whois_toolbox_btn" onclick="add_target('${response['ip_domain']}')">Add ${response['ip_domain']} as target</button> </div>` } $('#modal-content').append(content); $('[data-toggle="tooltip"]').tooltip(); } function show_quick_add_target_modal() { // this function will display the modal to add target $('#modal_title').html('Add target'); $('#modal-content').empty(); $('#modal-content').append(` If you would like to add IP/CIDRs, multiple domain, Please <a href="/target/add/target">click here.</a> <div class="mb-3"> <label for="target_name_modal" class="form-label">Target Name</label> <input class="form-control" type="text" id="target_name_modal" required="" placeholder="yourdomain.com"> </div> <div class="mb-3"> <label for="target_description_modal" class="form-label">Description (Optional)</label> <input class="form-control" type="text" id="target_description_modal" required="" placeholder="Target Description"> </div> <div class="mb-3"> <label for="h1_handle_modal" class="form-label">Hackerone Target Team Handle (Optional)</label> <input class="form-control" type="text" id="h1_handle_modal" placeholder="hackerone.com/team_handle, Only enter team_handle after /"> </div> <div class="mb-3 text-center"> <button class="btn btn-primary float-end" type="submit" id="add_target_modal" onclick="add_quick_target()">Add Target</button> </div> `); $('#modal_dialog').modal('show'); } function add_quick_target() { // this function will be a onclick for add target button on add_target modal $('#modal_dialog').modal('hide'); var domain_name = $('#target_name_modal').val(); var description = $('#target_description_modal').val(); var h1_handle = $('#h1_handle_modal').val(); const data = { 'domain_name': domain_name, 'h1_team_handle': h1_handle, 'description': description }; add_target(domain_name, h1_handle = h1_handle, description = description); } function add_target(domain_name, h1_handle = null, description = null) { // this function will add domain_name as target const add_api = '/api/add/target/?format=json'; const data = { 'domain_name': domain_name, 'h1_team_handle': h1_handle, 'description': description }; swal.queue([{ title: 'Add Target', text: `Would you like to add ${domain_name} as target?`, icon: 'info', showCancelButton: true, confirmButtonText: 'Add Target', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { return fetch(add_api, { method: 'POST', credentials: "same-origin", headers: { 'X-CSRFToken': getCookie("csrftoken"), 'Content-Type': 'application/json' }, body: JSON.stringify(data) }).then(function(response) { return response.json(); }).then(function(data) { if (data.status) { swal.queue([{ title: 'Target Successfully added!', text: `Do you wish to initiate the scan on new target?`, icon: 'success', showCancelButton: true, confirmButtonText: 'Initiate Scan', padding: '2em', showLoaderOnConfirm: true, preConfirm: function() { window.location = `/scan/start/${data.domain_id}`; } }]); } else { swal.insertQueueStep({ icon: 'error', title: data.message }); } }).catch(function() { swal.insertQueueStep({ icon: 'error', title: 'Oops! Unable to delete the scan history!' }); }) } }]); } function loadSubscanHistoryWidget(scan_history_id = null, domain_id = null) { // This function will load the subscan history widget if (scan_history_id) { var data = { 'scan_history_id': scan_history_id } } if (domain_id) { var data = { 'domain_id': domain_id } } fetch('/api/listSubScans/?format=json', { method: 'POST', credentials: "same-origin", body: JSON.stringify(data), headers: { "X-CSRFToken": getCookie("csrftoken"), "Content-Type": 'application/json', } }).then(function(response) { return response.json(); }).then(function(data) { console.log(data); $('#subscan_history_widget').empty(); if (data['status']) { $('#sub_scan_history_count').append(` <span class="badge badge-soft-primary me-1">${data['results'].length}</span> `) for (var result in data['results']) { var error_message = ''; var result_obj = data['results'][result]; var task_name = get_task_name(result_obj); if (result_obj.status == 0) { color = 'danger'; bg_color = 'bg-soft-danger'; status_badge = '<span class="float-end badge bg-danger">Failed</span>'; error_message = `</br><span class="text-danger">Error: ${result_obj.error_message}`; } else if (result_obj.status == 3) { color = 'danger'; bg_color = 'bg-soft-danger'; status_badge = '<span class="float-end badge bg-danger">Aborted</span>'; } else if (result_obj.status == 2) { color = 'success'; bg_color = 'bg-soft-success'; status_badge = '<span class="float-end badge bg-success">Task Completed</span>'; } else if (result_obj.status == 1) { color = 'primary'; bg_color = 'bg-soft-primary'; status_badge = '<span class="float-end badge bg-primary">Running</span>'; } $('#subscan_history_widget').append(` <div class="card border-${color} border mini-card"> <a href="#" class="text-reset item-hovered" onclick="show_subscan_results(${result_obj['id']})"> <div class="card-header ${bg_color} text-${color} mini-card-header"> ${task_name} on <b>${result_obj.subdomain_name}</b> </div> <div class="card-body mini-card-body"> <p class="card-text"> ${status_badge} <span class=""> Task Completed ${result_obj.completed_ago} ago </span> Took ${result_obj.time_taken} ${error_message} </p> </div> </a> </div> `); } } else { $('#sub_scan_history_count').append(` <span class="badge badge-soft-primary me-1">0</span> `) $('#subscan_history_widget').append(` <div class="alert alert-warning alert-dismissible fade show mt-2" role="alert"> <button type="button" class="btn-close" data-bs-dismiss="alert" aria-label="Close"></button> No Subscans has been initiated for any subdomains. You can select individual subdomains and initiate subscans like Directory Fuzzing, Vulnerability Scan etc. </div> `); } }); } function get_ips(scan_id=null, domain_id=null){ // this function will fetch and render ips in widget var url = '/api/queryIps/?'; if (scan_id) { url += `scan_id=${scan_id}`; } if (domain_id) { url += `target_id=${domain_id}`; } url += `&format=json`; $.getJSON(url, function(data) { $('#ip-address-count').empty(); for (var val in data['ips']){ ip = data['ips'][val] badge_color = ip['is_cdn'] ? 'warning' : 'primary'; if (scan_id) { $("#ip-address").append(`<span class='badge badge-soft-${badge_color} m-1 badge-link' data-toggle="tooltip" title="${ip['ports'].length} Ports Open." onclick="get_ip_details('${ip['address']}', scan_id=${scan_id}, domain_id=null)">${ip['address']}</span>`); } else if (domain_id) { $("#ip-address").append(`<span class='badge badge-soft-${badge_color} m-1 badge-link' data-toggle="tooltip" title="${ip['ports'].length} Ports Open." onclick="get_ip_details('${ip['address']}', scan_id=null, domain_id=${domain_id})">${ip['address']}</span>`); } // $("#ip-address").append(`<span class='badge badge-soft-${badge_color} m-1' data-toggle="modal" data-target="#tabsModal">${ip['address']}</span>`); } $('#ip-address-count').html(`<span class="badge badge-soft-primary me-1">${data['ips'].length}</span>`); $("body").tooltip({ selector: '[data-toggle=tooltip]' }); }); } function get_technologies(scan_id=null, domain_id=null){ // this function will fetch and render tech in widget var url = '/api/queryTechnologies/?'; if (scan_id) { url += `scan_id=${scan_id}`; } if (domain_id) { url += `target_id=${domain_id}`; } url += `&format=json`; $.getJSON(url, function(data) { $('#technologies-count').empty(); for (var val in data['technologies']){ tech = data['technologies'][val] if (scan_id) { $("#technologies").append(`<span class='badge badge-soft-primary m-1 badge-link' data-toggle="tooltip" title="${tech['count']} Subdomains use this technology." onclick="get_tech_details('${tech['name']}', scan_id=${scan_id}, domain_id=null)">${tech['name']}</span>`); } else if (domain_id) { $("#technologies").append(`<span class='badge badge-soft-primary m-1 badge-link' data-toggle="tooltip" title="${tech['count']} Subdomains use this technology." onclick="get_tech_details('${tech['name']}', scan_id=null, domain_id=${domain_id})">${tech['name']}</span>`); } } $('#technologies-count').html(`<span class="badge badge-soft-primary me-1">${data['technologies'].length}</span>`); $("body").tooltip({ selector: '[data-toggle=tooltip]' }); }); } function get_ports(scan_id=null, domain_id=null){ // this function will fetch and render ports in widget var url = '/api/queryPorts/?'; if (scan_id) { url += `scan_id=${scan_id}`; } if (domain_id) { url += `target_id=${domain_id}`; } url += `&format=json`; $.getJSON(url, function(data) { $('#ports-count').empty(); for (var val in data['ports']){ port = data['ports'][val] badge_color = port['is_uncommon'] ? 'danger' : 'primary'; if (scan_id) { $("#ports").append(`<span class='badge badge-soft-${badge_color} m-1 badge-link' data-toggle="tooltip" title="${port['description']}" onclick="get_port_details('${port['number']}', scan_id=${scan_id}, domain_id=null)">${port['number']}/${port['service_name']}</span>`); } else if (domain_id){ $("#ports").append(`<span class='badge badge-soft-${badge_color} m-1 badge-link' data-toggle="tooltip" title="${port['description']}" onclick="get_port_details('${port['number']}', scan_id=null, domain_id=${domain_id})">${port['number']}/${port['service_name']}</span>`); } } $('#ports-count').html(`<span class="badge badge-soft-primary me-1">${data['ports'].length}</span>`); $("body").tooltip({ selector: '[data-toggle=tooltip]' }); }); } function get_ip_details(ip_address, scan_id=null, domain_id=null){ var port_url = `/api/queryPorts/?ip_address=${ip_address}`; var subdomain_url = `/api/querySubdomains/?ip_address=${ip_address}`; if (scan_id) { port_url += `&scan_id=${scan_id}`; subdomain_url += `&scan_id=${scan_id}`; } else if(domain_id){ port_url += `&target_id=${domain_id}`; subdomain_url += `&target_id=${domain_id}`; } port_url += `&format=json`; subdomain_url += `&format=json`; var interesting_badge = `<span class="m-1 badge badge-soft-danger bs-tooltip" title="Interesting Subdomain">Interesting</span>`; var port_loader = `<span class="inner-div spinner-border text-primary align-self-center loader-sm" id="port-modal-loader"></span>`; var subdomain_loader = `<span class="inner-div spinner-border text-primary align-self-center loader-sm" id="subdomain-modal-loader"></span>`; // add tab modal title $('#modal_title').html('Details for IP: <b>' + ip_address + '</b>'); $('#modal-content').empty(); $('#modal-tabs').empty(); $('#modal-content').append(`<ul class='nav nav-tabs nav-bordered' id="modal_tab_nav"></ul><div id="modal_tab_content" class="tab-content"></div>`); $('#modal_tab_nav').append(`<li class="nav-item"><a class="nav-link active" data-bs-toggle="tab" href="#modal_content_port" aria-expanded="true"><span id="modal-open-ports-count"></span>Open Ports &nbsp;${port_loader}</a></li>`); $('#modal_tab_nav').append(`<li class="nav-item"><a class="nav-link" data-bs-toggle="tab" href="#modal_content_subdomain" aria-expanded="false"><span id="modal-subdomain-count"></span>Subdomains &nbsp;${subdomain_loader}</a></li>`) // add content area $('#modal_tab_content').empty(); $('#modal_tab_content').append(`<div class="tab-pane show active" id="modal_content_port"></div><div class="tab-pane" id="modal_content_subdomain"></div>`); $('#modal-open-ports').append(`<div class="modal-text" id="modal-text-open-port"></div>`); $('#modal-text-open-port').append(`<ul id="modal-open-port-text"></ul>`); $('#modal_content_port').append(`<ul id="modal_port_ul"></ul>`); $('#modal_content_subdomain').append(`<ul id="modal_subdomain_ul"></ul>`); $.getJSON(port_url, function(data) { $('#modal_content_port').empty(); $('#modal_content_port').append(`<p> IP Addresses ${ip_address} has ${data['ports'].length} Open Ports`); $('#modal-open-ports-count').html(`<b>${data['ports'].length}</b>&nbsp;&nbsp;`); for (port in data['ports']){ port_obj = data['ports'][port]; badge_color = port_obj['is_uncommon'] ? 'danger' : 'info'; $("#modal_content_port").append(`<li class="mt-1">${port_obj['description']} <b class="text-${badge_color}">(${port_obj['number']}/${port_obj['service_name']})</b></li>`) } $("#port-modal-loader").remove(); }); $('#modal_dialog').modal('show'); // query subdomains $.getJSON(subdomain_url, function(data) { $('#modal_content_subdomain').empty(); $('#modal_content_subdomain').append(`<p>${data['subdomains'].length} Subdomains are associated with IP ${ip_address}`); $('#modal-subdomain-count').html(`<b>${data['subdomains'].length}</b>&nbsp;&nbsp;`); for (subdomain in data['subdomains']){ subdomain_obj = data['subdomains'][subdomain]; badge_color = subdomain_obj['http_status'] >= 400 ? 'danger' : ''; li_id = get_randid(); if (subdomain_obj['http_url']) { $("#modal_content_subdomain").append(`<li class="mt-1" id="${li_id}"><a href='${subdomain_obj['http_url']}' target="_blank" class="text-${badge_color}">${subdomain_obj['name']}</a></li>`) } else { $("#modal_content_subdomain").append(`<li class="mt-1 text-${badge_color}" id="${li_id}">${subdomain_obj['name']}</li>`); } if (subdomain_obj['http_status']) { $("#"+li_id).append(get_http_badge(subdomain_obj['http_status'])); $('.bs-tooltip').tooltip(); } if (subdomain_obj['is_interesting']) { $("#"+li_id).append(interesting_badge) } } $("#modal-text-subdomain").append(`<span class="float-end text-danger">*Subdomains highlighted are 40X HTTP Status</span>`); $("#subdomain-modal-loader").remove(); }); } function get_port_details(port, scan_id=null, domain_id=null){ var ip_url = `/api/queryIps/?port=${port}`; var subdomain_url = `/api/querySubdomains/?port=${port}`; if (scan_id) { ip_url += `&scan_id=${scan_id}`; subdomain_url += `&scan_id=${scan_id}`; } else if(domain_id){ ip_url += `&target_id=${domain_id}`; subdomain_url += `&target_id=${domain_id}`; } ip_url += `&format=json`; subdomain_url += `&format=json`; var interesting_badge = `<span class="m-1 badge badge-soft-danger bs-tooltip" title="Interesting Subdomain">Interesting</span>`; var ip_spinner = `<span class="spinner-border spinner-border-sm me-1" id="ip-modal-loader"></span>`; var subdomain_spinner = `<span class="spinner-border spinner-border-sm me-1" id="subdomain-modal-loader"></span>`; $('#modal_title').html('Details for Port: <b>' + port + '</b>'); $('#modal-content').empty(); $('#modal-tabs').empty(); $('#modal-content').append(`<ul class='nav nav-tabs nav-bordered' id="modal_tab_nav"></ul><div id="modal_tab_content" class="tab-content"></div>`); $('#modal_tab_nav').append(`<li class="nav-item"><a class="nav-link active" data-bs-toggle="tab" href="#modal_content_ip" aria-expanded="true"><span id="modal-ip-count"></span>IP Address&nbsp;${ip_spinner}</a></li>`); $('#modal_tab_nav').append(`<li class="nav-item"><a class="nav-link" data-bs-toggle="tab" href="#modal_content_subdomain" aria-expanded="false"><span id="modal-subdomain-count"></span>Subdomains&nbsp;${subdomain_spinner}</a></li>`) // add content area $('#modal_tab_content').append(`<div class="tab-pane show active" id="modal_content_ip"></div><div class="tab-pane" id="modal_content_subdomain"></div>`); $('#modal_content_ip').append(`<ul id="modal_ip_ul"></ul>`); $('#modal_content_subdomain').append(`<ul id="modal_subdomain_ul"></ul>`); $('#modal_dialog').modal('show'); $.getJSON(ip_url, function(data) { $('#modal_ip_ul').empty(); $('#modal_ip_ul').append(`<p>${data['ips'].length} IP Addresses have Port ${port} Open`); $('#modal-ip-count').html(`<b>${data['ips'].length}</b>&nbsp;&nbsp;`); for (ip in data['ips']){ ip_obj = data['ips'][ip]; text_color = ip_obj['is_cdn'] ? 'warning' : ''; $("#modal_ip_ul").append(`<li class='mt-1 text-${text_color}'>${ip_obj['address']}</li>`) } $('#modal_ip_ul').append(`<span class="float-end text-warning">*IP Address highlighted are CDN IP Address</span>`); $("#ip-modal-loader").remove(); }); // query subdomains $.getJSON(subdomain_url, function(data) { $('#modal_subdomain_ul').empty(); $('#modal_subdomain_ul').append(`<p>${data['subdomains'].length} Subdomains have Port ${port} Open`); $('#modal-subdomain-count').html(`<b>${data['subdomains'].length}</b>&nbsp;&nbsp;`); for (subdomain in data['subdomains']){ subdomain_obj = data['subdomains'][subdomain]; badge_color = subdomain_obj['http_status'] >= 400 ? 'danger' : ''; li_id = get_randid(); if (subdomain_obj['http_url']) { $("#modal_subdomain_ul").append(`<li id="${li_id}" class="mt-1"><a href='${subdomain_obj['http_url']}' target="_blank" class="text-${badge_color}">${subdomain_obj['name']}</a></li>`) } else { $("#modal_subdomain_ul").append(`<li class="mt-1 text-${badge_color}" id="${li_id}">${subdomain_obj['name']}</li>`); } if (subdomain_obj['http_status']) { $("#"+li_id).append(get_http_badge(subdomain_obj['http_status'])); $('.bs-tooltip').tooltip(); } if (subdomain_obj['is_interesting']) { $("#"+li_id).append(interesting_badge) } } $("#modal_subdomain_ul").append(`<span class="float-end text-danger">*Subdomains highlighted are 40X HTTP Status</span>`); $("#subdomain-modal-loader").remove(); }); } function get_tech_details(tech, scan_id=null, domain_id=null){ var url = `/api/querySubdomains/?tech=${tech}`; if (scan_id) { url += `&scan_id=${scan_id}`; } else if(domain_id){ url += `&target_id=${domain_id}`; } url += `&format=json`; var interesting_badge = `<span class="m-1 badge badge-soft-danger bs-tooltip" title="Interesting Subdomain">Interesting</span>`; // render tab modal $('.modal-title').html('Details for Technology: <b>' + tech + '</b>'); $('#modal_dialog').modal('show'); $('.modal-text').empty(); $('#modal-footer').empty(); $('.modal-text').append(`<div class='outer-div' id="modal-loader"><span class="inner-div spinner-border text-primary align-self-center loader-sm"></span></div>`); // query subdomains $.getJSON(url, function(data) { $('#modal-loader').empty(); $('#modal-content').empty(); $('#modal-content').append(`${data['subdomains'].length} Subdomains are using ${tech}`); for (subdomain in data['subdomains']){ subdomain_obj = data['subdomains'][subdomain]; badge_color = subdomain_obj['http_status'] >= 400 ? 'danger' : ''; li_id = get_randid(); if (subdomain_obj['http_url']) { $("#modal-content").append(`<li id="${li_id}"><a href='${subdomain_obj['http_url']}' target="_blank" class="text-${badge_color}">${subdomain_obj['name']}</a></li>`) } else { $("#modal-content").append(`<li class="text-${badge_color}" id="${li_id}">${subdomain_obj['name']}</li>`); } if (subdomain_obj['http_status']) { $("#"+li_id).append(get_http_badge(subdomain_obj['http_status'])); $('.bs-tooltip').tooltip(); } if (subdomain_obj['is_interesting']) { $("#"+li_id).append(interesting_badge) } } $("#modal-content").append(`<span class="float-end text-danger">*Subdomains highlighted are 40X HTTP Status</span>`); $("#subdomain-modal-loader").remove(); }).fail(function(){ $('#modal-loader').empty(); }); } function get_http_badge(http_status){ switch (true) { case (http_status >= 400): badge_color = 'danger' break; case (http_status >= 300): badge_color = 'warning' break; case (http_status >= 200): badge_color = 'success' break; default: badge_color = 'danger' } if (http_status) { badge = `<span class="badge badge-soft-${badge_color} me-1 ms-1 bs-tooltip" data-placement="top" title="HTTP Status">${http_status}</span>`; return badge } } function get_and_render_cve_details(cve_id){ var api_url = `/api/tools/cve_details/?cve_id=${cve_id}&format=json`; Swal.fire({ title: 'Fetching CVE Details...' }); swal.showLoading(); fetch(api_url, { method: 'GET', credentials: "same-origin", headers: { "X-CSRFToken": getCookie("csrftoken"), "Content-Type": "application/json" }, }).then(response => response.json()).then(function(response) { console.log(response); swal.close(); if (response.status) { $('#xl-modal-title').empty(); $('#xl-modal-content').empty(); $('#xl-modal-footer').empty(); $('#xl-modal_title').html(`CVE Details of ${cve_id}`); var cvss_score_badge = 'danger'; if (response.result.cvss > 0.1 && response.result.cvss <= 3.9) { cvss_score_badge = 'info'; } else if (response.result.cvss > 3.9 && response.result.cvss <= 6.9) { cvss_score_badge = 'warning'; } content = `<div class="row mt-3"> <div class="col-sm-3"> <div class="nav flex-column nav-pills nav-pills-tab" id="v-pills-tab" role="tablist" aria-orientation="vertical"> <a class="nav-link active show mb-1" id="v-pills-cve-details-tab" data-bs-toggle="pill" href="#v-pills-cve-details" role="tab" aria-controls="v-pills-cve-details-tab" aria-selected="true">CVE Details</a> <a class="nav-link mb-1" id="v-pills-affected-products-tab" data-bs-toggle="pill" href="#v-pills-affected-products" role="tab" aria-controls="v-pills-affected-products-tab" aria-selected="true">Affected Products</a> <a class="nav-link mb-1" id="v-pills-affected-versions-tab" data-bs-toggle="pill" href="#v-pills-affected-versions" role="tab" aria-controls="v-pills-affected-versions-tab" aria-selected="true">Affected Versions</a> <a class="nav-link mb-1" id="v-pills-cve-references-tab" data-bs-toggle="pill" href="#v-pills-cve-references" role="tab" aria-controls="v-pills-cve-references-tab" aria-selected="true">References</a> </div> </div> <div class="col-sm-9"> <div class="tab-content pt-0">`; content += ` <div class="tab-pane fade active show" id="v-pills-cve-details" role="tabpanel" aria-labelledby="v-pills-cve-details-tab" data-simplebar style="max-height: 600px; min-height: 600px;"> <h4 class="header-title">${cve_id}</h4> <div class="alert alert-warning" role="alert"> ${response.result.summary} </div> <span class="badge badge-soft-primary">Assigner: ${response.result.assigner}</span> <span class="badge badge-outline-primary">CVSS Vector: ${response.result['cvss-vector']}</span> <table class="domain_details_table table table-hover table-borderless"> <tr style="display: none"> <th>&nbsp;</th> <th>&nbsp;</th> </tr> <tr> <td>CVSS Score</td> <td><span class="badge badge-soft-${cvss_score_badge}">${response.result.cvss ? response.result.cvss: "-"}</span></td> </tr> <tr> <td>Confidentiality Impact</td> <td>${response.result.impact.confidentiality ? response.result.impact.confidentiality: "N/A"}</td> </tr> <tr> <td>Integrity Impact</td> <td>${response.result.impact.integrity ? response.result.impact.integrity: "N/A"}</td> </tr> <tr> <td>Availability Impact</td> <td>${response.result.impact.availability ? response.result.impact.availability: "N/A"}</td> </tr> <tr> <td>Access Complexity</td> <td>${response.result.access.complexity ? response.result.access.complexity: "N/A"}</td> </tr> <tr> <td>Authentication</td> <td>${response.result.access.authentication ? response.result.access.authentication: "N/A"}</td> </tr> <tr> <td>CWE ID</td> <td><span class="badge badge-outline-danger">${response.result.cwe ? response.result.cwe: "N/A"}</span></td> </tr> </table> </div> `; content += `<div class="tab-pane fade" id="v-pills-cve-references" role="tabpanel" aria-labelledby="v-pills-cve-references-tab" data-simplebar style="max-height: 600px; min-height: 600px;"> <ul>`; for (var reference in response.result.references) { content += `<li><a href="${response.result.references[reference]}" target="_blank">${response.result.references[reference]}</a></li>`; } content += `</ul></div>`; content += `<div class="tab-pane fade" id="v-pills-affected-products" role="tabpanel" aria-labelledby="v-pills-affected-products-tab" data-simplebar style="max-height: 600px; min-height: 600px;"> <ul>`; for (var prod in response.result.vulnerable_product) { content += `<li>${response.result.vulnerable_product[prod]}</li>`; } content += `</ul></div>`; content += `<div class="tab-pane fade" id="v-pills-affected-versions" role="tabpanel" aria-labelledby="v-pills-affected-versions-tab" data-simplebar style="max-height: 600px; min-height: 600px;"> <ul>`; for (var conf in response.result.vulnerable_configuration) { content += `<li>${response.result.vulnerable_configuration[conf]['id']}</li>`; } content += `</ul></div>`; content += `</div></div></div>`; $('#xl-modal-content').append(content); $('#modal_xl_scroll_dialog').modal('show'); $("body").tooltip({ selector: '[data-toggle=tooltip]' }); } else{ swal.fire("Error!", response.message, "error", { button: "Okay", }); } }); } function get_most_vulnerable_target(scan_id=null, target_id=null, ignore_info=false, limit=50){ $('#most_vulnerable_target_div').empty(); $('#most_vulnerable_spinner').append(`<div class="spinner-border text-primary m-2" role="status"></div>`); var data = {}; if (scan_id) { data['scan_history_id'] = scan_id; } else if (target_id) { data['target_id'] = target_id; } data['ignore_info'] = ignore_info; data['limit'] = limit; fetch('/api/fetch/most_vulnerable/?format=json', { method: 'POST', credentials: "same-origin", body: JSON.stringify(data), headers: { "X-CSRFToken": getCookie("csrftoken"), "Content-Type": 'application/json', } }).then(function(response) { return response.json(); }).then(function(response) { $('#most_vulnerable_spinner').empty(); if (response.status) { $('#most_vulnerable_target_div').append(` <table class="table table-borderless table-nowrap table-hover table-centered m-0"> <thead> <tr> <th style="width: 60%">Target</th> <th style="width: 30%">Vulnerabilities Count</th> </tr> </thead> <tbody id="most_vulnerable_target_tbody"> </tbody> </table> `); for (var res in response.result) { var targ_obj = response.result[res]; var tr = `<tr onclick="window.location='/scan/detail/vuln?domain=${targ_obj.name}';" style="cursor: pointer;">`; if (scan_id || target_id) { tr = `<tr onclick="window.location='/scan/detail/vuln?subdomain=${targ_obj.name}';" style="cursor: pointer;">`; } $('#most_vulnerable_target_tbody').append(` ${tr} <td> <h5 class="m-0 fw-normal">${targ_obj.name}</h5> </td> <td> <span class="badge badge-outline-danger">${targ_obj.vuln_count} Vulnerabilities</span> </td> </tr> `); } } else{ $('#most_vulnerable_target_div').append(` <div class="mt-4 alert alert-warning"> Could not find most vulnerable targets. </br> Once the vulnerability scan is performed, reNgine will identify the most vulnerable targets.</div> `); } }); } function get_most_common_vulnerability(scan_id=null, target_id=null, ignore_info=false, limit=50){ $('#most_common_vuln_div').empty(); $('#most_common_vuln_spinner').append(`<div class="spinner-border text-primary m-2" role="status"></div>`); var data = {}; if (scan_id) { data['scan_history_id'] = scan_id; } else if (target_id) { data['target_id'] = target_id; } data['ignore_info'] = ignore_info; data['limit'] = limit; fetch('/api/fetch/most_common_vulnerability/?format=json', { method: 'POST', credentials: "same-origin", body: JSON.stringify(data), headers: { "X-CSRFToken": getCookie("csrftoken"), "Content-Type": 'application/json', } }).then(function(response) { return response.json(); }).then(function(response) { $('#most_common_vuln_spinner').empty(); if (response.status) { $('#most_common_vuln_div').append(` <table class="table table-borderless table-nowrap table-hover table-centered m-0"> <thead> <tr> <th style="width: 60%">Vulnerability Name</th> <th style="width: 20%">Count</th> <th style="width: 20%">Severity</th> </tr> </thead> <tbody id="most_common_vuln_tbody"> </tbody> </table> `); for (var res in response.result) { var vuln_obj = response.result[res]; var vuln_badge = ''; switch (vuln_obj.severity) { case -1: vuln_badge = get_severity_badge('Unknown'); break; case 0: vuln_badge = get_severity_badge('Info'); break; case 1: vuln_badge = get_severity_badge('Low'); break; case 2: vuln_badge = get_severity_badge('Medium'); break; case 3: vuln_badge = get_severity_badge('High'); break; case 4: vuln_badge = get_severity_badge('Critical'); break; default: vuln_badge = get_severity_badge('Unknown'); } $('#most_common_vuln_tbody').append(` <tr onclick="window.location='/scan/detail/vuln?vulnerability_name=${vuln_obj.name}';" style="cursor: pointer;"> <td> <h5 class="m-0 fw-normal">${vuln_obj.name}</h5> </td> <td> <span class="badge badge-outline-danger">${vuln_obj.count}</span> </td> <td> ${vuln_badge} </td> </tr> `); } } else{ $('#most_common_vuln_div').append(` <div class="mt-4 alert alert-warning"> Could not find Most Common Vulnerabilities. </br> Once the vulnerability scan is performed, reNgine will identify the Most Common Vulnerabilities.</div> `); } }); } function highlight_search(search_keyword, content){ // this function will send the highlighted text from search keyword var reg = new RegExp('('+search_keyword+')', 'gi'); return content.replace(reg, '<mark>$1</mark>'); } function validURL(str) { // checks for valid http url var pattern = new RegExp('^(https?:\\/\\/)?'+ // protocol '((([a-z\\d]([a-z\\d-]*[a-z\\d])*)\\.)+[a-z]{2,}|'+ // domain name '((\\d{1,3}\\.){3}\\d{1,3}))'+ // OR ip (v4) address '(\\:\\d+)?(\\/[-a-z\\d%_.~+]*)*'+ // port and path '(\\?[;&a-z\\d%_.~+=-]*)?'+ // query string '(\\#[-a-z\\d_]*)?$','i'); // fragment locator return !!pattern.test(str); } function shadeColor(color, percent) { //https://stackoverflow.com/a/13532993 var R = parseInt(color.substring(1,3),16); var G = parseInt(color.substring(3,5),16); var B = parseInt(color.substring(5,7),16); R = parseInt(R * (100 + percent) / 100); G = parseInt(G * (100 + percent) / 100); B = parseInt(B * (100 + percent) / 100); R = (R<255)?R:255; G = (G<255)?G:255; B = (B<255)?B:255; var RR = ((R.toString(16).length==1)?"0"+R.toString(16):R.toString(16)); var GG = ((G.toString(16).length==1)?"0"+G.toString(16):G.toString(16)); var BB = ((B.toString(16).length==1)?"0"+B.toString(16):B.toString(16)); return "#"+RR+GG+BB; }

yogeshojha

9fb69660d763b79e6ab505099c7e4cd58f19761c

18be197fed32ce87979564bb50f002e46290bc3f

## Inefficient regular expression This part of the regular expression may cause exponential backtracking on strings starting with '0' and containing many repetitions of '0'. [Show more details](https://github.com/yogeshojha/rengine/security/code-scanning/134)

github-advanced-security[bot]

yogeshojha/rengine

Fix #529

Nuclei returns the response to stdout: `{"template-id":"tech-detect","info":{"name":"Wappalyzer Technology Detection","author":["hakluke"],"tags":["tech"],"reference":null,"severity":"info"},"matcher-name":"nginx","type":"http","host":"https://example.com:443","matched-at":"https://example.com:443","timestamp":"2021-10-31T09:39:47.1571248Z","curl-command":"curl -X 'GET' -d '' -H 'Accept: */*' -H 'Accept-Language: en' -H 'User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_9_2) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/36.0.1944.0 Safari/537.36' 'https://example.com'"}` It needs to read host_url from matched-at, not from matched.

2021-10-31 10:27:33+00:00

2021-11-01 16:58:16+00:00

web/reNgine/tasks.py

import os import traceback import yaml import json import csv import validators import random import requests import logging import metafinder.extractor as metadata_extractor import whatportis import subprocess from selenium.webdriver.firefox.options import Options as FirefoxOptions from selenium import webdriver from emailfinder.extractor import * from dotted_dict import DottedDict from celery import shared_task from discord_webhook import DiscordWebhook from reNgine.celery import app from startScan.models import * from targetApp.models import Domain from scanEngine.models import EngineType from django.conf import settings from django.shortcuts import get_object_or_404 from celery import shared_task from datetime import datetime from degoogle import degoogle from django.conf import settings from django.utils import timezone, dateformat from django.shortcuts import get_object_or_404 from django.core.exceptions import ObjectDoesNotExist from reNgine.celery import app from reNgine.definitions import * from startScan.models import * from targetApp.models import Domain from scanEngine.models import EngineType, Configuration, Wordlist from .common_func import * ''' task for background scan ''' @app.task def initiate_scan( domain_id, scan_history_id, scan_type, engine_type, imported_subdomains=None, out_of_scope_subdomains=[] ): ''' scan_type = 0 -> immediate scan, need not create scan object scan_type = 1 -> scheduled scan ''' engine_object = EngineType.objects.get(pk=engine_type) domain = Domain.objects.get(pk=domain_id) if scan_type == 1: task = ScanHistory() task.scan_status = -1 elif scan_type == 0: task = ScanHistory.objects.get(pk=scan_history_id) # save the last scan date for domain model domain.last_scan_date = timezone.now() domain.save() # once the celery task starts, change the task status to Started task.scan_type = engine_object task.celery_id = initiate_scan.request.id task.domain = domain task.scan_status = 1 task.start_scan_date = timezone.now() task.subdomain_discovery = True if engine_object.subdomain_discovery else False task.dir_file_search = True if engine_object.dir_file_search else False task.port_scan = True if engine_object.port_scan else False task.fetch_url = True if engine_object.fetch_url else False task.osint = True if engine_object.osint else False task.screenshot = True if engine_object.screenshot else False task.vulnerability_scan = True if engine_object.vulnerability_scan else False task.save() activity_id = create_scan_activity(task, "Scanning Started", 2) results_dir = '/usr/src/scan_results/' os.chdir(results_dir) notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine has initiated recon for target {} with engine type {}'.format(domain.name, engine_object.engine_name)) try: current_scan_dir = domain.name + '_' + str(random.randint(100000000000, 999999999999)) os.mkdir(current_scan_dir) task.results_dir = current_scan_dir task.save() except Exception as exception: logger.error(exception) scan_failed(task) yaml_configuration = None excluded_subdomains = '' try: yaml_configuration = yaml.load( task.scan_type.yaml_configuration, Loader=yaml.FullLoader) except Exception as exception: logger.error(exception) # TODO: Put failed reason on db ''' Add GF patterns name to db for dynamic URLs menu ''' if engine_object.fetch_url and GF_PATTERNS in yaml_configuration[FETCH_URL]: task.used_gf_patterns = ','.join( pattern for pattern in yaml_configuration[FETCH_URL][GF_PATTERNS]) task.save() results_dir = results_dir + current_scan_dir # put all imported subdomains into txt file and also in Subdomain model if imported_subdomains: extract_imported_subdomain( imported_subdomains, task, domain, results_dir) if yaml_configuration: ''' a target in itself is a subdomain, some tool give subdomains as www.yogeshojha.com but url and everything else resolves to yogeshojha.com In that case, we would already need to store target itself as subdomain ''' initial_subdomain_file = '/target_domain.txt' if task.subdomain_discovery else '/sorted_subdomain_collection.txt' subdomain_file = open(results_dir + initial_subdomain_file, "w") subdomain_file.write(domain.name + "\n") subdomain_file.close() if(task.subdomain_discovery): activity_id = create_scan_activity(task, "Subdomain Scanning", 1) subdomain_scan( task, domain, yaml_configuration, results_dir, activity_id, out_of_scope_subdomains ) else: skip_subdomain_scan(task, domain, results_dir) update_last_activity(activity_id, 2) activity_id = create_scan_activity(task, "HTTP Crawler", 1) http_crawler( task, domain, results_dir, activity_id) update_last_activity(activity_id, 2) try: if task.screenshot: activity_id = create_scan_activity( task, "Visual Recon - Screenshot", 1) grab_screenshot( task, domain, yaml_configuration, current_scan_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if(task.port_scan): activity_id = create_scan_activity(task, "Port Scanning", 1) port_scanning(task, domain, yaml_configuration, results_dir) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.osint: activity_id = create_scan_activity(task, "OSINT Running", 1) perform_osint(task, domain, yaml_configuration, results_dir) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.dir_file_search: activity_id = create_scan_activity(task, "Directory Search", 1) directory_brute( task, domain, yaml_configuration, results_dir, activity_id ) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.fetch_url: activity_id = create_scan_activity(task, "Fetching endpoints", 1) fetch_endpoints( task, domain, yaml_configuration, results_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.vulnerability_scan: activity_id = create_scan_activity(task, "Vulnerability Scan", 1) vulnerability_scan( task, domain, yaml_configuration, results_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) activity_id = create_scan_activity(task, "Scan Completed", 2) if notification and notification[0].send_scan_status_notif: send_notification('*Scan Completed*\nreNgine has finished performing recon on target {}.'.format(domain.name)) ''' Once the scan is completed, save the status to successful ''' if ScanActivity.objects.filter(scan_of=task).filter(status=0).all(): task.scan_status = 0 else: task.scan_status = 2 task.stop_scan_date = timezone.now() task.save() # cleanup results delete_scan_data(results_dir) return {"status": True} def skip_subdomain_scan(task, domain, results_dir): # store default target as subdomain ''' If the imported subdomain already has target domain saved, we can skip this ''' if not Subdomain.objects.filter( scan_history=task, name=domain.name).exists(): subdomain_dict = DottedDict({ 'name': domain.name, 'scan_history': task, 'target_domain': domain }) save_subdomain(subdomain_dict) # Save target into target_domain.txt with open('{}/target_domain.txt'.format(results_dir), 'w+') as file: file.write(domain.name + '\n') file.close() ''' We can have two conditions, either subdomain scan happens, or subdomain scan does not happen, in either cases, because we are using import subdomain, we need to collect and sort all the subdomains Write target domain into subdomain_collection ''' os.system( 'cat {0}/target_domain.txt > {0}/subdomain_collection.txt'.format(results_dir)) os.system( 'cat {0}/from_imported.txt > {0}/subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/from_imported.txt'.format(results_dir)) ''' Sort all Subdomains ''' os.system( 'sort -u {0}/subdomain_collection.txt -o {0}/sorted_subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/subdomain_collection.txt'.format(results_dir)) def extract_imported_subdomain(imported_subdomains, task, domain, results_dir): valid_imported_subdomains = [subdomain for subdomain in imported_subdomains if validators.domain( subdomain) and domain.name == get_domain_from_subdomain(subdomain)] # remove any duplicate valid_imported_subdomains = list(set(valid_imported_subdomains)) with open('{}/from_imported.txt'.format(results_dir), 'w+') as file: for subdomain_name in valid_imported_subdomains: # save _subdomain to Subdomain model db if not Subdomain.objects.filter( scan_history=task, name=subdomain_name).exists(): subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': subdomain_name, 'is_imported_subdomain': True }) save_subdomain(subdomain_dict) # save subdomain to file file.write('{}\n'.format(subdomain_name)) file.close() def subdomain_scan(task, domain, yaml_configuration, results_dir, activity_id, out_of_scope_subdomains=None): ''' This function is responsible for performing subdomain enumeration ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Subdomain Gathering for target {} has been started'.format(domain.name)) subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' # check for all the tools and add them into string # if tool selected is all then make string, no need for loop if ALL in yaml_configuration[SUBDOMAIN_DISCOVERY][USES_TOOLS]: tools = 'amass-active amass-passive assetfinder sublist3r subfinder oneforall' else: tools = ' '.join( str(tool) for tool in yaml_configuration[SUBDOMAIN_DISCOVERY][USES_TOOLS]) logging.info(tools) # check for THREADS, by default 10 threads = 10 if THREADS in yaml_configuration[SUBDOMAIN_DISCOVERY]: _threads = yaml_configuration[SUBDOMAIN_DISCOVERY][THREADS] if _threads > 0: threads = _threads if 'amass' in tools: if 'amass-passive' in tools: amass_command = 'amass enum -passive -d {} -o {}/from_amass.txt'.format( domain.name, results_dir) if USE_AMASS_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_AMASS_CONFIG]: amass_command += ' -config /root/.config/amass.ini' # Run Amass Passive logging.info(amass_command) os.system(amass_command) if 'amass-active' in tools: amass_command = 'amass enum -active -d {} -o {}/from_amass_active.txt'.format( domain.name, results_dir) if USE_AMASS_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_AMASS_CONFIG]: amass_command += ' -config /root/.config/amass.ini' if AMASS_WORDLIST in yaml_configuration[SUBDOMAIN_DISCOVERY]: wordlist = yaml_configuration[SUBDOMAIN_DISCOVERY][AMASS_WORDLIST] if wordlist == 'default': wordlist_path = '/usr/src/wordlist/deepmagic.com-prefixes-top50000.txt' else: wordlist_path = '/usr/src/wordlist/' + wordlist + '.txt' if not os.path.exists(wordlist_path): wordlist_path = '/usr/src/' + AMASS_WORDLIST amass_command = amass_command + \ ' -brute -w {}'.format(wordlist_path) if amass_config_path: amass_command = amass_command + \ ' -config {}'.format('/usr/src/scan_results/' + amass_config_path) # Run Amass Active logging.info(amass_command) os.system(amass_command) if 'assetfinder' in tools: assetfinder_command = 'assetfinder --subs-only {} > {}/from_assetfinder.txt'.format( domain.name, results_dir) # Run Assetfinder logging.info(assetfinder_command) os.system(assetfinder_command) if 'sublist3r' in tools: sublist3r_command = 'python3 /usr/src/github/Sublist3r/sublist3r.py -d {} -t {} -o {}/from_sublister.txt'.format( domain.name, threads, results_dir) # Run sublist3r logging.info(sublist3r_command) os.system(sublist3r_command) if 'subfinder' in tools: subfinder_command = 'subfinder -d {} -t {} -o {}/from_subfinder.txt'.format( domain.name, threads, results_dir) if USE_SUBFINDER_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_SUBFINDER_CONFIG]: subfinder_command += ' -config /root/.config/subfinder/config.yaml' # Run Subfinder logging.info(subfinder_command) os.system(subfinder_command) if 'oneforall' in tools: oneforall_command = 'python3 /usr/src/github/OneForAll/oneforall.py --target {} run'.format( domain.name, results_dir) # Run OneForAll logging.info(oneforall_command) os.system(oneforall_command) extract_subdomain = "cut -d',' -f6 /usr/src/github/OneForAll/results/{}.csv >> {}/from_oneforall.txt".format( domain.name, results_dir) os.system(extract_subdomain) # remove the results from oneforall directory os.system( 'rm -rf /usr/src/github/OneForAll/results/{}.*'.format(domain.name)) ''' All tools have gathered the list of subdomains with filename initials as from_* We will gather all the results in one single file, sort them and remove the older results from_* ''' os.system( 'cat {0}/*.txt > {0}/subdomain_collection.txt'.format(results_dir)) ''' Write target domain into subdomain_collection ''' os.system( 'cat {0}/target_domain.txt >> {0}/subdomain_collection.txt'.format(results_dir)) ''' Remove all the from_* files ''' os.system('rm -f {}/from*'.format(results_dir)) ''' Sort all Subdomains ''' os.system( 'sort -u {0}/subdomain_collection.txt -o {0}/sorted_subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/subdomain_collection.txt'.format(results_dir)) ''' The final results will be stored in sorted_subdomain_collection. ''' # parse the subdomain list file and store in db with open(subdomain_scan_results_file) as subdomain_list: for _subdomain in subdomain_list: __subdomain = _subdomain.rstrip('\n') if not Subdomain.objects.filter(scan_history=task, name=__subdomain).exists( ) and validators.domain(__subdomain) and __subdomain not in out_of_scope_subdomains: subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': __subdomain, }) save_subdomain(subdomain_dict) notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: subdomains_count = Subdomain.objects.filter(scan_history=task).count() send_notification('Subdomain Gathering for target {} has been completed and has discovered *{}* subdomains.'.format(domain.name, subdomains_count)) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/sorted_subdomain_collection.txt') # check for any subdomain changes and send notif if any if notification and notification[0].send_subdomain_changes_notif: newly_added_subdomain = get_new_added_subdomain(task.id, domain.id) if newly_added_subdomain: message = "**{} New Subdomains Discovered on domain {}**".format(newly_added_subdomain.count(), domain.name) for subdomain in newly_added_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) removed_subdomain = get_removed_subdomain(task.id, domain.id) if removed_subdomain: message = "**{} Subdomains are no longer available on domain {}**".format(removed_subdomain.count(), domain.name) for subdomain in removed_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) # check for interesting subdomains and send notif if any if notification and notification[0].send_interesting_notif: interesting_subdomain = get_interesting_subdomains(task.id, domain.id) print(interesting_subdomain) if interesting_subdomain: message = "**{} Interesting Subdomains Found on domain {}**".format(interesting_subdomain.count(), domain.name) for subdomain in interesting_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) def get_new_added_subdomain(scan_id, domain_id): scan_history = ScanHistory.objects.filter( domain=domain_id).filter( subdomain_discovery=True).filter( id__lte=scan_id) if scan_history.count() > 1: last_scan = scan_history.order_by('-start_scan_date')[1] scanned_host_q1 = Subdomain.objects.filter( scan_history__id=scan_id).values('name') scanned_host_q2 = Subdomain.objects.filter( scan_history__id=last_scan.id).values('name') added_subdomain = scanned_host_q1.difference(scanned_host_q2) return Subdomain.objects.filter( scan_history=scan_id).filter( name__in=added_subdomain) def get_removed_subdomain(scan_id, domain_id): scan_history = ScanHistory.objects.filter( domain=domain_id).filter( subdomain_discovery=True).filter( id__lte=scan_id) if scan_history.count() > 1: last_scan = scan_history.order_by('-start_scan_date')[1] scanned_host_q1 = Subdomain.objects.filter( scan_history__id=scan_id).values('name') scanned_host_q2 = Subdomain.objects.filter( scan_history__id=last_scan.id).values('name') removed_subdomains = scanned_host_q2.difference(scanned_host_q1) print() return Subdomain.objects.filter( scan_history=last_scan).filter( name__in=removed_subdomains) def http_crawler(task, domain, results_dir, activity_id): ''' This function is runs right after subdomain gathering, and gathers important like page title, http status, etc HTTP Crawler runs by default ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('HTTP Crawler for target {} has been initiated.'.format(domain.name)) alive_file_location = results_dir + '/alive.txt' httpx_results_file = results_dir + '/httpx.json' subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' httpx_command = 'httpx -status-code -content-length -title -tech-detect -cdn -ip -follow-host-redirects -random-agent' proxy = get_random_proxy() if proxy: httpx_command += " --http-proxy '{}'".format(proxy) httpx_command += ' -json -o {}'.format( httpx_results_file ) httpx_command = 'cat {} | {}'.format(subdomain_scan_results_file, httpx_command) print(httpx_command) os.system(httpx_command) # alive subdomains from httpx alive_file = open(alive_file_location, 'w') # writing httpx results if os.path.isfile(httpx_results_file): httpx_json_result = open(httpx_results_file, 'r') lines = httpx_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) try: # fallback for older versions of httpx if 'url' in json_st: subdomain = Subdomain.objects.get( scan_history=task, name=json_st['input']) else: subdomain = Subdomain.objects.get( scan_history=task, name=json_st['url'].split("//")[-1]) ''' Saving Default http urls to EndPoint ''' endpoint = EndPoint() endpoint.scan_history = task endpoint.target_domain = domain endpoint.subdomain = subdomain if 'url' in json_st: endpoint.http_url = json_st['url'] subdomain.http_url = json_st['url'] if 'status-code' in json_st: endpoint.http_status = json_st['status-code'] subdomain.http_status = json_st['status-code'] if 'title' in json_st: endpoint.page_title = json_st['title'] subdomain.page_title = json_st['title'] if 'content-length' in json_st: endpoint.content_length = json_st['content-length'] subdomain.content_length = json_st['content-length'] if 'content-type' in json_st: endpoint.content_type = json_st['content-type'] subdomain.content_type = json_st['content-type'] if 'webserver' in json_st: endpoint.webserver = json_st['webserver'] subdomain.webserver = json_st['webserver'] if 'response-time' in json_st: response_time = float( ''.join( ch for ch in json_st['response-time'] if not ch.isalpha())) if json_st['response-time'][-2:] == 'ms': response_time = response_time / 1000 endpoint.response_time = response_time subdomain.response_time = response_time if 'cnames' in json_st: cname_list = ','.join(json_st['cnames']) subdomain.cname = cname_list discovered_date = timezone.now() endpoint.discovered_date = discovered_date subdomain.discovered_date = discovered_date endpoint.is_default = True endpoint.save() subdomain.save() if 'technologies' in json_st: for _tech in json_st['technologies']: if Technology.objects.filter(name=_tech).exists(): tech = Technology.objects.get(name=_tech) else: tech = Technology(name=_tech) tech.save() subdomain.technologies.add(tech) endpoint.technologies.add(tech) if 'a' in json_st: for _ip in json_st['a']: if IpAddress.objects.filter(address=_ip).exists(): ip = IpAddress.objects.get(address=_ip) else: ip = IpAddress(address=_ip) if 'cdn' in json_st: ip.is_cdn = json_st['cdn'] ip.save() subdomain.ip_addresses.add(ip) # see if to ignore 404 or 5xx alive_file.write(json_st['url'] + '\n') subdomain.save() endpoint.save() except Exception as exception: logging.error(exception) alive_file.close() if notification and notification[0].send_scan_status_notif: alive_count = Subdomain.objects.filter( scan_history__id=task.id).values('name').distinct().filter( http_status__exact=200).count() send_notification('HTTP Crawler for target {} has been completed.\n\n {} subdomains were alive (http status 200).'.format(domain.name, alive_count)) def grab_screenshot(task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for taking screenshots ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine is currently gathering screenshots for {}'.format(domain.name)) output_screenshots_path = results_dir + '/screenshots' result_csv_path = results_dir + '/screenshots/Requests.csv' alive_subdomains_path = results_dir + '/alive.txt' eyewitness_command = 'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py' eyewitness_command += ' -f {} -d {} --no-prompt'.format( alive_subdomains_path, output_screenshots_path ) if EYEWITNESS in yaml_configuration \ and TIMEOUT in yaml_configuration[EYEWITNESS] \ and yaml_configuration[EYEWITNESS][TIMEOUT] > 0: eyewitness_command += ' --timeout {}'.format( yaml_configuration[EYEWITNESS][TIMEOUT] ) if EYEWITNESS in yaml_configuration \ and THREADS in yaml_configuration[EYEWITNESS] \ and yaml_configuration[EYEWITNESS][THREADS] > 0: eyewitness_command += ' --threads {}'.format( yaml_configuration[EYEWITNESS][THREADS] ) logger.info(eyewitness_command) os.system(eyewitness_command) if os.path.isfile(result_csv_path): logger.info('Gathering Eyewitness results') with open(result_csv_path, 'r') as file: reader = csv.reader(file) for row in reader: if row[3] == 'Successful' \ and Subdomain.objects.filter( scan_history__id=task.id).filter(name=row[2]).exists(): subdomain = Subdomain.objects.get( scan_history__id=task.id, name=row[2] ) subdomain.screenshot_path = row[4].replace( '/usr/src/scan_results/', '' ) subdomain.save() # remove all db, html extra files in screenshot results os.system('rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format( output_screenshots_path, )) os.system('rm -rf {0}/source'.format( output_screenshots_path, )) if notification and notification[0].send_scan_status_notif: send_notification('reNgine has finished gathering screenshots for {}'.format(domain.name)) def port_scanning(task, domain, yaml_configuration, results_dir): ''' This function is responsible for running the port scan ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Port Scan initiated for {}'.format(domain.name)) subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' port_results_file = results_dir + '/ports.json' # check the yaml_configuration and choose the ports to be scanned scan_ports = '-' # default port scan everything if PORTS in yaml_configuration[PORT_SCAN]: # TODO: legacy code, remove top-100 in future versions all_ports = yaml_configuration[PORT_SCAN][PORTS] if 'full' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -p {}'.format( subdomain_scan_results_file, port_results_file, '-') elif 'top-100' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -top-ports 100'.format( subdomain_scan_results_file, port_results_file) elif 'top-1000' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -top-ports 1000'.format( subdomain_scan_results_file, port_results_file) else: scan_ports = ','.join( str(port) for port in all_ports) naabu_command = 'cat {} | naabu -json -o {} -p {}'.format( subdomain_scan_results_file, port_results_file, scan_ports) # check for exclude ports if EXCLUDE_PORTS in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][EXCLUDE_PORTS]: exclude_ports = ','.join( str(port) for port in yaml_configuration['port_scan']['exclude_ports']) naabu_command = naabu_command + \ ' -exclude-ports {}'.format(exclude_ports) if NAABU_RATE in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][NAABU_RATE] > 0: naabu_command = naabu_command + \ ' -rate {}'.format( yaml_configuration[PORT_SCAN][NAABU_RATE]) if USE_NAABU_CONFIG in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][USE_NAABU_CONFIG]: naabu_command += ' -config /root/.config/naabu/naabu.conf' # run naabu os.system(naabu_command) # writing port results try: port_json_result = open(port_results_file, 'r') lines = port_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) port_number = json_st['port'] ip_address = json_st['ip'] # see if port already exists if Port.objects.filter(number__exact=port_number).exists(): port = Port.objects.get(number=port_number) else: port = Port() port.number = port_number if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port_detail = whatportis.get_ports(str(port_number)) if len(port_detail): port.service_name = port_detail[0].name port.description = port_detail[0].description port.save() if IpAddress.objects.filter(address=json_st['ip']).exists(): ip = IpAddress.objects.get(address=json_st['ip']) ip.ports.add(port) ip.save() except BaseException as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: port_count = Port.objects.filter( ports__in=IpAddress.objects.filter( ip_addresses__in=Subdomain.objects.filter( scan_history__id=task.id))).distinct().count() send_notification('reNgine has finished Port Scanning on {} and has identified {} ports.'.format(domain.name, port_count)) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/ports.json') def check_waf(): ''' This function will check for the WAF being used in subdomains using wafw00f ''' pass def directory_brute(task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for performing directory scan ''' # scan directories for all the alive subdomain with http status > # 200 notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Directory Bruteforce has been initiated for {}.'.format(domain.name)) alive_subdomains = Subdomain.objects.filter( scan_history__id=task.id).exclude(http_url__isnull=True) dirs_results = results_dir + '/dirs.json' # check the yaml settings if EXTENSIONS in yaml_configuration[DIR_FILE_SEARCH]: extensions = ','.join( str(ext) for ext in yaml_configuration[DIR_FILE_SEARCH][EXTENSIONS]) else: extensions = 'php,git,yaml,conf,db,mysql,bak,txt' # Threads if THREADS in yaml_configuration[DIR_FILE_SEARCH] \ and yaml_configuration[DIR_FILE_SEARCH][THREADS] > 0: threads = yaml_configuration[DIR_FILE_SEARCH][THREADS] else: threads = 10 for subdomain in alive_subdomains: # delete any existing dirs.json if os.path.isfile(dirs_results): os.system('rm -rf {}'.format(dirs_results)) dirsearch_command = 'python3 /usr/src/github/dirsearch/dirsearch.py' dirsearch_command += ' -u {}'.format(subdomain.http_url) if (WORDLIST not in yaml_configuration[DIR_FILE_SEARCH] or not yaml_configuration[DIR_FILE_SEARCH][WORDLIST] or 'default' in yaml_configuration[DIR_FILE_SEARCH][WORDLIST]): wordlist_location = '/usr/src/github/dirsearch/db/dicc.txt' else: wordlist_location = '/usr/src/wordlist/' + \ yaml_configuration[DIR_FILE_SEARCH][WORDLIST] + '.txt' dirsearch_command += ' -w {}'.format(wordlist_location) dirsearch_command += ' --format json -o {}'.format(dirs_results) dirsearch_command += ' -e {}'.format(extensions) dirsearch_command += ' -t {}'.format(threads) dirsearch_command += ' --random-agent --follow-redirects --exclude-status 403,401,404' if EXCLUDE_EXTENSIONS in yaml_configuration[DIR_FILE_SEARCH]: exclude_extensions = ','.join( str(ext) for ext in yaml_configuration[DIR_FILE_SEARCH][EXCLUDE_EXTENSIONS]) dirsearch_command += ' -X {}'.format(exclude_extensions) if EXCLUDE_TEXT in yaml_configuration[DIR_FILE_SEARCH]: exclude_text = ','.join( str(text) for text in yaml_configuration[DIR_FILE_SEARCH][EXCLUDE_TEXT]) dirsearch_command += ' -exclude-texts {}'.format(exclude_text) # check if recursive strategy is set to on if RECURSIVE_LEVEL in yaml_configuration[DIR_FILE_SEARCH]: dirsearch_command += ' --recursion-depth {}'.format(yaml_configuration[DIR_FILE_SEARCH][RECURSIVE_LEVEL]) if RECURSIVE_LEVEL in yaml_configuration[DIR_FILE_SEARCH]: dirsearch_command += ' --recursion-depth {}'.format(yaml_configuration[DIR_FILE_SEARCH][RECURSIVE_LEVEL]) # proxy proxy = get_random_proxy() if proxy: dirsearch_command += " --proxy '{}'".format(proxy) print(dirsearch_command) os.system(dirsearch_command) try: if os.path.isfile(dirs_results): with open(dirs_results, "r") as json_file: json_string = json_file.read() subdomain = Subdomain.objects.get( scan_history__id=task.id, http_url=subdomain.http_url) subdomain.directory_json = json_string subdomain.save() except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: send_notification('Directory Bruteforce has been completed for {}.'.format(domain.name)) def fetch_endpoints( task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for fetching all the urls associated with target and run HTTP probe It first runs gau to gather all urls from wayback, then we will use hakrawler to identify more urls ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine is currently gathering endpoints for {}.'.format(domain.name)) # check yaml settings if ALL in yaml_configuration[FETCH_URL][USES_TOOLS]: tools = 'gauplus hakrawler waybackurls gospider' else: tools = ' '.join( str(tool) for tool in yaml_configuration[FETCH_URL][USES_TOOLS]) if INTENSITY in yaml_configuration[FETCH_URL]: scan_type = yaml_configuration[FETCH_URL][INTENSITY] else: scan_type = 'normal' domain_regex = "\'https?://([a-z0-9]+[.])*{}.*\'".format(domain.name) if 'deep' in scan_type: # performs deep url gathering for all the subdomains present - # RECOMMENDED logger.info('Deep URLS Fetch') os.system(settings.TOOL_LOCATION + 'get_urls.sh %s %s %s %s %s' % ("None", results_dir, scan_type, domain_regex, tools)) else: # perform url gathering only for main domain - USE only for quick scan logger.info('Non Deep URLS Fetch') os.system( settings.TOOL_LOCATION + 'get_urls.sh %s %s %s %s %s' % ( domain.name, results_dir, scan_type, domain_regex, tools )) if IGNORE_FILE_EXTENSION in yaml_configuration[FETCH_URL]: ignore_extension = '|'.join( yaml_configuration[FETCH_URL][IGNORE_FILE_EXTENSION]) logger.info('Ignore extensions' + ignore_extension) os.system( 'cat {0}/all_urls.txt | grep -Eiv "\\.({1}).*" > {0}/temp_urls.txt'.format( results_dir, ignore_extension)) os.system( 'rm {0}/all_urls.txt && mv {0}/temp_urls.txt {0}/all_urls.txt'.format(results_dir)) ''' Store all the endpoints and then run the httpx ''' try: endpoint_final_url = results_dir + '/all_urls.txt' if os.path.isfile(endpoint_final_url): with open(endpoint_final_url) as endpoint_list: for url in endpoint_list: http_url = url.rstrip('\n') if not EndPoint.objects.filter(scan_history=task, http_url=http_url).exists(): _subdomain = get_subdomain_from_url(http_url) if Subdomain.objects.filter( scan_history=task).filter( name=_subdomain).exists(): subdomain = Subdomain.objects.get( scan_history=task, name=_subdomain) else: ''' gau or gosppider can gather interesting endpoints which when parsed can give subdomains that were not existent from subdomain scan. so storing them ''' logger.error( 'Subdomain {} not found, adding...'.format(_subdomain)) subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': _subdomain, }) subdomain = save_subdomain(subdomain_dict) endpoint_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'subdomain': subdomain, 'http_url': http_url, }) save_endpoint(endpoint_dict) except Exception as e: logger.error(e) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/all_urls.txt') ''' TODO: Go spider & waybackurls accumulates a lot of urls, which is good but nuclei takes forever to scan even a simple website, so we will do http probing and filter HTTP status 404, this way we can reduce the number of Non Existent URLS ''' logger.info('HTTP Probing on collected endpoints') httpx_command = 'httpx -l {0}/all_urls.txt -status-code -content-length -ip -cdn -title -tech-detect -json -follow-redirects -random-agent -o {0}/final_httpx_urls.json'.format(results_dir) proxy = get_random_proxy() if proxy: httpx_command += " --http-proxy '{}'".format(proxy) os.system(httpx_command) url_results_file = results_dir + '/final_httpx_urls.json' try: urls_json_result = open(url_results_file, 'r') lines = urls_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) http_url = json_st['url'] _subdomain = get_subdomain_from_url(http_url) if Subdomain.objects.filter( scan_history=task).filter( name=_subdomain).exists(): subdomain_obj = Subdomain.objects.get( scan_history=task, name=_subdomain) else: subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': _subdomain, }) subdomain_obj = save_subdomain(subdomain_dict) if EndPoint.objects.filter( scan_history=task).filter( http_url=http_url).exists(): endpoint = EndPoint.objects.get( scan_history=task, http_url=http_url) else: endpoint = EndPoint() endpoint_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'http_url': http_url, 'subdomain': subdomain_obj }) endpoint = save_endpoint(endpoint_dict) if 'title' in json_st: endpoint.page_title = json_st['title'] if 'webserver' in json_st: endpoint.webserver = json_st['webserver'] if 'content-length' in json_st: endpoint.content_length = json_st['content-length'] if 'content-type' in json_st: endpoint.content_type = json_st['content-type'] if 'status-code' in json_st: endpoint.http_status = json_st['status-code'] if 'response-time' in json_st: response_time = float(''.join(ch for ch in json_st['response-time'] if not ch.isalpha())) if json_st['response-time'][-2:] == 'ms': response_time = response_time / 1000 endpoint.response_time = response_time endpoint.save() if 'technologies' in json_st: for _tech in json_st['technologies']: if Technology.objects.filter(name=_tech).exists(): tech = Technology.objects.get(name=_tech) else: tech = Technology(name=_tech) tech.save() endpoint.technologies.add(tech) # get subdomain object subdomain = Subdomain.objects.get(scan_history=task, name=_subdomain) subdomain.technologies.add(tech) subdomain.save() except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: endpoint_count = EndPoint.objects.filter( scan_history__id=task.id).values('http_url').distinct().count() endpoint_alive_count = EndPoint.objects.filter( scan_history__id=task.id, http_status__exact=200).values('http_url').distinct().count() send_notification('reNgine has finished gathering endpoints for {} and has discovered *{}* unique endpoints.\n\n{} of those endpoints reported HTTP status 200.'.format( domain.name, endpoint_count, endpoint_alive_count )) # once endpoint is saved, run gf patterns TODO: run threads if GF_PATTERNS in yaml_configuration[FETCH_URL]: for pattern in yaml_configuration[FETCH_URL][GF_PATTERNS]: logger.info('Running GF for {}'.format(pattern)) gf_output_file_path = '{0}/gf_patterns_{1}.txt'.format( results_dir, pattern) gf_command = 'cat {0}/all_urls.txt | gf {1} >> {2}'.format( results_dir, pattern, gf_output_file_path) os.system(gf_command) if os.path.exists(gf_output_file_path): with open(gf_output_file_path) as gf_output: for line in gf_output: url = line.rstrip('\n') try: endpoint = EndPoint.objects.get( scan_history=task, http_url=url) earlier_pattern = endpoint.matched_gf_patterns new_pattern = earlier_pattern + ',' + pattern if earlier_pattern else pattern endpoint.matched_gf_patterns = new_pattern except Exception as e: # add the url in db logger.error(e) logger.info('Adding URL' + url) endpoint = EndPoint() endpoint.http_url = url endpoint.target_domain = domain endpoint.scan_history = task try: _subdomain = Subdomain.objects.get( scan_history=task, name=get_subdomain_from_url(url)) endpoint.subdomain = _subdomain except Exception as e: continue endpoint.matched_gf_patterns = pattern finally: endpoint.save() def vulnerability_scan( task, domain, yaml_configuration, results_dir, activity_id): notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Vulnerability scan has been initiated for {}.'.format(domain.name)) ''' This function will run nuclei as a vulnerability scanner ---- unfurl the urls to keep only domain and path, this will be sent to vuln scan ignore certain file extensions Thanks: https://github.com/six2dez/reconftw ''' urls_path = '/alive.txt' if task.scan_type.fetch_url: os.system('cat {0}/all_urls.txt | grep -Eiv "\\.(eot|jpg|jpeg|gif|css|tif|tiff|png|ttf|otf|woff|woff2|ico|pdf|svg|txt|js|doc|docx)$" | unfurl -u format %s://%d%p >> {0}/unfurl_urls.txt'.format(results_dir)) os.system( 'sort -u {0}/unfurl_urls.txt -o {0}/unfurl_urls.txt'.format(results_dir)) urls_path = '/unfurl_urls.txt' vulnerability_result_path = results_dir + '/vulnerability.json' vulnerability_scan_input_file = results_dir + urls_path nuclei_command = 'nuclei -json -l {} -o {}'.format( vulnerability_scan_input_file, vulnerability_result_path) # check nuclei config if USE_NUCLEI_CONFIG in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[VULNERABILITY_SCAN][USE_NUCLEI_CONFIG]: nuclei_command += ' -config /root/.config/nuclei/config.yaml' ''' Nuclei Templates Either custom template has to be supplied or default template, if neither has been supplied then use all templates including custom templates ''' if CUSTOM_NUCLEI_TEMPLATE in yaml_configuration[ VULNERABILITY_SCAN] or NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: # check yaml settings for templates if NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: if ALL in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_TEMPLATE]: template = NUCLEI_TEMPLATES_PATH else: _template = ','.join([NUCLEI_TEMPLATES_PATH + str(element) for element in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_TEMPLATE]]) template = _template.replace(',', ' -t ') # Update nuclei command with templates nuclei_command = nuclei_command + ' -t ' + template if CUSTOM_NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: # add .yaml to the custom template extensions _template = ','.join( [str(element) + '.yaml' for element in yaml_configuration[VULNERABILITY_SCAN][CUSTOM_NUCLEI_TEMPLATE]]) template = _template.replace(',', ' -t ') # Update nuclei command with templates nuclei_command = nuclei_command + ' -t ' + template else: nuclei_command = nuclei_command + ' -t /root/nuclei-templates' # check yaml settings for concurrency if NUCLEI_CONCURRENCY in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][NUCLEI_CONCURRENCY] > 0: concurrency = yaml_configuration[VULNERABILITY_SCAN][NUCLEI_CONCURRENCY] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -c ' + str(concurrency) if RATE_LIMIT in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][RATE_LIMIT] > 0: rate_limit = yaml_configuration[VULNERABILITY_SCAN][RATE_LIMIT] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -rl ' + str(rate_limit) if TIMEOUT in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][TIMEOUT] > 0: timeout = yaml_configuration[VULNERABILITY_SCAN][TIMEOUT] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -timeout ' + str(timeout) if RETRIES in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][RETRIES] > 0: retries = yaml_configuration[VULNERABILITY_SCAN][RETRIES] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -retries ' + str(retries) # for severity if NUCLEI_SEVERITY in yaml_configuration[VULNERABILITY_SCAN] and ALL not in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_SEVERITY]: _severity = ','.join( [str(element) for element in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_SEVERITY]]) severity = _severity.replace(" ", "") else: severity = "critical, high, medium, low, info" # update nuclei templates before running scan os.system('nuclei -update-templates') for _severity in severity.split(","): # delete any existing vulnerability.json file if os.path.isfile(vulnerability_result_path): os.system('rm {}'.format(vulnerability_result_path)) # run nuclei final_nuclei_command = nuclei_command + ' -severity ' + _severity proxy = get_random_proxy() if proxy: final_nuclei_command += " --proxy-url '{}'".format(proxy) logger.info(final_nuclei_command) os.system(final_nuclei_command) try: if os.path.isfile(vulnerability_result_path): urls_json_result = open(vulnerability_result_path, 'r') lines = urls_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) host = json_st['host'] _subdomain = get_subdomain_from_url(host) try: subdomain = Subdomain.objects.get( name=_subdomain, scan_history=task) vulnerability = Vulnerability() vulnerability.subdomain = subdomain vulnerability.scan_history = task vulnerability.target_domain = domain try: endpoint = EndPoint.objects.get( scan_history=task, target_domain=domain, http_url=host) vulnerability.endpoint = endpoint except Exception as exception: logger.error(exception) if 'name' in json_st['info']: vulnerability.name = json_st['info']['name'] if 'severity' in json_st['info']: if json_st['info']['severity'] == 'info': severity = 0 elif json_st['info']['severity'] == 'low': severity = 1 elif json_st['info']['severity'] == 'medium': severity = 2 elif json_st['info']['severity'] == 'high': severity = 3 elif json_st['info']['severity'] == 'critical': severity = 4 else: severity = 0 else: severity = 0 vulnerability.severity = severity if 'tags' in json_st['info']: vulnerability.tags = json_st['info']['tags'] if 'description' in json_st['info']: vulnerability.description = json_st['info']['description'] if 'reference' in json_st['info']: vulnerability.reference = json_st['info']['reference'] if 'matched' in json_st: vulnerability.http_url = json_st['matched'] if 'templateID' in json_st: vulnerability.template_used = json_st['templateID'] if 'description' in json_st: vulnerability.description = json_st['description'] if 'matcher_name' in json_st: vulnerability.matcher_name = json_st['matcher_name'] if 'extracted_results' in json_st: vulnerability.extracted_results = json_st['extracted_results'] vulnerability.discovered_date = timezone.now() vulnerability.open_status = True vulnerability.save() # send notification for all vulnerabilities except info if json_st['info']['severity'] != "info" and notification and notification[0].send_vuln_notif: message = "*Alert: Vulnerability Identified*" message += "\n\n" message += "A *{}* severity vulnerability has been identified.".format(json_st['info']['severity']) message += "\nVulnerability Name: {}".format(json_st['info']['name']) message += "\nVulnerable URL: {}".format(json_st['host']) send_notification(message) # send report to hackerone if Hackerone.objects.all().exists() and json_st['info']['severity'] != 'info' and json_st['info']['severity'] \ != 'low' and vulnerability.target_domain.h1_team_handle: hackerone = Hackerone.objects.all()[0] if hackerone.send_critical and json_st['info']['severity'] == 'critical': send_hackerone_report(vulnerability.id) elif hackerone.send_high and json_st['info']['severity'] == 'high': send_hackerone_report(vulnerability.id) elif hackerone.send_medium and json_st['info']['severity'] == 'medium': send_hackerone_report(vulnerability.id) except ObjectDoesNotExist: logger.error('Object not found') continue except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: info_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=0).count() low_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=1).count() medium_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=2).count() high_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=3).count() critical_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=4).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count message = 'Vulnerability scan has been completed for {} and discovered {} vulnerabilities.'.format( domain.name, vulnerability_count ) message += '\n\n*Vulnerability Stats:*' message += '\nCritical: {}'.format(critical_count) message += '\nHigh: {}'.format(high_count) message += '\nMedium: {}'.format(medium_count) message += '\nLow: {}'.format(low_count) message += '\nInfo: {}'.format(info_count) send_notification(message) def scan_failed(task): task.scan_status = 0 task.stop_scan_date = timezone.now() task.save() def create_scan_activity(task, message, status): scan_activity = ScanActivity() scan_activity.scan_of = task scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id def update_last_activity(id, activity_status): ScanActivity.objects.filter( id=id).update( status=activity_status, time=timezone.now()) def delete_scan_data(results_dir): # remove all txt,html,json files os.system('find {} -name "*.txt" -type f -delete'.format(results_dir)) os.system('find {} -name "*.html" -type f -delete'.format(results_dir)) os.system('find {} -name "*.json" -type f -delete'.format(results_dir)) def save_subdomain(subdomain_dict): subdomain = Subdomain() subdomain.discovered_date = timezone.now() subdomain.target_domain = subdomain_dict.get('target_domain') subdomain.scan_history = subdomain_dict.get('scan_history') subdomain.name = subdomain_dict.get('name') subdomain.http_url = subdomain_dict.get('http_url') subdomain.screenshot_path = subdomain_dict.get('screenshot_path') subdomain.http_header_path = subdomain_dict.get('http_header_path') subdomain.cname = subdomain_dict.get('cname') subdomain.is_cdn = subdomain_dict.get('is_cdn') subdomain.content_type = subdomain_dict.get('content_type') subdomain.webserver = subdomain_dict.get('webserver') subdomain.page_title = subdomain_dict.get('page_title') subdomain.is_imported_subdomain = subdomain_dict.get( 'is_imported_subdomain') if 'is_imported_subdomain' in subdomain_dict else False if 'http_status' in subdomain_dict: subdomain.http_status = subdomain_dict.get('http_status') if 'response_time' in subdomain_dict: subdomain.response_time = subdomain_dict.get('response_time') if 'content_length' in subdomain_dict: subdomain.content_length = subdomain_dict.get('content_length') subdomain.save() return subdomain def save_endpoint(endpoint_dict): endpoint = EndPoint() endpoint.discovered_date = timezone.now() endpoint.scan_history = endpoint_dict.get('scan_history') endpoint.target_domain = endpoint_dict.get('target_domain') if 'target_domain' in endpoint_dict else None endpoint.subdomain = endpoint_dict.get('subdomain') if 'target_domain' in endpoint_dict else None endpoint.http_url = endpoint_dict.get('http_url') endpoint.page_title = endpoint_dict.get('page_title') if 'page_title' in endpoint_dict else None endpoint.content_type = endpoint_dict.get('content_type') if 'content_type' in endpoint_dict else None endpoint.webserver = endpoint_dict.get('webserver') if 'webserver' in endpoint_dict else None endpoint.response_time = endpoint_dict.get('response_time') if 'response_time' in endpoint_dict else 0 endpoint.http_status = endpoint_dict.get('http_status') if 'http_status' in endpoint_dict else 0 endpoint.content_length = endpoint_dict.get('content_length') if 'content_length' in endpoint_dict else 0 endpoint.is_default = endpoint_dict.get('is_default') if 'is_default' in endpoint_dict else False endpoint.save() return endpoint def perform_osint(task, domain, yaml_configuration, results_dir): notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine has initiated OSINT on target {}'.format(domain.name)) if 'discover' in yaml_configuration[OSINT]: osint_discovery(task, domain, yaml_configuration, results_dir) if 'dork' in yaml_configuration[OSINT]: dorking(task, yaml_configuration) if notification and notification[0].send_scan_status_notif: send_notification('reNgine has completed performing OSINT on target {}'.format(domain.name)) def osint_discovery(task, domain, yaml_configuration, results_dir): if ALL in yaml_configuration[OSINT][OSINT_DISCOVER]: osint_lookup = 'emails metainfo employees' else: osint_lookup = ' '.join( str(lookup) for lookup in yaml_configuration[OSINT][OSINT_DISCOVER]) if 'metainfo' in osint_lookup: if INTENSITY in yaml_configuration[OSINT]: osint_intensity = yaml_configuration[OSINT][INTENSITY] else: osint_intensity = 'normal' if OSINT_DOCUMENTS_LIMIT in yaml_configuration[OSINT]: documents_limit = yaml_configuration[OSINT][OSINT_DOCUMENTS_LIMIT] else: documents_limit = 50 if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': domain.name, 'domain': domain, 'scan_id': task, 'documents_limit': documents_limit }) get_and_save_meta_info(meta_dict) elif osint_intensity == 'deep': # get all subdomains in scan_id subdomains = Subdomain.objects.filter(scan_history=task) for subdomain in subdomains: meta_dict = DottedDict({ 'osint_target': subdomain.name, 'domain': domain, 'scan_id': task, 'documents_limit': documents_limit }) get_and_save_meta_info(meta_dict) if 'emails' in osint_lookup: get_and_save_emails(task, results_dir) get_and_save_leaked_credentials(task, results_dir) if 'employees' in osint_lookup: get_and_save_employees(task, results_dir) def dorking(scan_history, yaml_configuration): # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh # look in stackoverflow if ALL in yaml_configuration[OSINT][OSINT_DORK]: dork_lookup = 'stackoverflow, 3rdparty, social_media, project_management, code_sharing, config_files, jenkins, cloud_buckets, php_error, exposed_documents, struts_rce, db_files, traefik, git_exposed' else: dork_lookup = ' '.join( str(lookup) for lookup in yaml_configuration[OSINT][OSINT_DORK]) if 'stackoverflow' in dork_lookup: dork = 'site:stackoverflow.com' dork_type = 'stackoverflow' get_and_save_dork_results( dork, dork_type, scan_history, in_target=False ) if '3rdparty' in dork_lookup: # look in 3rd party sitee dork_type = '3rdparty' lookup_websites = [ 'gitter.im', 'papaly.com', 'productforums.google.com', 'coggle.it', 'replt.it', 'ycombinator.com', 'libraries.io', 'npm.runkit.com', 'npmjs.com', 'scribd.com', 'gitter.im' ] dork = '' for website in lookup_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'social_media' in dork_lookup: dork_type = 'Social Media' social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'pinterest.com', 'tumblr.com', 'reddit.com' ] dork = '' for website in social_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'project_management' in dork_lookup: dork_type = 'Project Management' project_websites = [ 'trello.com', '*.atlassian.net' ] dork = '' for website in project_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'code_sharing' in dork_lookup: dork_type = 'Code Sharing Sites' code_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] dork = '' for website in code_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'config_files' in dork_lookup: dork_type = 'Config Files' config_file_ext = [ 'env', 'xml', 'conf', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] dork = '' for extension in config_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'jenkins' in dork_lookup: dork_type = 'Jenkins' dork = 'intitle:\"Dashboard [Jenkins]\"' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) if 'wordpress_files' in dork_lookup: dork_type = 'Wordpress Files' inurl_lookup = [ 'wp-content', 'wp-includes' ] dork = '' for lookup in inurl_lookup: dork = dork + ' | ' + 'inurl:' + lookup get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'cloud_buckets' in dork_lookup: dork_type = 'Cloud Buckets' cloud_websites = [ '.s3.amazonaws.com', 'storage.googleapis.com', 'amazonaws.com' ] dork = '' for website in cloud_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'php_error' in dork_lookup: dork_type = 'PHP Error' error_words = [ '\"PHP Parse error\"', '\"PHP Warning\"', '\"PHP Error\"' ] dork = '' for word in error_words: dork = dork + ' | ' + word get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'exposed_documents' in dork_lookup: dork_type = 'Exposed Documents' docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] dork = '' for extension in docs_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'struts_rce' in dork_lookup: dork_type = 'Apache Struts RCE' struts_file_ext = [ 'action', 'struts', 'do' ] dork = '' for extension in struts_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'db_files' in dork_lookup: dork_type = 'Database Files' db_file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] dork = '' for extension in db_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'traefik' in dork_lookup: dork = 'intitle:traefik inurl:8080/dashboard' dork_type = 'Traefik' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) if 'git_exposed' in dork_lookup: dork = 'inurl:\"/.git\"' dork_type = '.git Exposed' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) def get_and_save_dork_results(dork, type, scan_history, in_target=False): degoogle_obj = degoogle.dg() proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy if in_target: query = dork + " site:" + scan_history.domain.name else: query = dork + " \"{}\"".format(scan_history.domain.name) logger.info(query) degoogle_obj.query = query results = degoogle_obj.run() logger.info(results) for result in results: dork, _ = Dork.objects.get_or_create( type=type, description=result['desc'], url=result['url'] ) scan_history.dorks.add(dork) def get_and_save_employees(scan_history, results_dir): theHarvester_location = '/usr/src/github/theHarvester' # update proxies.yaml if Proxy.objects.all().exists(): proxy = Proxy.objects.all()[0] if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(theHarvester_location + '/proxies.yaml', 'w') as file: documents = yaml.dump(yaml_data, file) os.system('cd {} && python3 theHarvester.py -d {} -b all -f {}/theHarvester.html'.format( theHarvester_location, scan_history.domain.name, results_dir )) file_location = results_dir + '/theHarvester.html' print(file_location) # delete proxy environ var if os.environ.get(('https_proxy')): del os.environ['https_proxy'] if os.environ.get(('HTTPS_PROXY')): del os.environ['HTTPS_PROXY'] if os.path.isfile(file_location): logger.info('Parsing theHarvester results') options = FirefoxOptions() options.add_argument("--headless") driver = webdriver.Firefox(options=options) driver.get('file://'+file_location) tabledata = driver.execute_script('return tabledata') # save email addresses and linkedin employees for data in tabledata: if data['record'] == 'email': _email = data['result'] email, _ = Email.objects.get_or_create(address=_email) scan_history.emails.add(email) elif data['record'] == 'people': _employee = data['result'] split_val = _employee.split('-') name = split_val[0] if len(split_val) == 2: designation = split_val[1] else: designation = "" employee, _ = Employee.objects.get_or_create(name=name, designation=designation) scan_history.employees.add(employee) driver.quit() print(tabledata) def get_and_save_emails(scan_history, results_dir): leak_target_path = '{}/creds_target.txt'.format(results_dir) # get email address proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy emails = [] try: logger.info('OSINT: Getting emails from Google') email_from_google = get_emails_from_google(scan_history.domain.name) logger.info('OSINT: Getting emails from Bing') email_from_bing = get_emails_from_bing(scan_history.domain.name) logger.info('OSINT: Getting emails from Baidu') email_from_baidu = get_emails_from_baidu(scan_history.domain.name) emails = list(set(email_from_google + email_from_bing + email_from_baidu)) logger.info(emails) except Exception as e: logger.error(e) leak_target_file = open(leak_target_path, 'w') for _email in emails: email, _ = Email.objects.get_or_create(address=_email) scan_history.emails.add(email) leak_target_file.write('{}\n'.format(_email)) # fill leak_target_file with possible email address leak_target_file.write('%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.write('%.%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%.%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.write('%_%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%_%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.close() def get_and_save_leaked_credentials(scan_history, results_dir): logger.info('OSINT: Getting leaked credentials...') leak_target_file = '{}/creds_target.txt'.format(results_dir) leak_output_file = '{}/pwndb.json'.format(results_dir) pwndb_command = 'python3 /usr/src/github/pwndb/pwndb.py --proxy tor:9150 --output json --list {}'.format( leak_target_file ) try: pwndb_output = subprocess.getoutput(pwndb_command) creds = json.loads(pwndb_output) for cred in creds: if cred['username'] != 'donate': email_id = "{}@{}".format(cred['username'], cred['domain']) email_obj, _ = Email.objects.get_or_create( address=email_id, ) email_obj.password = cred['password'] email_obj.save() scan_history.emails.add(email_obj) except Exception as e: logger.error(e) pass def get_and_save_meta_info(meta_dict): logger.info('Getting METADATA for {}'.format(meta_dict.osint_target)) proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy result = metadata_extractor.extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if result: results = result.get_metadata() for meta in results: meta_finder_document = MetaFinderDocument() subdomain = Subdomain.objects.get(scan_history=meta_dict.scan_id, name=meta_dict.osint_target) meta_finder_document.subdomain = subdomain meta_finder_document.target_domain = meta_dict.domain meta_finder_document.scan_history = meta_dict.scan_id item = DottedDict(results[meta]) meta_finder_document.url = item.url meta_finder_document.doc_name = meta meta_finder_document.http_status = item.status_code metadata = results[meta]['metadata'] for data in metadata: if 'Producer' in metadata and metadata['Producer']: meta_finder_document.producer = metadata['Producer'].rstrip('\x00') if 'Creator' in metadata and metadata['Creator']: meta_finder_document.creator = metadata['Creator'].rstrip('\x00') if 'CreationDate' in metadata and metadata['CreationDate']: meta_finder_document.creation_date = metadata['CreationDate'].rstrip('\x00') if 'ModDate' in metadata and metadata['ModDate']: meta_finder_document.modified_date = metadata['ModDate'].rstrip('\x00') if 'Author' in metadata and metadata['Author']: meta_finder_document.author = metadata['Author'].rstrip('\x00') if 'Title' in metadata and metadata['Title']: meta_finder_document.title = metadata['Title'].rstrip('\x00') if 'OSInfo' in metadata and metadata['OSInfo']: meta_finder_document.os = metadata['OSInfo'].rstrip('\x00') meta_finder_document.save() @app.task(bind=True) def test_task(self): print('*' * 40) print('test task run') print('*' * 40)

import os import traceback import yaml import json import csv import validators import random import requests import logging import metafinder.extractor as metadata_extractor import whatportis import subprocess from selenium.webdriver.firefox.options import Options as FirefoxOptions from selenium import webdriver from emailfinder.extractor import * from dotted_dict import DottedDict from celery import shared_task from discord_webhook import DiscordWebhook from reNgine.celery import app from startScan.models import * from targetApp.models import Domain from scanEngine.models import EngineType from django.conf import settings from django.shortcuts import get_object_or_404 from celery import shared_task from datetime import datetime from degoogle import degoogle from django.conf import settings from django.utils import timezone, dateformat from django.shortcuts import get_object_or_404 from django.core.exceptions import ObjectDoesNotExist from reNgine.celery import app from reNgine.definitions import * from startScan.models import * from targetApp.models import Domain from scanEngine.models import EngineType, Configuration, Wordlist from .common_func import * ''' task for background scan ''' @app.task def initiate_scan( domain_id, scan_history_id, scan_type, engine_type, imported_subdomains=None, out_of_scope_subdomains=[] ): ''' scan_type = 0 -> immediate scan, need not create scan object scan_type = 1 -> scheduled scan ''' engine_object = EngineType.objects.get(pk=engine_type) domain = Domain.objects.get(pk=domain_id) if scan_type == 1: task = ScanHistory() task.scan_status = -1 elif scan_type == 0: task = ScanHistory.objects.get(pk=scan_history_id) # save the last scan date for domain model domain.last_scan_date = timezone.now() domain.save() # once the celery task starts, change the task status to Started task.scan_type = engine_object task.celery_id = initiate_scan.request.id task.domain = domain task.scan_status = 1 task.start_scan_date = timezone.now() task.subdomain_discovery = True if engine_object.subdomain_discovery else False task.dir_file_search = True if engine_object.dir_file_search else False task.port_scan = True if engine_object.port_scan else False task.fetch_url = True if engine_object.fetch_url else False task.osint = True if engine_object.osint else False task.screenshot = True if engine_object.screenshot else False task.vulnerability_scan = True if engine_object.vulnerability_scan else False task.save() activity_id = create_scan_activity(task, "Scanning Started", 2) results_dir = '/usr/src/scan_results/' os.chdir(results_dir) notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine has initiated recon for target {} with engine type {}'.format(domain.name, engine_object.engine_name)) try: current_scan_dir = domain.name + '_' + str(random.randint(100000000000, 999999999999)) os.mkdir(current_scan_dir) task.results_dir = current_scan_dir task.save() except Exception as exception: logger.error(exception) scan_failed(task) yaml_configuration = None excluded_subdomains = '' try: yaml_configuration = yaml.load( task.scan_type.yaml_configuration, Loader=yaml.FullLoader) except Exception as exception: logger.error(exception) # TODO: Put failed reason on db ''' Add GF patterns name to db for dynamic URLs menu ''' if engine_object.fetch_url and GF_PATTERNS in yaml_configuration[FETCH_URL]: task.used_gf_patterns = ','.join( pattern for pattern in yaml_configuration[FETCH_URL][GF_PATTERNS]) task.save() results_dir = results_dir + current_scan_dir # put all imported subdomains into txt file and also in Subdomain model if imported_subdomains: extract_imported_subdomain( imported_subdomains, task, domain, results_dir) if yaml_configuration: ''' a target in itself is a subdomain, some tool give subdomains as www.yogeshojha.com but url and everything else resolves to yogeshojha.com In that case, we would already need to store target itself as subdomain ''' initial_subdomain_file = '/target_domain.txt' if task.subdomain_discovery else '/sorted_subdomain_collection.txt' subdomain_file = open(results_dir + initial_subdomain_file, "w") subdomain_file.write(domain.name + "\n") subdomain_file.close() if(task.subdomain_discovery): activity_id = create_scan_activity(task, "Subdomain Scanning", 1) subdomain_scan( task, domain, yaml_configuration, results_dir, activity_id, out_of_scope_subdomains ) else: skip_subdomain_scan(task, domain, results_dir) update_last_activity(activity_id, 2) activity_id = create_scan_activity(task, "HTTP Crawler", 1) http_crawler( task, domain, results_dir, activity_id) update_last_activity(activity_id, 2) try: if task.screenshot: activity_id = create_scan_activity( task, "Visual Recon - Screenshot", 1) grab_screenshot( task, domain, yaml_configuration, current_scan_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if(task.port_scan): activity_id = create_scan_activity(task, "Port Scanning", 1) port_scanning(task, domain, yaml_configuration, results_dir) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.osint: activity_id = create_scan_activity(task, "OSINT Running", 1) perform_osint(task, domain, yaml_configuration, results_dir) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.dir_file_search: activity_id = create_scan_activity(task, "Directory Search", 1) directory_brute( task, domain, yaml_configuration, results_dir, activity_id ) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.fetch_url: activity_id = create_scan_activity(task, "Fetching endpoints", 1) fetch_endpoints( task, domain, yaml_configuration, results_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.vulnerability_scan: activity_id = create_scan_activity(task, "Vulnerability Scan", 1) vulnerability_scan( task, domain, yaml_configuration, results_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) activity_id = create_scan_activity(task, "Scan Completed", 2) if notification and notification[0].send_scan_status_notif: send_notification('*Scan Completed*\nreNgine has finished performing recon on target {}.'.format(domain.name)) ''' Once the scan is completed, save the status to successful ''' if ScanActivity.objects.filter(scan_of=task).filter(status=0).all(): task.scan_status = 0 else: task.scan_status = 2 task.stop_scan_date = timezone.now() task.save() # cleanup results delete_scan_data(results_dir) return {"status": True} def skip_subdomain_scan(task, domain, results_dir): # store default target as subdomain ''' If the imported subdomain already has target domain saved, we can skip this ''' if not Subdomain.objects.filter( scan_history=task, name=domain.name).exists(): subdomain_dict = DottedDict({ 'name': domain.name, 'scan_history': task, 'target_domain': domain }) save_subdomain(subdomain_dict) # Save target into target_domain.txt with open('{}/target_domain.txt'.format(results_dir), 'w+') as file: file.write(domain.name + '\n') file.close() ''' We can have two conditions, either subdomain scan happens, or subdomain scan does not happen, in either cases, because we are using import subdomain, we need to collect and sort all the subdomains Write target domain into subdomain_collection ''' os.system( 'cat {0}/target_domain.txt > {0}/subdomain_collection.txt'.format(results_dir)) os.system( 'cat {0}/from_imported.txt > {0}/subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/from_imported.txt'.format(results_dir)) ''' Sort all Subdomains ''' os.system( 'sort -u {0}/subdomain_collection.txt -o {0}/sorted_subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/subdomain_collection.txt'.format(results_dir)) def extract_imported_subdomain(imported_subdomains, task, domain, results_dir): valid_imported_subdomains = [subdomain for subdomain in imported_subdomains if validators.domain( subdomain) and domain.name == get_domain_from_subdomain(subdomain)] # remove any duplicate valid_imported_subdomains = list(set(valid_imported_subdomains)) with open('{}/from_imported.txt'.format(results_dir), 'w+') as file: for subdomain_name in valid_imported_subdomains: # save _subdomain to Subdomain model db if not Subdomain.objects.filter( scan_history=task, name=subdomain_name).exists(): subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': subdomain_name, 'is_imported_subdomain': True }) save_subdomain(subdomain_dict) # save subdomain to file file.write('{}\n'.format(subdomain_name)) file.close() def subdomain_scan(task, domain, yaml_configuration, results_dir, activity_id, out_of_scope_subdomains=None): ''' This function is responsible for performing subdomain enumeration ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Subdomain Gathering for target {} has been started'.format(domain.name)) subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' # check for all the tools and add them into string # if tool selected is all then make string, no need for loop if ALL in yaml_configuration[SUBDOMAIN_DISCOVERY][USES_TOOLS]: tools = 'amass-active amass-passive assetfinder sublist3r subfinder oneforall' else: tools = ' '.join( str(tool) for tool in yaml_configuration[SUBDOMAIN_DISCOVERY][USES_TOOLS]) logging.info(tools) # check for THREADS, by default 10 threads = 10 if THREADS in yaml_configuration[SUBDOMAIN_DISCOVERY]: _threads = yaml_configuration[SUBDOMAIN_DISCOVERY][THREADS] if _threads > 0: threads = _threads if 'amass' in tools: if 'amass-passive' in tools: amass_command = 'amass enum -passive -d {} -o {}/from_amass.txt'.format( domain.name, results_dir) if USE_AMASS_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_AMASS_CONFIG]: amass_command += ' -config /root/.config/amass.ini' # Run Amass Passive logging.info(amass_command) os.system(amass_command) if 'amass-active' in tools: amass_command = 'amass enum -active -d {} -o {}/from_amass_active.txt'.format( domain.name, results_dir) if USE_AMASS_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_AMASS_CONFIG]: amass_command += ' -config /root/.config/amass.ini' if AMASS_WORDLIST in yaml_configuration[SUBDOMAIN_DISCOVERY]: wordlist = yaml_configuration[SUBDOMAIN_DISCOVERY][AMASS_WORDLIST] if wordlist == 'default': wordlist_path = '/usr/src/wordlist/deepmagic.com-prefixes-top50000.txt' else: wordlist_path = '/usr/src/wordlist/' + wordlist + '.txt' if not os.path.exists(wordlist_path): wordlist_path = '/usr/src/' + AMASS_WORDLIST amass_command = amass_command + \ ' -brute -w {}'.format(wordlist_path) if amass_config_path: amass_command = amass_command + \ ' -config {}'.format('/usr/src/scan_results/' + amass_config_path) # Run Amass Active logging.info(amass_command) os.system(amass_command) if 'assetfinder' in tools: assetfinder_command = 'assetfinder --subs-only {} > {}/from_assetfinder.txt'.format( domain.name, results_dir) # Run Assetfinder logging.info(assetfinder_command) os.system(assetfinder_command) if 'sublist3r' in tools: sublist3r_command = 'python3 /usr/src/github/Sublist3r/sublist3r.py -d {} -t {} -o {}/from_sublister.txt'.format( domain.name, threads, results_dir) # Run sublist3r logging.info(sublist3r_command) os.system(sublist3r_command) if 'subfinder' in tools: subfinder_command = 'subfinder -d {} -t {} -o {}/from_subfinder.txt'.format( domain.name, threads, results_dir) if USE_SUBFINDER_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_SUBFINDER_CONFIG]: subfinder_command += ' -config /root/.config/subfinder/config.yaml' # Run Subfinder logging.info(subfinder_command) os.system(subfinder_command) if 'oneforall' in tools: oneforall_command = 'python3 /usr/src/github/OneForAll/oneforall.py --target {} run'.format( domain.name, results_dir) # Run OneForAll logging.info(oneforall_command) os.system(oneforall_command) extract_subdomain = "cut -d',' -f6 /usr/src/github/OneForAll/results/{}.csv >> {}/from_oneforall.txt".format( domain.name, results_dir) os.system(extract_subdomain) # remove the results from oneforall directory os.system( 'rm -rf /usr/src/github/OneForAll/results/{}.*'.format(domain.name)) ''' All tools have gathered the list of subdomains with filename initials as from_* We will gather all the results in one single file, sort them and remove the older results from_* ''' os.system( 'cat {0}/*.txt > {0}/subdomain_collection.txt'.format(results_dir)) ''' Write target domain into subdomain_collection ''' os.system( 'cat {0}/target_domain.txt >> {0}/subdomain_collection.txt'.format(results_dir)) ''' Remove all the from_* files ''' os.system('rm -f {}/from*'.format(results_dir)) ''' Sort all Subdomains ''' os.system( 'sort -u {0}/subdomain_collection.txt -o {0}/sorted_subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/subdomain_collection.txt'.format(results_dir)) ''' The final results will be stored in sorted_subdomain_collection. ''' # parse the subdomain list file and store in db with open(subdomain_scan_results_file) as subdomain_list: for _subdomain in subdomain_list: __subdomain = _subdomain.rstrip('\n') if not Subdomain.objects.filter(scan_history=task, name=__subdomain).exists( ) and validators.domain(__subdomain) and __subdomain not in out_of_scope_subdomains: subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': __subdomain, }) save_subdomain(subdomain_dict) notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: subdomains_count = Subdomain.objects.filter(scan_history=task).count() send_notification('Subdomain Gathering for target {} has been completed and has discovered *{}* subdomains.'.format(domain.name, subdomains_count)) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/sorted_subdomain_collection.txt') # check for any subdomain changes and send notif if any if notification and notification[0].send_subdomain_changes_notif: newly_added_subdomain = get_new_added_subdomain(task.id, domain.id) if newly_added_subdomain: message = "**{} New Subdomains Discovered on domain {}**".format(newly_added_subdomain.count(), domain.name) for subdomain in newly_added_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) removed_subdomain = get_removed_subdomain(task.id, domain.id) if removed_subdomain: message = "**{} Subdomains are no longer available on domain {}**".format(removed_subdomain.count(), domain.name) for subdomain in removed_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) # check for interesting subdomains and send notif if any if notification and notification[0].send_interesting_notif: interesting_subdomain = get_interesting_subdomains(task.id, domain.id) print(interesting_subdomain) if interesting_subdomain: message = "**{} Interesting Subdomains Found on domain {}**".format(interesting_subdomain.count(), domain.name) for subdomain in interesting_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) def get_new_added_subdomain(scan_id, domain_id): scan_history = ScanHistory.objects.filter( domain=domain_id).filter( subdomain_discovery=True).filter( id__lte=scan_id) if scan_history.count() > 1: last_scan = scan_history.order_by('-start_scan_date')[1] scanned_host_q1 = Subdomain.objects.filter( scan_history__id=scan_id).values('name') scanned_host_q2 = Subdomain.objects.filter( scan_history__id=last_scan.id).values('name') added_subdomain = scanned_host_q1.difference(scanned_host_q2) return Subdomain.objects.filter( scan_history=scan_id).filter( name__in=added_subdomain) def get_removed_subdomain(scan_id, domain_id): scan_history = ScanHistory.objects.filter( domain=domain_id).filter( subdomain_discovery=True).filter( id__lte=scan_id) if scan_history.count() > 1: last_scan = scan_history.order_by('-start_scan_date')[1] scanned_host_q1 = Subdomain.objects.filter( scan_history__id=scan_id).values('name') scanned_host_q2 = Subdomain.objects.filter( scan_history__id=last_scan.id).values('name') removed_subdomains = scanned_host_q2.difference(scanned_host_q1) print() return Subdomain.objects.filter( scan_history=last_scan).filter( name__in=removed_subdomains) def http_crawler(task, domain, results_dir, activity_id): ''' This function is runs right after subdomain gathering, and gathers important like page title, http status, etc HTTP Crawler runs by default ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('HTTP Crawler for target {} has been initiated.'.format(domain.name)) alive_file_location = results_dir + '/alive.txt' httpx_results_file = results_dir + '/httpx.json' subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' httpx_command = 'httpx -status-code -content-length -title -tech-detect -cdn -ip -follow-host-redirects -random-agent' proxy = get_random_proxy() if proxy: httpx_command += " --http-proxy '{}'".format(proxy) httpx_command += ' -json -o {}'.format( httpx_results_file ) httpx_command = 'cat {} | {}'.format(subdomain_scan_results_file, httpx_command) print(httpx_command) os.system(httpx_command) # alive subdomains from httpx alive_file = open(alive_file_location, 'w') # writing httpx results if os.path.isfile(httpx_results_file): httpx_json_result = open(httpx_results_file, 'r') lines = httpx_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) try: # fallback for older versions of httpx if 'url' in json_st: subdomain = Subdomain.objects.get( scan_history=task, name=json_st['input']) else: subdomain = Subdomain.objects.get( scan_history=task, name=json_st['url'].split("//")[-1]) ''' Saving Default http urls to EndPoint ''' endpoint = EndPoint() endpoint.scan_history = task endpoint.target_domain = domain endpoint.subdomain = subdomain if 'url' in json_st: endpoint.http_url = json_st['url'] subdomain.http_url = json_st['url'] if 'status-code' in json_st: endpoint.http_status = json_st['status-code'] subdomain.http_status = json_st['status-code'] if 'title' in json_st: endpoint.page_title = json_st['title'] subdomain.page_title = json_st['title'] if 'content-length' in json_st: endpoint.content_length = json_st['content-length'] subdomain.content_length = json_st['content-length'] if 'content-type' in json_st: endpoint.content_type = json_st['content-type'] subdomain.content_type = json_st['content-type'] if 'webserver' in json_st: endpoint.webserver = json_st['webserver'] subdomain.webserver = json_st['webserver'] if 'response-time' in json_st: response_time = float( ''.join( ch for ch in json_st['response-time'] if not ch.isalpha())) if json_st['response-time'][-2:] == 'ms': response_time = response_time / 1000 endpoint.response_time = response_time subdomain.response_time = response_time if 'cnames' in json_st: cname_list = ','.join(json_st['cnames']) subdomain.cname = cname_list discovered_date = timezone.now() endpoint.discovered_date = discovered_date subdomain.discovered_date = discovered_date endpoint.is_default = True endpoint.save() subdomain.save() if 'technologies' in json_st: for _tech in json_st['technologies']: if Technology.objects.filter(name=_tech).exists(): tech = Technology.objects.get(name=_tech) else: tech = Technology(name=_tech) tech.save() subdomain.technologies.add(tech) endpoint.technologies.add(tech) if 'a' in json_st: for _ip in json_st['a']: if IpAddress.objects.filter(address=_ip).exists(): ip = IpAddress.objects.get(address=_ip) else: ip = IpAddress(address=_ip) if 'cdn' in json_st: ip.is_cdn = json_st['cdn'] ip.save() subdomain.ip_addresses.add(ip) # see if to ignore 404 or 5xx alive_file.write(json_st['url'] + '\n') subdomain.save() endpoint.save() except Exception as exception: logging.error(exception) alive_file.close() if notification and notification[0].send_scan_status_notif: alive_count = Subdomain.objects.filter( scan_history__id=task.id).values('name').distinct().filter( http_status__exact=200).count() send_notification('HTTP Crawler for target {} has been completed.\n\n {} subdomains were alive (http status 200).'.format(domain.name, alive_count)) def grab_screenshot(task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for taking screenshots ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine is currently gathering screenshots for {}'.format(domain.name)) output_screenshots_path = results_dir + '/screenshots' result_csv_path = results_dir + '/screenshots/Requests.csv' alive_subdomains_path = results_dir + '/alive.txt' eyewitness_command = 'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py' eyewitness_command += ' -f {} -d {} --no-prompt'.format( alive_subdomains_path, output_screenshots_path ) if EYEWITNESS in yaml_configuration \ and TIMEOUT in yaml_configuration[EYEWITNESS] \ and yaml_configuration[EYEWITNESS][TIMEOUT] > 0: eyewitness_command += ' --timeout {}'.format( yaml_configuration[EYEWITNESS][TIMEOUT] ) if EYEWITNESS in yaml_configuration \ and THREADS in yaml_configuration[EYEWITNESS] \ and yaml_configuration[EYEWITNESS][THREADS] > 0: eyewitness_command += ' --threads {}'.format( yaml_configuration[EYEWITNESS][THREADS] ) logger.info(eyewitness_command) os.system(eyewitness_command) if os.path.isfile(result_csv_path): logger.info('Gathering Eyewitness results') with open(result_csv_path, 'r') as file: reader = csv.reader(file) for row in reader: if row[3] == 'Successful' \ and Subdomain.objects.filter( scan_history__id=task.id).filter(name=row[2]).exists(): subdomain = Subdomain.objects.get( scan_history__id=task.id, name=row[2] ) subdomain.screenshot_path = row[4].replace( '/usr/src/scan_results/', '' ) subdomain.save() # remove all db, html extra files in screenshot results os.system('rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format( output_screenshots_path, )) os.system('rm -rf {0}/source'.format( output_screenshots_path, )) if notification and notification[0].send_scan_status_notif: send_notification('reNgine has finished gathering screenshots for {}'.format(domain.name)) def port_scanning(task, domain, yaml_configuration, results_dir): ''' This function is responsible for running the port scan ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Port Scan initiated for {}'.format(domain.name)) subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' port_results_file = results_dir + '/ports.json' # check the yaml_configuration and choose the ports to be scanned scan_ports = '-' # default port scan everything if PORTS in yaml_configuration[PORT_SCAN]: # TODO: legacy code, remove top-100 in future versions all_ports = yaml_configuration[PORT_SCAN][PORTS] if 'full' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -p {}'.format( subdomain_scan_results_file, port_results_file, '-') elif 'top-100' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -top-ports 100'.format( subdomain_scan_results_file, port_results_file) elif 'top-1000' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -top-ports 1000'.format( subdomain_scan_results_file, port_results_file) else: scan_ports = ','.join( str(port) for port in all_ports) naabu_command = 'cat {} | naabu -json -o {} -p {}'.format( subdomain_scan_results_file, port_results_file, scan_ports) # check for exclude ports if EXCLUDE_PORTS in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][EXCLUDE_PORTS]: exclude_ports = ','.join( str(port) for port in yaml_configuration['port_scan']['exclude_ports']) naabu_command = naabu_command + \ ' -exclude-ports {}'.format(exclude_ports) if NAABU_RATE in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][NAABU_RATE] > 0: naabu_command = naabu_command + \ ' -rate {}'.format( yaml_configuration[PORT_SCAN][NAABU_RATE]) if USE_NAABU_CONFIG in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][USE_NAABU_CONFIG]: naabu_command += ' -config /root/.config/naabu/naabu.conf' # run naabu os.system(naabu_command) # writing port results try: port_json_result = open(port_results_file, 'r') lines = port_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) port_number = json_st['port'] ip_address = json_st['ip'] # see if port already exists if Port.objects.filter(number__exact=port_number).exists(): port = Port.objects.get(number=port_number) else: port = Port() port.number = port_number if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port_detail = whatportis.get_ports(str(port_number)) if len(port_detail): port.service_name = port_detail[0].name port.description = port_detail[0].description port.save() if IpAddress.objects.filter(address=json_st['ip']).exists(): ip = IpAddress.objects.get(address=json_st['ip']) ip.ports.add(port) ip.save() except BaseException as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: port_count = Port.objects.filter( ports__in=IpAddress.objects.filter( ip_addresses__in=Subdomain.objects.filter( scan_history__id=task.id))).distinct().count() send_notification('reNgine has finished Port Scanning on {} and has identified {} ports.'.format(domain.name, port_count)) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/ports.json') def check_waf(): ''' This function will check for the WAF being used in subdomains using wafw00f ''' pass def directory_brute(task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for performing directory scan ''' # scan directories for all the alive subdomain with http status > # 200 notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Directory Bruteforce has been initiated for {}.'.format(domain.name)) alive_subdomains = Subdomain.objects.filter( scan_history__id=task.id).exclude(http_url__isnull=True) dirs_results = results_dir + '/dirs.json' # check the yaml settings if EXTENSIONS in yaml_configuration[DIR_FILE_SEARCH]: extensions = ','.join( str(ext) for ext in yaml_configuration[DIR_FILE_SEARCH][EXTENSIONS]) else: extensions = 'php,git,yaml,conf,db,mysql,bak,txt' # Threads if THREADS in yaml_configuration[DIR_FILE_SEARCH] \ and yaml_configuration[DIR_FILE_SEARCH][THREADS] > 0: threads = yaml_configuration[DIR_FILE_SEARCH][THREADS] else: threads = 10 for subdomain in alive_subdomains: # delete any existing dirs.json if os.path.isfile(dirs_results): os.system('rm -rf {}'.format(dirs_results)) dirsearch_command = 'python3 /usr/src/github/dirsearch/dirsearch.py' dirsearch_command += ' -u {}'.format(subdomain.http_url) if (WORDLIST not in yaml_configuration[DIR_FILE_SEARCH] or not yaml_configuration[DIR_FILE_SEARCH][WORDLIST] or 'default' in yaml_configuration[DIR_FILE_SEARCH][WORDLIST]): wordlist_location = '/usr/src/github/dirsearch/db/dicc.txt' else: wordlist_location = '/usr/src/wordlist/' + \ yaml_configuration[DIR_FILE_SEARCH][WORDLIST] + '.txt' dirsearch_command += ' -w {}'.format(wordlist_location) dirsearch_command += ' --format json -o {}'.format(dirs_results) dirsearch_command += ' -e {}'.format(extensions) dirsearch_command += ' -t {}'.format(threads) dirsearch_command += ' --random-agent --follow-redirects --exclude-status 403,401,404' if EXCLUDE_EXTENSIONS in yaml_configuration[DIR_FILE_SEARCH]: exclude_extensions = ','.join( str(ext) for ext in yaml_configuration[DIR_FILE_SEARCH][EXCLUDE_EXTENSIONS]) dirsearch_command += ' -X {}'.format(exclude_extensions) if EXCLUDE_TEXT in yaml_configuration[DIR_FILE_SEARCH]: exclude_text = ','.join( str(text) for text in yaml_configuration[DIR_FILE_SEARCH][EXCLUDE_TEXT]) dirsearch_command += ' -exclude-texts {}'.format(exclude_text) # check if recursive strategy is set to on if RECURSIVE_LEVEL in yaml_configuration[DIR_FILE_SEARCH]: dirsearch_command += ' --recursion-depth {}'.format(yaml_configuration[DIR_FILE_SEARCH][RECURSIVE_LEVEL]) if RECURSIVE_LEVEL in yaml_configuration[DIR_FILE_SEARCH]: dirsearch_command += ' --recursion-depth {}'.format(yaml_configuration[DIR_FILE_SEARCH][RECURSIVE_LEVEL]) # proxy proxy = get_random_proxy() if proxy: dirsearch_command += " --proxy '{}'".format(proxy) print(dirsearch_command) os.system(dirsearch_command) try: if os.path.isfile(dirs_results): with open(dirs_results, "r") as json_file: json_string = json_file.read() subdomain = Subdomain.objects.get( scan_history__id=task.id, http_url=subdomain.http_url) subdomain.directory_json = json_string subdomain.save() except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: send_notification('Directory Bruteforce has been completed for {}.'.format(domain.name)) def fetch_endpoints( task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for fetching all the urls associated with target and run HTTP probe It first runs gau to gather all urls from wayback, then we will use hakrawler to identify more urls ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine is currently gathering endpoints for {}.'.format(domain.name)) # check yaml settings if ALL in yaml_configuration[FETCH_URL][USES_TOOLS]: tools = 'gauplus hakrawler waybackurls gospider' else: tools = ' '.join( str(tool) for tool in yaml_configuration[FETCH_URL][USES_TOOLS]) if INTENSITY in yaml_configuration[FETCH_URL]: scan_type = yaml_configuration[FETCH_URL][INTENSITY] else: scan_type = 'normal' domain_regex = "\'https?://([a-z0-9]+[.])*{}.*\'".format(domain.name) if 'deep' in scan_type: # performs deep url gathering for all the subdomains present - # RECOMMENDED logger.info('Deep URLS Fetch') os.system(settings.TOOL_LOCATION + 'get_urls.sh %s %s %s %s %s' % ("None", results_dir, scan_type, domain_regex, tools)) else: # perform url gathering only for main domain - USE only for quick scan logger.info('Non Deep URLS Fetch') os.system( settings.TOOL_LOCATION + 'get_urls.sh %s %s %s %s %s' % ( domain.name, results_dir, scan_type, domain_regex, tools )) if IGNORE_FILE_EXTENSION in yaml_configuration[FETCH_URL]: ignore_extension = '|'.join( yaml_configuration[FETCH_URL][IGNORE_FILE_EXTENSION]) logger.info('Ignore extensions' + ignore_extension) os.system( 'cat {0}/all_urls.txt | grep -Eiv "\\.({1}).*" > {0}/temp_urls.txt'.format( results_dir, ignore_extension)) os.system( 'rm {0}/all_urls.txt && mv {0}/temp_urls.txt {0}/all_urls.txt'.format(results_dir)) ''' Store all the endpoints and then run the httpx ''' try: endpoint_final_url = results_dir + '/all_urls.txt' if os.path.isfile(endpoint_final_url): with open(endpoint_final_url) as endpoint_list: for url in endpoint_list: http_url = url.rstrip('\n') if not EndPoint.objects.filter(scan_history=task, http_url=http_url).exists(): _subdomain = get_subdomain_from_url(http_url) if Subdomain.objects.filter( scan_history=task).filter( name=_subdomain).exists(): subdomain = Subdomain.objects.get( scan_history=task, name=_subdomain) else: ''' gau or gosppider can gather interesting endpoints which when parsed can give subdomains that were not existent from subdomain scan. so storing them ''' logger.error( 'Subdomain {} not found, adding...'.format(_subdomain)) subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': _subdomain, }) subdomain = save_subdomain(subdomain_dict) endpoint_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'subdomain': subdomain, 'http_url': http_url, }) save_endpoint(endpoint_dict) except Exception as e: logger.error(e) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/all_urls.txt') ''' TODO: Go spider & waybackurls accumulates a lot of urls, which is good but nuclei takes forever to scan even a simple website, so we will do http probing and filter HTTP status 404, this way we can reduce the number of Non Existent URLS ''' logger.info('HTTP Probing on collected endpoints') httpx_command = 'httpx -l {0}/all_urls.txt -status-code -content-length -ip -cdn -title -tech-detect -json -follow-redirects -random-agent -o {0}/final_httpx_urls.json'.format(results_dir) proxy = get_random_proxy() if proxy: httpx_command += " --http-proxy '{}'".format(proxy) os.system(httpx_command) url_results_file = results_dir + '/final_httpx_urls.json' try: urls_json_result = open(url_results_file, 'r') lines = urls_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) http_url = json_st['url'] _subdomain = get_subdomain_from_url(http_url) if Subdomain.objects.filter( scan_history=task).filter( name=_subdomain).exists(): subdomain_obj = Subdomain.objects.get( scan_history=task, name=_subdomain) else: subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': _subdomain, }) subdomain_obj = save_subdomain(subdomain_dict) if EndPoint.objects.filter( scan_history=task).filter( http_url=http_url).exists(): endpoint = EndPoint.objects.get( scan_history=task, http_url=http_url) else: endpoint = EndPoint() endpoint_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'http_url': http_url, 'subdomain': subdomain_obj }) endpoint = save_endpoint(endpoint_dict) if 'title' in json_st: endpoint.page_title = json_st['title'] if 'webserver' in json_st: endpoint.webserver = json_st['webserver'] if 'content-length' in json_st: endpoint.content_length = json_st['content-length'] if 'content-type' in json_st: endpoint.content_type = json_st['content-type'] if 'status-code' in json_st: endpoint.http_status = json_st['status-code'] if 'response-time' in json_st: response_time = float(''.join(ch for ch in json_st['response-time'] if not ch.isalpha())) if json_st['response-time'][-2:] == 'ms': response_time = response_time / 1000 endpoint.response_time = response_time endpoint.save() if 'technologies' in json_st: for _tech in json_st['technologies']: if Technology.objects.filter(name=_tech).exists(): tech = Technology.objects.get(name=_tech) else: tech = Technology(name=_tech) tech.save() endpoint.technologies.add(tech) # get subdomain object subdomain = Subdomain.objects.get(scan_history=task, name=_subdomain) subdomain.technologies.add(tech) subdomain.save() except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: endpoint_count = EndPoint.objects.filter( scan_history__id=task.id).values('http_url').distinct().count() endpoint_alive_count = EndPoint.objects.filter( scan_history__id=task.id, http_status__exact=200).values('http_url').distinct().count() send_notification('reNgine has finished gathering endpoints for {} and has discovered *{}* unique endpoints.\n\n{} of those endpoints reported HTTP status 200.'.format( domain.name, endpoint_count, endpoint_alive_count )) # once endpoint is saved, run gf patterns TODO: run threads if GF_PATTERNS in yaml_configuration[FETCH_URL]: for pattern in yaml_configuration[FETCH_URL][GF_PATTERNS]: logger.info('Running GF for {}'.format(pattern)) gf_output_file_path = '{0}/gf_patterns_{1}.txt'.format( results_dir, pattern) gf_command = 'cat {0}/all_urls.txt | gf {1} >> {2}'.format( results_dir, pattern, gf_output_file_path) os.system(gf_command) if os.path.exists(gf_output_file_path): with open(gf_output_file_path) as gf_output: for line in gf_output: url = line.rstrip('\n') try: endpoint = EndPoint.objects.get( scan_history=task, http_url=url) earlier_pattern = endpoint.matched_gf_patterns new_pattern = earlier_pattern + ',' + pattern if earlier_pattern else pattern endpoint.matched_gf_patterns = new_pattern except Exception as e: # add the url in db logger.error(e) logger.info('Adding URL' + url) endpoint = EndPoint() endpoint.http_url = url endpoint.target_domain = domain endpoint.scan_history = task try: _subdomain = Subdomain.objects.get( scan_history=task, name=get_subdomain_from_url(url)) endpoint.subdomain = _subdomain except Exception as e: continue endpoint.matched_gf_patterns = pattern finally: endpoint.save() def vulnerability_scan( task, domain, yaml_configuration, results_dir, activity_id): notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Vulnerability scan has been initiated for {}.'.format(domain.name)) ''' This function will run nuclei as a vulnerability scanner ---- unfurl the urls to keep only domain and path, this will be sent to vuln scan ignore certain file extensions Thanks: https://github.com/six2dez/reconftw ''' urls_path = '/alive.txt' if task.scan_type.fetch_url: os.system('cat {0}/all_urls.txt | grep -Eiv "\\.(eot|jpg|jpeg|gif|css|tif|tiff|png|ttf|otf|woff|woff2|ico|pdf|svg|txt|js|doc|docx)$" | unfurl -u format %s://%d%p >> {0}/unfurl_urls.txt'.format(results_dir)) os.system( 'sort -u {0}/unfurl_urls.txt -o {0}/unfurl_urls.txt'.format(results_dir)) urls_path = '/unfurl_urls.txt' vulnerability_result_path = results_dir + '/vulnerability.json' vulnerability_scan_input_file = results_dir + urls_path nuclei_command = 'nuclei -json -l {} -o {}'.format( vulnerability_scan_input_file, vulnerability_result_path) # check nuclei config if USE_NUCLEI_CONFIG in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[VULNERABILITY_SCAN][USE_NUCLEI_CONFIG]: nuclei_command += ' -config /root/.config/nuclei/config.yaml' ''' Nuclei Templates Either custom template has to be supplied or default template, if neither has been supplied then use all templates including custom templates ''' if CUSTOM_NUCLEI_TEMPLATE in yaml_configuration[ VULNERABILITY_SCAN] or NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: # check yaml settings for templates if NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: if ALL in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_TEMPLATE]: template = NUCLEI_TEMPLATES_PATH else: _template = ','.join([NUCLEI_TEMPLATES_PATH + str(element) for element in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_TEMPLATE]]) template = _template.replace(',', ' -t ') # Update nuclei command with templates nuclei_command = nuclei_command + ' -t ' + template if CUSTOM_NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: # add .yaml to the custom template extensions _template = ','.join( [str(element) + '.yaml' for element in yaml_configuration[VULNERABILITY_SCAN][CUSTOM_NUCLEI_TEMPLATE]]) template = _template.replace(',', ' -t ') # Update nuclei command with templates nuclei_command = nuclei_command + ' -t ' + template else: nuclei_command = nuclei_command + ' -t /root/nuclei-templates' # check yaml settings for concurrency if NUCLEI_CONCURRENCY in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][NUCLEI_CONCURRENCY] > 0: concurrency = yaml_configuration[VULNERABILITY_SCAN][NUCLEI_CONCURRENCY] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -c ' + str(concurrency) if RATE_LIMIT in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][RATE_LIMIT] > 0: rate_limit = yaml_configuration[VULNERABILITY_SCAN][RATE_LIMIT] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -rl ' + str(rate_limit) if TIMEOUT in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][TIMEOUT] > 0: timeout = yaml_configuration[VULNERABILITY_SCAN][TIMEOUT] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -timeout ' + str(timeout) if RETRIES in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][RETRIES] > 0: retries = yaml_configuration[VULNERABILITY_SCAN][RETRIES] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -retries ' + str(retries) # for severity if NUCLEI_SEVERITY in yaml_configuration[VULNERABILITY_SCAN] and ALL not in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_SEVERITY]: _severity = ','.join( [str(element) for element in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_SEVERITY]]) severity = _severity.replace(" ", "") else: severity = "critical, high, medium, low, info" # update nuclei templates before running scan os.system('nuclei -update-templates') for _severity in severity.split(","): # delete any existing vulnerability.json file if os.path.isfile(vulnerability_result_path): os.system('rm {}'.format(vulnerability_result_path)) # run nuclei final_nuclei_command = nuclei_command + ' -severity ' + _severity proxy = get_random_proxy() if proxy: final_nuclei_command += " --proxy-url '{}'".format(proxy) logger.info(final_nuclei_command) os.system(final_nuclei_command) try: if os.path.isfile(vulnerability_result_path): urls_json_result = open(vulnerability_result_path, 'r') lines = urls_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) host = json_st['host'] _subdomain = get_subdomain_from_url(host) try: subdomain = Subdomain.objects.get( name=_subdomain, scan_history=task) vulnerability = Vulnerability() vulnerability.subdomain = subdomain vulnerability.scan_history = task vulnerability.target_domain = domain try: endpoint = EndPoint.objects.get( scan_history=task, target_domain=domain, http_url=host) vulnerability.endpoint = endpoint except Exception as exception: logger.error(exception) if 'name' in json_st['info']: vulnerability.name = json_st['info']['name'] if 'severity' in json_st['info']: if json_st['info']['severity'] == 'info': severity = 0 elif json_st['info']['severity'] == 'low': severity = 1 elif json_st['info']['severity'] == 'medium': severity = 2 elif json_st['info']['severity'] == 'high': severity = 3 elif json_st['info']['severity'] == 'critical': severity = 4 else: severity = 0 else: severity = 0 vulnerability.severity = severity if 'tags' in json_st['info']: vulnerability.tags = json_st['info']['tags'] if 'description' in json_st['info']: vulnerability.description = json_st['info']['description'] if 'reference' in json_st['info']: vulnerability.reference = json_st['info']['reference'] if 'matched' in json_st: # TODO remove in rengine 1.1. 'matched' isn't used in nuclei 2.5.3 vulnerability.http_url = json_st['matched'] if 'matched-at' in json_st: vulnerability.http_url = json_st['matched-at'] if 'templateID' in json_st: vulnerability.template_used = json_st['templateID'] if 'description' in json_st: vulnerability.description = json_st['description'] if 'matcher_name' in json_st: vulnerability.matcher_name = json_st['matcher_name'] if 'extracted_results' in json_st: vulnerability.extracted_results = json_st['extracted_results'] vulnerability.discovered_date = timezone.now() vulnerability.open_status = True vulnerability.save() # send notification for all vulnerabilities except info if json_st['info']['severity'] != "info" and notification and notification[0].send_vuln_notif: message = "*Alert: Vulnerability Identified*" message += "\n\n" message += "A *{}* severity vulnerability has been identified.".format(json_st['info']['severity']) message += "\nVulnerability Name: {}".format(json_st['info']['name']) message += "\nVulnerable URL: {}".format(json_st['host']) send_notification(message) # send report to hackerone if Hackerone.objects.all().exists() and json_st['info']['severity'] != 'info' and json_st['info']['severity'] \ != 'low' and vulnerability.target_domain.h1_team_handle: hackerone = Hackerone.objects.all()[0] if hackerone.send_critical and json_st['info']['severity'] == 'critical': send_hackerone_report(vulnerability.id) elif hackerone.send_high and json_st['info']['severity'] == 'high': send_hackerone_report(vulnerability.id) elif hackerone.send_medium and json_st['info']['severity'] == 'medium': send_hackerone_report(vulnerability.id) except ObjectDoesNotExist: logger.error('Object not found') continue except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: info_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=0).count() low_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=1).count() medium_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=2).count() high_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=3).count() critical_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=4).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count message = 'Vulnerability scan has been completed for {} and discovered {} vulnerabilities.'.format( domain.name, vulnerability_count ) message += '\n\n*Vulnerability Stats:*' message += '\nCritical: {}'.format(critical_count) message += '\nHigh: {}'.format(high_count) message += '\nMedium: {}'.format(medium_count) message += '\nLow: {}'.format(low_count) message += '\nInfo: {}'.format(info_count) send_notification(message) def scan_failed(task): task.scan_status = 0 task.stop_scan_date = timezone.now() task.save() def create_scan_activity(task, message, status): scan_activity = ScanActivity() scan_activity.scan_of = task scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id def update_last_activity(id, activity_status): ScanActivity.objects.filter( id=id).update( status=activity_status, time=timezone.now()) def delete_scan_data(results_dir): # remove all txt,html,json files os.system('find {} -name "*.txt" -type f -delete'.format(results_dir)) os.system('find {} -name "*.html" -type f -delete'.format(results_dir)) os.system('find {} -name "*.json" -type f -delete'.format(results_dir)) def save_subdomain(subdomain_dict): subdomain = Subdomain() subdomain.discovered_date = timezone.now() subdomain.target_domain = subdomain_dict.get('target_domain') subdomain.scan_history = subdomain_dict.get('scan_history') subdomain.name = subdomain_dict.get('name') subdomain.http_url = subdomain_dict.get('http_url') subdomain.screenshot_path = subdomain_dict.get('screenshot_path') subdomain.http_header_path = subdomain_dict.get('http_header_path') subdomain.cname = subdomain_dict.get('cname') subdomain.is_cdn = subdomain_dict.get('is_cdn') subdomain.content_type = subdomain_dict.get('content_type') subdomain.webserver = subdomain_dict.get('webserver') subdomain.page_title = subdomain_dict.get('page_title') subdomain.is_imported_subdomain = subdomain_dict.get( 'is_imported_subdomain') if 'is_imported_subdomain' in subdomain_dict else False if 'http_status' in subdomain_dict: subdomain.http_status = subdomain_dict.get('http_status') if 'response_time' in subdomain_dict: subdomain.response_time = subdomain_dict.get('response_time') if 'content_length' in subdomain_dict: subdomain.content_length = subdomain_dict.get('content_length') subdomain.save() return subdomain def save_endpoint(endpoint_dict): endpoint = EndPoint() endpoint.discovered_date = timezone.now() endpoint.scan_history = endpoint_dict.get('scan_history') endpoint.target_domain = endpoint_dict.get('target_domain') if 'target_domain' in endpoint_dict else None endpoint.subdomain = endpoint_dict.get('subdomain') if 'target_domain' in endpoint_dict else None endpoint.http_url = endpoint_dict.get('http_url') endpoint.page_title = endpoint_dict.get('page_title') if 'page_title' in endpoint_dict else None endpoint.content_type = endpoint_dict.get('content_type') if 'content_type' in endpoint_dict else None endpoint.webserver = endpoint_dict.get('webserver') if 'webserver' in endpoint_dict else None endpoint.response_time = endpoint_dict.get('response_time') if 'response_time' in endpoint_dict else 0 endpoint.http_status = endpoint_dict.get('http_status') if 'http_status' in endpoint_dict else 0 endpoint.content_length = endpoint_dict.get('content_length') if 'content_length' in endpoint_dict else 0 endpoint.is_default = endpoint_dict.get('is_default') if 'is_default' in endpoint_dict else False endpoint.save() return endpoint def perform_osint(task, domain, yaml_configuration, results_dir): notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine has initiated OSINT on target {}'.format(domain.name)) if 'discover' in yaml_configuration[OSINT]: osint_discovery(task, domain, yaml_configuration, results_dir) if 'dork' in yaml_configuration[OSINT]: dorking(task, yaml_configuration) if notification and notification[0].send_scan_status_notif: send_notification('reNgine has completed performing OSINT on target {}'.format(domain.name)) def osint_discovery(task, domain, yaml_configuration, results_dir): if ALL in yaml_configuration[OSINT][OSINT_DISCOVER]: osint_lookup = 'emails metainfo employees' else: osint_lookup = ' '.join( str(lookup) for lookup in yaml_configuration[OSINT][OSINT_DISCOVER]) if 'metainfo' in osint_lookup: if INTENSITY in yaml_configuration[OSINT]: osint_intensity = yaml_configuration[OSINT][INTENSITY] else: osint_intensity = 'normal' if OSINT_DOCUMENTS_LIMIT in yaml_configuration[OSINT]: documents_limit = yaml_configuration[OSINT][OSINT_DOCUMENTS_LIMIT] else: documents_limit = 50 if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': domain.name, 'domain': domain, 'scan_id': task, 'documents_limit': documents_limit }) get_and_save_meta_info(meta_dict) elif osint_intensity == 'deep': # get all subdomains in scan_id subdomains = Subdomain.objects.filter(scan_history=task) for subdomain in subdomains: meta_dict = DottedDict({ 'osint_target': subdomain.name, 'domain': domain, 'scan_id': task, 'documents_limit': documents_limit }) get_and_save_meta_info(meta_dict) if 'emails' in osint_lookup: get_and_save_emails(task, results_dir) get_and_save_leaked_credentials(task, results_dir) if 'employees' in osint_lookup: get_and_save_employees(task, results_dir) def dorking(scan_history, yaml_configuration): # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh # look in stackoverflow if ALL in yaml_configuration[OSINT][OSINT_DORK]: dork_lookup = 'stackoverflow, 3rdparty, social_media, project_management, code_sharing, config_files, jenkins, cloud_buckets, php_error, exposed_documents, struts_rce, db_files, traefik, git_exposed' else: dork_lookup = ' '.join( str(lookup) for lookup in yaml_configuration[OSINT][OSINT_DORK]) if 'stackoverflow' in dork_lookup: dork = 'site:stackoverflow.com' dork_type = 'stackoverflow' get_and_save_dork_results( dork, dork_type, scan_history, in_target=False ) if '3rdparty' in dork_lookup: # look in 3rd party sitee dork_type = '3rdparty' lookup_websites = [ 'gitter.im', 'papaly.com', 'productforums.google.com', 'coggle.it', 'replt.it', 'ycombinator.com', 'libraries.io', 'npm.runkit.com', 'npmjs.com', 'scribd.com', 'gitter.im' ] dork = '' for website in lookup_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'social_media' in dork_lookup: dork_type = 'Social Media' social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'pinterest.com', 'tumblr.com', 'reddit.com' ] dork = '' for website in social_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'project_management' in dork_lookup: dork_type = 'Project Management' project_websites = [ 'trello.com', '*.atlassian.net' ] dork = '' for website in project_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'code_sharing' in dork_lookup: dork_type = 'Code Sharing Sites' code_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] dork = '' for website in code_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'config_files' in dork_lookup: dork_type = 'Config Files' config_file_ext = [ 'env', 'xml', 'conf', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] dork = '' for extension in config_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'jenkins' in dork_lookup: dork_type = 'Jenkins' dork = 'intitle:\"Dashboard [Jenkins]\"' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) if 'wordpress_files' in dork_lookup: dork_type = 'Wordpress Files' inurl_lookup = [ 'wp-content', 'wp-includes' ] dork = '' for lookup in inurl_lookup: dork = dork + ' | ' + 'inurl:' + lookup get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'cloud_buckets' in dork_lookup: dork_type = 'Cloud Buckets' cloud_websites = [ '.s3.amazonaws.com', 'storage.googleapis.com', 'amazonaws.com' ] dork = '' for website in cloud_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'php_error' in dork_lookup: dork_type = 'PHP Error' error_words = [ '\"PHP Parse error\"', '\"PHP Warning\"', '\"PHP Error\"' ] dork = '' for word in error_words: dork = dork + ' | ' + word get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'exposed_documents' in dork_lookup: dork_type = 'Exposed Documents' docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] dork = '' for extension in docs_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'struts_rce' in dork_lookup: dork_type = 'Apache Struts RCE' struts_file_ext = [ 'action', 'struts', 'do' ] dork = '' for extension in struts_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'db_files' in dork_lookup: dork_type = 'Database Files' db_file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] dork = '' for extension in db_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'traefik' in dork_lookup: dork = 'intitle:traefik inurl:8080/dashboard' dork_type = 'Traefik' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) if 'git_exposed' in dork_lookup: dork = 'inurl:\"/.git\"' dork_type = '.git Exposed' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) def get_and_save_dork_results(dork, type, scan_history, in_target=False): degoogle_obj = degoogle.dg() proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy if in_target: query = dork + " site:" + scan_history.domain.name else: query = dork + " \"{}\"".format(scan_history.domain.name) logger.info(query) degoogle_obj.query = query results = degoogle_obj.run() logger.info(results) for result in results: dork, _ = Dork.objects.get_or_create( type=type, description=result['desc'], url=result['url'] ) scan_history.dorks.add(dork) def get_and_save_employees(scan_history, results_dir): theHarvester_location = '/usr/src/github/theHarvester' # update proxies.yaml if Proxy.objects.all().exists(): proxy = Proxy.objects.all()[0] if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(theHarvester_location + '/proxies.yaml', 'w') as file: documents = yaml.dump(yaml_data, file) os.system('cd {} && python3 theHarvester.py -d {} -b all -f {}/theHarvester.html'.format( theHarvester_location, scan_history.domain.name, results_dir )) file_location = results_dir + '/theHarvester.html' print(file_location) # delete proxy environ var if os.environ.get(('https_proxy')): del os.environ['https_proxy'] if os.environ.get(('HTTPS_PROXY')): del os.environ['HTTPS_PROXY'] if os.path.isfile(file_location): logger.info('Parsing theHarvester results') options = FirefoxOptions() options.add_argument("--headless") driver = webdriver.Firefox(options=options) driver.get('file://'+file_location) tabledata = driver.execute_script('return tabledata') # save email addresses and linkedin employees for data in tabledata: if data['record'] == 'email': _email = data['result'] email, _ = Email.objects.get_or_create(address=_email) scan_history.emails.add(email) elif data['record'] == 'people': _employee = data['result'] split_val = _employee.split('-') name = split_val[0] if len(split_val) == 2: designation = split_val[1] else: designation = "" employee, _ = Employee.objects.get_or_create(name=name, designation=designation) scan_history.employees.add(employee) driver.quit() print(tabledata) def get_and_save_emails(scan_history, results_dir): leak_target_path = '{}/creds_target.txt'.format(results_dir) # get email address proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy emails = [] try: logger.info('OSINT: Getting emails from Google') email_from_google = get_emails_from_google(scan_history.domain.name) logger.info('OSINT: Getting emails from Bing') email_from_bing = get_emails_from_bing(scan_history.domain.name) logger.info('OSINT: Getting emails from Baidu') email_from_baidu = get_emails_from_baidu(scan_history.domain.name) emails = list(set(email_from_google + email_from_bing + email_from_baidu)) logger.info(emails) except Exception as e: logger.error(e) leak_target_file = open(leak_target_path, 'w') for _email in emails: email, _ = Email.objects.get_or_create(address=_email) scan_history.emails.add(email) leak_target_file.write('{}\n'.format(_email)) # fill leak_target_file with possible email address leak_target_file.write('%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.write('%.%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%.%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.write('%_%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%_%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.close() def get_and_save_leaked_credentials(scan_history, results_dir): logger.info('OSINT: Getting leaked credentials...') leak_target_file = '{}/creds_target.txt'.format(results_dir) leak_output_file = '{}/pwndb.json'.format(results_dir) pwndb_command = 'python3 /usr/src/github/pwndb/pwndb.py --proxy tor:9150 --output json --list {}'.format( leak_target_file ) try: pwndb_output = subprocess.getoutput(pwndb_command) creds = json.loads(pwndb_output) for cred in creds: if cred['username'] != 'donate': email_id = "{}@{}".format(cred['username'], cred['domain']) email_obj, _ = Email.objects.get_or_create( address=email_id, ) email_obj.password = cred['password'] email_obj.save() scan_history.emails.add(email_obj) except Exception as e: logger.error(e) pass def get_and_save_meta_info(meta_dict): logger.info('Getting METADATA for {}'.format(meta_dict.osint_target)) proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy result = metadata_extractor.extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if result: results = result.get_metadata() for meta in results: meta_finder_document = MetaFinderDocument() subdomain = Subdomain.objects.get(scan_history=meta_dict.scan_id, name=meta_dict.osint_target) meta_finder_document.subdomain = subdomain meta_finder_document.target_domain = meta_dict.domain meta_finder_document.scan_history = meta_dict.scan_id item = DottedDict(results[meta]) meta_finder_document.url = item.url meta_finder_document.doc_name = meta meta_finder_document.http_status = item.status_code metadata = results[meta]['metadata'] for data in metadata: if 'Producer' in metadata and metadata['Producer']: meta_finder_document.producer = metadata['Producer'].rstrip('\x00') if 'Creator' in metadata and metadata['Creator']: meta_finder_document.creator = metadata['Creator'].rstrip('\x00') if 'CreationDate' in metadata and metadata['CreationDate']: meta_finder_document.creation_date = metadata['CreationDate'].rstrip('\x00') if 'ModDate' in metadata and metadata['ModDate']: meta_finder_document.modified_date = metadata['ModDate'].rstrip('\x00') if 'Author' in metadata and metadata['Author']: meta_finder_document.author = metadata['Author'].rstrip('\x00') if 'Title' in metadata and metadata['Title']: meta_finder_document.title = metadata['Title'].rstrip('\x00') if 'OSInfo' in metadata and metadata['OSInfo']: meta_finder_document.os = metadata['OSInfo'].rstrip('\x00') meta_finder_document.save() @app.task(bind=True) def test_task(self): print('*' * 40) print('test task run') print('*' * 40)

radaram

43af3a6aecdece4923ee74b108853f7b9c51ed12

27d6ec5827a51fd74e3ab97a5cef38fc7f5d9168

But we can remove this, not sure if this `matched` is returned to some specific version.

yogeshojha

yogeshojha/rengine

Fix #529

Nuclei returns the response to stdout: `{"template-id":"tech-detect","info":{"name":"Wappalyzer Technology Detection","author":["hakluke"],"tags":["tech"],"reference":null,"severity":"info"},"matcher-name":"nginx","type":"http","host":"https://example.com:443","matched-at":"https://example.com:443","timestamp":"2021-10-31T09:39:47.1571248Z","curl-command":"curl -X 'GET' -d '' -H 'Accept: */*' -H 'Accept-Language: en' -H 'User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_9_2) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/36.0.1944.0 Safari/537.36' 'https://example.com'"}` It needs to read host_url from matched-at, not from matched.

2021-10-31 10:27:33+00:00

2021-11-01 16:58:16+00:00

web/reNgine/tasks.py

import os import traceback import yaml import json import csv import validators import random import requests import logging import metafinder.extractor as metadata_extractor import whatportis import subprocess from selenium.webdriver.firefox.options import Options as FirefoxOptions from selenium import webdriver from emailfinder.extractor import * from dotted_dict import DottedDict from celery import shared_task from discord_webhook import DiscordWebhook from reNgine.celery import app from startScan.models import * from targetApp.models import Domain from scanEngine.models import EngineType from django.conf import settings from django.shortcuts import get_object_or_404 from celery import shared_task from datetime import datetime from degoogle import degoogle from django.conf import settings from django.utils import timezone, dateformat from django.shortcuts import get_object_or_404 from django.core.exceptions import ObjectDoesNotExist from reNgine.celery import app from reNgine.definitions import * from startScan.models import * from targetApp.models import Domain from scanEngine.models import EngineType, Configuration, Wordlist from .common_func import * ''' task for background scan ''' @app.task def initiate_scan( domain_id, scan_history_id, scan_type, engine_type, imported_subdomains=None, out_of_scope_subdomains=[] ): ''' scan_type = 0 -> immediate scan, need not create scan object scan_type = 1 -> scheduled scan ''' engine_object = EngineType.objects.get(pk=engine_type) domain = Domain.objects.get(pk=domain_id) if scan_type == 1: task = ScanHistory() task.scan_status = -1 elif scan_type == 0: task = ScanHistory.objects.get(pk=scan_history_id) # save the last scan date for domain model domain.last_scan_date = timezone.now() domain.save() # once the celery task starts, change the task status to Started task.scan_type = engine_object task.celery_id = initiate_scan.request.id task.domain = domain task.scan_status = 1 task.start_scan_date = timezone.now() task.subdomain_discovery = True if engine_object.subdomain_discovery else False task.dir_file_search = True if engine_object.dir_file_search else False task.port_scan = True if engine_object.port_scan else False task.fetch_url = True if engine_object.fetch_url else False task.osint = True if engine_object.osint else False task.screenshot = True if engine_object.screenshot else False task.vulnerability_scan = True if engine_object.vulnerability_scan else False task.save() activity_id = create_scan_activity(task, "Scanning Started", 2) results_dir = '/usr/src/scan_results/' os.chdir(results_dir) notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine has initiated recon for target {} with engine type {}'.format(domain.name, engine_object.engine_name)) try: current_scan_dir = domain.name + '_' + str(random.randint(100000000000, 999999999999)) os.mkdir(current_scan_dir) task.results_dir = current_scan_dir task.save() except Exception as exception: logger.error(exception) scan_failed(task) yaml_configuration = None excluded_subdomains = '' try: yaml_configuration = yaml.load( task.scan_type.yaml_configuration, Loader=yaml.FullLoader) except Exception as exception: logger.error(exception) # TODO: Put failed reason on db ''' Add GF patterns name to db for dynamic URLs menu ''' if engine_object.fetch_url and GF_PATTERNS in yaml_configuration[FETCH_URL]: task.used_gf_patterns = ','.join( pattern for pattern in yaml_configuration[FETCH_URL][GF_PATTERNS]) task.save() results_dir = results_dir + current_scan_dir # put all imported subdomains into txt file and also in Subdomain model if imported_subdomains: extract_imported_subdomain( imported_subdomains, task, domain, results_dir) if yaml_configuration: ''' a target in itself is a subdomain, some tool give subdomains as www.yogeshojha.com but url and everything else resolves to yogeshojha.com In that case, we would already need to store target itself as subdomain ''' initial_subdomain_file = '/target_domain.txt' if task.subdomain_discovery else '/sorted_subdomain_collection.txt' subdomain_file = open(results_dir + initial_subdomain_file, "w") subdomain_file.write(domain.name + "\n") subdomain_file.close() if(task.subdomain_discovery): activity_id = create_scan_activity(task, "Subdomain Scanning", 1) subdomain_scan( task, domain, yaml_configuration, results_dir, activity_id, out_of_scope_subdomains ) else: skip_subdomain_scan(task, domain, results_dir) update_last_activity(activity_id, 2) activity_id = create_scan_activity(task, "HTTP Crawler", 1) http_crawler( task, domain, results_dir, activity_id) update_last_activity(activity_id, 2) try: if task.screenshot: activity_id = create_scan_activity( task, "Visual Recon - Screenshot", 1) grab_screenshot( task, domain, yaml_configuration, current_scan_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if(task.port_scan): activity_id = create_scan_activity(task, "Port Scanning", 1) port_scanning(task, domain, yaml_configuration, results_dir) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.osint: activity_id = create_scan_activity(task, "OSINT Running", 1) perform_osint(task, domain, yaml_configuration, results_dir) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.dir_file_search: activity_id = create_scan_activity(task, "Directory Search", 1) directory_brute( task, domain, yaml_configuration, results_dir, activity_id ) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.fetch_url: activity_id = create_scan_activity(task, "Fetching endpoints", 1) fetch_endpoints( task, domain, yaml_configuration, results_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.vulnerability_scan: activity_id = create_scan_activity(task, "Vulnerability Scan", 1) vulnerability_scan( task, domain, yaml_configuration, results_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) activity_id = create_scan_activity(task, "Scan Completed", 2) if notification and notification[0].send_scan_status_notif: send_notification('*Scan Completed*\nreNgine has finished performing recon on target {}.'.format(domain.name)) ''' Once the scan is completed, save the status to successful ''' if ScanActivity.objects.filter(scan_of=task).filter(status=0).all(): task.scan_status = 0 else: task.scan_status = 2 task.stop_scan_date = timezone.now() task.save() # cleanup results delete_scan_data(results_dir) return {"status": True} def skip_subdomain_scan(task, domain, results_dir): # store default target as subdomain ''' If the imported subdomain already has target domain saved, we can skip this ''' if not Subdomain.objects.filter( scan_history=task, name=domain.name).exists(): subdomain_dict = DottedDict({ 'name': domain.name, 'scan_history': task, 'target_domain': domain }) save_subdomain(subdomain_dict) # Save target into target_domain.txt with open('{}/target_domain.txt'.format(results_dir), 'w+') as file: file.write(domain.name + '\n') file.close() ''' We can have two conditions, either subdomain scan happens, or subdomain scan does not happen, in either cases, because we are using import subdomain, we need to collect and sort all the subdomains Write target domain into subdomain_collection ''' os.system( 'cat {0}/target_domain.txt > {0}/subdomain_collection.txt'.format(results_dir)) os.system( 'cat {0}/from_imported.txt > {0}/subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/from_imported.txt'.format(results_dir)) ''' Sort all Subdomains ''' os.system( 'sort -u {0}/subdomain_collection.txt -o {0}/sorted_subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/subdomain_collection.txt'.format(results_dir)) def extract_imported_subdomain(imported_subdomains, task, domain, results_dir): valid_imported_subdomains = [subdomain for subdomain in imported_subdomains if validators.domain( subdomain) and domain.name == get_domain_from_subdomain(subdomain)] # remove any duplicate valid_imported_subdomains = list(set(valid_imported_subdomains)) with open('{}/from_imported.txt'.format(results_dir), 'w+') as file: for subdomain_name in valid_imported_subdomains: # save _subdomain to Subdomain model db if not Subdomain.objects.filter( scan_history=task, name=subdomain_name).exists(): subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': subdomain_name, 'is_imported_subdomain': True }) save_subdomain(subdomain_dict) # save subdomain to file file.write('{}\n'.format(subdomain_name)) file.close() def subdomain_scan(task, domain, yaml_configuration, results_dir, activity_id, out_of_scope_subdomains=None): ''' This function is responsible for performing subdomain enumeration ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Subdomain Gathering for target {} has been started'.format(domain.name)) subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' # check for all the tools and add them into string # if tool selected is all then make string, no need for loop if ALL in yaml_configuration[SUBDOMAIN_DISCOVERY][USES_TOOLS]: tools = 'amass-active amass-passive assetfinder sublist3r subfinder oneforall' else: tools = ' '.join( str(tool) for tool in yaml_configuration[SUBDOMAIN_DISCOVERY][USES_TOOLS]) logging.info(tools) # check for THREADS, by default 10 threads = 10 if THREADS in yaml_configuration[SUBDOMAIN_DISCOVERY]: _threads = yaml_configuration[SUBDOMAIN_DISCOVERY][THREADS] if _threads > 0: threads = _threads if 'amass' in tools: if 'amass-passive' in tools: amass_command = 'amass enum -passive -d {} -o {}/from_amass.txt'.format( domain.name, results_dir) if USE_AMASS_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_AMASS_CONFIG]: amass_command += ' -config /root/.config/amass.ini' # Run Amass Passive logging.info(amass_command) os.system(amass_command) if 'amass-active' in tools: amass_command = 'amass enum -active -d {} -o {}/from_amass_active.txt'.format( domain.name, results_dir) if USE_AMASS_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_AMASS_CONFIG]: amass_command += ' -config /root/.config/amass.ini' if AMASS_WORDLIST in yaml_configuration[SUBDOMAIN_DISCOVERY]: wordlist = yaml_configuration[SUBDOMAIN_DISCOVERY][AMASS_WORDLIST] if wordlist == 'default': wordlist_path = '/usr/src/wordlist/deepmagic.com-prefixes-top50000.txt' else: wordlist_path = '/usr/src/wordlist/' + wordlist + '.txt' if not os.path.exists(wordlist_path): wordlist_path = '/usr/src/' + AMASS_WORDLIST amass_command = amass_command + \ ' -brute -w {}'.format(wordlist_path) if amass_config_path: amass_command = amass_command + \ ' -config {}'.format('/usr/src/scan_results/' + amass_config_path) # Run Amass Active logging.info(amass_command) os.system(amass_command) if 'assetfinder' in tools: assetfinder_command = 'assetfinder --subs-only {} > {}/from_assetfinder.txt'.format( domain.name, results_dir) # Run Assetfinder logging.info(assetfinder_command) os.system(assetfinder_command) if 'sublist3r' in tools: sublist3r_command = 'python3 /usr/src/github/Sublist3r/sublist3r.py -d {} -t {} -o {}/from_sublister.txt'.format( domain.name, threads, results_dir) # Run sublist3r logging.info(sublist3r_command) os.system(sublist3r_command) if 'subfinder' in tools: subfinder_command = 'subfinder -d {} -t {} -o {}/from_subfinder.txt'.format( domain.name, threads, results_dir) if USE_SUBFINDER_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_SUBFINDER_CONFIG]: subfinder_command += ' -config /root/.config/subfinder/config.yaml' # Run Subfinder logging.info(subfinder_command) os.system(subfinder_command) if 'oneforall' in tools: oneforall_command = 'python3 /usr/src/github/OneForAll/oneforall.py --target {} run'.format( domain.name, results_dir) # Run OneForAll logging.info(oneforall_command) os.system(oneforall_command) extract_subdomain = "cut -d',' -f6 /usr/src/github/OneForAll/results/{}.csv >> {}/from_oneforall.txt".format( domain.name, results_dir) os.system(extract_subdomain) # remove the results from oneforall directory os.system( 'rm -rf /usr/src/github/OneForAll/results/{}.*'.format(domain.name)) ''' All tools have gathered the list of subdomains with filename initials as from_* We will gather all the results in one single file, sort them and remove the older results from_* ''' os.system( 'cat {0}/*.txt > {0}/subdomain_collection.txt'.format(results_dir)) ''' Write target domain into subdomain_collection ''' os.system( 'cat {0}/target_domain.txt >> {0}/subdomain_collection.txt'.format(results_dir)) ''' Remove all the from_* files ''' os.system('rm -f {}/from*'.format(results_dir)) ''' Sort all Subdomains ''' os.system( 'sort -u {0}/subdomain_collection.txt -o {0}/sorted_subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/subdomain_collection.txt'.format(results_dir)) ''' The final results will be stored in sorted_subdomain_collection. ''' # parse the subdomain list file and store in db with open(subdomain_scan_results_file) as subdomain_list: for _subdomain in subdomain_list: __subdomain = _subdomain.rstrip('\n') if not Subdomain.objects.filter(scan_history=task, name=__subdomain).exists( ) and validators.domain(__subdomain) and __subdomain not in out_of_scope_subdomains: subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': __subdomain, }) save_subdomain(subdomain_dict) notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: subdomains_count = Subdomain.objects.filter(scan_history=task).count() send_notification('Subdomain Gathering for target {} has been completed and has discovered *{}* subdomains.'.format(domain.name, subdomains_count)) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/sorted_subdomain_collection.txt') # check for any subdomain changes and send notif if any if notification and notification[0].send_subdomain_changes_notif: newly_added_subdomain = get_new_added_subdomain(task.id, domain.id) if newly_added_subdomain: message = "**{} New Subdomains Discovered on domain {}**".format(newly_added_subdomain.count(), domain.name) for subdomain in newly_added_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) removed_subdomain = get_removed_subdomain(task.id, domain.id) if removed_subdomain: message = "**{} Subdomains are no longer available on domain {}**".format(removed_subdomain.count(), domain.name) for subdomain in removed_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) # check for interesting subdomains and send notif if any if notification and notification[0].send_interesting_notif: interesting_subdomain = get_interesting_subdomains(task.id, domain.id) print(interesting_subdomain) if interesting_subdomain: message = "**{} Interesting Subdomains Found on domain {}**".format(interesting_subdomain.count(), domain.name) for subdomain in interesting_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) def get_new_added_subdomain(scan_id, domain_id): scan_history = ScanHistory.objects.filter( domain=domain_id).filter( subdomain_discovery=True).filter( id__lte=scan_id) if scan_history.count() > 1: last_scan = scan_history.order_by('-start_scan_date')[1] scanned_host_q1 = Subdomain.objects.filter( scan_history__id=scan_id).values('name') scanned_host_q2 = Subdomain.objects.filter( scan_history__id=last_scan.id).values('name') added_subdomain = scanned_host_q1.difference(scanned_host_q2) return Subdomain.objects.filter( scan_history=scan_id).filter( name__in=added_subdomain) def get_removed_subdomain(scan_id, domain_id): scan_history = ScanHistory.objects.filter( domain=domain_id).filter( subdomain_discovery=True).filter( id__lte=scan_id) if scan_history.count() > 1: last_scan = scan_history.order_by('-start_scan_date')[1] scanned_host_q1 = Subdomain.objects.filter( scan_history__id=scan_id).values('name') scanned_host_q2 = Subdomain.objects.filter( scan_history__id=last_scan.id).values('name') removed_subdomains = scanned_host_q2.difference(scanned_host_q1) print() return Subdomain.objects.filter( scan_history=last_scan).filter( name__in=removed_subdomains) def http_crawler(task, domain, results_dir, activity_id): ''' This function is runs right after subdomain gathering, and gathers important like page title, http status, etc HTTP Crawler runs by default ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('HTTP Crawler for target {} has been initiated.'.format(domain.name)) alive_file_location = results_dir + '/alive.txt' httpx_results_file = results_dir + '/httpx.json' subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' httpx_command = 'httpx -status-code -content-length -title -tech-detect -cdn -ip -follow-host-redirects -random-agent' proxy = get_random_proxy() if proxy: httpx_command += " --http-proxy '{}'".format(proxy) httpx_command += ' -json -o {}'.format( httpx_results_file ) httpx_command = 'cat {} | {}'.format(subdomain_scan_results_file, httpx_command) print(httpx_command) os.system(httpx_command) # alive subdomains from httpx alive_file = open(alive_file_location, 'w') # writing httpx results if os.path.isfile(httpx_results_file): httpx_json_result = open(httpx_results_file, 'r') lines = httpx_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) try: # fallback for older versions of httpx if 'url' in json_st: subdomain = Subdomain.objects.get( scan_history=task, name=json_st['input']) else: subdomain = Subdomain.objects.get( scan_history=task, name=json_st['url'].split("//")[-1]) ''' Saving Default http urls to EndPoint ''' endpoint = EndPoint() endpoint.scan_history = task endpoint.target_domain = domain endpoint.subdomain = subdomain if 'url' in json_st: endpoint.http_url = json_st['url'] subdomain.http_url = json_st['url'] if 'status-code' in json_st: endpoint.http_status = json_st['status-code'] subdomain.http_status = json_st['status-code'] if 'title' in json_st: endpoint.page_title = json_st['title'] subdomain.page_title = json_st['title'] if 'content-length' in json_st: endpoint.content_length = json_st['content-length'] subdomain.content_length = json_st['content-length'] if 'content-type' in json_st: endpoint.content_type = json_st['content-type'] subdomain.content_type = json_st['content-type'] if 'webserver' in json_st: endpoint.webserver = json_st['webserver'] subdomain.webserver = json_st['webserver'] if 'response-time' in json_st: response_time = float( ''.join( ch for ch in json_st['response-time'] if not ch.isalpha())) if json_st['response-time'][-2:] == 'ms': response_time = response_time / 1000 endpoint.response_time = response_time subdomain.response_time = response_time if 'cnames' in json_st: cname_list = ','.join(json_st['cnames']) subdomain.cname = cname_list discovered_date = timezone.now() endpoint.discovered_date = discovered_date subdomain.discovered_date = discovered_date endpoint.is_default = True endpoint.save() subdomain.save() if 'technologies' in json_st: for _tech in json_st['technologies']: if Technology.objects.filter(name=_tech).exists(): tech = Technology.objects.get(name=_tech) else: tech = Technology(name=_tech) tech.save() subdomain.technologies.add(tech) endpoint.technologies.add(tech) if 'a' in json_st: for _ip in json_st['a']: if IpAddress.objects.filter(address=_ip).exists(): ip = IpAddress.objects.get(address=_ip) else: ip = IpAddress(address=_ip) if 'cdn' in json_st: ip.is_cdn = json_st['cdn'] ip.save() subdomain.ip_addresses.add(ip) # see if to ignore 404 or 5xx alive_file.write(json_st['url'] + '\n') subdomain.save() endpoint.save() except Exception as exception: logging.error(exception) alive_file.close() if notification and notification[0].send_scan_status_notif: alive_count = Subdomain.objects.filter( scan_history__id=task.id).values('name').distinct().filter( http_status__exact=200).count() send_notification('HTTP Crawler for target {} has been completed.\n\n {} subdomains were alive (http status 200).'.format(domain.name, alive_count)) def grab_screenshot(task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for taking screenshots ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine is currently gathering screenshots for {}'.format(domain.name)) output_screenshots_path = results_dir + '/screenshots' result_csv_path = results_dir + '/screenshots/Requests.csv' alive_subdomains_path = results_dir + '/alive.txt' eyewitness_command = 'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py' eyewitness_command += ' -f {} -d {} --no-prompt'.format( alive_subdomains_path, output_screenshots_path ) if EYEWITNESS in yaml_configuration \ and TIMEOUT in yaml_configuration[EYEWITNESS] \ and yaml_configuration[EYEWITNESS][TIMEOUT] > 0: eyewitness_command += ' --timeout {}'.format( yaml_configuration[EYEWITNESS][TIMEOUT] ) if EYEWITNESS in yaml_configuration \ and THREADS in yaml_configuration[EYEWITNESS] \ and yaml_configuration[EYEWITNESS][THREADS] > 0: eyewitness_command += ' --threads {}'.format( yaml_configuration[EYEWITNESS][THREADS] ) logger.info(eyewitness_command) os.system(eyewitness_command) if os.path.isfile(result_csv_path): logger.info('Gathering Eyewitness results') with open(result_csv_path, 'r') as file: reader = csv.reader(file) for row in reader: if row[3] == 'Successful' \ and Subdomain.objects.filter( scan_history__id=task.id).filter(name=row[2]).exists(): subdomain = Subdomain.objects.get( scan_history__id=task.id, name=row[2] ) subdomain.screenshot_path = row[4].replace( '/usr/src/scan_results/', '' ) subdomain.save() # remove all db, html extra files in screenshot results os.system('rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format( output_screenshots_path, )) os.system('rm -rf {0}/source'.format( output_screenshots_path, )) if notification and notification[0].send_scan_status_notif: send_notification('reNgine has finished gathering screenshots for {}'.format(domain.name)) def port_scanning(task, domain, yaml_configuration, results_dir): ''' This function is responsible for running the port scan ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Port Scan initiated for {}'.format(domain.name)) subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' port_results_file = results_dir + '/ports.json' # check the yaml_configuration and choose the ports to be scanned scan_ports = '-' # default port scan everything if PORTS in yaml_configuration[PORT_SCAN]: # TODO: legacy code, remove top-100 in future versions all_ports = yaml_configuration[PORT_SCAN][PORTS] if 'full' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -p {}'.format( subdomain_scan_results_file, port_results_file, '-') elif 'top-100' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -top-ports 100'.format( subdomain_scan_results_file, port_results_file) elif 'top-1000' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -top-ports 1000'.format( subdomain_scan_results_file, port_results_file) else: scan_ports = ','.join( str(port) for port in all_ports) naabu_command = 'cat {} | naabu -json -o {} -p {}'.format( subdomain_scan_results_file, port_results_file, scan_ports) # check for exclude ports if EXCLUDE_PORTS in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][EXCLUDE_PORTS]: exclude_ports = ','.join( str(port) for port in yaml_configuration['port_scan']['exclude_ports']) naabu_command = naabu_command + \ ' -exclude-ports {}'.format(exclude_ports) if NAABU_RATE in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][NAABU_RATE] > 0: naabu_command = naabu_command + \ ' -rate {}'.format( yaml_configuration[PORT_SCAN][NAABU_RATE]) if USE_NAABU_CONFIG in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][USE_NAABU_CONFIG]: naabu_command += ' -config /root/.config/naabu/naabu.conf' # run naabu os.system(naabu_command) # writing port results try: port_json_result = open(port_results_file, 'r') lines = port_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) port_number = json_st['port'] ip_address = json_st['ip'] # see if port already exists if Port.objects.filter(number__exact=port_number).exists(): port = Port.objects.get(number=port_number) else: port = Port() port.number = port_number if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port_detail = whatportis.get_ports(str(port_number)) if len(port_detail): port.service_name = port_detail[0].name port.description = port_detail[0].description port.save() if IpAddress.objects.filter(address=json_st['ip']).exists(): ip = IpAddress.objects.get(address=json_st['ip']) ip.ports.add(port) ip.save() except BaseException as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: port_count = Port.objects.filter( ports__in=IpAddress.objects.filter( ip_addresses__in=Subdomain.objects.filter( scan_history__id=task.id))).distinct().count() send_notification('reNgine has finished Port Scanning on {} and has identified {} ports.'.format(domain.name, port_count)) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/ports.json') def check_waf(): ''' This function will check for the WAF being used in subdomains using wafw00f ''' pass def directory_brute(task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for performing directory scan ''' # scan directories for all the alive subdomain with http status > # 200 notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Directory Bruteforce has been initiated for {}.'.format(domain.name)) alive_subdomains = Subdomain.objects.filter( scan_history__id=task.id).exclude(http_url__isnull=True) dirs_results = results_dir + '/dirs.json' # check the yaml settings if EXTENSIONS in yaml_configuration[DIR_FILE_SEARCH]: extensions = ','.join( str(ext) for ext in yaml_configuration[DIR_FILE_SEARCH][EXTENSIONS]) else: extensions = 'php,git,yaml,conf,db,mysql,bak,txt' # Threads if THREADS in yaml_configuration[DIR_FILE_SEARCH] \ and yaml_configuration[DIR_FILE_SEARCH][THREADS] > 0: threads = yaml_configuration[DIR_FILE_SEARCH][THREADS] else: threads = 10 for subdomain in alive_subdomains: # delete any existing dirs.json if os.path.isfile(dirs_results): os.system('rm -rf {}'.format(dirs_results)) dirsearch_command = 'python3 /usr/src/github/dirsearch/dirsearch.py' dirsearch_command += ' -u {}'.format(subdomain.http_url) if (WORDLIST not in yaml_configuration[DIR_FILE_SEARCH] or not yaml_configuration[DIR_FILE_SEARCH][WORDLIST] or 'default' in yaml_configuration[DIR_FILE_SEARCH][WORDLIST]): wordlist_location = '/usr/src/github/dirsearch/db/dicc.txt' else: wordlist_location = '/usr/src/wordlist/' + \ yaml_configuration[DIR_FILE_SEARCH][WORDLIST] + '.txt' dirsearch_command += ' -w {}'.format(wordlist_location) dirsearch_command += ' --format json -o {}'.format(dirs_results) dirsearch_command += ' -e {}'.format(extensions) dirsearch_command += ' -t {}'.format(threads) dirsearch_command += ' --random-agent --follow-redirects --exclude-status 403,401,404' if EXCLUDE_EXTENSIONS in yaml_configuration[DIR_FILE_SEARCH]: exclude_extensions = ','.join( str(ext) for ext in yaml_configuration[DIR_FILE_SEARCH][EXCLUDE_EXTENSIONS]) dirsearch_command += ' -X {}'.format(exclude_extensions) if EXCLUDE_TEXT in yaml_configuration[DIR_FILE_SEARCH]: exclude_text = ','.join( str(text) for text in yaml_configuration[DIR_FILE_SEARCH][EXCLUDE_TEXT]) dirsearch_command += ' -exclude-texts {}'.format(exclude_text) # check if recursive strategy is set to on if RECURSIVE_LEVEL in yaml_configuration[DIR_FILE_SEARCH]: dirsearch_command += ' --recursion-depth {}'.format(yaml_configuration[DIR_FILE_SEARCH][RECURSIVE_LEVEL]) if RECURSIVE_LEVEL in yaml_configuration[DIR_FILE_SEARCH]: dirsearch_command += ' --recursion-depth {}'.format(yaml_configuration[DIR_FILE_SEARCH][RECURSIVE_LEVEL]) # proxy proxy = get_random_proxy() if proxy: dirsearch_command += " --proxy '{}'".format(proxy) print(dirsearch_command) os.system(dirsearch_command) try: if os.path.isfile(dirs_results): with open(dirs_results, "r") as json_file: json_string = json_file.read() subdomain = Subdomain.objects.get( scan_history__id=task.id, http_url=subdomain.http_url) subdomain.directory_json = json_string subdomain.save() except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: send_notification('Directory Bruteforce has been completed for {}.'.format(domain.name)) def fetch_endpoints( task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for fetching all the urls associated with target and run HTTP probe It first runs gau to gather all urls from wayback, then we will use hakrawler to identify more urls ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine is currently gathering endpoints for {}.'.format(domain.name)) # check yaml settings if ALL in yaml_configuration[FETCH_URL][USES_TOOLS]: tools = 'gauplus hakrawler waybackurls gospider' else: tools = ' '.join( str(tool) for tool in yaml_configuration[FETCH_URL][USES_TOOLS]) if INTENSITY in yaml_configuration[FETCH_URL]: scan_type = yaml_configuration[FETCH_URL][INTENSITY] else: scan_type = 'normal' domain_regex = "\'https?://([a-z0-9]+[.])*{}.*\'".format(domain.name) if 'deep' in scan_type: # performs deep url gathering for all the subdomains present - # RECOMMENDED logger.info('Deep URLS Fetch') os.system(settings.TOOL_LOCATION + 'get_urls.sh %s %s %s %s %s' % ("None", results_dir, scan_type, domain_regex, tools)) else: # perform url gathering only for main domain - USE only for quick scan logger.info('Non Deep URLS Fetch') os.system( settings.TOOL_LOCATION + 'get_urls.sh %s %s %s %s %s' % ( domain.name, results_dir, scan_type, domain_regex, tools )) if IGNORE_FILE_EXTENSION in yaml_configuration[FETCH_URL]: ignore_extension = '|'.join( yaml_configuration[FETCH_URL][IGNORE_FILE_EXTENSION]) logger.info('Ignore extensions' + ignore_extension) os.system( 'cat {0}/all_urls.txt | grep -Eiv "\\.({1}).*" > {0}/temp_urls.txt'.format( results_dir, ignore_extension)) os.system( 'rm {0}/all_urls.txt && mv {0}/temp_urls.txt {0}/all_urls.txt'.format(results_dir)) ''' Store all the endpoints and then run the httpx ''' try: endpoint_final_url = results_dir + '/all_urls.txt' if os.path.isfile(endpoint_final_url): with open(endpoint_final_url) as endpoint_list: for url in endpoint_list: http_url = url.rstrip('\n') if not EndPoint.objects.filter(scan_history=task, http_url=http_url).exists(): _subdomain = get_subdomain_from_url(http_url) if Subdomain.objects.filter( scan_history=task).filter( name=_subdomain).exists(): subdomain = Subdomain.objects.get( scan_history=task, name=_subdomain) else: ''' gau or gosppider can gather interesting endpoints which when parsed can give subdomains that were not existent from subdomain scan. so storing them ''' logger.error( 'Subdomain {} not found, adding...'.format(_subdomain)) subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': _subdomain, }) subdomain = save_subdomain(subdomain_dict) endpoint_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'subdomain': subdomain, 'http_url': http_url, }) save_endpoint(endpoint_dict) except Exception as e: logger.error(e) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/all_urls.txt') ''' TODO: Go spider & waybackurls accumulates a lot of urls, which is good but nuclei takes forever to scan even a simple website, so we will do http probing and filter HTTP status 404, this way we can reduce the number of Non Existent URLS ''' logger.info('HTTP Probing on collected endpoints') httpx_command = 'httpx -l {0}/all_urls.txt -status-code -content-length -ip -cdn -title -tech-detect -json -follow-redirects -random-agent -o {0}/final_httpx_urls.json'.format(results_dir) proxy = get_random_proxy() if proxy: httpx_command += " --http-proxy '{}'".format(proxy) os.system(httpx_command) url_results_file = results_dir + '/final_httpx_urls.json' try: urls_json_result = open(url_results_file, 'r') lines = urls_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) http_url = json_st['url'] _subdomain = get_subdomain_from_url(http_url) if Subdomain.objects.filter( scan_history=task).filter( name=_subdomain).exists(): subdomain_obj = Subdomain.objects.get( scan_history=task, name=_subdomain) else: subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': _subdomain, }) subdomain_obj = save_subdomain(subdomain_dict) if EndPoint.objects.filter( scan_history=task).filter( http_url=http_url).exists(): endpoint = EndPoint.objects.get( scan_history=task, http_url=http_url) else: endpoint = EndPoint() endpoint_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'http_url': http_url, 'subdomain': subdomain_obj }) endpoint = save_endpoint(endpoint_dict) if 'title' in json_st: endpoint.page_title = json_st['title'] if 'webserver' in json_st: endpoint.webserver = json_st['webserver'] if 'content-length' in json_st: endpoint.content_length = json_st['content-length'] if 'content-type' in json_st: endpoint.content_type = json_st['content-type'] if 'status-code' in json_st: endpoint.http_status = json_st['status-code'] if 'response-time' in json_st: response_time = float(''.join(ch for ch in json_st['response-time'] if not ch.isalpha())) if json_st['response-time'][-2:] == 'ms': response_time = response_time / 1000 endpoint.response_time = response_time endpoint.save() if 'technologies' in json_st: for _tech in json_st['technologies']: if Technology.objects.filter(name=_tech).exists(): tech = Technology.objects.get(name=_tech) else: tech = Technology(name=_tech) tech.save() endpoint.technologies.add(tech) # get subdomain object subdomain = Subdomain.objects.get(scan_history=task, name=_subdomain) subdomain.technologies.add(tech) subdomain.save() except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: endpoint_count = EndPoint.objects.filter( scan_history__id=task.id).values('http_url').distinct().count() endpoint_alive_count = EndPoint.objects.filter( scan_history__id=task.id, http_status__exact=200).values('http_url').distinct().count() send_notification('reNgine has finished gathering endpoints for {} and has discovered *{}* unique endpoints.\n\n{} of those endpoints reported HTTP status 200.'.format( domain.name, endpoint_count, endpoint_alive_count )) # once endpoint is saved, run gf patterns TODO: run threads if GF_PATTERNS in yaml_configuration[FETCH_URL]: for pattern in yaml_configuration[FETCH_URL][GF_PATTERNS]: logger.info('Running GF for {}'.format(pattern)) gf_output_file_path = '{0}/gf_patterns_{1}.txt'.format( results_dir, pattern) gf_command = 'cat {0}/all_urls.txt | gf {1} >> {2}'.format( results_dir, pattern, gf_output_file_path) os.system(gf_command) if os.path.exists(gf_output_file_path): with open(gf_output_file_path) as gf_output: for line in gf_output: url = line.rstrip('\n') try: endpoint = EndPoint.objects.get( scan_history=task, http_url=url) earlier_pattern = endpoint.matched_gf_patterns new_pattern = earlier_pattern + ',' + pattern if earlier_pattern else pattern endpoint.matched_gf_patterns = new_pattern except Exception as e: # add the url in db logger.error(e) logger.info('Adding URL' + url) endpoint = EndPoint() endpoint.http_url = url endpoint.target_domain = domain endpoint.scan_history = task try: _subdomain = Subdomain.objects.get( scan_history=task, name=get_subdomain_from_url(url)) endpoint.subdomain = _subdomain except Exception as e: continue endpoint.matched_gf_patterns = pattern finally: endpoint.save() def vulnerability_scan( task, domain, yaml_configuration, results_dir, activity_id): notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Vulnerability scan has been initiated for {}.'.format(domain.name)) ''' This function will run nuclei as a vulnerability scanner ---- unfurl the urls to keep only domain and path, this will be sent to vuln scan ignore certain file extensions Thanks: https://github.com/six2dez/reconftw ''' urls_path = '/alive.txt' if task.scan_type.fetch_url: os.system('cat {0}/all_urls.txt | grep -Eiv "\\.(eot|jpg|jpeg|gif|css|tif|tiff|png|ttf|otf|woff|woff2|ico|pdf|svg|txt|js|doc|docx)$" | unfurl -u format %s://%d%p >> {0}/unfurl_urls.txt'.format(results_dir)) os.system( 'sort -u {0}/unfurl_urls.txt -o {0}/unfurl_urls.txt'.format(results_dir)) urls_path = '/unfurl_urls.txt' vulnerability_result_path = results_dir + '/vulnerability.json' vulnerability_scan_input_file = results_dir + urls_path nuclei_command = 'nuclei -json -l {} -o {}'.format( vulnerability_scan_input_file, vulnerability_result_path) # check nuclei config if USE_NUCLEI_CONFIG in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[VULNERABILITY_SCAN][USE_NUCLEI_CONFIG]: nuclei_command += ' -config /root/.config/nuclei/config.yaml' ''' Nuclei Templates Either custom template has to be supplied or default template, if neither has been supplied then use all templates including custom templates ''' if CUSTOM_NUCLEI_TEMPLATE in yaml_configuration[ VULNERABILITY_SCAN] or NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: # check yaml settings for templates if NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: if ALL in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_TEMPLATE]: template = NUCLEI_TEMPLATES_PATH else: _template = ','.join([NUCLEI_TEMPLATES_PATH + str(element) for element in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_TEMPLATE]]) template = _template.replace(',', ' -t ') # Update nuclei command with templates nuclei_command = nuclei_command + ' -t ' + template if CUSTOM_NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: # add .yaml to the custom template extensions _template = ','.join( [str(element) + '.yaml' for element in yaml_configuration[VULNERABILITY_SCAN][CUSTOM_NUCLEI_TEMPLATE]]) template = _template.replace(',', ' -t ') # Update nuclei command with templates nuclei_command = nuclei_command + ' -t ' + template else: nuclei_command = nuclei_command + ' -t /root/nuclei-templates' # check yaml settings for concurrency if NUCLEI_CONCURRENCY in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][NUCLEI_CONCURRENCY] > 0: concurrency = yaml_configuration[VULNERABILITY_SCAN][NUCLEI_CONCURRENCY] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -c ' + str(concurrency) if RATE_LIMIT in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][RATE_LIMIT] > 0: rate_limit = yaml_configuration[VULNERABILITY_SCAN][RATE_LIMIT] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -rl ' + str(rate_limit) if TIMEOUT in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][TIMEOUT] > 0: timeout = yaml_configuration[VULNERABILITY_SCAN][TIMEOUT] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -timeout ' + str(timeout) if RETRIES in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][RETRIES] > 0: retries = yaml_configuration[VULNERABILITY_SCAN][RETRIES] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -retries ' + str(retries) # for severity if NUCLEI_SEVERITY in yaml_configuration[VULNERABILITY_SCAN] and ALL not in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_SEVERITY]: _severity = ','.join( [str(element) for element in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_SEVERITY]]) severity = _severity.replace(" ", "") else: severity = "critical, high, medium, low, info" # update nuclei templates before running scan os.system('nuclei -update-templates') for _severity in severity.split(","): # delete any existing vulnerability.json file if os.path.isfile(vulnerability_result_path): os.system('rm {}'.format(vulnerability_result_path)) # run nuclei final_nuclei_command = nuclei_command + ' -severity ' + _severity proxy = get_random_proxy() if proxy: final_nuclei_command += " --proxy-url '{}'".format(proxy) logger.info(final_nuclei_command) os.system(final_nuclei_command) try: if os.path.isfile(vulnerability_result_path): urls_json_result = open(vulnerability_result_path, 'r') lines = urls_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) host = json_st['host'] _subdomain = get_subdomain_from_url(host) try: subdomain = Subdomain.objects.get( name=_subdomain, scan_history=task) vulnerability = Vulnerability() vulnerability.subdomain = subdomain vulnerability.scan_history = task vulnerability.target_domain = domain try: endpoint = EndPoint.objects.get( scan_history=task, target_domain=domain, http_url=host) vulnerability.endpoint = endpoint except Exception as exception: logger.error(exception) if 'name' in json_st['info']: vulnerability.name = json_st['info']['name'] if 'severity' in json_st['info']: if json_st['info']['severity'] == 'info': severity = 0 elif json_st['info']['severity'] == 'low': severity = 1 elif json_st['info']['severity'] == 'medium': severity = 2 elif json_st['info']['severity'] == 'high': severity = 3 elif json_st['info']['severity'] == 'critical': severity = 4 else: severity = 0 else: severity = 0 vulnerability.severity = severity if 'tags' in json_st['info']: vulnerability.tags = json_st['info']['tags'] if 'description' in json_st['info']: vulnerability.description = json_st['info']['description'] if 'reference' in json_st['info']: vulnerability.reference = json_st['info']['reference'] if 'matched' in json_st: vulnerability.http_url = json_st['matched'] if 'templateID' in json_st: vulnerability.template_used = json_st['templateID'] if 'description' in json_st: vulnerability.description = json_st['description'] if 'matcher_name' in json_st: vulnerability.matcher_name = json_st['matcher_name'] if 'extracted_results' in json_st: vulnerability.extracted_results = json_st['extracted_results'] vulnerability.discovered_date = timezone.now() vulnerability.open_status = True vulnerability.save() # send notification for all vulnerabilities except info if json_st['info']['severity'] != "info" and notification and notification[0].send_vuln_notif: message = "*Alert: Vulnerability Identified*" message += "\n\n" message += "A *{}* severity vulnerability has been identified.".format(json_st['info']['severity']) message += "\nVulnerability Name: {}".format(json_st['info']['name']) message += "\nVulnerable URL: {}".format(json_st['host']) send_notification(message) # send report to hackerone if Hackerone.objects.all().exists() and json_st['info']['severity'] != 'info' and json_st['info']['severity'] \ != 'low' and vulnerability.target_domain.h1_team_handle: hackerone = Hackerone.objects.all()[0] if hackerone.send_critical and json_st['info']['severity'] == 'critical': send_hackerone_report(vulnerability.id) elif hackerone.send_high and json_st['info']['severity'] == 'high': send_hackerone_report(vulnerability.id) elif hackerone.send_medium and json_st['info']['severity'] == 'medium': send_hackerone_report(vulnerability.id) except ObjectDoesNotExist: logger.error('Object not found') continue except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: info_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=0).count() low_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=1).count() medium_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=2).count() high_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=3).count() critical_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=4).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count message = 'Vulnerability scan has been completed for {} and discovered {} vulnerabilities.'.format( domain.name, vulnerability_count ) message += '\n\n*Vulnerability Stats:*' message += '\nCritical: {}'.format(critical_count) message += '\nHigh: {}'.format(high_count) message += '\nMedium: {}'.format(medium_count) message += '\nLow: {}'.format(low_count) message += '\nInfo: {}'.format(info_count) send_notification(message) def scan_failed(task): task.scan_status = 0 task.stop_scan_date = timezone.now() task.save() def create_scan_activity(task, message, status): scan_activity = ScanActivity() scan_activity.scan_of = task scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id def update_last_activity(id, activity_status): ScanActivity.objects.filter( id=id).update( status=activity_status, time=timezone.now()) def delete_scan_data(results_dir): # remove all txt,html,json files os.system('find {} -name "*.txt" -type f -delete'.format(results_dir)) os.system('find {} -name "*.html" -type f -delete'.format(results_dir)) os.system('find {} -name "*.json" -type f -delete'.format(results_dir)) def save_subdomain(subdomain_dict): subdomain = Subdomain() subdomain.discovered_date = timezone.now() subdomain.target_domain = subdomain_dict.get('target_domain') subdomain.scan_history = subdomain_dict.get('scan_history') subdomain.name = subdomain_dict.get('name') subdomain.http_url = subdomain_dict.get('http_url') subdomain.screenshot_path = subdomain_dict.get('screenshot_path') subdomain.http_header_path = subdomain_dict.get('http_header_path') subdomain.cname = subdomain_dict.get('cname') subdomain.is_cdn = subdomain_dict.get('is_cdn') subdomain.content_type = subdomain_dict.get('content_type') subdomain.webserver = subdomain_dict.get('webserver') subdomain.page_title = subdomain_dict.get('page_title') subdomain.is_imported_subdomain = subdomain_dict.get( 'is_imported_subdomain') if 'is_imported_subdomain' in subdomain_dict else False if 'http_status' in subdomain_dict: subdomain.http_status = subdomain_dict.get('http_status') if 'response_time' in subdomain_dict: subdomain.response_time = subdomain_dict.get('response_time') if 'content_length' in subdomain_dict: subdomain.content_length = subdomain_dict.get('content_length') subdomain.save() return subdomain def save_endpoint(endpoint_dict): endpoint = EndPoint() endpoint.discovered_date = timezone.now() endpoint.scan_history = endpoint_dict.get('scan_history') endpoint.target_domain = endpoint_dict.get('target_domain') if 'target_domain' in endpoint_dict else None endpoint.subdomain = endpoint_dict.get('subdomain') if 'target_domain' in endpoint_dict else None endpoint.http_url = endpoint_dict.get('http_url') endpoint.page_title = endpoint_dict.get('page_title') if 'page_title' in endpoint_dict else None endpoint.content_type = endpoint_dict.get('content_type') if 'content_type' in endpoint_dict else None endpoint.webserver = endpoint_dict.get('webserver') if 'webserver' in endpoint_dict else None endpoint.response_time = endpoint_dict.get('response_time') if 'response_time' in endpoint_dict else 0 endpoint.http_status = endpoint_dict.get('http_status') if 'http_status' in endpoint_dict else 0 endpoint.content_length = endpoint_dict.get('content_length') if 'content_length' in endpoint_dict else 0 endpoint.is_default = endpoint_dict.get('is_default') if 'is_default' in endpoint_dict else False endpoint.save() return endpoint def perform_osint(task, domain, yaml_configuration, results_dir): notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine has initiated OSINT on target {}'.format(domain.name)) if 'discover' in yaml_configuration[OSINT]: osint_discovery(task, domain, yaml_configuration, results_dir) if 'dork' in yaml_configuration[OSINT]: dorking(task, yaml_configuration) if notification and notification[0].send_scan_status_notif: send_notification('reNgine has completed performing OSINT on target {}'.format(domain.name)) def osint_discovery(task, domain, yaml_configuration, results_dir): if ALL in yaml_configuration[OSINT][OSINT_DISCOVER]: osint_lookup = 'emails metainfo employees' else: osint_lookup = ' '.join( str(lookup) for lookup in yaml_configuration[OSINT][OSINT_DISCOVER]) if 'metainfo' in osint_lookup: if INTENSITY in yaml_configuration[OSINT]: osint_intensity = yaml_configuration[OSINT][INTENSITY] else: osint_intensity = 'normal' if OSINT_DOCUMENTS_LIMIT in yaml_configuration[OSINT]: documents_limit = yaml_configuration[OSINT][OSINT_DOCUMENTS_LIMIT] else: documents_limit = 50 if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': domain.name, 'domain': domain, 'scan_id': task, 'documents_limit': documents_limit }) get_and_save_meta_info(meta_dict) elif osint_intensity == 'deep': # get all subdomains in scan_id subdomains = Subdomain.objects.filter(scan_history=task) for subdomain in subdomains: meta_dict = DottedDict({ 'osint_target': subdomain.name, 'domain': domain, 'scan_id': task, 'documents_limit': documents_limit }) get_and_save_meta_info(meta_dict) if 'emails' in osint_lookup: get_and_save_emails(task, results_dir) get_and_save_leaked_credentials(task, results_dir) if 'employees' in osint_lookup: get_and_save_employees(task, results_dir) def dorking(scan_history, yaml_configuration): # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh # look in stackoverflow if ALL in yaml_configuration[OSINT][OSINT_DORK]: dork_lookup = 'stackoverflow, 3rdparty, social_media, project_management, code_sharing, config_files, jenkins, cloud_buckets, php_error, exposed_documents, struts_rce, db_files, traefik, git_exposed' else: dork_lookup = ' '.join( str(lookup) for lookup in yaml_configuration[OSINT][OSINT_DORK]) if 'stackoverflow' in dork_lookup: dork = 'site:stackoverflow.com' dork_type = 'stackoverflow' get_and_save_dork_results( dork, dork_type, scan_history, in_target=False ) if '3rdparty' in dork_lookup: # look in 3rd party sitee dork_type = '3rdparty' lookup_websites = [ 'gitter.im', 'papaly.com', 'productforums.google.com', 'coggle.it', 'replt.it', 'ycombinator.com', 'libraries.io', 'npm.runkit.com', 'npmjs.com', 'scribd.com', 'gitter.im' ] dork = '' for website in lookup_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'social_media' in dork_lookup: dork_type = 'Social Media' social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'pinterest.com', 'tumblr.com', 'reddit.com' ] dork = '' for website in social_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'project_management' in dork_lookup: dork_type = 'Project Management' project_websites = [ 'trello.com', '*.atlassian.net' ] dork = '' for website in project_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'code_sharing' in dork_lookup: dork_type = 'Code Sharing Sites' code_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] dork = '' for website in code_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'config_files' in dork_lookup: dork_type = 'Config Files' config_file_ext = [ 'env', 'xml', 'conf', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] dork = '' for extension in config_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'jenkins' in dork_lookup: dork_type = 'Jenkins' dork = 'intitle:\"Dashboard [Jenkins]\"' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) if 'wordpress_files' in dork_lookup: dork_type = 'Wordpress Files' inurl_lookup = [ 'wp-content', 'wp-includes' ] dork = '' for lookup in inurl_lookup: dork = dork + ' | ' + 'inurl:' + lookup get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'cloud_buckets' in dork_lookup: dork_type = 'Cloud Buckets' cloud_websites = [ '.s3.amazonaws.com', 'storage.googleapis.com', 'amazonaws.com' ] dork = '' for website in cloud_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'php_error' in dork_lookup: dork_type = 'PHP Error' error_words = [ '\"PHP Parse error\"', '\"PHP Warning\"', '\"PHP Error\"' ] dork = '' for word in error_words: dork = dork + ' | ' + word get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'exposed_documents' in dork_lookup: dork_type = 'Exposed Documents' docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] dork = '' for extension in docs_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'struts_rce' in dork_lookup: dork_type = 'Apache Struts RCE' struts_file_ext = [ 'action', 'struts', 'do' ] dork = '' for extension in struts_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'db_files' in dork_lookup: dork_type = 'Database Files' db_file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] dork = '' for extension in db_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'traefik' in dork_lookup: dork = 'intitle:traefik inurl:8080/dashboard' dork_type = 'Traefik' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) if 'git_exposed' in dork_lookup: dork = 'inurl:\"/.git\"' dork_type = '.git Exposed' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) def get_and_save_dork_results(dork, type, scan_history, in_target=False): degoogle_obj = degoogle.dg() proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy if in_target: query = dork + " site:" + scan_history.domain.name else: query = dork + " \"{}\"".format(scan_history.domain.name) logger.info(query) degoogle_obj.query = query results = degoogle_obj.run() logger.info(results) for result in results: dork, _ = Dork.objects.get_or_create( type=type, description=result['desc'], url=result['url'] ) scan_history.dorks.add(dork) def get_and_save_employees(scan_history, results_dir): theHarvester_location = '/usr/src/github/theHarvester' # update proxies.yaml if Proxy.objects.all().exists(): proxy = Proxy.objects.all()[0] if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(theHarvester_location + '/proxies.yaml', 'w') as file: documents = yaml.dump(yaml_data, file) os.system('cd {} && python3 theHarvester.py -d {} -b all -f {}/theHarvester.html'.format( theHarvester_location, scan_history.domain.name, results_dir )) file_location = results_dir + '/theHarvester.html' print(file_location) # delete proxy environ var if os.environ.get(('https_proxy')): del os.environ['https_proxy'] if os.environ.get(('HTTPS_PROXY')): del os.environ['HTTPS_PROXY'] if os.path.isfile(file_location): logger.info('Parsing theHarvester results') options = FirefoxOptions() options.add_argument("--headless") driver = webdriver.Firefox(options=options) driver.get('file://'+file_location) tabledata = driver.execute_script('return tabledata') # save email addresses and linkedin employees for data in tabledata: if data['record'] == 'email': _email = data['result'] email, _ = Email.objects.get_or_create(address=_email) scan_history.emails.add(email) elif data['record'] == 'people': _employee = data['result'] split_val = _employee.split('-') name = split_val[0] if len(split_val) == 2: designation = split_val[1] else: designation = "" employee, _ = Employee.objects.get_or_create(name=name, designation=designation) scan_history.employees.add(employee) driver.quit() print(tabledata) def get_and_save_emails(scan_history, results_dir): leak_target_path = '{}/creds_target.txt'.format(results_dir) # get email address proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy emails = [] try: logger.info('OSINT: Getting emails from Google') email_from_google = get_emails_from_google(scan_history.domain.name) logger.info('OSINT: Getting emails from Bing') email_from_bing = get_emails_from_bing(scan_history.domain.name) logger.info('OSINT: Getting emails from Baidu') email_from_baidu = get_emails_from_baidu(scan_history.domain.name) emails = list(set(email_from_google + email_from_bing + email_from_baidu)) logger.info(emails) except Exception as e: logger.error(e) leak_target_file = open(leak_target_path, 'w') for _email in emails: email, _ = Email.objects.get_or_create(address=_email) scan_history.emails.add(email) leak_target_file.write('{}\n'.format(_email)) # fill leak_target_file with possible email address leak_target_file.write('%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.write('%.%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%.%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.write('%_%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%_%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.close() def get_and_save_leaked_credentials(scan_history, results_dir): logger.info('OSINT: Getting leaked credentials...') leak_target_file = '{}/creds_target.txt'.format(results_dir) leak_output_file = '{}/pwndb.json'.format(results_dir) pwndb_command = 'python3 /usr/src/github/pwndb/pwndb.py --proxy tor:9150 --output json --list {}'.format( leak_target_file ) try: pwndb_output = subprocess.getoutput(pwndb_command) creds = json.loads(pwndb_output) for cred in creds: if cred['username'] != 'donate': email_id = "{}@{}".format(cred['username'], cred['domain']) email_obj, _ = Email.objects.get_or_create( address=email_id, ) email_obj.password = cred['password'] email_obj.save() scan_history.emails.add(email_obj) except Exception as e: logger.error(e) pass def get_and_save_meta_info(meta_dict): logger.info('Getting METADATA for {}'.format(meta_dict.osint_target)) proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy result = metadata_extractor.extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if result: results = result.get_metadata() for meta in results: meta_finder_document = MetaFinderDocument() subdomain = Subdomain.objects.get(scan_history=meta_dict.scan_id, name=meta_dict.osint_target) meta_finder_document.subdomain = subdomain meta_finder_document.target_domain = meta_dict.domain meta_finder_document.scan_history = meta_dict.scan_id item = DottedDict(results[meta]) meta_finder_document.url = item.url meta_finder_document.doc_name = meta meta_finder_document.http_status = item.status_code metadata = results[meta]['metadata'] for data in metadata: if 'Producer' in metadata and metadata['Producer']: meta_finder_document.producer = metadata['Producer'].rstrip('\x00') if 'Creator' in metadata and metadata['Creator']: meta_finder_document.creator = metadata['Creator'].rstrip('\x00') if 'CreationDate' in metadata and metadata['CreationDate']: meta_finder_document.creation_date = metadata['CreationDate'].rstrip('\x00') if 'ModDate' in metadata and metadata['ModDate']: meta_finder_document.modified_date = metadata['ModDate'].rstrip('\x00') if 'Author' in metadata and metadata['Author']: meta_finder_document.author = metadata['Author'].rstrip('\x00') if 'Title' in metadata and metadata['Title']: meta_finder_document.title = metadata['Title'].rstrip('\x00') if 'OSInfo' in metadata and metadata['OSInfo']: meta_finder_document.os = metadata['OSInfo'].rstrip('\x00') meta_finder_document.save() @app.task(bind=True) def test_task(self): print('*' * 40) print('test task run') print('*' * 40)

import os import traceback import yaml import json import csv import validators import random import requests import logging import metafinder.extractor as metadata_extractor import whatportis import subprocess from selenium.webdriver.firefox.options import Options as FirefoxOptions from selenium import webdriver from emailfinder.extractor import * from dotted_dict import DottedDict from celery import shared_task from discord_webhook import DiscordWebhook from reNgine.celery import app from startScan.models import * from targetApp.models import Domain from scanEngine.models import EngineType from django.conf import settings from django.shortcuts import get_object_or_404 from celery import shared_task from datetime import datetime from degoogle import degoogle from django.conf import settings from django.utils import timezone, dateformat from django.shortcuts import get_object_or_404 from django.core.exceptions import ObjectDoesNotExist from reNgine.celery import app from reNgine.definitions import * from startScan.models import * from targetApp.models import Domain from scanEngine.models import EngineType, Configuration, Wordlist from .common_func import * ''' task for background scan ''' @app.task def initiate_scan( domain_id, scan_history_id, scan_type, engine_type, imported_subdomains=None, out_of_scope_subdomains=[] ): ''' scan_type = 0 -> immediate scan, need not create scan object scan_type = 1 -> scheduled scan ''' engine_object = EngineType.objects.get(pk=engine_type) domain = Domain.objects.get(pk=domain_id) if scan_type == 1: task = ScanHistory() task.scan_status = -1 elif scan_type == 0: task = ScanHistory.objects.get(pk=scan_history_id) # save the last scan date for domain model domain.last_scan_date = timezone.now() domain.save() # once the celery task starts, change the task status to Started task.scan_type = engine_object task.celery_id = initiate_scan.request.id task.domain = domain task.scan_status = 1 task.start_scan_date = timezone.now() task.subdomain_discovery = True if engine_object.subdomain_discovery else False task.dir_file_search = True if engine_object.dir_file_search else False task.port_scan = True if engine_object.port_scan else False task.fetch_url = True if engine_object.fetch_url else False task.osint = True if engine_object.osint else False task.screenshot = True if engine_object.screenshot else False task.vulnerability_scan = True if engine_object.vulnerability_scan else False task.save() activity_id = create_scan_activity(task, "Scanning Started", 2) results_dir = '/usr/src/scan_results/' os.chdir(results_dir) notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine has initiated recon for target {} with engine type {}'.format(domain.name, engine_object.engine_name)) try: current_scan_dir = domain.name + '_' + str(random.randint(100000000000, 999999999999)) os.mkdir(current_scan_dir) task.results_dir = current_scan_dir task.save() except Exception as exception: logger.error(exception) scan_failed(task) yaml_configuration = None excluded_subdomains = '' try: yaml_configuration = yaml.load( task.scan_type.yaml_configuration, Loader=yaml.FullLoader) except Exception as exception: logger.error(exception) # TODO: Put failed reason on db ''' Add GF patterns name to db for dynamic URLs menu ''' if engine_object.fetch_url and GF_PATTERNS in yaml_configuration[FETCH_URL]: task.used_gf_patterns = ','.join( pattern for pattern in yaml_configuration[FETCH_URL][GF_PATTERNS]) task.save() results_dir = results_dir + current_scan_dir # put all imported subdomains into txt file and also in Subdomain model if imported_subdomains: extract_imported_subdomain( imported_subdomains, task, domain, results_dir) if yaml_configuration: ''' a target in itself is a subdomain, some tool give subdomains as www.yogeshojha.com but url and everything else resolves to yogeshojha.com In that case, we would already need to store target itself as subdomain ''' initial_subdomain_file = '/target_domain.txt' if task.subdomain_discovery else '/sorted_subdomain_collection.txt' subdomain_file = open(results_dir + initial_subdomain_file, "w") subdomain_file.write(domain.name + "\n") subdomain_file.close() if(task.subdomain_discovery): activity_id = create_scan_activity(task, "Subdomain Scanning", 1) subdomain_scan( task, domain, yaml_configuration, results_dir, activity_id, out_of_scope_subdomains ) else: skip_subdomain_scan(task, domain, results_dir) update_last_activity(activity_id, 2) activity_id = create_scan_activity(task, "HTTP Crawler", 1) http_crawler( task, domain, results_dir, activity_id) update_last_activity(activity_id, 2) try: if task.screenshot: activity_id = create_scan_activity( task, "Visual Recon - Screenshot", 1) grab_screenshot( task, domain, yaml_configuration, current_scan_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if(task.port_scan): activity_id = create_scan_activity(task, "Port Scanning", 1) port_scanning(task, domain, yaml_configuration, results_dir) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.osint: activity_id = create_scan_activity(task, "OSINT Running", 1) perform_osint(task, domain, yaml_configuration, results_dir) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.dir_file_search: activity_id = create_scan_activity(task, "Directory Search", 1) directory_brute( task, domain, yaml_configuration, results_dir, activity_id ) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.fetch_url: activity_id = create_scan_activity(task, "Fetching endpoints", 1) fetch_endpoints( task, domain, yaml_configuration, results_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.vulnerability_scan: activity_id = create_scan_activity(task, "Vulnerability Scan", 1) vulnerability_scan( task, domain, yaml_configuration, results_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) activity_id = create_scan_activity(task, "Scan Completed", 2) if notification and notification[0].send_scan_status_notif: send_notification('*Scan Completed*\nreNgine has finished performing recon on target {}.'.format(domain.name)) ''' Once the scan is completed, save the status to successful ''' if ScanActivity.objects.filter(scan_of=task).filter(status=0).all(): task.scan_status = 0 else: task.scan_status = 2 task.stop_scan_date = timezone.now() task.save() # cleanup results delete_scan_data(results_dir) return {"status": True} def skip_subdomain_scan(task, domain, results_dir): # store default target as subdomain ''' If the imported subdomain already has target domain saved, we can skip this ''' if not Subdomain.objects.filter( scan_history=task, name=domain.name).exists(): subdomain_dict = DottedDict({ 'name': domain.name, 'scan_history': task, 'target_domain': domain }) save_subdomain(subdomain_dict) # Save target into target_domain.txt with open('{}/target_domain.txt'.format(results_dir), 'w+') as file: file.write(domain.name + '\n') file.close() ''' We can have two conditions, either subdomain scan happens, or subdomain scan does not happen, in either cases, because we are using import subdomain, we need to collect and sort all the subdomains Write target domain into subdomain_collection ''' os.system( 'cat {0}/target_domain.txt > {0}/subdomain_collection.txt'.format(results_dir)) os.system( 'cat {0}/from_imported.txt > {0}/subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/from_imported.txt'.format(results_dir)) ''' Sort all Subdomains ''' os.system( 'sort -u {0}/subdomain_collection.txt -o {0}/sorted_subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/subdomain_collection.txt'.format(results_dir)) def extract_imported_subdomain(imported_subdomains, task, domain, results_dir): valid_imported_subdomains = [subdomain for subdomain in imported_subdomains if validators.domain( subdomain) and domain.name == get_domain_from_subdomain(subdomain)] # remove any duplicate valid_imported_subdomains = list(set(valid_imported_subdomains)) with open('{}/from_imported.txt'.format(results_dir), 'w+') as file: for subdomain_name in valid_imported_subdomains: # save _subdomain to Subdomain model db if not Subdomain.objects.filter( scan_history=task, name=subdomain_name).exists(): subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': subdomain_name, 'is_imported_subdomain': True }) save_subdomain(subdomain_dict) # save subdomain to file file.write('{}\n'.format(subdomain_name)) file.close() def subdomain_scan(task, domain, yaml_configuration, results_dir, activity_id, out_of_scope_subdomains=None): ''' This function is responsible for performing subdomain enumeration ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Subdomain Gathering for target {} has been started'.format(domain.name)) subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' # check for all the tools and add them into string # if tool selected is all then make string, no need for loop if ALL in yaml_configuration[SUBDOMAIN_DISCOVERY][USES_TOOLS]: tools = 'amass-active amass-passive assetfinder sublist3r subfinder oneforall' else: tools = ' '.join( str(tool) for tool in yaml_configuration[SUBDOMAIN_DISCOVERY][USES_TOOLS]) logging.info(tools) # check for THREADS, by default 10 threads = 10 if THREADS in yaml_configuration[SUBDOMAIN_DISCOVERY]: _threads = yaml_configuration[SUBDOMAIN_DISCOVERY][THREADS] if _threads > 0: threads = _threads if 'amass' in tools: if 'amass-passive' in tools: amass_command = 'amass enum -passive -d {} -o {}/from_amass.txt'.format( domain.name, results_dir) if USE_AMASS_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_AMASS_CONFIG]: amass_command += ' -config /root/.config/amass.ini' # Run Amass Passive logging.info(amass_command) os.system(amass_command) if 'amass-active' in tools: amass_command = 'amass enum -active -d {} -o {}/from_amass_active.txt'.format( domain.name, results_dir) if USE_AMASS_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_AMASS_CONFIG]: amass_command += ' -config /root/.config/amass.ini' if AMASS_WORDLIST in yaml_configuration[SUBDOMAIN_DISCOVERY]: wordlist = yaml_configuration[SUBDOMAIN_DISCOVERY][AMASS_WORDLIST] if wordlist == 'default': wordlist_path = '/usr/src/wordlist/deepmagic.com-prefixes-top50000.txt' else: wordlist_path = '/usr/src/wordlist/' + wordlist + '.txt' if not os.path.exists(wordlist_path): wordlist_path = '/usr/src/' + AMASS_WORDLIST amass_command = amass_command + \ ' -brute -w {}'.format(wordlist_path) if amass_config_path: amass_command = amass_command + \ ' -config {}'.format('/usr/src/scan_results/' + amass_config_path) # Run Amass Active logging.info(amass_command) os.system(amass_command) if 'assetfinder' in tools: assetfinder_command = 'assetfinder --subs-only {} > {}/from_assetfinder.txt'.format( domain.name, results_dir) # Run Assetfinder logging.info(assetfinder_command) os.system(assetfinder_command) if 'sublist3r' in tools: sublist3r_command = 'python3 /usr/src/github/Sublist3r/sublist3r.py -d {} -t {} -o {}/from_sublister.txt'.format( domain.name, threads, results_dir) # Run sublist3r logging.info(sublist3r_command) os.system(sublist3r_command) if 'subfinder' in tools: subfinder_command = 'subfinder -d {} -t {} -o {}/from_subfinder.txt'.format( domain.name, threads, results_dir) if USE_SUBFINDER_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_SUBFINDER_CONFIG]: subfinder_command += ' -config /root/.config/subfinder/config.yaml' # Run Subfinder logging.info(subfinder_command) os.system(subfinder_command) if 'oneforall' in tools: oneforall_command = 'python3 /usr/src/github/OneForAll/oneforall.py --target {} run'.format( domain.name, results_dir) # Run OneForAll logging.info(oneforall_command) os.system(oneforall_command) extract_subdomain = "cut -d',' -f6 /usr/src/github/OneForAll/results/{}.csv >> {}/from_oneforall.txt".format( domain.name, results_dir) os.system(extract_subdomain) # remove the results from oneforall directory os.system( 'rm -rf /usr/src/github/OneForAll/results/{}.*'.format(domain.name)) ''' All tools have gathered the list of subdomains with filename initials as from_* We will gather all the results in one single file, sort them and remove the older results from_* ''' os.system( 'cat {0}/*.txt > {0}/subdomain_collection.txt'.format(results_dir)) ''' Write target domain into subdomain_collection ''' os.system( 'cat {0}/target_domain.txt >> {0}/subdomain_collection.txt'.format(results_dir)) ''' Remove all the from_* files ''' os.system('rm -f {}/from*'.format(results_dir)) ''' Sort all Subdomains ''' os.system( 'sort -u {0}/subdomain_collection.txt -o {0}/sorted_subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/subdomain_collection.txt'.format(results_dir)) ''' The final results will be stored in sorted_subdomain_collection. ''' # parse the subdomain list file and store in db with open(subdomain_scan_results_file) as subdomain_list: for _subdomain in subdomain_list: __subdomain = _subdomain.rstrip('\n') if not Subdomain.objects.filter(scan_history=task, name=__subdomain).exists( ) and validators.domain(__subdomain) and __subdomain not in out_of_scope_subdomains: subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': __subdomain, }) save_subdomain(subdomain_dict) notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: subdomains_count = Subdomain.objects.filter(scan_history=task).count() send_notification('Subdomain Gathering for target {} has been completed and has discovered *{}* subdomains.'.format(domain.name, subdomains_count)) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/sorted_subdomain_collection.txt') # check for any subdomain changes and send notif if any if notification and notification[0].send_subdomain_changes_notif: newly_added_subdomain = get_new_added_subdomain(task.id, domain.id) if newly_added_subdomain: message = "**{} New Subdomains Discovered on domain {}**".format(newly_added_subdomain.count(), domain.name) for subdomain in newly_added_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) removed_subdomain = get_removed_subdomain(task.id, domain.id) if removed_subdomain: message = "**{} Subdomains are no longer available on domain {}**".format(removed_subdomain.count(), domain.name) for subdomain in removed_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) # check for interesting subdomains and send notif if any if notification and notification[0].send_interesting_notif: interesting_subdomain = get_interesting_subdomains(task.id, domain.id) print(interesting_subdomain) if interesting_subdomain: message = "**{} Interesting Subdomains Found on domain {}**".format(interesting_subdomain.count(), domain.name) for subdomain in interesting_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) def get_new_added_subdomain(scan_id, domain_id): scan_history = ScanHistory.objects.filter( domain=domain_id).filter( subdomain_discovery=True).filter( id__lte=scan_id) if scan_history.count() > 1: last_scan = scan_history.order_by('-start_scan_date')[1] scanned_host_q1 = Subdomain.objects.filter( scan_history__id=scan_id).values('name') scanned_host_q2 = Subdomain.objects.filter( scan_history__id=last_scan.id).values('name') added_subdomain = scanned_host_q1.difference(scanned_host_q2) return Subdomain.objects.filter( scan_history=scan_id).filter( name__in=added_subdomain) def get_removed_subdomain(scan_id, domain_id): scan_history = ScanHistory.objects.filter( domain=domain_id).filter( subdomain_discovery=True).filter( id__lte=scan_id) if scan_history.count() > 1: last_scan = scan_history.order_by('-start_scan_date')[1] scanned_host_q1 = Subdomain.objects.filter( scan_history__id=scan_id).values('name') scanned_host_q2 = Subdomain.objects.filter( scan_history__id=last_scan.id).values('name') removed_subdomains = scanned_host_q2.difference(scanned_host_q1) print() return Subdomain.objects.filter( scan_history=last_scan).filter( name__in=removed_subdomains) def http_crawler(task, domain, results_dir, activity_id): ''' This function is runs right after subdomain gathering, and gathers important like page title, http status, etc HTTP Crawler runs by default ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('HTTP Crawler for target {} has been initiated.'.format(domain.name)) alive_file_location = results_dir + '/alive.txt' httpx_results_file = results_dir + '/httpx.json' subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' httpx_command = 'httpx -status-code -content-length -title -tech-detect -cdn -ip -follow-host-redirects -random-agent' proxy = get_random_proxy() if proxy: httpx_command += " --http-proxy '{}'".format(proxy) httpx_command += ' -json -o {}'.format( httpx_results_file ) httpx_command = 'cat {} | {}'.format(subdomain_scan_results_file, httpx_command) print(httpx_command) os.system(httpx_command) # alive subdomains from httpx alive_file = open(alive_file_location, 'w') # writing httpx results if os.path.isfile(httpx_results_file): httpx_json_result = open(httpx_results_file, 'r') lines = httpx_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) try: # fallback for older versions of httpx if 'url' in json_st: subdomain = Subdomain.objects.get( scan_history=task, name=json_st['input']) else: subdomain = Subdomain.objects.get( scan_history=task, name=json_st['url'].split("//")[-1]) ''' Saving Default http urls to EndPoint ''' endpoint = EndPoint() endpoint.scan_history = task endpoint.target_domain = domain endpoint.subdomain = subdomain if 'url' in json_st: endpoint.http_url = json_st['url'] subdomain.http_url = json_st['url'] if 'status-code' in json_st: endpoint.http_status = json_st['status-code'] subdomain.http_status = json_st['status-code'] if 'title' in json_st: endpoint.page_title = json_st['title'] subdomain.page_title = json_st['title'] if 'content-length' in json_st: endpoint.content_length = json_st['content-length'] subdomain.content_length = json_st['content-length'] if 'content-type' in json_st: endpoint.content_type = json_st['content-type'] subdomain.content_type = json_st['content-type'] if 'webserver' in json_st: endpoint.webserver = json_st['webserver'] subdomain.webserver = json_st['webserver'] if 'response-time' in json_st: response_time = float( ''.join( ch for ch in json_st['response-time'] if not ch.isalpha())) if json_st['response-time'][-2:] == 'ms': response_time = response_time / 1000 endpoint.response_time = response_time subdomain.response_time = response_time if 'cnames' in json_st: cname_list = ','.join(json_st['cnames']) subdomain.cname = cname_list discovered_date = timezone.now() endpoint.discovered_date = discovered_date subdomain.discovered_date = discovered_date endpoint.is_default = True endpoint.save() subdomain.save() if 'technologies' in json_st: for _tech in json_st['technologies']: if Technology.objects.filter(name=_tech).exists(): tech = Technology.objects.get(name=_tech) else: tech = Technology(name=_tech) tech.save() subdomain.technologies.add(tech) endpoint.technologies.add(tech) if 'a' in json_st: for _ip in json_st['a']: if IpAddress.objects.filter(address=_ip).exists(): ip = IpAddress.objects.get(address=_ip) else: ip = IpAddress(address=_ip) if 'cdn' in json_st: ip.is_cdn = json_st['cdn'] ip.save() subdomain.ip_addresses.add(ip) # see if to ignore 404 or 5xx alive_file.write(json_st['url'] + '\n') subdomain.save() endpoint.save() except Exception as exception: logging.error(exception) alive_file.close() if notification and notification[0].send_scan_status_notif: alive_count = Subdomain.objects.filter( scan_history__id=task.id).values('name').distinct().filter( http_status__exact=200).count() send_notification('HTTP Crawler for target {} has been completed.\n\n {} subdomains were alive (http status 200).'.format(domain.name, alive_count)) def grab_screenshot(task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for taking screenshots ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine is currently gathering screenshots for {}'.format(domain.name)) output_screenshots_path = results_dir + '/screenshots' result_csv_path = results_dir + '/screenshots/Requests.csv' alive_subdomains_path = results_dir + '/alive.txt' eyewitness_command = 'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py' eyewitness_command += ' -f {} -d {} --no-prompt'.format( alive_subdomains_path, output_screenshots_path ) if EYEWITNESS in yaml_configuration \ and TIMEOUT in yaml_configuration[EYEWITNESS] \ and yaml_configuration[EYEWITNESS][TIMEOUT] > 0: eyewitness_command += ' --timeout {}'.format( yaml_configuration[EYEWITNESS][TIMEOUT] ) if EYEWITNESS in yaml_configuration \ and THREADS in yaml_configuration[EYEWITNESS] \ and yaml_configuration[EYEWITNESS][THREADS] > 0: eyewitness_command += ' --threads {}'.format( yaml_configuration[EYEWITNESS][THREADS] ) logger.info(eyewitness_command) os.system(eyewitness_command) if os.path.isfile(result_csv_path): logger.info('Gathering Eyewitness results') with open(result_csv_path, 'r') as file: reader = csv.reader(file) for row in reader: if row[3] == 'Successful' \ and Subdomain.objects.filter( scan_history__id=task.id).filter(name=row[2]).exists(): subdomain = Subdomain.objects.get( scan_history__id=task.id, name=row[2] ) subdomain.screenshot_path = row[4].replace( '/usr/src/scan_results/', '' ) subdomain.save() # remove all db, html extra files in screenshot results os.system('rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format( output_screenshots_path, )) os.system('rm -rf {0}/source'.format( output_screenshots_path, )) if notification and notification[0].send_scan_status_notif: send_notification('reNgine has finished gathering screenshots for {}'.format(domain.name)) def port_scanning(task, domain, yaml_configuration, results_dir): ''' This function is responsible for running the port scan ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Port Scan initiated for {}'.format(domain.name)) subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' port_results_file = results_dir + '/ports.json' # check the yaml_configuration and choose the ports to be scanned scan_ports = '-' # default port scan everything if PORTS in yaml_configuration[PORT_SCAN]: # TODO: legacy code, remove top-100 in future versions all_ports = yaml_configuration[PORT_SCAN][PORTS] if 'full' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -p {}'.format( subdomain_scan_results_file, port_results_file, '-') elif 'top-100' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -top-ports 100'.format( subdomain_scan_results_file, port_results_file) elif 'top-1000' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -top-ports 1000'.format( subdomain_scan_results_file, port_results_file) else: scan_ports = ','.join( str(port) for port in all_ports) naabu_command = 'cat {} | naabu -json -o {} -p {}'.format( subdomain_scan_results_file, port_results_file, scan_ports) # check for exclude ports if EXCLUDE_PORTS in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][EXCLUDE_PORTS]: exclude_ports = ','.join( str(port) for port in yaml_configuration['port_scan']['exclude_ports']) naabu_command = naabu_command + \ ' -exclude-ports {}'.format(exclude_ports) if NAABU_RATE in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][NAABU_RATE] > 0: naabu_command = naabu_command + \ ' -rate {}'.format( yaml_configuration[PORT_SCAN][NAABU_RATE]) if USE_NAABU_CONFIG in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][USE_NAABU_CONFIG]: naabu_command += ' -config /root/.config/naabu/naabu.conf' # run naabu os.system(naabu_command) # writing port results try: port_json_result = open(port_results_file, 'r') lines = port_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) port_number = json_st['port'] ip_address = json_st['ip'] # see if port already exists if Port.objects.filter(number__exact=port_number).exists(): port = Port.objects.get(number=port_number) else: port = Port() port.number = port_number if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port_detail = whatportis.get_ports(str(port_number)) if len(port_detail): port.service_name = port_detail[0].name port.description = port_detail[0].description port.save() if IpAddress.objects.filter(address=json_st['ip']).exists(): ip = IpAddress.objects.get(address=json_st['ip']) ip.ports.add(port) ip.save() except BaseException as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: port_count = Port.objects.filter( ports__in=IpAddress.objects.filter( ip_addresses__in=Subdomain.objects.filter( scan_history__id=task.id))).distinct().count() send_notification('reNgine has finished Port Scanning on {} and has identified {} ports.'.format(domain.name, port_count)) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/ports.json') def check_waf(): ''' This function will check for the WAF being used in subdomains using wafw00f ''' pass def directory_brute(task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for performing directory scan ''' # scan directories for all the alive subdomain with http status > # 200 notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Directory Bruteforce has been initiated for {}.'.format(domain.name)) alive_subdomains = Subdomain.objects.filter( scan_history__id=task.id).exclude(http_url__isnull=True) dirs_results = results_dir + '/dirs.json' # check the yaml settings if EXTENSIONS in yaml_configuration[DIR_FILE_SEARCH]: extensions = ','.join( str(ext) for ext in yaml_configuration[DIR_FILE_SEARCH][EXTENSIONS]) else: extensions = 'php,git,yaml,conf,db,mysql,bak,txt' # Threads if THREADS in yaml_configuration[DIR_FILE_SEARCH] \ and yaml_configuration[DIR_FILE_SEARCH][THREADS] > 0: threads = yaml_configuration[DIR_FILE_SEARCH][THREADS] else: threads = 10 for subdomain in alive_subdomains: # delete any existing dirs.json if os.path.isfile(dirs_results): os.system('rm -rf {}'.format(dirs_results)) dirsearch_command = 'python3 /usr/src/github/dirsearch/dirsearch.py' dirsearch_command += ' -u {}'.format(subdomain.http_url) if (WORDLIST not in yaml_configuration[DIR_FILE_SEARCH] or not yaml_configuration[DIR_FILE_SEARCH][WORDLIST] or 'default' in yaml_configuration[DIR_FILE_SEARCH][WORDLIST]): wordlist_location = '/usr/src/github/dirsearch/db/dicc.txt' else: wordlist_location = '/usr/src/wordlist/' + \ yaml_configuration[DIR_FILE_SEARCH][WORDLIST] + '.txt' dirsearch_command += ' -w {}'.format(wordlist_location) dirsearch_command += ' --format json -o {}'.format(dirs_results) dirsearch_command += ' -e {}'.format(extensions) dirsearch_command += ' -t {}'.format(threads) dirsearch_command += ' --random-agent --follow-redirects --exclude-status 403,401,404' if EXCLUDE_EXTENSIONS in yaml_configuration[DIR_FILE_SEARCH]: exclude_extensions = ','.join( str(ext) for ext in yaml_configuration[DIR_FILE_SEARCH][EXCLUDE_EXTENSIONS]) dirsearch_command += ' -X {}'.format(exclude_extensions) if EXCLUDE_TEXT in yaml_configuration[DIR_FILE_SEARCH]: exclude_text = ','.join( str(text) for text in yaml_configuration[DIR_FILE_SEARCH][EXCLUDE_TEXT]) dirsearch_command += ' -exclude-texts {}'.format(exclude_text) # check if recursive strategy is set to on if RECURSIVE_LEVEL in yaml_configuration[DIR_FILE_SEARCH]: dirsearch_command += ' --recursion-depth {}'.format(yaml_configuration[DIR_FILE_SEARCH][RECURSIVE_LEVEL]) if RECURSIVE_LEVEL in yaml_configuration[DIR_FILE_SEARCH]: dirsearch_command += ' --recursion-depth {}'.format(yaml_configuration[DIR_FILE_SEARCH][RECURSIVE_LEVEL]) # proxy proxy = get_random_proxy() if proxy: dirsearch_command += " --proxy '{}'".format(proxy) print(dirsearch_command) os.system(dirsearch_command) try: if os.path.isfile(dirs_results): with open(dirs_results, "r") as json_file: json_string = json_file.read() subdomain = Subdomain.objects.get( scan_history__id=task.id, http_url=subdomain.http_url) subdomain.directory_json = json_string subdomain.save() except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: send_notification('Directory Bruteforce has been completed for {}.'.format(domain.name)) def fetch_endpoints( task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for fetching all the urls associated with target and run HTTP probe It first runs gau to gather all urls from wayback, then we will use hakrawler to identify more urls ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine is currently gathering endpoints for {}.'.format(domain.name)) # check yaml settings if ALL in yaml_configuration[FETCH_URL][USES_TOOLS]: tools = 'gauplus hakrawler waybackurls gospider' else: tools = ' '.join( str(tool) for tool in yaml_configuration[FETCH_URL][USES_TOOLS]) if INTENSITY in yaml_configuration[FETCH_URL]: scan_type = yaml_configuration[FETCH_URL][INTENSITY] else: scan_type = 'normal' domain_regex = "\'https?://([a-z0-9]+[.])*{}.*\'".format(domain.name) if 'deep' in scan_type: # performs deep url gathering for all the subdomains present - # RECOMMENDED logger.info('Deep URLS Fetch') os.system(settings.TOOL_LOCATION + 'get_urls.sh %s %s %s %s %s' % ("None", results_dir, scan_type, domain_regex, tools)) else: # perform url gathering only for main domain - USE only for quick scan logger.info('Non Deep URLS Fetch') os.system( settings.TOOL_LOCATION + 'get_urls.sh %s %s %s %s %s' % ( domain.name, results_dir, scan_type, domain_regex, tools )) if IGNORE_FILE_EXTENSION in yaml_configuration[FETCH_URL]: ignore_extension = '|'.join( yaml_configuration[FETCH_URL][IGNORE_FILE_EXTENSION]) logger.info('Ignore extensions' + ignore_extension) os.system( 'cat {0}/all_urls.txt | grep -Eiv "\\.({1}).*" > {0}/temp_urls.txt'.format( results_dir, ignore_extension)) os.system( 'rm {0}/all_urls.txt && mv {0}/temp_urls.txt {0}/all_urls.txt'.format(results_dir)) ''' Store all the endpoints and then run the httpx ''' try: endpoint_final_url = results_dir + '/all_urls.txt' if os.path.isfile(endpoint_final_url): with open(endpoint_final_url) as endpoint_list: for url in endpoint_list: http_url = url.rstrip('\n') if not EndPoint.objects.filter(scan_history=task, http_url=http_url).exists(): _subdomain = get_subdomain_from_url(http_url) if Subdomain.objects.filter( scan_history=task).filter( name=_subdomain).exists(): subdomain = Subdomain.objects.get( scan_history=task, name=_subdomain) else: ''' gau or gosppider can gather interesting endpoints which when parsed can give subdomains that were not existent from subdomain scan. so storing them ''' logger.error( 'Subdomain {} not found, adding...'.format(_subdomain)) subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': _subdomain, }) subdomain = save_subdomain(subdomain_dict) endpoint_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'subdomain': subdomain, 'http_url': http_url, }) save_endpoint(endpoint_dict) except Exception as e: logger.error(e) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/all_urls.txt') ''' TODO: Go spider & waybackurls accumulates a lot of urls, which is good but nuclei takes forever to scan even a simple website, so we will do http probing and filter HTTP status 404, this way we can reduce the number of Non Existent URLS ''' logger.info('HTTP Probing on collected endpoints') httpx_command = 'httpx -l {0}/all_urls.txt -status-code -content-length -ip -cdn -title -tech-detect -json -follow-redirects -random-agent -o {0}/final_httpx_urls.json'.format(results_dir) proxy = get_random_proxy() if proxy: httpx_command += " --http-proxy '{}'".format(proxy) os.system(httpx_command) url_results_file = results_dir + '/final_httpx_urls.json' try: urls_json_result = open(url_results_file, 'r') lines = urls_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) http_url = json_st['url'] _subdomain = get_subdomain_from_url(http_url) if Subdomain.objects.filter( scan_history=task).filter( name=_subdomain).exists(): subdomain_obj = Subdomain.objects.get( scan_history=task, name=_subdomain) else: subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': _subdomain, }) subdomain_obj = save_subdomain(subdomain_dict) if EndPoint.objects.filter( scan_history=task).filter( http_url=http_url).exists(): endpoint = EndPoint.objects.get( scan_history=task, http_url=http_url) else: endpoint = EndPoint() endpoint_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'http_url': http_url, 'subdomain': subdomain_obj }) endpoint = save_endpoint(endpoint_dict) if 'title' in json_st: endpoint.page_title = json_st['title'] if 'webserver' in json_st: endpoint.webserver = json_st['webserver'] if 'content-length' in json_st: endpoint.content_length = json_st['content-length'] if 'content-type' in json_st: endpoint.content_type = json_st['content-type'] if 'status-code' in json_st: endpoint.http_status = json_st['status-code'] if 'response-time' in json_st: response_time = float(''.join(ch for ch in json_st['response-time'] if not ch.isalpha())) if json_st['response-time'][-2:] == 'ms': response_time = response_time / 1000 endpoint.response_time = response_time endpoint.save() if 'technologies' in json_st: for _tech in json_st['technologies']: if Technology.objects.filter(name=_tech).exists(): tech = Technology.objects.get(name=_tech) else: tech = Technology(name=_tech) tech.save() endpoint.technologies.add(tech) # get subdomain object subdomain = Subdomain.objects.get(scan_history=task, name=_subdomain) subdomain.technologies.add(tech) subdomain.save() except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: endpoint_count = EndPoint.objects.filter( scan_history__id=task.id).values('http_url').distinct().count() endpoint_alive_count = EndPoint.objects.filter( scan_history__id=task.id, http_status__exact=200).values('http_url').distinct().count() send_notification('reNgine has finished gathering endpoints for {} and has discovered *{}* unique endpoints.\n\n{} of those endpoints reported HTTP status 200.'.format( domain.name, endpoint_count, endpoint_alive_count )) # once endpoint is saved, run gf patterns TODO: run threads if GF_PATTERNS in yaml_configuration[FETCH_URL]: for pattern in yaml_configuration[FETCH_URL][GF_PATTERNS]: logger.info('Running GF for {}'.format(pattern)) gf_output_file_path = '{0}/gf_patterns_{1}.txt'.format( results_dir, pattern) gf_command = 'cat {0}/all_urls.txt | gf {1} >> {2}'.format( results_dir, pattern, gf_output_file_path) os.system(gf_command) if os.path.exists(gf_output_file_path): with open(gf_output_file_path) as gf_output: for line in gf_output: url = line.rstrip('\n') try: endpoint = EndPoint.objects.get( scan_history=task, http_url=url) earlier_pattern = endpoint.matched_gf_patterns new_pattern = earlier_pattern + ',' + pattern if earlier_pattern else pattern endpoint.matched_gf_patterns = new_pattern except Exception as e: # add the url in db logger.error(e) logger.info('Adding URL' + url) endpoint = EndPoint() endpoint.http_url = url endpoint.target_domain = domain endpoint.scan_history = task try: _subdomain = Subdomain.objects.get( scan_history=task, name=get_subdomain_from_url(url)) endpoint.subdomain = _subdomain except Exception as e: continue endpoint.matched_gf_patterns = pattern finally: endpoint.save() def vulnerability_scan( task, domain, yaml_configuration, results_dir, activity_id): notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Vulnerability scan has been initiated for {}.'.format(domain.name)) ''' This function will run nuclei as a vulnerability scanner ---- unfurl the urls to keep only domain and path, this will be sent to vuln scan ignore certain file extensions Thanks: https://github.com/six2dez/reconftw ''' urls_path = '/alive.txt' if task.scan_type.fetch_url: os.system('cat {0}/all_urls.txt | grep -Eiv "\\.(eot|jpg|jpeg|gif|css|tif|tiff|png|ttf|otf|woff|woff2|ico|pdf|svg|txt|js|doc|docx)$" | unfurl -u format %s://%d%p >> {0}/unfurl_urls.txt'.format(results_dir)) os.system( 'sort -u {0}/unfurl_urls.txt -o {0}/unfurl_urls.txt'.format(results_dir)) urls_path = '/unfurl_urls.txt' vulnerability_result_path = results_dir + '/vulnerability.json' vulnerability_scan_input_file = results_dir + urls_path nuclei_command = 'nuclei -json -l {} -o {}'.format( vulnerability_scan_input_file, vulnerability_result_path) # check nuclei config if USE_NUCLEI_CONFIG in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[VULNERABILITY_SCAN][USE_NUCLEI_CONFIG]: nuclei_command += ' -config /root/.config/nuclei/config.yaml' ''' Nuclei Templates Either custom template has to be supplied or default template, if neither has been supplied then use all templates including custom templates ''' if CUSTOM_NUCLEI_TEMPLATE in yaml_configuration[ VULNERABILITY_SCAN] or NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: # check yaml settings for templates if NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: if ALL in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_TEMPLATE]: template = NUCLEI_TEMPLATES_PATH else: _template = ','.join([NUCLEI_TEMPLATES_PATH + str(element) for element in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_TEMPLATE]]) template = _template.replace(',', ' -t ') # Update nuclei command with templates nuclei_command = nuclei_command + ' -t ' + template if CUSTOM_NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: # add .yaml to the custom template extensions _template = ','.join( [str(element) + '.yaml' for element in yaml_configuration[VULNERABILITY_SCAN][CUSTOM_NUCLEI_TEMPLATE]]) template = _template.replace(',', ' -t ') # Update nuclei command with templates nuclei_command = nuclei_command + ' -t ' + template else: nuclei_command = nuclei_command + ' -t /root/nuclei-templates' # check yaml settings for concurrency if NUCLEI_CONCURRENCY in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][NUCLEI_CONCURRENCY] > 0: concurrency = yaml_configuration[VULNERABILITY_SCAN][NUCLEI_CONCURRENCY] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -c ' + str(concurrency) if RATE_LIMIT in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][RATE_LIMIT] > 0: rate_limit = yaml_configuration[VULNERABILITY_SCAN][RATE_LIMIT] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -rl ' + str(rate_limit) if TIMEOUT in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][TIMEOUT] > 0: timeout = yaml_configuration[VULNERABILITY_SCAN][TIMEOUT] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -timeout ' + str(timeout) if RETRIES in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][RETRIES] > 0: retries = yaml_configuration[VULNERABILITY_SCAN][RETRIES] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -retries ' + str(retries) # for severity if NUCLEI_SEVERITY in yaml_configuration[VULNERABILITY_SCAN] and ALL not in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_SEVERITY]: _severity = ','.join( [str(element) for element in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_SEVERITY]]) severity = _severity.replace(" ", "") else: severity = "critical, high, medium, low, info" # update nuclei templates before running scan os.system('nuclei -update-templates') for _severity in severity.split(","): # delete any existing vulnerability.json file if os.path.isfile(vulnerability_result_path): os.system('rm {}'.format(vulnerability_result_path)) # run nuclei final_nuclei_command = nuclei_command + ' -severity ' + _severity proxy = get_random_proxy() if proxy: final_nuclei_command += " --proxy-url '{}'".format(proxy) logger.info(final_nuclei_command) os.system(final_nuclei_command) try: if os.path.isfile(vulnerability_result_path): urls_json_result = open(vulnerability_result_path, 'r') lines = urls_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) host = json_st['host'] _subdomain = get_subdomain_from_url(host) try: subdomain = Subdomain.objects.get( name=_subdomain, scan_history=task) vulnerability = Vulnerability() vulnerability.subdomain = subdomain vulnerability.scan_history = task vulnerability.target_domain = domain try: endpoint = EndPoint.objects.get( scan_history=task, target_domain=domain, http_url=host) vulnerability.endpoint = endpoint except Exception as exception: logger.error(exception) if 'name' in json_st['info']: vulnerability.name = json_st['info']['name'] if 'severity' in json_st['info']: if json_st['info']['severity'] == 'info': severity = 0 elif json_st['info']['severity'] == 'low': severity = 1 elif json_st['info']['severity'] == 'medium': severity = 2 elif json_st['info']['severity'] == 'high': severity = 3 elif json_st['info']['severity'] == 'critical': severity = 4 else: severity = 0 else: severity = 0 vulnerability.severity = severity if 'tags' in json_st['info']: vulnerability.tags = json_st['info']['tags'] if 'description' in json_st['info']: vulnerability.description = json_st['info']['description'] if 'reference' in json_st['info']: vulnerability.reference = json_st['info']['reference'] if 'matched' in json_st: # TODO remove in rengine 1.1. 'matched' isn't used in nuclei 2.5.3 vulnerability.http_url = json_st['matched'] if 'matched-at' in json_st: vulnerability.http_url = json_st['matched-at'] if 'templateID' in json_st: vulnerability.template_used = json_st['templateID'] if 'description' in json_st: vulnerability.description = json_st['description'] if 'matcher_name' in json_st: vulnerability.matcher_name = json_st['matcher_name'] if 'extracted_results' in json_st: vulnerability.extracted_results = json_st['extracted_results'] vulnerability.discovered_date = timezone.now() vulnerability.open_status = True vulnerability.save() # send notification for all vulnerabilities except info if json_st['info']['severity'] != "info" and notification and notification[0].send_vuln_notif: message = "*Alert: Vulnerability Identified*" message += "\n\n" message += "A *{}* severity vulnerability has been identified.".format(json_st['info']['severity']) message += "\nVulnerability Name: {}".format(json_st['info']['name']) message += "\nVulnerable URL: {}".format(json_st['host']) send_notification(message) # send report to hackerone if Hackerone.objects.all().exists() and json_st['info']['severity'] != 'info' and json_st['info']['severity'] \ != 'low' and vulnerability.target_domain.h1_team_handle: hackerone = Hackerone.objects.all()[0] if hackerone.send_critical and json_st['info']['severity'] == 'critical': send_hackerone_report(vulnerability.id) elif hackerone.send_high and json_st['info']['severity'] == 'high': send_hackerone_report(vulnerability.id) elif hackerone.send_medium and json_st['info']['severity'] == 'medium': send_hackerone_report(vulnerability.id) except ObjectDoesNotExist: logger.error('Object not found') continue except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: info_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=0).count() low_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=1).count() medium_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=2).count() high_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=3).count() critical_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=4).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count message = 'Vulnerability scan has been completed for {} and discovered {} vulnerabilities.'.format( domain.name, vulnerability_count ) message += '\n\n*Vulnerability Stats:*' message += '\nCritical: {}'.format(critical_count) message += '\nHigh: {}'.format(high_count) message += '\nMedium: {}'.format(medium_count) message += '\nLow: {}'.format(low_count) message += '\nInfo: {}'.format(info_count) send_notification(message) def scan_failed(task): task.scan_status = 0 task.stop_scan_date = timezone.now() task.save() def create_scan_activity(task, message, status): scan_activity = ScanActivity() scan_activity.scan_of = task scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id def update_last_activity(id, activity_status): ScanActivity.objects.filter( id=id).update( status=activity_status, time=timezone.now()) def delete_scan_data(results_dir): # remove all txt,html,json files os.system('find {} -name "*.txt" -type f -delete'.format(results_dir)) os.system('find {} -name "*.html" -type f -delete'.format(results_dir)) os.system('find {} -name "*.json" -type f -delete'.format(results_dir)) def save_subdomain(subdomain_dict): subdomain = Subdomain() subdomain.discovered_date = timezone.now() subdomain.target_domain = subdomain_dict.get('target_domain') subdomain.scan_history = subdomain_dict.get('scan_history') subdomain.name = subdomain_dict.get('name') subdomain.http_url = subdomain_dict.get('http_url') subdomain.screenshot_path = subdomain_dict.get('screenshot_path') subdomain.http_header_path = subdomain_dict.get('http_header_path') subdomain.cname = subdomain_dict.get('cname') subdomain.is_cdn = subdomain_dict.get('is_cdn') subdomain.content_type = subdomain_dict.get('content_type') subdomain.webserver = subdomain_dict.get('webserver') subdomain.page_title = subdomain_dict.get('page_title') subdomain.is_imported_subdomain = subdomain_dict.get( 'is_imported_subdomain') if 'is_imported_subdomain' in subdomain_dict else False if 'http_status' in subdomain_dict: subdomain.http_status = subdomain_dict.get('http_status') if 'response_time' in subdomain_dict: subdomain.response_time = subdomain_dict.get('response_time') if 'content_length' in subdomain_dict: subdomain.content_length = subdomain_dict.get('content_length') subdomain.save() return subdomain def save_endpoint(endpoint_dict): endpoint = EndPoint() endpoint.discovered_date = timezone.now() endpoint.scan_history = endpoint_dict.get('scan_history') endpoint.target_domain = endpoint_dict.get('target_domain') if 'target_domain' in endpoint_dict else None endpoint.subdomain = endpoint_dict.get('subdomain') if 'target_domain' in endpoint_dict else None endpoint.http_url = endpoint_dict.get('http_url') endpoint.page_title = endpoint_dict.get('page_title') if 'page_title' in endpoint_dict else None endpoint.content_type = endpoint_dict.get('content_type') if 'content_type' in endpoint_dict else None endpoint.webserver = endpoint_dict.get('webserver') if 'webserver' in endpoint_dict else None endpoint.response_time = endpoint_dict.get('response_time') if 'response_time' in endpoint_dict else 0 endpoint.http_status = endpoint_dict.get('http_status') if 'http_status' in endpoint_dict else 0 endpoint.content_length = endpoint_dict.get('content_length') if 'content_length' in endpoint_dict else 0 endpoint.is_default = endpoint_dict.get('is_default') if 'is_default' in endpoint_dict else False endpoint.save() return endpoint def perform_osint(task, domain, yaml_configuration, results_dir): notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine has initiated OSINT on target {}'.format(domain.name)) if 'discover' in yaml_configuration[OSINT]: osint_discovery(task, domain, yaml_configuration, results_dir) if 'dork' in yaml_configuration[OSINT]: dorking(task, yaml_configuration) if notification and notification[0].send_scan_status_notif: send_notification('reNgine has completed performing OSINT on target {}'.format(domain.name)) def osint_discovery(task, domain, yaml_configuration, results_dir): if ALL in yaml_configuration[OSINT][OSINT_DISCOVER]: osint_lookup = 'emails metainfo employees' else: osint_lookup = ' '.join( str(lookup) for lookup in yaml_configuration[OSINT][OSINT_DISCOVER]) if 'metainfo' in osint_lookup: if INTENSITY in yaml_configuration[OSINT]: osint_intensity = yaml_configuration[OSINT][INTENSITY] else: osint_intensity = 'normal' if OSINT_DOCUMENTS_LIMIT in yaml_configuration[OSINT]: documents_limit = yaml_configuration[OSINT][OSINT_DOCUMENTS_LIMIT] else: documents_limit = 50 if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': domain.name, 'domain': domain, 'scan_id': task, 'documents_limit': documents_limit }) get_and_save_meta_info(meta_dict) elif osint_intensity == 'deep': # get all subdomains in scan_id subdomains = Subdomain.objects.filter(scan_history=task) for subdomain in subdomains: meta_dict = DottedDict({ 'osint_target': subdomain.name, 'domain': domain, 'scan_id': task, 'documents_limit': documents_limit }) get_and_save_meta_info(meta_dict) if 'emails' in osint_lookup: get_and_save_emails(task, results_dir) get_and_save_leaked_credentials(task, results_dir) if 'employees' in osint_lookup: get_and_save_employees(task, results_dir) def dorking(scan_history, yaml_configuration): # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh # look in stackoverflow if ALL in yaml_configuration[OSINT][OSINT_DORK]: dork_lookup = 'stackoverflow, 3rdparty, social_media, project_management, code_sharing, config_files, jenkins, cloud_buckets, php_error, exposed_documents, struts_rce, db_files, traefik, git_exposed' else: dork_lookup = ' '.join( str(lookup) for lookup in yaml_configuration[OSINT][OSINT_DORK]) if 'stackoverflow' in dork_lookup: dork = 'site:stackoverflow.com' dork_type = 'stackoverflow' get_and_save_dork_results( dork, dork_type, scan_history, in_target=False ) if '3rdparty' in dork_lookup: # look in 3rd party sitee dork_type = '3rdparty' lookup_websites = [ 'gitter.im', 'papaly.com', 'productforums.google.com', 'coggle.it', 'replt.it', 'ycombinator.com', 'libraries.io', 'npm.runkit.com', 'npmjs.com', 'scribd.com', 'gitter.im' ] dork = '' for website in lookup_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'social_media' in dork_lookup: dork_type = 'Social Media' social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'pinterest.com', 'tumblr.com', 'reddit.com' ] dork = '' for website in social_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'project_management' in dork_lookup: dork_type = 'Project Management' project_websites = [ 'trello.com', '*.atlassian.net' ] dork = '' for website in project_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'code_sharing' in dork_lookup: dork_type = 'Code Sharing Sites' code_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] dork = '' for website in code_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'config_files' in dork_lookup: dork_type = 'Config Files' config_file_ext = [ 'env', 'xml', 'conf', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] dork = '' for extension in config_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'jenkins' in dork_lookup: dork_type = 'Jenkins' dork = 'intitle:\"Dashboard [Jenkins]\"' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) if 'wordpress_files' in dork_lookup: dork_type = 'Wordpress Files' inurl_lookup = [ 'wp-content', 'wp-includes' ] dork = '' for lookup in inurl_lookup: dork = dork + ' | ' + 'inurl:' + lookup get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'cloud_buckets' in dork_lookup: dork_type = 'Cloud Buckets' cloud_websites = [ '.s3.amazonaws.com', 'storage.googleapis.com', 'amazonaws.com' ] dork = '' for website in cloud_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'php_error' in dork_lookup: dork_type = 'PHP Error' error_words = [ '\"PHP Parse error\"', '\"PHP Warning\"', '\"PHP Error\"' ] dork = '' for word in error_words: dork = dork + ' | ' + word get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'exposed_documents' in dork_lookup: dork_type = 'Exposed Documents' docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] dork = '' for extension in docs_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'struts_rce' in dork_lookup: dork_type = 'Apache Struts RCE' struts_file_ext = [ 'action', 'struts', 'do' ] dork = '' for extension in struts_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'db_files' in dork_lookup: dork_type = 'Database Files' db_file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] dork = '' for extension in db_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'traefik' in dork_lookup: dork = 'intitle:traefik inurl:8080/dashboard' dork_type = 'Traefik' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) if 'git_exposed' in dork_lookup: dork = 'inurl:\"/.git\"' dork_type = '.git Exposed' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) def get_and_save_dork_results(dork, type, scan_history, in_target=False): degoogle_obj = degoogle.dg() proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy if in_target: query = dork + " site:" + scan_history.domain.name else: query = dork + " \"{}\"".format(scan_history.domain.name) logger.info(query) degoogle_obj.query = query results = degoogle_obj.run() logger.info(results) for result in results: dork, _ = Dork.objects.get_or_create( type=type, description=result['desc'], url=result['url'] ) scan_history.dorks.add(dork) def get_and_save_employees(scan_history, results_dir): theHarvester_location = '/usr/src/github/theHarvester' # update proxies.yaml if Proxy.objects.all().exists(): proxy = Proxy.objects.all()[0] if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(theHarvester_location + '/proxies.yaml', 'w') as file: documents = yaml.dump(yaml_data, file) os.system('cd {} && python3 theHarvester.py -d {} -b all -f {}/theHarvester.html'.format( theHarvester_location, scan_history.domain.name, results_dir )) file_location = results_dir + '/theHarvester.html' print(file_location) # delete proxy environ var if os.environ.get(('https_proxy')): del os.environ['https_proxy'] if os.environ.get(('HTTPS_PROXY')): del os.environ['HTTPS_PROXY'] if os.path.isfile(file_location): logger.info('Parsing theHarvester results') options = FirefoxOptions() options.add_argument("--headless") driver = webdriver.Firefox(options=options) driver.get('file://'+file_location) tabledata = driver.execute_script('return tabledata') # save email addresses and linkedin employees for data in tabledata: if data['record'] == 'email': _email = data['result'] email, _ = Email.objects.get_or_create(address=_email) scan_history.emails.add(email) elif data['record'] == 'people': _employee = data['result'] split_val = _employee.split('-') name = split_val[0] if len(split_val) == 2: designation = split_val[1] else: designation = "" employee, _ = Employee.objects.get_or_create(name=name, designation=designation) scan_history.employees.add(employee) driver.quit() print(tabledata) def get_and_save_emails(scan_history, results_dir): leak_target_path = '{}/creds_target.txt'.format(results_dir) # get email address proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy emails = [] try: logger.info('OSINT: Getting emails from Google') email_from_google = get_emails_from_google(scan_history.domain.name) logger.info('OSINT: Getting emails from Bing') email_from_bing = get_emails_from_bing(scan_history.domain.name) logger.info('OSINT: Getting emails from Baidu') email_from_baidu = get_emails_from_baidu(scan_history.domain.name) emails = list(set(email_from_google + email_from_bing + email_from_baidu)) logger.info(emails) except Exception as e: logger.error(e) leak_target_file = open(leak_target_path, 'w') for _email in emails: email, _ = Email.objects.get_or_create(address=_email) scan_history.emails.add(email) leak_target_file.write('{}\n'.format(_email)) # fill leak_target_file with possible email address leak_target_file.write('%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.write('%.%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%.%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.write('%_%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%_%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.close() def get_and_save_leaked_credentials(scan_history, results_dir): logger.info('OSINT: Getting leaked credentials...') leak_target_file = '{}/creds_target.txt'.format(results_dir) leak_output_file = '{}/pwndb.json'.format(results_dir) pwndb_command = 'python3 /usr/src/github/pwndb/pwndb.py --proxy tor:9150 --output json --list {}'.format( leak_target_file ) try: pwndb_output = subprocess.getoutput(pwndb_command) creds = json.loads(pwndb_output) for cred in creds: if cred['username'] != 'donate': email_id = "{}@{}".format(cred['username'], cred['domain']) email_obj, _ = Email.objects.get_or_create( address=email_id, ) email_obj.password = cred['password'] email_obj.save() scan_history.emails.add(email_obj) except Exception as e: logger.error(e) pass def get_and_save_meta_info(meta_dict): logger.info('Getting METADATA for {}'.format(meta_dict.osint_target)) proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy result = metadata_extractor.extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if result: results = result.get_metadata() for meta in results: meta_finder_document = MetaFinderDocument() subdomain = Subdomain.objects.get(scan_history=meta_dict.scan_id, name=meta_dict.osint_target) meta_finder_document.subdomain = subdomain meta_finder_document.target_domain = meta_dict.domain meta_finder_document.scan_history = meta_dict.scan_id item = DottedDict(results[meta]) meta_finder_document.url = item.url meta_finder_document.doc_name = meta meta_finder_document.http_status = item.status_code metadata = results[meta]['metadata'] for data in metadata: if 'Producer' in metadata and metadata['Producer']: meta_finder_document.producer = metadata['Producer'].rstrip('\x00') if 'Creator' in metadata and metadata['Creator']: meta_finder_document.creator = metadata['Creator'].rstrip('\x00') if 'CreationDate' in metadata and metadata['CreationDate']: meta_finder_document.creation_date = metadata['CreationDate'].rstrip('\x00') if 'ModDate' in metadata and metadata['ModDate']: meta_finder_document.modified_date = metadata['ModDate'].rstrip('\x00') if 'Author' in metadata and metadata['Author']: meta_finder_document.author = metadata['Author'].rstrip('\x00') if 'Title' in metadata and metadata['Title']: meta_finder_document.title = metadata['Title'].rstrip('\x00') if 'OSInfo' in metadata and metadata['OSInfo']: meta_finder_document.os = metadata['OSInfo'].rstrip('\x00') meta_finder_document.save() @app.task(bind=True) def test_task(self): print('*' * 40) print('test task run') print('*' * 40)

radaram

43af3a6aecdece4923ee74b108853f7b9c51ed12

27d6ec5827a51fd74e3ab97a5cef38fc7f5d9168

Can you please confirm, if it is always returning `matched-at` on the latest Nuclei release? Thanks

yogeshojha

yogeshojha/rengine

Fix #529

Nuclei returns the response to stdout: `{"template-id":"tech-detect","info":{"name":"Wappalyzer Technology Detection","author":["hakluke"],"tags":["tech"],"reference":null,"severity":"info"},"matcher-name":"nginx","type":"http","host":"https://example.com:443","matched-at":"https://example.com:443","timestamp":"2021-10-31T09:39:47.1571248Z","curl-command":"curl -X 'GET' -d '' -H 'Accept: */*' -H 'Accept-Language: en' -H 'User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_9_2) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/36.0.1944.0 Safari/537.36' 'https://example.com'"}` It needs to read host_url from matched-at, not from matched.

2021-10-31 10:27:33+00:00

2021-11-01 16:58:16+00:00

web/reNgine/tasks.py

import os import traceback import yaml import json import csv import validators import random import requests import logging import metafinder.extractor as metadata_extractor import whatportis import subprocess from selenium.webdriver.firefox.options import Options as FirefoxOptions from selenium import webdriver from emailfinder.extractor import * from dotted_dict import DottedDict from celery import shared_task from discord_webhook import DiscordWebhook from reNgine.celery import app from startScan.models import * from targetApp.models import Domain from scanEngine.models import EngineType from django.conf import settings from django.shortcuts import get_object_or_404 from celery import shared_task from datetime import datetime from degoogle import degoogle from django.conf import settings from django.utils import timezone, dateformat from django.shortcuts import get_object_or_404 from django.core.exceptions import ObjectDoesNotExist from reNgine.celery import app from reNgine.definitions import * from startScan.models import * from targetApp.models import Domain from scanEngine.models import EngineType, Configuration, Wordlist from .common_func import * ''' task for background scan ''' @app.task def initiate_scan( domain_id, scan_history_id, scan_type, engine_type, imported_subdomains=None, out_of_scope_subdomains=[] ): ''' scan_type = 0 -> immediate scan, need not create scan object scan_type = 1 -> scheduled scan ''' engine_object = EngineType.objects.get(pk=engine_type) domain = Domain.objects.get(pk=domain_id) if scan_type == 1: task = ScanHistory() task.scan_status = -1 elif scan_type == 0: task = ScanHistory.objects.get(pk=scan_history_id) # save the last scan date for domain model domain.last_scan_date = timezone.now() domain.save() # once the celery task starts, change the task status to Started task.scan_type = engine_object task.celery_id = initiate_scan.request.id task.domain = domain task.scan_status = 1 task.start_scan_date = timezone.now() task.subdomain_discovery = True if engine_object.subdomain_discovery else False task.dir_file_search = True if engine_object.dir_file_search else False task.port_scan = True if engine_object.port_scan else False task.fetch_url = True if engine_object.fetch_url else False task.osint = True if engine_object.osint else False task.screenshot = True if engine_object.screenshot else False task.vulnerability_scan = True if engine_object.vulnerability_scan else False task.save() activity_id = create_scan_activity(task, "Scanning Started", 2) results_dir = '/usr/src/scan_results/' os.chdir(results_dir) notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine has initiated recon for target {} with engine type {}'.format(domain.name, engine_object.engine_name)) try: current_scan_dir = domain.name + '_' + str(random.randint(100000000000, 999999999999)) os.mkdir(current_scan_dir) task.results_dir = current_scan_dir task.save() except Exception as exception: logger.error(exception) scan_failed(task) yaml_configuration = None excluded_subdomains = '' try: yaml_configuration = yaml.load( task.scan_type.yaml_configuration, Loader=yaml.FullLoader) except Exception as exception: logger.error(exception) # TODO: Put failed reason on db ''' Add GF patterns name to db for dynamic URLs menu ''' if engine_object.fetch_url and GF_PATTERNS in yaml_configuration[FETCH_URL]: task.used_gf_patterns = ','.join( pattern for pattern in yaml_configuration[FETCH_URL][GF_PATTERNS]) task.save() results_dir = results_dir + current_scan_dir # put all imported subdomains into txt file and also in Subdomain model if imported_subdomains: extract_imported_subdomain( imported_subdomains, task, domain, results_dir) if yaml_configuration: ''' a target in itself is a subdomain, some tool give subdomains as www.yogeshojha.com but url and everything else resolves to yogeshojha.com In that case, we would already need to store target itself as subdomain ''' initial_subdomain_file = '/target_domain.txt' if task.subdomain_discovery else '/sorted_subdomain_collection.txt' subdomain_file = open(results_dir + initial_subdomain_file, "w") subdomain_file.write(domain.name + "\n") subdomain_file.close() if(task.subdomain_discovery): activity_id = create_scan_activity(task, "Subdomain Scanning", 1) subdomain_scan( task, domain, yaml_configuration, results_dir, activity_id, out_of_scope_subdomains ) else: skip_subdomain_scan(task, domain, results_dir) update_last_activity(activity_id, 2) activity_id = create_scan_activity(task, "HTTP Crawler", 1) http_crawler( task, domain, results_dir, activity_id) update_last_activity(activity_id, 2) try: if task.screenshot: activity_id = create_scan_activity( task, "Visual Recon - Screenshot", 1) grab_screenshot( task, domain, yaml_configuration, current_scan_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if(task.port_scan): activity_id = create_scan_activity(task, "Port Scanning", 1) port_scanning(task, domain, yaml_configuration, results_dir) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.osint: activity_id = create_scan_activity(task, "OSINT Running", 1) perform_osint(task, domain, yaml_configuration, results_dir) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.dir_file_search: activity_id = create_scan_activity(task, "Directory Search", 1) directory_brute( task, domain, yaml_configuration, results_dir, activity_id ) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.fetch_url: activity_id = create_scan_activity(task, "Fetching endpoints", 1) fetch_endpoints( task, domain, yaml_configuration, results_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.vulnerability_scan: activity_id = create_scan_activity(task, "Vulnerability Scan", 1) vulnerability_scan( task, domain, yaml_configuration, results_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) activity_id = create_scan_activity(task, "Scan Completed", 2) if notification and notification[0].send_scan_status_notif: send_notification('*Scan Completed*\nreNgine has finished performing recon on target {}.'.format(domain.name)) ''' Once the scan is completed, save the status to successful ''' if ScanActivity.objects.filter(scan_of=task).filter(status=0).all(): task.scan_status = 0 else: task.scan_status = 2 task.stop_scan_date = timezone.now() task.save() # cleanup results delete_scan_data(results_dir) return {"status": True} def skip_subdomain_scan(task, domain, results_dir): # store default target as subdomain ''' If the imported subdomain already has target domain saved, we can skip this ''' if not Subdomain.objects.filter( scan_history=task, name=domain.name).exists(): subdomain_dict = DottedDict({ 'name': domain.name, 'scan_history': task, 'target_domain': domain }) save_subdomain(subdomain_dict) # Save target into target_domain.txt with open('{}/target_domain.txt'.format(results_dir), 'w+') as file: file.write(domain.name + '\n') file.close() ''' We can have two conditions, either subdomain scan happens, or subdomain scan does not happen, in either cases, because we are using import subdomain, we need to collect and sort all the subdomains Write target domain into subdomain_collection ''' os.system( 'cat {0}/target_domain.txt > {0}/subdomain_collection.txt'.format(results_dir)) os.system( 'cat {0}/from_imported.txt > {0}/subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/from_imported.txt'.format(results_dir)) ''' Sort all Subdomains ''' os.system( 'sort -u {0}/subdomain_collection.txt -o {0}/sorted_subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/subdomain_collection.txt'.format(results_dir)) def extract_imported_subdomain(imported_subdomains, task, domain, results_dir): valid_imported_subdomains = [subdomain for subdomain in imported_subdomains if validators.domain( subdomain) and domain.name == get_domain_from_subdomain(subdomain)] # remove any duplicate valid_imported_subdomains = list(set(valid_imported_subdomains)) with open('{}/from_imported.txt'.format(results_dir), 'w+') as file: for subdomain_name in valid_imported_subdomains: # save _subdomain to Subdomain model db if not Subdomain.objects.filter( scan_history=task, name=subdomain_name).exists(): subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': subdomain_name, 'is_imported_subdomain': True }) save_subdomain(subdomain_dict) # save subdomain to file file.write('{}\n'.format(subdomain_name)) file.close() def subdomain_scan(task, domain, yaml_configuration, results_dir, activity_id, out_of_scope_subdomains=None): ''' This function is responsible for performing subdomain enumeration ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Subdomain Gathering for target {} has been started'.format(domain.name)) subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' # check for all the tools and add them into string # if tool selected is all then make string, no need for loop if ALL in yaml_configuration[SUBDOMAIN_DISCOVERY][USES_TOOLS]: tools = 'amass-active amass-passive assetfinder sublist3r subfinder oneforall' else: tools = ' '.join( str(tool) for tool in yaml_configuration[SUBDOMAIN_DISCOVERY][USES_TOOLS]) logging.info(tools) # check for THREADS, by default 10 threads = 10 if THREADS in yaml_configuration[SUBDOMAIN_DISCOVERY]: _threads = yaml_configuration[SUBDOMAIN_DISCOVERY][THREADS] if _threads > 0: threads = _threads if 'amass' in tools: if 'amass-passive' in tools: amass_command = 'amass enum -passive -d {} -o {}/from_amass.txt'.format( domain.name, results_dir) if USE_AMASS_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_AMASS_CONFIG]: amass_command += ' -config /root/.config/amass.ini' # Run Amass Passive logging.info(amass_command) os.system(amass_command) if 'amass-active' in tools: amass_command = 'amass enum -active -d {} -o {}/from_amass_active.txt'.format( domain.name, results_dir) if USE_AMASS_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_AMASS_CONFIG]: amass_command += ' -config /root/.config/amass.ini' if AMASS_WORDLIST in yaml_configuration[SUBDOMAIN_DISCOVERY]: wordlist = yaml_configuration[SUBDOMAIN_DISCOVERY][AMASS_WORDLIST] if wordlist == 'default': wordlist_path = '/usr/src/wordlist/deepmagic.com-prefixes-top50000.txt' else: wordlist_path = '/usr/src/wordlist/' + wordlist + '.txt' if not os.path.exists(wordlist_path): wordlist_path = '/usr/src/' + AMASS_WORDLIST amass_command = amass_command + \ ' -brute -w {}'.format(wordlist_path) if amass_config_path: amass_command = amass_command + \ ' -config {}'.format('/usr/src/scan_results/' + amass_config_path) # Run Amass Active logging.info(amass_command) os.system(amass_command) if 'assetfinder' in tools: assetfinder_command = 'assetfinder --subs-only {} > {}/from_assetfinder.txt'.format( domain.name, results_dir) # Run Assetfinder logging.info(assetfinder_command) os.system(assetfinder_command) if 'sublist3r' in tools: sublist3r_command = 'python3 /usr/src/github/Sublist3r/sublist3r.py -d {} -t {} -o {}/from_sublister.txt'.format( domain.name, threads, results_dir) # Run sublist3r logging.info(sublist3r_command) os.system(sublist3r_command) if 'subfinder' in tools: subfinder_command = 'subfinder -d {} -t {} -o {}/from_subfinder.txt'.format( domain.name, threads, results_dir) if USE_SUBFINDER_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_SUBFINDER_CONFIG]: subfinder_command += ' -config /root/.config/subfinder/config.yaml' # Run Subfinder logging.info(subfinder_command) os.system(subfinder_command) if 'oneforall' in tools: oneforall_command = 'python3 /usr/src/github/OneForAll/oneforall.py --target {} run'.format( domain.name, results_dir) # Run OneForAll logging.info(oneforall_command) os.system(oneforall_command) extract_subdomain = "cut -d',' -f6 /usr/src/github/OneForAll/results/{}.csv >> {}/from_oneforall.txt".format( domain.name, results_dir) os.system(extract_subdomain) # remove the results from oneforall directory os.system( 'rm -rf /usr/src/github/OneForAll/results/{}.*'.format(domain.name)) ''' All tools have gathered the list of subdomains with filename initials as from_* We will gather all the results in one single file, sort them and remove the older results from_* ''' os.system( 'cat {0}/*.txt > {0}/subdomain_collection.txt'.format(results_dir)) ''' Write target domain into subdomain_collection ''' os.system( 'cat {0}/target_domain.txt >> {0}/subdomain_collection.txt'.format(results_dir)) ''' Remove all the from_* files ''' os.system('rm -f {}/from*'.format(results_dir)) ''' Sort all Subdomains ''' os.system( 'sort -u {0}/subdomain_collection.txt -o {0}/sorted_subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/subdomain_collection.txt'.format(results_dir)) ''' The final results will be stored in sorted_subdomain_collection. ''' # parse the subdomain list file and store in db with open(subdomain_scan_results_file) as subdomain_list: for _subdomain in subdomain_list: __subdomain = _subdomain.rstrip('\n') if not Subdomain.objects.filter(scan_history=task, name=__subdomain).exists( ) and validators.domain(__subdomain) and __subdomain not in out_of_scope_subdomains: subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': __subdomain, }) save_subdomain(subdomain_dict) notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: subdomains_count = Subdomain.objects.filter(scan_history=task).count() send_notification('Subdomain Gathering for target {} has been completed and has discovered *{}* subdomains.'.format(domain.name, subdomains_count)) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/sorted_subdomain_collection.txt') # check for any subdomain changes and send notif if any if notification and notification[0].send_subdomain_changes_notif: newly_added_subdomain = get_new_added_subdomain(task.id, domain.id) if newly_added_subdomain: message = "**{} New Subdomains Discovered on domain {}**".format(newly_added_subdomain.count(), domain.name) for subdomain in newly_added_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) removed_subdomain = get_removed_subdomain(task.id, domain.id) if removed_subdomain: message = "**{} Subdomains are no longer available on domain {}**".format(removed_subdomain.count(), domain.name) for subdomain in removed_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) # check for interesting subdomains and send notif if any if notification and notification[0].send_interesting_notif: interesting_subdomain = get_interesting_subdomains(task.id, domain.id) print(interesting_subdomain) if interesting_subdomain: message = "**{} Interesting Subdomains Found on domain {}**".format(interesting_subdomain.count(), domain.name) for subdomain in interesting_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) def get_new_added_subdomain(scan_id, domain_id): scan_history = ScanHistory.objects.filter( domain=domain_id).filter( subdomain_discovery=True).filter( id__lte=scan_id) if scan_history.count() > 1: last_scan = scan_history.order_by('-start_scan_date')[1] scanned_host_q1 = Subdomain.objects.filter( scan_history__id=scan_id).values('name') scanned_host_q2 = Subdomain.objects.filter( scan_history__id=last_scan.id).values('name') added_subdomain = scanned_host_q1.difference(scanned_host_q2) return Subdomain.objects.filter( scan_history=scan_id).filter( name__in=added_subdomain) def get_removed_subdomain(scan_id, domain_id): scan_history = ScanHistory.objects.filter( domain=domain_id).filter( subdomain_discovery=True).filter( id__lte=scan_id) if scan_history.count() > 1: last_scan = scan_history.order_by('-start_scan_date')[1] scanned_host_q1 = Subdomain.objects.filter( scan_history__id=scan_id).values('name') scanned_host_q2 = Subdomain.objects.filter( scan_history__id=last_scan.id).values('name') removed_subdomains = scanned_host_q2.difference(scanned_host_q1) print() return Subdomain.objects.filter( scan_history=last_scan).filter( name__in=removed_subdomains) def http_crawler(task, domain, results_dir, activity_id): ''' This function is runs right after subdomain gathering, and gathers important like page title, http status, etc HTTP Crawler runs by default ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('HTTP Crawler for target {} has been initiated.'.format(domain.name)) alive_file_location = results_dir + '/alive.txt' httpx_results_file = results_dir + '/httpx.json' subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' httpx_command = 'httpx -status-code -content-length -title -tech-detect -cdn -ip -follow-host-redirects -random-agent' proxy = get_random_proxy() if proxy: httpx_command += " --http-proxy '{}'".format(proxy) httpx_command += ' -json -o {}'.format( httpx_results_file ) httpx_command = 'cat {} | {}'.format(subdomain_scan_results_file, httpx_command) print(httpx_command) os.system(httpx_command) # alive subdomains from httpx alive_file = open(alive_file_location, 'w') # writing httpx results if os.path.isfile(httpx_results_file): httpx_json_result = open(httpx_results_file, 'r') lines = httpx_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) try: # fallback for older versions of httpx if 'url' in json_st: subdomain = Subdomain.objects.get( scan_history=task, name=json_st['input']) else: subdomain = Subdomain.objects.get( scan_history=task, name=json_st['url'].split("//")[-1]) ''' Saving Default http urls to EndPoint ''' endpoint = EndPoint() endpoint.scan_history = task endpoint.target_domain = domain endpoint.subdomain = subdomain if 'url' in json_st: endpoint.http_url = json_st['url'] subdomain.http_url = json_st['url'] if 'status-code' in json_st: endpoint.http_status = json_st['status-code'] subdomain.http_status = json_st['status-code'] if 'title' in json_st: endpoint.page_title = json_st['title'] subdomain.page_title = json_st['title'] if 'content-length' in json_st: endpoint.content_length = json_st['content-length'] subdomain.content_length = json_st['content-length'] if 'content-type' in json_st: endpoint.content_type = json_st['content-type'] subdomain.content_type = json_st['content-type'] if 'webserver' in json_st: endpoint.webserver = json_st['webserver'] subdomain.webserver = json_st['webserver'] if 'response-time' in json_st: response_time = float( ''.join( ch for ch in json_st['response-time'] if not ch.isalpha())) if json_st['response-time'][-2:] == 'ms': response_time = response_time / 1000 endpoint.response_time = response_time subdomain.response_time = response_time if 'cnames' in json_st: cname_list = ','.join(json_st['cnames']) subdomain.cname = cname_list discovered_date = timezone.now() endpoint.discovered_date = discovered_date subdomain.discovered_date = discovered_date endpoint.is_default = True endpoint.save() subdomain.save() if 'technologies' in json_st: for _tech in json_st['technologies']: if Technology.objects.filter(name=_tech).exists(): tech = Technology.objects.get(name=_tech) else: tech = Technology(name=_tech) tech.save() subdomain.technologies.add(tech) endpoint.technologies.add(tech) if 'a' in json_st: for _ip in json_st['a']: if IpAddress.objects.filter(address=_ip).exists(): ip = IpAddress.objects.get(address=_ip) else: ip = IpAddress(address=_ip) if 'cdn' in json_st: ip.is_cdn = json_st['cdn'] ip.save() subdomain.ip_addresses.add(ip) # see if to ignore 404 or 5xx alive_file.write(json_st['url'] + '\n') subdomain.save() endpoint.save() except Exception as exception: logging.error(exception) alive_file.close() if notification and notification[0].send_scan_status_notif: alive_count = Subdomain.objects.filter( scan_history__id=task.id).values('name').distinct().filter( http_status__exact=200).count() send_notification('HTTP Crawler for target {} has been completed.\n\n {} subdomains were alive (http status 200).'.format(domain.name, alive_count)) def grab_screenshot(task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for taking screenshots ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine is currently gathering screenshots for {}'.format(domain.name)) output_screenshots_path = results_dir + '/screenshots' result_csv_path = results_dir + '/screenshots/Requests.csv' alive_subdomains_path = results_dir + '/alive.txt' eyewitness_command = 'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py' eyewitness_command += ' -f {} -d {} --no-prompt'.format( alive_subdomains_path, output_screenshots_path ) if EYEWITNESS in yaml_configuration \ and TIMEOUT in yaml_configuration[EYEWITNESS] \ and yaml_configuration[EYEWITNESS][TIMEOUT] > 0: eyewitness_command += ' --timeout {}'.format( yaml_configuration[EYEWITNESS][TIMEOUT] ) if EYEWITNESS in yaml_configuration \ and THREADS in yaml_configuration[EYEWITNESS] \ and yaml_configuration[EYEWITNESS][THREADS] > 0: eyewitness_command += ' --threads {}'.format( yaml_configuration[EYEWITNESS][THREADS] ) logger.info(eyewitness_command) os.system(eyewitness_command) if os.path.isfile(result_csv_path): logger.info('Gathering Eyewitness results') with open(result_csv_path, 'r') as file: reader = csv.reader(file) for row in reader: if row[3] == 'Successful' \ and Subdomain.objects.filter( scan_history__id=task.id).filter(name=row[2]).exists(): subdomain = Subdomain.objects.get( scan_history__id=task.id, name=row[2] ) subdomain.screenshot_path = row[4].replace( '/usr/src/scan_results/', '' ) subdomain.save() # remove all db, html extra files in screenshot results os.system('rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format( output_screenshots_path, )) os.system('rm -rf {0}/source'.format( output_screenshots_path, )) if notification and notification[0].send_scan_status_notif: send_notification('reNgine has finished gathering screenshots for {}'.format(domain.name)) def port_scanning(task, domain, yaml_configuration, results_dir): ''' This function is responsible for running the port scan ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Port Scan initiated for {}'.format(domain.name)) subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' port_results_file = results_dir + '/ports.json' # check the yaml_configuration and choose the ports to be scanned scan_ports = '-' # default port scan everything if PORTS in yaml_configuration[PORT_SCAN]: # TODO: legacy code, remove top-100 in future versions all_ports = yaml_configuration[PORT_SCAN][PORTS] if 'full' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -p {}'.format( subdomain_scan_results_file, port_results_file, '-') elif 'top-100' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -top-ports 100'.format( subdomain_scan_results_file, port_results_file) elif 'top-1000' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -top-ports 1000'.format( subdomain_scan_results_file, port_results_file) else: scan_ports = ','.join( str(port) for port in all_ports) naabu_command = 'cat {} | naabu -json -o {} -p {}'.format( subdomain_scan_results_file, port_results_file, scan_ports) # check for exclude ports if EXCLUDE_PORTS in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][EXCLUDE_PORTS]: exclude_ports = ','.join( str(port) for port in yaml_configuration['port_scan']['exclude_ports']) naabu_command = naabu_command + \ ' -exclude-ports {}'.format(exclude_ports) if NAABU_RATE in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][NAABU_RATE] > 0: naabu_command = naabu_command + \ ' -rate {}'.format( yaml_configuration[PORT_SCAN][NAABU_RATE]) if USE_NAABU_CONFIG in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][USE_NAABU_CONFIG]: naabu_command += ' -config /root/.config/naabu/naabu.conf' # run naabu os.system(naabu_command) # writing port results try: port_json_result = open(port_results_file, 'r') lines = port_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) port_number = json_st['port'] ip_address = json_st['ip'] # see if port already exists if Port.objects.filter(number__exact=port_number).exists(): port = Port.objects.get(number=port_number) else: port = Port() port.number = port_number if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port_detail = whatportis.get_ports(str(port_number)) if len(port_detail): port.service_name = port_detail[0].name port.description = port_detail[0].description port.save() if IpAddress.objects.filter(address=json_st['ip']).exists(): ip = IpAddress.objects.get(address=json_st['ip']) ip.ports.add(port) ip.save() except BaseException as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: port_count = Port.objects.filter( ports__in=IpAddress.objects.filter( ip_addresses__in=Subdomain.objects.filter( scan_history__id=task.id))).distinct().count() send_notification('reNgine has finished Port Scanning on {} and has identified {} ports.'.format(domain.name, port_count)) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/ports.json') def check_waf(): ''' This function will check for the WAF being used in subdomains using wafw00f ''' pass def directory_brute(task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for performing directory scan ''' # scan directories for all the alive subdomain with http status > # 200 notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Directory Bruteforce has been initiated for {}.'.format(domain.name)) alive_subdomains = Subdomain.objects.filter( scan_history__id=task.id).exclude(http_url__isnull=True) dirs_results = results_dir + '/dirs.json' # check the yaml settings if EXTENSIONS in yaml_configuration[DIR_FILE_SEARCH]: extensions = ','.join( str(ext) for ext in yaml_configuration[DIR_FILE_SEARCH][EXTENSIONS]) else: extensions = 'php,git,yaml,conf,db,mysql,bak,txt' # Threads if THREADS in yaml_configuration[DIR_FILE_SEARCH] \ and yaml_configuration[DIR_FILE_SEARCH][THREADS] > 0: threads = yaml_configuration[DIR_FILE_SEARCH][THREADS] else: threads = 10 for subdomain in alive_subdomains: # delete any existing dirs.json if os.path.isfile(dirs_results): os.system('rm -rf {}'.format(dirs_results)) dirsearch_command = 'python3 /usr/src/github/dirsearch/dirsearch.py' dirsearch_command += ' -u {}'.format(subdomain.http_url) if (WORDLIST not in yaml_configuration[DIR_FILE_SEARCH] or not yaml_configuration[DIR_FILE_SEARCH][WORDLIST] or 'default' in yaml_configuration[DIR_FILE_SEARCH][WORDLIST]): wordlist_location = '/usr/src/github/dirsearch/db/dicc.txt' else: wordlist_location = '/usr/src/wordlist/' + \ yaml_configuration[DIR_FILE_SEARCH][WORDLIST] + '.txt' dirsearch_command += ' -w {}'.format(wordlist_location) dirsearch_command += ' --format json -o {}'.format(dirs_results) dirsearch_command += ' -e {}'.format(extensions) dirsearch_command += ' -t {}'.format(threads) dirsearch_command += ' --random-agent --follow-redirects --exclude-status 403,401,404' if EXCLUDE_EXTENSIONS in yaml_configuration[DIR_FILE_SEARCH]: exclude_extensions = ','.join( str(ext) for ext in yaml_configuration[DIR_FILE_SEARCH][EXCLUDE_EXTENSIONS]) dirsearch_command += ' -X {}'.format(exclude_extensions) if EXCLUDE_TEXT in yaml_configuration[DIR_FILE_SEARCH]: exclude_text = ','.join( str(text) for text in yaml_configuration[DIR_FILE_SEARCH][EXCLUDE_TEXT]) dirsearch_command += ' -exclude-texts {}'.format(exclude_text) # check if recursive strategy is set to on if RECURSIVE_LEVEL in yaml_configuration[DIR_FILE_SEARCH]: dirsearch_command += ' --recursion-depth {}'.format(yaml_configuration[DIR_FILE_SEARCH][RECURSIVE_LEVEL]) if RECURSIVE_LEVEL in yaml_configuration[DIR_FILE_SEARCH]: dirsearch_command += ' --recursion-depth {}'.format(yaml_configuration[DIR_FILE_SEARCH][RECURSIVE_LEVEL]) # proxy proxy = get_random_proxy() if proxy: dirsearch_command += " --proxy '{}'".format(proxy) print(dirsearch_command) os.system(dirsearch_command) try: if os.path.isfile(dirs_results): with open(dirs_results, "r") as json_file: json_string = json_file.read() subdomain = Subdomain.objects.get( scan_history__id=task.id, http_url=subdomain.http_url) subdomain.directory_json = json_string subdomain.save() except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: send_notification('Directory Bruteforce has been completed for {}.'.format(domain.name)) def fetch_endpoints( task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for fetching all the urls associated with target and run HTTP probe It first runs gau to gather all urls from wayback, then we will use hakrawler to identify more urls ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine is currently gathering endpoints for {}.'.format(domain.name)) # check yaml settings if ALL in yaml_configuration[FETCH_URL][USES_TOOLS]: tools = 'gauplus hakrawler waybackurls gospider' else: tools = ' '.join( str(tool) for tool in yaml_configuration[FETCH_URL][USES_TOOLS]) if INTENSITY in yaml_configuration[FETCH_URL]: scan_type = yaml_configuration[FETCH_URL][INTENSITY] else: scan_type = 'normal' domain_regex = "\'https?://([a-z0-9]+[.])*{}.*\'".format(domain.name) if 'deep' in scan_type: # performs deep url gathering for all the subdomains present - # RECOMMENDED logger.info('Deep URLS Fetch') os.system(settings.TOOL_LOCATION + 'get_urls.sh %s %s %s %s %s' % ("None", results_dir, scan_type, domain_regex, tools)) else: # perform url gathering only for main domain - USE only for quick scan logger.info('Non Deep URLS Fetch') os.system( settings.TOOL_LOCATION + 'get_urls.sh %s %s %s %s %s' % ( domain.name, results_dir, scan_type, domain_regex, tools )) if IGNORE_FILE_EXTENSION in yaml_configuration[FETCH_URL]: ignore_extension = '|'.join( yaml_configuration[FETCH_URL][IGNORE_FILE_EXTENSION]) logger.info('Ignore extensions' + ignore_extension) os.system( 'cat {0}/all_urls.txt | grep -Eiv "\\.({1}).*" > {0}/temp_urls.txt'.format( results_dir, ignore_extension)) os.system( 'rm {0}/all_urls.txt && mv {0}/temp_urls.txt {0}/all_urls.txt'.format(results_dir)) ''' Store all the endpoints and then run the httpx ''' try: endpoint_final_url = results_dir + '/all_urls.txt' if os.path.isfile(endpoint_final_url): with open(endpoint_final_url) as endpoint_list: for url in endpoint_list: http_url = url.rstrip('\n') if not EndPoint.objects.filter(scan_history=task, http_url=http_url).exists(): _subdomain = get_subdomain_from_url(http_url) if Subdomain.objects.filter( scan_history=task).filter( name=_subdomain).exists(): subdomain = Subdomain.objects.get( scan_history=task, name=_subdomain) else: ''' gau or gosppider can gather interesting endpoints which when parsed can give subdomains that were not existent from subdomain scan. so storing them ''' logger.error( 'Subdomain {} not found, adding...'.format(_subdomain)) subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': _subdomain, }) subdomain = save_subdomain(subdomain_dict) endpoint_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'subdomain': subdomain, 'http_url': http_url, }) save_endpoint(endpoint_dict) except Exception as e: logger.error(e) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/all_urls.txt') ''' TODO: Go spider & waybackurls accumulates a lot of urls, which is good but nuclei takes forever to scan even a simple website, so we will do http probing and filter HTTP status 404, this way we can reduce the number of Non Existent URLS ''' logger.info('HTTP Probing on collected endpoints') httpx_command = 'httpx -l {0}/all_urls.txt -status-code -content-length -ip -cdn -title -tech-detect -json -follow-redirects -random-agent -o {0}/final_httpx_urls.json'.format(results_dir) proxy = get_random_proxy() if proxy: httpx_command += " --http-proxy '{}'".format(proxy) os.system(httpx_command) url_results_file = results_dir + '/final_httpx_urls.json' try: urls_json_result = open(url_results_file, 'r') lines = urls_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) http_url = json_st['url'] _subdomain = get_subdomain_from_url(http_url) if Subdomain.objects.filter( scan_history=task).filter( name=_subdomain).exists(): subdomain_obj = Subdomain.objects.get( scan_history=task, name=_subdomain) else: subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': _subdomain, }) subdomain_obj = save_subdomain(subdomain_dict) if EndPoint.objects.filter( scan_history=task).filter( http_url=http_url).exists(): endpoint = EndPoint.objects.get( scan_history=task, http_url=http_url) else: endpoint = EndPoint() endpoint_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'http_url': http_url, 'subdomain': subdomain_obj }) endpoint = save_endpoint(endpoint_dict) if 'title' in json_st: endpoint.page_title = json_st['title'] if 'webserver' in json_st: endpoint.webserver = json_st['webserver'] if 'content-length' in json_st: endpoint.content_length = json_st['content-length'] if 'content-type' in json_st: endpoint.content_type = json_st['content-type'] if 'status-code' in json_st: endpoint.http_status = json_st['status-code'] if 'response-time' in json_st: response_time = float(''.join(ch for ch in json_st['response-time'] if not ch.isalpha())) if json_st['response-time'][-2:] == 'ms': response_time = response_time / 1000 endpoint.response_time = response_time endpoint.save() if 'technologies' in json_st: for _tech in json_st['technologies']: if Technology.objects.filter(name=_tech).exists(): tech = Technology.objects.get(name=_tech) else: tech = Technology(name=_tech) tech.save() endpoint.technologies.add(tech) # get subdomain object subdomain = Subdomain.objects.get(scan_history=task, name=_subdomain) subdomain.technologies.add(tech) subdomain.save() except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: endpoint_count = EndPoint.objects.filter( scan_history__id=task.id).values('http_url').distinct().count() endpoint_alive_count = EndPoint.objects.filter( scan_history__id=task.id, http_status__exact=200).values('http_url').distinct().count() send_notification('reNgine has finished gathering endpoints for {} and has discovered *{}* unique endpoints.\n\n{} of those endpoints reported HTTP status 200.'.format( domain.name, endpoint_count, endpoint_alive_count )) # once endpoint is saved, run gf patterns TODO: run threads if GF_PATTERNS in yaml_configuration[FETCH_URL]: for pattern in yaml_configuration[FETCH_URL][GF_PATTERNS]: logger.info('Running GF for {}'.format(pattern)) gf_output_file_path = '{0}/gf_patterns_{1}.txt'.format( results_dir, pattern) gf_command = 'cat {0}/all_urls.txt | gf {1} >> {2}'.format( results_dir, pattern, gf_output_file_path) os.system(gf_command) if os.path.exists(gf_output_file_path): with open(gf_output_file_path) as gf_output: for line in gf_output: url = line.rstrip('\n') try: endpoint = EndPoint.objects.get( scan_history=task, http_url=url) earlier_pattern = endpoint.matched_gf_patterns new_pattern = earlier_pattern + ',' + pattern if earlier_pattern else pattern endpoint.matched_gf_patterns = new_pattern except Exception as e: # add the url in db logger.error(e) logger.info('Adding URL' + url) endpoint = EndPoint() endpoint.http_url = url endpoint.target_domain = domain endpoint.scan_history = task try: _subdomain = Subdomain.objects.get( scan_history=task, name=get_subdomain_from_url(url)) endpoint.subdomain = _subdomain except Exception as e: continue endpoint.matched_gf_patterns = pattern finally: endpoint.save() def vulnerability_scan( task, domain, yaml_configuration, results_dir, activity_id): notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Vulnerability scan has been initiated for {}.'.format(domain.name)) ''' This function will run nuclei as a vulnerability scanner ---- unfurl the urls to keep only domain and path, this will be sent to vuln scan ignore certain file extensions Thanks: https://github.com/six2dez/reconftw ''' urls_path = '/alive.txt' if task.scan_type.fetch_url: os.system('cat {0}/all_urls.txt | grep -Eiv "\\.(eot|jpg|jpeg|gif|css|tif|tiff|png|ttf|otf|woff|woff2|ico|pdf|svg|txt|js|doc|docx)$" | unfurl -u format %s://%d%p >> {0}/unfurl_urls.txt'.format(results_dir)) os.system( 'sort -u {0}/unfurl_urls.txt -o {0}/unfurl_urls.txt'.format(results_dir)) urls_path = '/unfurl_urls.txt' vulnerability_result_path = results_dir + '/vulnerability.json' vulnerability_scan_input_file = results_dir + urls_path nuclei_command = 'nuclei -json -l {} -o {}'.format( vulnerability_scan_input_file, vulnerability_result_path) # check nuclei config if USE_NUCLEI_CONFIG in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[VULNERABILITY_SCAN][USE_NUCLEI_CONFIG]: nuclei_command += ' -config /root/.config/nuclei/config.yaml' ''' Nuclei Templates Either custom template has to be supplied or default template, if neither has been supplied then use all templates including custom templates ''' if CUSTOM_NUCLEI_TEMPLATE in yaml_configuration[ VULNERABILITY_SCAN] or NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: # check yaml settings for templates if NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: if ALL in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_TEMPLATE]: template = NUCLEI_TEMPLATES_PATH else: _template = ','.join([NUCLEI_TEMPLATES_PATH + str(element) for element in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_TEMPLATE]]) template = _template.replace(',', ' -t ') # Update nuclei command with templates nuclei_command = nuclei_command + ' -t ' + template if CUSTOM_NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: # add .yaml to the custom template extensions _template = ','.join( [str(element) + '.yaml' for element in yaml_configuration[VULNERABILITY_SCAN][CUSTOM_NUCLEI_TEMPLATE]]) template = _template.replace(',', ' -t ') # Update nuclei command with templates nuclei_command = nuclei_command + ' -t ' + template else: nuclei_command = nuclei_command + ' -t /root/nuclei-templates' # check yaml settings for concurrency if NUCLEI_CONCURRENCY in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][NUCLEI_CONCURRENCY] > 0: concurrency = yaml_configuration[VULNERABILITY_SCAN][NUCLEI_CONCURRENCY] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -c ' + str(concurrency) if RATE_LIMIT in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][RATE_LIMIT] > 0: rate_limit = yaml_configuration[VULNERABILITY_SCAN][RATE_LIMIT] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -rl ' + str(rate_limit) if TIMEOUT in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][TIMEOUT] > 0: timeout = yaml_configuration[VULNERABILITY_SCAN][TIMEOUT] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -timeout ' + str(timeout) if RETRIES in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][RETRIES] > 0: retries = yaml_configuration[VULNERABILITY_SCAN][RETRIES] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -retries ' + str(retries) # for severity if NUCLEI_SEVERITY in yaml_configuration[VULNERABILITY_SCAN] and ALL not in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_SEVERITY]: _severity = ','.join( [str(element) for element in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_SEVERITY]]) severity = _severity.replace(" ", "") else: severity = "critical, high, medium, low, info" # update nuclei templates before running scan os.system('nuclei -update-templates') for _severity in severity.split(","): # delete any existing vulnerability.json file if os.path.isfile(vulnerability_result_path): os.system('rm {}'.format(vulnerability_result_path)) # run nuclei final_nuclei_command = nuclei_command + ' -severity ' + _severity proxy = get_random_proxy() if proxy: final_nuclei_command += " --proxy-url '{}'".format(proxy) logger.info(final_nuclei_command) os.system(final_nuclei_command) try: if os.path.isfile(vulnerability_result_path): urls_json_result = open(vulnerability_result_path, 'r') lines = urls_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) host = json_st['host'] _subdomain = get_subdomain_from_url(host) try: subdomain = Subdomain.objects.get( name=_subdomain, scan_history=task) vulnerability = Vulnerability() vulnerability.subdomain = subdomain vulnerability.scan_history = task vulnerability.target_domain = domain try: endpoint = EndPoint.objects.get( scan_history=task, target_domain=domain, http_url=host) vulnerability.endpoint = endpoint except Exception as exception: logger.error(exception) if 'name' in json_st['info']: vulnerability.name = json_st['info']['name'] if 'severity' in json_st['info']: if json_st['info']['severity'] == 'info': severity = 0 elif json_st['info']['severity'] == 'low': severity = 1 elif json_st['info']['severity'] == 'medium': severity = 2 elif json_st['info']['severity'] == 'high': severity = 3 elif json_st['info']['severity'] == 'critical': severity = 4 else: severity = 0 else: severity = 0 vulnerability.severity = severity if 'tags' in json_st['info']: vulnerability.tags = json_st['info']['tags'] if 'description' in json_st['info']: vulnerability.description = json_st['info']['description'] if 'reference' in json_st['info']: vulnerability.reference = json_st['info']['reference'] if 'matched' in json_st: vulnerability.http_url = json_st['matched'] if 'templateID' in json_st: vulnerability.template_used = json_st['templateID'] if 'description' in json_st: vulnerability.description = json_st['description'] if 'matcher_name' in json_st: vulnerability.matcher_name = json_st['matcher_name'] if 'extracted_results' in json_st: vulnerability.extracted_results = json_st['extracted_results'] vulnerability.discovered_date = timezone.now() vulnerability.open_status = True vulnerability.save() # send notification for all vulnerabilities except info if json_st['info']['severity'] != "info" and notification and notification[0].send_vuln_notif: message = "*Alert: Vulnerability Identified*" message += "\n\n" message += "A *{}* severity vulnerability has been identified.".format(json_st['info']['severity']) message += "\nVulnerability Name: {}".format(json_st['info']['name']) message += "\nVulnerable URL: {}".format(json_st['host']) send_notification(message) # send report to hackerone if Hackerone.objects.all().exists() and json_st['info']['severity'] != 'info' and json_st['info']['severity'] \ != 'low' and vulnerability.target_domain.h1_team_handle: hackerone = Hackerone.objects.all()[0] if hackerone.send_critical and json_st['info']['severity'] == 'critical': send_hackerone_report(vulnerability.id) elif hackerone.send_high and json_st['info']['severity'] == 'high': send_hackerone_report(vulnerability.id) elif hackerone.send_medium and json_st['info']['severity'] == 'medium': send_hackerone_report(vulnerability.id) except ObjectDoesNotExist: logger.error('Object not found') continue except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: info_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=0).count() low_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=1).count() medium_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=2).count() high_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=3).count() critical_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=4).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count message = 'Vulnerability scan has been completed for {} and discovered {} vulnerabilities.'.format( domain.name, vulnerability_count ) message += '\n\n*Vulnerability Stats:*' message += '\nCritical: {}'.format(critical_count) message += '\nHigh: {}'.format(high_count) message += '\nMedium: {}'.format(medium_count) message += '\nLow: {}'.format(low_count) message += '\nInfo: {}'.format(info_count) send_notification(message) def scan_failed(task): task.scan_status = 0 task.stop_scan_date = timezone.now() task.save() def create_scan_activity(task, message, status): scan_activity = ScanActivity() scan_activity.scan_of = task scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id def update_last_activity(id, activity_status): ScanActivity.objects.filter( id=id).update( status=activity_status, time=timezone.now()) def delete_scan_data(results_dir): # remove all txt,html,json files os.system('find {} -name "*.txt" -type f -delete'.format(results_dir)) os.system('find {} -name "*.html" -type f -delete'.format(results_dir)) os.system('find {} -name "*.json" -type f -delete'.format(results_dir)) def save_subdomain(subdomain_dict): subdomain = Subdomain() subdomain.discovered_date = timezone.now() subdomain.target_domain = subdomain_dict.get('target_domain') subdomain.scan_history = subdomain_dict.get('scan_history') subdomain.name = subdomain_dict.get('name') subdomain.http_url = subdomain_dict.get('http_url') subdomain.screenshot_path = subdomain_dict.get('screenshot_path') subdomain.http_header_path = subdomain_dict.get('http_header_path') subdomain.cname = subdomain_dict.get('cname') subdomain.is_cdn = subdomain_dict.get('is_cdn') subdomain.content_type = subdomain_dict.get('content_type') subdomain.webserver = subdomain_dict.get('webserver') subdomain.page_title = subdomain_dict.get('page_title') subdomain.is_imported_subdomain = subdomain_dict.get( 'is_imported_subdomain') if 'is_imported_subdomain' in subdomain_dict else False if 'http_status' in subdomain_dict: subdomain.http_status = subdomain_dict.get('http_status') if 'response_time' in subdomain_dict: subdomain.response_time = subdomain_dict.get('response_time') if 'content_length' in subdomain_dict: subdomain.content_length = subdomain_dict.get('content_length') subdomain.save() return subdomain def save_endpoint(endpoint_dict): endpoint = EndPoint() endpoint.discovered_date = timezone.now() endpoint.scan_history = endpoint_dict.get('scan_history') endpoint.target_domain = endpoint_dict.get('target_domain') if 'target_domain' in endpoint_dict else None endpoint.subdomain = endpoint_dict.get('subdomain') if 'target_domain' in endpoint_dict else None endpoint.http_url = endpoint_dict.get('http_url') endpoint.page_title = endpoint_dict.get('page_title') if 'page_title' in endpoint_dict else None endpoint.content_type = endpoint_dict.get('content_type') if 'content_type' in endpoint_dict else None endpoint.webserver = endpoint_dict.get('webserver') if 'webserver' in endpoint_dict else None endpoint.response_time = endpoint_dict.get('response_time') if 'response_time' in endpoint_dict else 0 endpoint.http_status = endpoint_dict.get('http_status') if 'http_status' in endpoint_dict else 0 endpoint.content_length = endpoint_dict.get('content_length') if 'content_length' in endpoint_dict else 0 endpoint.is_default = endpoint_dict.get('is_default') if 'is_default' in endpoint_dict else False endpoint.save() return endpoint def perform_osint(task, domain, yaml_configuration, results_dir): notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine has initiated OSINT on target {}'.format(domain.name)) if 'discover' in yaml_configuration[OSINT]: osint_discovery(task, domain, yaml_configuration, results_dir) if 'dork' in yaml_configuration[OSINT]: dorking(task, yaml_configuration) if notification and notification[0].send_scan_status_notif: send_notification('reNgine has completed performing OSINT on target {}'.format(domain.name)) def osint_discovery(task, domain, yaml_configuration, results_dir): if ALL in yaml_configuration[OSINT][OSINT_DISCOVER]: osint_lookup = 'emails metainfo employees' else: osint_lookup = ' '.join( str(lookup) for lookup in yaml_configuration[OSINT][OSINT_DISCOVER]) if 'metainfo' in osint_lookup: if INTENSITY in yaml_configuration[OSINT]: osint_intensity = yaml_configuration[OSINT][INTENSITY] else: osint_intensity = 'normal' if OSINT_DOCUMENTS_LIMIT in yaml_configuration[OSINT]: documents_limit = yaml_configuration[OSINT][OSINT_DOCUMENTS_LIMIT] else: documents_limit = 50 if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': domain.name, 'domain': domain, 'scan_id': task, 'documents_limit': documents_limit }) get_and_save_meta_info(meta_dict) elif osint_intensity == 'deep': # get all subdomains in scan_id subdomains = Subdomain.objects.filter(scan_history=task) for subdomain in subdomains: meta_dict = DottedDict({ 'osint_target': subdomain.name, 'domain': domain, 'scan_id': task, 'documents_limit': documents_limit }) get_and_save_meta_info(meta_dict) if 'emails' in osint_lookup: get_and_save_emails(task, results_dir) get_and_save_leaked_credentials(task, results_dir) if 'employees' in osint_lookup: get_and_save_employees(task, results_dir) def dorking(scan_history, yaml_configuration): # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh # look in stackoverflow if ALL in yaml_configuration[OSINT][OSINT_DORK]: dork_lookup = 'stackoverflow, 3rdparty, social_media, project_management, code_sharing, config_files, jenkins, cloud_buckets, php_error, exposed_documents, struts_rce, db_files, traefik, git_exposed' else: dork_lookup = ' '.join( str(lookup) for lookup in yaml_configuration[OSINT][OSINT_DORK]) if 'stackoverflow' in dork_lookup: dork = 'site:stackoverflow.com' dork_type = 'stackoverflow' get_and_save_dork_results( dork, dork_type, scan_history, in_target=False ) if '3rdparty' in dork_lookup: # look in 3rd party sitee dork_type = '3rdparty' lookup_websites = [ 'gitter.im', 'papaly.com', 'productforums.google.com', 'coggle.it', 'replt.it', 'ycombinator.com', 'libraries.io', 'npm.runkit.com', 'npmjs.com', 'scribd.com', 'gitter.im' ] dork = '' for website in lookup_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'social_media' in dork_lookup: dork_type = 'Social Media' social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'pinterest.com', 'tumblr.com', 'reddit.com' ] dork = '' for website in social_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'project_management' in dork_lookup: dork_type = 'Project Management' project_websites = [ 'trello.com', '*.atlassian.net' ] dork = '' for website in project_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'code_sharing' in dork_lookup: dork_type = 'Code Sharing Sites' code_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] dork = '' for website in code_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'config_files' in dork_lookup: dork_type = 'Config Files' config_file_ext = [ 'env', 'xml', 'conf', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] dork = '' for extension in config_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'jenkins' in dork_lookup: dork_type = 'Jenkins' dork = 'intitle:\"Dashboard [Jenkins]\"' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) if 'wordpress_files' in dork_lookup: dork_type = 'Wordpress Files' inurl_lookup = [ 'wp-content', 'wp-includes' ] dork = '' for lookup in inurl_lookup: dork = dork + ' | ' + 'inurl:' + lookup get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'cloud_buckets' in dork_lookup: dork_type = 'Cloud Buckets' cloud_websites = [ '.s3.amazonaws.com', 'storage.googleapis.com', 'amazonaws.com' ] dork = '' for website in cloud_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'php_error' in dork_lookup: dork_type = 'PHP Error' error_words = [ '\"PHP Parse error\"', '\"PHP Warning\"', '\"PHP Error\"' ] dork = '' for word in error_words: dork = dork + ' | ' + word get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'exposed_documents' in dork_lookup: dork_type = 'Exposed Documents' docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] dork = '' for extension in docs_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'struts_rce' in dork_lookup: dork_type = 'Apache Struts RCE' struts_file_ext = [ 'action', 'struts', 'do' ] dork = '' for extension in struts_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'db_files' in dork_lookup: dork_type = 'Database Files' db_file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] dork = '' for extension in db_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'traefik' in dork_lookup: dork = 'intitle:traefik inurl:8080/dashboard' dork_type = 'Traefik' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) if 'git_exposed' in dork_lookup: dork = 'inurl:\"/.git\"' dork_type = '.git Exposed' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) def get_and_save_dork_results(dork, type, scan_history, in_target=False): degoogle_obj = degoogle.dg() proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy if in_target: query = dork + " site:" + scan_history.domain.name else: query = dork + " \"{}\"".format(scan_history.domain.name) logger.info(query) degoogle_obj.query = query results = degoogle_obj.run() logger.info(results) for result in results: dork, _ = Dork.objects.get_or_create( type=type, description=result['desc'], url=result['url'] ) scan_history.dorks.add(dork) def get_and_save_employees(scan_history, results_dir): theHarvester_location = '/usr/src/github/theHarvester' # update proxies.yaml if Proxy.objects.all().exists(): proxy = Proxy.objects.all()[0] if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(theHarvester_location + '/proxies.yaml', 'w') as file: documents = yaml.dump(yaml_data, file) os.system('cd {} && python3 theHarvester.py -d {} -b all -f {}/theHarvester.html'.format( theHarvester_location, scan_history.domain.name, results_dir )) file_location = results_dir + '/theHarvester.html' print(file_location) # delete proxy environ var if os.environ.get(('https_proxy')): del os.environ['https_proxy'] if os.environ.get(('HTTPS_PROXY')): del os.environ['HTTPS_PROXY'] if os.path.isfile(file_location): logger.info('Parsing theHarvester results') options = FirefoxOptions() options.add_argument("--headless") driver = webdriver.Firefox(options=options) driver.get('file://'+file_location) tabledata = driver.execute_script('return tabledata') # save email addresses and linkedin employees for data in tabledata: if data['record'] == 'email': _email = data['result'] email, _ = Email.objects.get_or_create(address=_email) scan_history.emails.add(email) elif data['record'] == 'people': _employee = data['result'] split_val = _employee.split('-') name = split_val[0] if len(split_val) == 2: designation = split_val[1] else: designation = "" employee, _ = Employee.objects.get_or_create(name=name, designation=designation) scan_history.employees.add(employee) driver.quit() print(tabledata) def get_and_save_emails(scan_history, results_dir): leak_target_path = '{}/creds_target.txt'.format(results_dir) # get email address proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy emails = [] try: logger.info('OSINT: Getting emails from Google') email_from_google = get_emails_from_google(scan_history.domain.name) logger.info('OSINT: Getting emails from Bing') email_from_bing = get_emails_from_bing(scan_history.domain.name) logger.info('OSINT: Getting emails from Baidu') email_from_baidu = get_emails_from_baidu(scan_history.domain.name) emails = list(set(email_from_google + email_from_bing + email_from_baidu)) logger.info(emails) except Exception as e: logger.error(e) leak_target_file = open(leak_target_path, 'w') for _email in emails: email, _ = Email.objects.get_or_create(address=_email) scan_history.emails.add(email) leak_target_file.write('{}\n'.format(_email)) # fill leak_target_file with possible email address leak_target_file.write('%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.write('%.%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%.%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.write('%_%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%_%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.close() def get_and_save_leaked_credentials(scan_history, results_dir): logger.info('OSINT: Getting leaked credentials...') leak_target_file = '{}/creds_target.txt'.format(results_dir) leak_output_file = '{}/pwndb.json'.format(results_dir) pwndb_command = 'python3 /usr/src/github/pwndb/pwndb.py --proxy tor:9150 --output json --list {}'.format( leak_target_file ) try: pwndb_output = subprocess.getoutput(pwndb_command) creds = json.loads(pwndb_output) for cred in creds: if cred['username'] != 'donate': email_id = "{}@{}".format(cred['username'], cred['domain']) email_obj, _ = Email.objects.get_or_create( address=email_id, ) email_obj.password = cred['password'] email_obj.save() scan_history.emails.add(email_obj) except Exception as e: logger.error(e) pass def get_and_save_meta_info(meta_dict): logger.info('Getting METADATA for {}'.format(meta_dict.osint_target)) proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy result = metadata_extractor.extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if result: results = result.get_metadata() for meta in results: meta_finder_document = MetaFinderDocument() subdomain = Subdomain.objects.get(scan_history=meta_dict.scan_id, name=meta_dict.osint_target) meta_finder_document.subdomain = subdomain meta_finder_document.target_domain = meta_dict.domain meta_finder_document.scan_history = meta_dict.scan_id item = DottedDict(results[meta]) meta_finder_document.url = item.url meta_finder_document.doc_name = meta meta_finder_document.http_status = item.status_code metadata = results[meta]['metadata'] for data in metadata: if 'Producer' in metadata and metadata['Producer']: meta_finder_document.producer = metadata['Producer'].rstrip('\x00') if 'Creator' in metadata and metadata['Creator']: meta_finder_document.creator = metadata['Creator'].rstrip('\x00') if 'CreationDate' in metadata and metadata['CreationDate']: meta_finder_document.creation_date = metadata['CreationDate'].rstrip('\x00') if 'ModDate' in metadata and metadata['ModDate']: meta_finder_document.modified_date = metadata['ModDate'].rstrip('\x00') if 'Author' in metadata and metadata['Author']: meta_finder_document.author = metadata['Author'].rstrip('\x00') if 'Title' in metadata and metadata['Title']: meta_finder_document.title = metadata['Title'].rstrip('\x00') if 'OSInfo' in metadata and metadata['OSInfo']: meta_finder_document.os = metadata['OSInfo'].rstrip('\x00') meta_finder_document.save() @app.task(bind=True) def test_task(self): print('*' * 40) print('test task run') print('*' * 40)

import os import traceback import yaml import json import csv import validators import random import requests import logging import metafinder.extractor as metadata_extractor import whatportis import subprocess from selenium.webdriver.firefox.options import Options as FirefoxOptions from selenium import webdriver from emailfinder.extractor import * from dotted_dict import DottedDict from celery import shared_task from discord_webhook import DiscordWebhook from reNgine.celery import app from startScan.models import * from targetApp.models import Domain from scanEngine.models import EngineType from django.conf import settings from django.shortcuts import get_object_or_404 from celery import shared_task from datetime import datetime from degoogle import degoogle from django.conf import settings from django.utils import timezone, dateformat from django.shortcuts import get_object_or_404 from django.core.exceptions import ObjectDoesNotExist from reNgine.celery import app from reNgine.definitions import * from startScan.models import * from targetApp.models import Domain from scanEngine.models import EngineType, Configuration, Wordlist from .common_func import * ''' task for background scan ''' @app.task def initiate_scan( domain_id, scan_history_id, scan_type, engine_type, imported_subdomains=None, out_of_scope_subdomains=[] ): ''' scan_type = 0 -> immediate scan, need not create scan object scan_type = 1 -> scheduled scan ''' engine_object = EngineType.objects.get(pk=engine_type) domain = Domain.objects.get(pk=domain_id) if scan_type == 1: task = ScanHistory() task.scan_status = -1 elif scan_type == 0: task = ScanHistory.objects.get(pk=scan_history_id) # save the last scan date for domain model domain.last_scan_date = timezone.now() domain.save() # once the celery task starts, change the task status to Started task.scan_type = engine_object task.celery_id = initiate_scan.request.id task.domain = domain task.scan_status = 1 task.start_scan_date = timezone.now() task.subdomain_discovery = True if engine_object.subdomain_discovery else False task.dir_file_search = True if engine_object.dir_file_search else False task.port_scan = True if engine_object.port_scan else False task.fetch_url = True if engine_object.fetch_url else False task.osint = True if engine_object.osint else False task.screenshot = True if engine_object.screenshot else False task.vulnerability_scan = True if engine_object.vulnerability_scan else False task.save() activity_id = create_scan_activity(task, "Scanning Started", 2) results_dir = '/usr/src/scan_results/' os.chdir(results_dir) notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine has initiated recon for target {} with engine type {}'.format(domain.name, engine_object.engine_name)) try: current_scan_dir = domain.name + '_' + str(random.randint(100000000000, 999999999999)) os.mkdir(current_scan_dir) task.results_dir = current_scan_dir task.save() except Exception as exception: logger.error(exception) scan_failed(task) yaml_configuration = None excluded_subdomains = '' try: yaml_configuration = yaml.load( task.scan_type.yaml_configuration, Loader=yaml.FullLoader) except Exception as exception: logger.error(exception) # TODO: Put failed reason on db ''' Add GF patterns name to db for dynamic URLs menu ''' if engine_object.fetch_url and GF_PATTERNS in yaml_configuration[FETCH_URL]: task.used_gf_patterns = ','.join( pattern for pattern in yaml_configuration[FETCH_URL][GF_PATTERNS]) task.save() results_dir = results_dir + current_scan_dir # put all imported subdomains into txt file and also in Subdomain model if imported_subdomains: extract_imported_subdomain( imported_subdomains, task, domain, results_dir) if yaml_configuration: ''' a target in itself is a subdomain, some tool give subdomains as www.yogeshojha.com but url and everything else resolves to yogeshojha.com In that case, we would already need to store target itself as subdomain ''' initial_subdomain_file = '/target_domain.txt' if task.subdomain_discovery else '/sorted_subdomain_collection.txt' subdomain_file = open(results_dir + initial_subdomain_file, "w") subdomain_file.write(domain.name + "\n") subdomain_file.close() if(task.subdomain_discovery): activity_id = create_scan_activity(task, "Subdomain Scanning", 1) subdomain_scan( task, domain, yaml_configuration, results_dir, activity_id, out_of_scope_subdomains ) else: skip_subdomain_scan(task, domain, results_dir) update_last_activity(activity_id, 2) activity_id = create_scan_activity(task, "HTTP Crawler", 1) http_crawler( task, domain, results_dir, activity_id) update_last_activity(activity_id, 2) try: if task.screenshot: activity_id = create_scan_activity( task, "Visual Recon - Screenshot", 1) grab_screenshot( task, domain, yaml_configuration, current_scan_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if(task.port_scan): activity_id = create_scan_activity(task, "Port Scanning", 1) port_scanning(task, domain, yaml_configuration, results_dir) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.osint: activity_id = create_scan_activity(task, "OSINT Running", 1) perform_osint(task, domain, yaml_configuration, results_dir) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.dir_file_search: activity_id = create_scan_activity(task, "Directory Search", 1) directory_brute( task, domain, yaml_configuration, results_dir, activity_id ) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.fetch_url: activity_id = create_scan_activity(task, "Fetching endpoints", 1) fetch_endpoints( task, domain, yaml_configuration, results_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.vulnerability_scan: activity_id = create_scan_activity(task, "Vulnerability Scan", 1) vulnerability_scan( task, domain, yaml_configuration, results_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) activity_id = create_scan_activity(task, "Scan Completed", 2) if notification and notification[0].send_scan_status_notif: send_notification('*Scan Completed*\nreNgine has finished performing recon on target {}.'.format(domain.name)) ''' Once the scan is completed, save the status to successful ''' if ScanActivity.objects.filter(scan_of=task).filter(status=0).all(): task.scan_status = 0 else: task.scan_status = 2 task.stop_scan_date = timezone.now() task.save() # cleanup results delete_scan_data(results_dir) return {"status": True} def skip_subdomain_scan(task, domain, results_dir): # store default target as subdomain ''' If the imported subdomain already has target domain saved, we can skip this ''' if not Subdomain.objects.filter( scan_history=task, name=domain.name).exists(): subdomain_dict = DottedDict({ 'name': domain.name, 'scan_history': task, 'target_domain': domain }) save_subdomain(subdomain_dict) # Save target into target_domain.txt with open('{}/target_domain.txt'.format(results_dir), 'w+') as file: file.write(domain.name + '\n') file.close() ''' We can have two conditions, either subdomain scan happens, or subdomain scan does not happen, in either cases, because we are using import subdomain, we need to collect and sort all the subdomains Write target domain into subdomain_collection ''' os.system( 'cat {0}/target_domain.txt > {0}/subdomain_collection.txt'.format(results_dir)) os.system( 'cat {0}/from_imported.txt > {0}/subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/from_imported.txt'.format(results_dir)) ''' Sort all Subdomains ''' os.system( 'sort -u {0}/subdomain_collection.txt -o {0}/sorted_subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/subdomain_collection.txt'.format(results_dir)) def extract_imported_subdomain(imported_subdomains, task, domain, results_dir): valid_imported_subdomains = [subdomain for subdomain in imported_subdomains if validators.domain( subdomain) and domain.name == get_domain_from_subdomain(subdomain)] # remove any duplicate valid_imported_subdomains = list(set(valid_imported_subdomains)) with open('{}/from_imported.txt'.format(results_dir), 'w+') as file: for subdomain_name in valid_imported_subdomains: # save _subdomain to Subdomain model db if not Subdomain.objects.filter( scan_history=task, name=subdomain_name).exists(): subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': subdomain_name, 'is_imported_subdomain': True }) save_subdomain(subdomain_dict) # save subdomain to file file.write('{}\n'.format(subdomain_name)) file.close() def subdomain_scan(task, domain, yaml_configuration, results_dir, activity_id, out_of_scope_subdomains=None): ''' This function is responsible for performing subdomain enumeration ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Subdomain Gathering for target {} has been started'.format(domain.name)) subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' # check for all the tools and add them into string # if tool selected is all then make string, no need for loop if ALL in yaml_configuration[SUBDOMAIN_DISCOVERY][USES_TOOLS]: tools = 'amass-active amass-passive assetfinder sublist3r subfinder oneforall' else: tools = ' '.join( str(tool) for tool in yaml_configuration[SUBDOMAIN_DISCOVERY][USES_TOOLS]) logging.info(tools) # check for THREADS, by default 10 threads = 10 if THREADS in yaml_configuration[SUBDOMAIN_DISCOVERY]: _threads = yaml_configuration[SUBDOMAIN_DISCOVERY][THREADS] if _threads > 0: threads = _threads if 'amass' in tools: if 'amass-passive' in tools: amass_command = 'amass enum -passive -d {} -o {}/from_amass.txt'.format( domain.name, results_dir) if USE_AMASS_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_AMASS_CONFIG]: amass_command += ' -config /root/.config/amass.ini' # Run Amass Passive logging.info(amass_command) os.system(amass_command) if 'amass-active' in tools: amass_command = 'amass enum -active -d {} -o {}/from_amass_active.txt'.format( domain.name, results_dir) if USE_AMASS_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_AMASS_CONFIG]: amass_command += ' -config /root/.config/amass.ini' if AMASS_WORDLIST in yaml_configuration[SUBDOMAIN_DISCOVERY]: wordlist = yaml_configuration[SUBDOMAIN_DISCOVERY][AMASS_WORDLIST] if wordlist == 'default': wordlist_path = '/usr/src/wordlist/deepmagic.com-prefixes-top50000.txt' else: wordlist_path = '/usr/src/wordlist/' + wordlist + '.txt' if not os.path.exists(wordlist_path): wordlist_path = '/usr/src/' + AMASS_WORDLIST amass_command = amass_command + \ ' -brute -w {}'.format(wordlist_path) if amass_config_path: amass_command = amass_command + \ ' -config {}'.format('/usr/src/scan_results/' + amass_config_path) # Run Amass Active logging.info(amass_command) os.system(amass_command) if 'assetfinder' in tools: assetfinder_command = 'assetfinder --subs-only {} > {}/from_assetfinder.txt'.format( domain.name, results_dir) # Run Assetfinder logging.info(assetfinder_command) os.system(assetfinder_command) if 'sublist3r' in tools: sublist3r_command = 'python3 /usr/src/github/Sublist3r/sublist3r.py -d {} -t {} -o {}/from_sublister.txt'.format( domain.name, threads, results_dir) # Run sublist3r logging.info(sublist3r_command) os.system(sublist3r_command) if 'subfinder' in tools: subfinder_command = 'subfinder -d {} -t {} -o {}/from_subfinder.txt'.format( domain.name, threads, results_dir) if USE_SUBFINDER_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_SUBFINDER_CONFIG]: subfinder_command += ' -config /root/.config/subfinder/config.yaml' # Run Subfinder logging.info(subfinder_command) os.system(subfinder_command) if 'oneforall' in tools: oneforall_command = 'python3 /usr/src/github/OneForAll/oneforall.py --target {} run'.format( domain.name, results_dir) # Run OneForAll logging.info(oneforall_command) os.system(oneforall_command) extract_subdomain = "cut -d',' -f6 /usr/src/github/OneForAll/results/{}.csv >> {}/from_oneforall.txt".format( domain.name, results_dir) os.system(extract_subdomain) # remove the results from oneforall directory os.system( 'rm -rf /usr/src/github/OneForAll/results/{}.*'.format(domain.name)) ''' All tools have gathered the list of subdomains with filename initials as from_* We will gather all the results in one single file, sort them and remove the older results from_* ''' os.system( 'cat {0}/*.txt > {0}/subdomain_collection.txt'.format(results_dir)) ''' Write target domain into subdomain_collection ''' os.system( 'cat {0}/target_domain.txt >> {0}/subdomain_collection.txt'.format(results_dir)) ''' Remove all the from_* files ''' os.system('rm -f {}/from*'.format(results_dir)) ''' Sort all Subdomains ''' os.system( 'sort -u {0}/subdomain_collection.txt -o {0}/sorted_subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/subdomain_collection.txt'.format(results_dir)) ''' The final results will be stored in sorted_subdomain_collection. ''' # parse the subdomain list file and store in db with open(subdomain_scan_results_file) as subdomain_list: for _subdomain in subdomain_list: __subdomain = _subdomain.rstrip('\n') if not Subdomain.objects.filter(scan_history=task, name=__subdomain).exists( ) and validators.domain(__subdomain) and __subdomain not in out_of_scope_subdomains: subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': __subdomain, }) save_subdomain(subdomain_dict) notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: subdomains_count = Subdomain.objects.filter(scan_history=task).count() send_notification('Subdomain Gathering for target {} has been completed and has discovered *{}* subdomains.'.format(domain.name, subdomains_count)) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/sorted_subdomain_collection.txt') # check for any subdomain changes and send notif if any if notification and notification[0].send_subdomain_changes_notif: newly_added_subdomain = get_new_added_subdomain(task.id, domain.id) if newly_added_subdomain: message = "**{} New Subdomains Discovered on domain {}**".format(newly_added_subdomain.count(), domain.name) for subdomain in newly_added_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) removed_subdomain = get_removed_subdomain(task.id, domain.id) if removed_subdomain: message = "**{} Subdomains are no longer available on domain {}**".format(removed_subdomain.count(), domain.name) for subdomain in removed_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) # check for interesting subdomains and send notif if any if notification and notification[0].send_interesting_notif: interesting_subdomain = get_interesting_subdomains(task.id, domain.id) print(interesting_subdomain) if interesting_subdomain: message = "**{} Interesting Subdomains Found on domain {}**".format(interesting_subdomain.count(), domain.name) for subdomain in interesting_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) def get_new_added_subdomain(scan_id, domain_id): scan_history = ScanHistory.objects.filter( domain=domain_id).filter( subdomain_discovery=True).filter( id__lte=scan_id) if scan_history.count() > 1: last_scan = scan_history.order_by('-start_scan_date')[1] scanned_host_q1 = Subdomain.objects.filter( scan_history__id=scan_id).values('name') scanned_host_q2 = Subdomain.objects.filter( scan_history__id=last_scan.id).values('name') added_subdomain = scanned_host_q1.difference(scanned_host_q2) return Subdomain.objects.filter( scan_history=scan_id).filter( name__in=added_subdomain) def get_removed_subdomain(scan_id, domain_id): scan_history = ScanHistory.objects.filter( domain=domain_id).filter( subdomain_discovery=True).filter( id__lte=scan_id) if scan_history.count() > 1: last_scan = scan_history.order_by('-start_scan_date')[1] scanned_host_q1 = Subdomain.objects.filter( scan_history__id=scan_id).values('name') scanned_host_q2 = Subdomain.objects.filter( scan_history__id=last_scan.id).values('name') removed_subdomains = scanned_host_q2.difference(scanned_host_q1) print() return Subdomain.objects.filter( scan_history=last_scan).filter( name__in=removed_subdomains) def http_crawler(task, domain, results_dir, activity_id): ''' This function is runs right after subdomain gathering, and gathers important like page title, http status, etc HTTP Crawler runs by default ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('HTTP Crawler for target {} has been initiated.'.format(domain.name)) alive_file_location = results_dir + '/alive.txt' httpx_results_file = results_dir + '/httpx.json' subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' httpx_command = 'httpx -status-code -content-length -title -tech-detect -cdn -ip -follow-host-redirects -random-agent' proxy = get_random_proxy() if proxy: httpx_command += " --http-proxy '{}'".format(proxy) httpx_command += ' -json -o {}'.format( httpx_results_file ) httpx_command = 'cat {} | {}'.format(subdomain_scan_results_file, httpx_command) print(httpx_command) os.system(httpx_command) # alive subdomains from httpx alive_file = open(alive_file_location, 'w') # writing httpx results if os.path.isfile(httpx_results_file): httpx_json_result = open(httpx_results_file, 'r') lines = httpx_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) try: # fallback for older versions of httpx if 'url' in json_st: subdomain = Subdomain.objects.get( scan_history=task, name=json_st['input']) else: subdomain = Subdomain.objects.get( scan_history=task, name=json_st['url'].split("//")[-1]) ''' Saving Default http urls to EndPoint ''' endpoint = EndPoint() endpoint.scan_history = task endpoint.target_domain = domain endpoint.subdomain = subdomain if 'url' in json_st: endpoint.http_url = json_st['url'] subdomain.http_url = json_st['url'] if 'status-code' in json_st: endpoint.http_status = json_st['status-code'] subdomain.http_status = json_st['status-code'] if 'title' in json_st: endpoint.page_title = json_st['title'] subdomain.page_title = json_st['title'] if 'content-length' in json_st: endpoint.content_length = json_st['content-length'] subdomain.content_length = json_st['content-length'] if 'content-type' in json_st: endpoint.content_type = json_st['content-type'] subdomain.content_type = json_st['content-type'] if 'webserver' in json_st: endpoint.webserver = json_st['webserver'] subdomain.webserver = json_st['webserver'] if 'response-time' in json_st: response_time = float( ''.join( ch for ch in json_st['response-time'] if not ch.isalpha())) if json_st['response-time'][-2:] == 'ms': response_time = response_time / 1000 endpoint.response_time = response_time subdomain.response_time = response_time if 'cnames' in json_st: cname_list = ','.join(json_st['cnames']) subdomain.cname = cname_list discovered_date = timezone.now() endpoint.discovered_date = discovered_date subdomain.discovered_date = discovered_date endpoint.is_default = True endpoint.save() subdomain.save() if 'technologies' in json_st: for _tech in json_st['technologies']: if Technology.objects.filter(name=_tech).exists(): tech = Technology.objects.get(name=_tech) else: tech = Technology(name=_tech) tech.save() subdomain.technologies.add(tech) endpoint.technologies.add(tech) if 'a' in json_st: for _ip in json_st['a']: if IpAddress.objects.filter(address=_ip).exists(): ip = IpAddress.objects.get(address=_ip) else: ip = IpAddress(address=_ip) if 'cdn' in json_st: ip.is_cdn = json_st['cdn'] ip.save() subdomain.ip_addresses.add(ip) # see if to ignore 404 or 5xx alive_file.write(json_st['url'] + '\n') subdomain.save() endpoint.save() except Exception as exception: logging.error(exception) alive_file.close() if notification and notification[0].send_scan_status_notif: alive_count = Subdomain.objects.filter( scan_history__id=task.id).values('name').distinct().filter( http_status__exact=200).count() send_notification('HTTP Crawler for target {} has been completed.\n\n {} subdomains were alive (http status 200).'.format(domain.name, alive_count)) def grab_screenshot(task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for taking screenshots ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine is currently gathering screenshots for {}'.format(domain.name)) output_screenshots_path = results_dir + '/screenshots' result_csv_path = results_dir + '/screenshots/Requests.csv' alive_subdomains_path = results_dir + '/alive.txt' eyewitness_command = 'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py' eyewitness_command += ' -f {} -d {} --no-prompt'.format( alive_subdomains_path, output_screenshots_path ) if EYEWITNESS in yaml_configuration \ and TIMEOUT in yaml_configuration[EYEWITNESS] \ and yaml_configuration[EYEWITNESS][TIMEOUT] > 0: eyewitness_command += ' --timeout {}'.format( yaml_configuration[EYEWITNESS][TIMEOUT] ) if EYEWITNESS in yaml_configuration \ and THREADS in yaml_configuration[EYEWITNESS] \ and yaml_configuration[EYEWITNESS][THREADS] > 0: eyewitness_command += ' --threads {}'.format( yaml_configuration[EYEWITNESS][THREADS] ) logger.info(eyewitness_command) os.system(eyewitness_command) if os.path.isfile(result_csv_path): logger.info('Gathering Eyewitness results') with open(result_csv_path, 'r') as file: reader = csv.reader(file) for row in reader: if row[3] == 'Successful' \ and Subdomain.objects.filter( scan_history__id=task.id).filter(name=row[2]).exists(): subdomain = Subdomain.objects.get( scan_history__id=task.id, name=row[2] ) subdomain.screenshot_path = row[4].replace( '/usr/src/scan_results/', '' ) subdomain.save() # remove all db, html extra files in screenshot results os.system('rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format( output_screenshots_path, )) os.system('rm -rf {0}/source'.format( output_screenshots_path, )) if notification and notification[0].send_scan_status_notif: send_notification('reNgine has finished gathering screenshots for {}'.format(domain.name)) def port_scanning(task, domain, yaml_configuration, results_dir): ''' This function is responsible for running the port scan ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Port Scan initiated for {}'.format(domain.name)) subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' port_results_file = results_dir + '/ports.json' # check the yaml_configuration and choose the ports to be scanned scan_ports = '-' # default port scan everything if PORTS in yaml_configuration[PORT_SCAN]: # TODO: legacy code, remove top-100 in future versions all_ports = yaml_configuration[PORT_SCAN][PORTS] if 'full' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -p {}'.format( subdomain_scan_results_file, port_results_file, '-') elif 'top-100' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -top-ports 100'.format( subdomain_scan_results_file, port_results_file) elif 'top-1000' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -top-ports 1000'.format( subdomain_scan_results_file, port_results_file) else: scan_ports = ','.join( str(port) for port in all_ports) naabu_command = 'cat {} | naabu -json -o {} -p {}'.format( subdomain_scan_results_file, port_results_file, scan_ports) # check for exclude ports if EXCLUDE_PORTS in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][EXCLUDE_PORTS]: exclude_ports = ','.join( str(port) for port in yaml_configuration['port_scan']['exclude_ports']) naabu_command = naabu_command + \ ' -exclude-ports {}'.format(exclude_ports) if NAABU_RATE in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][NAABU_RATE] > 0: naabu_command = naabu_command + \ ' -rate {}'.format( yaml_configuration[PORT_SCAN][NAABU_RATE]) if USE_NAABU_CONFIG in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][USE_NAABU_CONFIG]: naabu_command += ' -config /root/.config/naabu/naabu.conf' # run naabu os.system(naabu_command) # writing port results try: port_json_result = open(port_results_file, 'r') lines = port_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) port_number = json_st['port'] ip_address = json_st['ip'] # see if port already exists if Port.objects.filter(number__exact=port_number).exists(): port = Port.objects.get(number=port_number) else: port = Port() port.number = port_number if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port_detail = whatportis.get_ports(str(port_number)) if len(port_detail): port.service_name = port_detail[0].name port.description = port_detail[0].description port.save() if IpAddress.objects.filter(address=json_st['ip']).exists(): ip = IpAddress.objects.get(address=json_st['ip']) ip.ports.add(port) ip.save() except BaseException as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: port_count = Port.objects.filter( ports__in=IpAddress.objects.filter( ip_addresses__in=Subdomain.objects.filter( scan_history__id=task.id))).distinct().count() send_notification('reNgine has finished Port Scanning on {} and has identified {} ports.'.format(domain.name, port_count)) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/ports.json') def check_waf(): ''' This function will check for the WAF being used in subdomains using wafw00f ''' pass def directory_brute(task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for performing directory scan ''' # scan directories for all the alive subdomain with http status > # 200 notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Directory Bruteforce has been initiated for {}.'.format(domain.name)) alive_subdomains = Subdomain.objects.filter( scan_history__id=task.id).exclude(http_url__isnull=True) dirs_results = results_dir + '/dirs.json' # check the yaml settings if EXTENSIONS in yaml_configuration[DIR_FILE_SEARCH]: extensions = ','.join( str(ext) for ext in yaml_configuration[DIR_FILE_SEARCH][EXTENSIONS]) else: extensions = 'php,git,yaml,conf,db,mysql,bak,txt' # Threads if THREADS in yaml_configuration[DIR_FILE_SEARCH] \ and yaml_configuration[DIR_FILE_SEARCH][THREADS] > 0: threads = yaml_configuration[DIR_FILE_SEARCH][THREADS] else: threads = 10 for subdomain in alive_subdomains: # delete any existing dirs.json if os.path.isfile(dirs_results): os.system('rm -rf {}'.format(dirs_results)) dirsearch_command = 'python3 /usr/src/github/dirsearch/dirsearch.py' dirsearch_command += ' -u {}'.format(subdomain.http_url) if (WORDLIST not in yaml_configuration[DIR_FILE_SEARCH] or not yaml_configuration[DIR_FILE_SEARCH][WORDLIST] or 'default' in yaml_configuration[DIR_FILE_SEARCH][WORDLIST]): wordlist_location = '/usr/src/github/dirsearch/db/dicc.txt' else: wordlist_location = '/usr/src/wordlist/' + \ yaml_configuration[DIR_FILE_SEARCH][WORDLIST] + '.txt' dirsearch_command += ' -w {}'.format(wordlist_location) dirsearch_command += ' --format json -o {}'.format(dirs_results) dirsearch_command += ' -e {}'.format(extensions) dirsearch_command += ' -t {}'.format(threads) dirsearch_command += ' --random-agent --follow-redirects --exclude-status 403,401,404' if EXCLUDE_EXTENSIONS in yaml_configuration[DIR_FILE_SEARCH]: exclude_extensions = ','.join( str(ext) for ext in yaml_configuration[DIR_FILE_SEARCH][EXCLUDE_EXTENSIONS]) dirsearch_command += ' -X {}'.format(exclude_extensions) if EXCLUDE_TEXT in yaml_configuration[DIR_FILE_SEARCH]: exclude_text = ','.join( str(text) for text in yaml_configuration[DIR_FILE_SEARCH][EXCLUDE_TEXT]) dirsearch_command += ' -exclude-texts {}'.format(exclude_text) # check if recursive strategy is set to on if RECURSIVE_LEVEL in yaml_configuration[DIR_FILE_SEARCH]: dirsearch_command += ' --recursion-depth {}'.format(yaml_configuration[DIR_FILE_SEARCH][RECURSIVE_LEVEL]) if RECURSIVE_LEVEL in yaml_configuration[DIR_FILE_SEARCH]: dirsearch_command += ' --recursion-depth {}'.format(yaml_configuration[DIR_FILE_SEARCH][RECURSIVE_LEVEL]) # proxy proxy = get_random_proxy() if proxy: dirsearch_command += " --proxy '{}'".format(proxy) print(dirsearch_command) os.system(dirsearch_command) try: if os.path.isfile(dirs_results): with open(dirs_results, "r") as json_file: json_string = json_file.read() subdomain = Subdomain.objects.get( scan_history__id=task.id, http_url=subdomain.http_url) subdomain.directory_json = json_string subdomain.save() except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: send_notification('Directory Bruteforce has been completed for {}.'.format(domain.name)) def fetch_endpoints( task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for fetching all the urls associated with target and run HTTP probe It first runs gau to gather all urls from wayback, then we will use hakrawler to identify more urls ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine is currently gathering endpoints for {}.'.format(domain.name)) # check yaml settings if ALL in yaml_configuration[FETCH_URL][USES_TOOLS]: tools = 'gauplus hakrawler waybackurls gospider' else: tools = ' '.join( str(tool) for tool in yaml_configuration[FETCH_URL][USES_TOOLS]) if INTENSITY in yaml_configuration[FETCH_URL]: scan_type = yaml_configuration[FETCH_URL][INTENSITY] else: scan_type = 'normal' domain_regex = "\'https?://([a-z0-9]+[.])*{}.*\'".format(domain.name) if 'deep' in scan_type: # performs deep url gathering for all the subdomains present - # RECOMMENDED logger.info('Deep URLS Fetch') os.system(settings.TOOL_LOCATION + 'get_urls.sh %s %s %s %s %s' % ("None", results_dir, scan_type, domain_regex, tools)) else: # perform url gathering only for main domain - USE only for quick scan logger.info('Non Deep URLS Fetch') os.system( settings.TOOL_LOCATION + 'get_urls.sh %s %s %s %s %s' % ( domain.name, results_dir, scan_type, domain_regex, tools )) if IGNORE_FILE_EXTENSION in yaml_configuration[FETCH_URL]: ignore_extension = '|'.join( yaml_configuration[FETCH_URL][IGNORE_FILE_EXTENSION]) logger.info('Ignore extensions' + ignore_extension) os.system( 'cat {0}/all_urls.txt | grep -Eiv "\\.({1}).*" > {0}/temp_urls.txt'.format( results_dir, ignore_extension)) os.system( 'rm {0}/all_urls.txt && mv {0}/temp_urls.txt {0}/all_urls.txt'.format(results_dir)) ''' Store all the endpoints and then run the httpx ''' try: endpoint_final_url = results_dir + '/all_urls.txt' if os.path.isfile(endpoint_final_url): with open(endpoint_final_url) as endpoint_list: for url in endpoint_list: http_url = url.rstrip('\n') if not EndPoint.objects.filter(scan_history=task, http_url=http_url).exists(): _subdomain = get_subdomain_from_url(http_url) if Subdomain.objects.filter( scan_history=task).filter( name=_subdomain).exists(): subdomain = Subdomain.objects.get( scan_history=task, name=_subdomain) else: ''' gau or gosppider can gather interesting endpoints which when parsed can give subdomains that were not existent from subdomain scan. so storing them ''' logger.error( 'Subdomain {} not found, adding...'.format(_subdomain)) subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': _subdomain, }) subdomain = save_subdomain(subdomain_dict) endpoint_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'subdomain': subdomain, 'http_url': http_url, }) save_endpoint(endpoint_dict) except Exception as e: logger.error(e) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/all_urls.txt') ''' TODO: Go spider & waybackurls accumulates a lot of urls, which is good but nuclei takes forever to scan even a simple website, so we will do http probing and filter HTTP status 404, this way we can reduce the number of Non Existent URLS ''' logger.info('HTTP Probing on collected endpoints') httpx_command = 'httpx -l {0}/all_urls.txt -status-code -content-length -ip -cdn -title -tech-detect -json -follow-redirects -random-agent -o {0}/final_httpx_urls.json'.format(results_dir) proxy = get_random_proxy() if proxy: httpx_command += " --http-proxy '{}'".format(proxy) os.system(httpx_command) url_results_file = results_dir + '/final_httpx_urls.json' try: urls_json_result = open(url_results_file, 'r') lines = urls_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) http_url = json_st['url'] _subdomain = get_subdomain_from_url(http_url) if Subdomain.objects.filter( scan_history=task).filter( name=_subdomain).exists(): subdomain_obj = Subdomain.objects.get( scan_history=task, name=_subdomain) else: subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': _subdomain, }) subdomain_obj = save_subdomain(subdomain_dict) if EndPoint.objects.filter( scan_history=task).filter( http_url=http_url).exists(): endpoint = EndPoint.objects.get( scan_history=task, http_url=http_url) else: endpoint = EndPoint() endpoint_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'http_url': http_url, 'subdomain': subdomain_obj }) endpoint = save_endpoint(endpoint_dict) if 'title' in json_st: endpoint.page_title = json_st['title'] if 'webserver' in json_st: endpoint.webserver = json_st['webserver'] if 'content-length' in json_st: endpoint.content_length = json_st['content-length'] if 'content-type' in json_st: endpoint.content_type = json_st['content-type'] if 'status-code' in json_st: endpoint.http_status = json_st['status-code'] if 'response-time' in json_st: response_time = float(''.join(ch for ch in json_st['response-time'] if not ch.isalpha())) if json_st['response-time'][-2:] == 'ms': response_time = response_time / 1000 endpoint.response_time = response_time endpoint.save() if 'technologies' in json_st: for _tech in json_st['technologies']: if Technology.objects.filter(name=_tech).exists(): tech = Technology.objects.get(name=_tech) else: tech = Technology(name=_tech) tech.save() endpoint.technologies.add(tech) # get subdomain object subdomain = Subdomain.objects.get(scan_history=task, name=_subdomain) subdomain.technologies.add(tech) subdomain.save() except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: endpoint_count = EndPoint.objects.filter( scan_history__id=task.id).values('http_url').distinct().count() endpoint_alive_count = EndPoint.objects.filter( scan_history__id=task.id, http_status__exact=200).values('http_url').distinct().count() send_notification('reNgine has finished gathering endpoints for {} and has discovered *{}* unique endpoints.\n\n{} of those endpoints reported HTTP status 200.'.format( domain.name, endpoint_count, endpoint_alive_count )) # once endpoint is saved, run gf patterns TODO: run threads if GF_PATTERNS in yaml_configuration[FETCH_URL]: for pattern in yaml_configuration[FETCH_URL][GF_PATTERNS]: logger.info('Running GF for {}'.format(pattern)) gf_output_file_path = '{0}/gf_patterns_{1}.txt'.format( results_dir, pattern) gf_command = 'cat {0}/all_urls.txt | gf {1} >> {2}'.format( results_dir, pattern, gf_output_file_path) os.system(gf_command) if os.path.exists(gf_output_file_path): with open(gf_output_file_path) as gf_output: for line in gf_output: url = line.rstrip('\n') try: endpoint = EndPoint.objects.get( scan_history=task, http_url=url) earlier_pattern = endpoint.matched_gf_patterns new_pattern = earlier_pattern + ',' + pattern if earlier_pattern else pattern endpoint.matched_gf_patterns = new_pattern except Exception as e: # add the url in db logger.error(e) logger.info('Adding URL' + url) endpoint = EndPoint() endpoint.http_url = url endpoint.target_domain = domain endpoint.scan_history = task try: _subdomain = Subdomain.objects.get( scan_history=task, name=get_subdomain_from_url(url)) endpoint.subdomain = _subdomain except Exception as e: continue endpoint.matched_gf_patterns = pattern finally: endpoint.save() def vulnerability_scan( task, domain, yaml_configuration, results_dir, activity_id): notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Vulnerability scan has been initiated for {}.'.format(domain.name)) ''' This function will run nuclei as a vulnerability scanner ---- unfurl the urls to keep only domain and path, this will be sent to vuln scan ignore certain file extensions Thanks: https://github.com/six2dez/reconftw ''' urls_path = '/alive.txt' if task.scan_type.fetch_url: os.system('cat {0}/all_urls.txt | grep -Eiv "\\.(eot|jpg|jpeg|gif|css|tif|tiff|png|ttf|otf|woff|woff2|ico|pdf|svg|txt|js|doc|docx)$" | unfurl -u format %s://%d%p >> {0}/unfurl_urls.txt'.format(results_dir)) os.system( 'sort -u {0}/unfurl_urls.txt -o {0}/unfurl_urls.txt'.format(results_dir)) urls_path = '/unfurl_urls.txt' vulnerability_result_path = results_dir + '/vulnerability.json' vulnerability_scan_input_file = results_dir + urls_path nuclei_command = 'nuclei -json -l {} -o {}'.format( vulnerability_scan_input_file, vulnerability_result_path) # check nuclei config if USE_NUCLEI_CONFIG in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[VULNERABILITY_SCAN][USE_NUCLEI_CONFIG]: nuclei_command += ' -config /root/.config/nuclei/config.yaml' ''' Nuclei Templates Either custom template has to be supplied or default template, if neither has been supplied then use all templates including custom templates ''' if CUSTOM_NUCLEI_TEMPLATE in yaml_configuration[ VULNERABILITY_SCAN] or NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: # check yaml settings for templates if NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: if ALL in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_TEMPLATE]: template = NUCLEI_TEMPLATES_PATH else: _template = ','.join([NUCLEI_TEMPLATES_PATH + str(element) for element in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_TEMPLATE]]) template = _template.replace(',', ' -t ') # Update nuclei command with templates nuclei_command = nuclei_command + ' -t ' + template if CUSTOM_NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: # add .yaml to the custom template extensions _template = ','.join( [str(element) + '.yaml' for element in yaml_configuration[VULNERABILITY_SCAN][CUSTOM_NUCLEI_TEMPLATE]]) template = _template.replace(',', ' -t ') # Update nuclei command with templates nuclei_command = nuclei_command + ' -t ' + template else: nuclei_command = nuclei_command + ' -t /root/nuclei-templates' # check yaml settings for concurrency if NUCLEI_CONCURRENCY in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][NUCLEI_CONCURRENCY] > 0: concurrency = yaml_configuration[VULNERABILITY_SCAN][NUCLEI_CONCURRENCY] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -c ' + str(concurrency) if RATE_LIMIT in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][RATE_LIMIT] > 0: rate_limit = yaml_configuration[VULNERABILITY_SCAN][RATE_LIMIT] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -rl ' + str(rate_limit) if TIMEOUT in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][TIMEOUT] > 0: timeout = yaml_configuration[VULNERABILITY_SCAN][TIMEOUT] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -timeout ' + str(timeout) if RETRIES in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][RETRIES] > 0: retries = yaml_configuration[VULNERABILITY_SCAN][RETRIES] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -retries ' + str(retries) # for severity if NUCLEI_SEVERITY in yaml_configuration[VULNERABILITY_SCAN] and ALL not in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_SEVERITY]: _severity = ','.join( [str(element) for element in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_SEVERITY]]) severity = _severity.replace(" ", "") else: severity = "critical, high, medium, low, info" # update nuclei templates before running scan os.system('nuclei -update-templates') for _severity in severity.split(","): # delete any existing vulnerability.json file if os.path.isfile(vulnerability_result_path): os.system('rm {}'.format(vulnerability_result_path)) # run nuclei final_nuclei_command = nuclei_command + ' -severity ' + _severity proxy = get_random_proxy() if proxy: final_nuclei_command += " --proxy-url '{}'".format(proxy) logger.info(final_nuclei_command) os.system(final_nuclei_command) try: if os.path.isfile(vulnerability_result_path): urls_json_result = open(vulnerability_result_path, 'r') lines = urls_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) host = json_st['host'] _subdomain = get_subdomain_from_url(host) try: subdomain = Subdomain.objects.get( name=_subdomain, scan_history=task) vulnerability = Vulnerability() vulnerability.subdomain = subdomain vulnerability.scan_history = task vulnerability.target_domain = domain try: endpoint = EndPoint.objects.get( scan_history=task, target_domain=domain, http_url=host) vulnerability.endpoint = endpoint except Exception as exception: logger.error(exception) if 'name' in json_st['info']: vulnerability.name = json_st['info']['name'] if 'severity' in json_st['info']: if json_st['info']['severity'] == 'info': severity = 0 elif json_st['info']['severity'] == 'low': severity = 1 elif json_st['info']['severity'] == 'medium': severity = 2 elif json_st['info']['severity'] == 'high': severity = 3 elif json_st['info']['severity'] == 'critical': severity = 4 else: severity = 0 else: severity = 0 vulnerability.severity = severity if 'tags' in json_st['info']: vulnerability.tags = json_st['info']['tags'] if 'description' in json_st['info']: vulnerability.description = json_st['info']['description'] if 'reference' in json_st['info']: vulnerability.reference = json_st['info']['reference'] if 'matched' in json_st: # TODO remove in rengine 1.1. 'matched' isn't used in nuclei 2.5.3 vulnerability.http_url = json_st['matched'] if 'matched-at' in json_st: vulnerability.http_url = json_st['matched-at'] if 'templateID' in json_st: vulnerability.template_used = json_st['templateID'] if 'description' in json_st: vulnerability.description = json_st['description'] if 'matcher_name' in json_st: vulnerability.matcher_name = json_st['matcher_name'] if 'extracted_results' in json_st: vulnerability.extracted_results = json_st['extracted_results'] vulnerability.discovered_date = timezone.now() vulnerability.open_status = True vulnerability.save() # send notification for all vulnerabilities except info if json_st['info']['severity'] != "info" and notification and notification[0].send_vuln_notif: message = "*Alert: Vulnerability Identified*" message += "\n\n" message += "A *{}* severity vulnerability has been identified.".format(json_st['info']['severity']) message += "\nVulnerability Name: {}".format(json_st['info']['name']) message += "\nVulnerable URL: {}".format(json_st['host']) send_notification(message) # send report to hackerone if Hackerone.objects.all().exists() and json_st['info']['severity'] != 'info' and json_st['info']['severity'] \ != 'low' and vulnerability.target_domain.h1_team_handle: hackerone = Hackerone.objects.all()[0] if hackerone.send_critical and json_st['info']['severity'] == 'critical': send_hackerone_report(vulnerability.id) elif hackerone.send_high and json_st['info']['severity'] == 'high': send_hackerone_report(vulnerability.id) elif hackerone.send_medium and json_st['info']['severity'] == 'medium': send_hackerone_report(vulnerability.id) except ObjectDoesNotExist: logger.error('Object not found') continue except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: info_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=0).count() low_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=1).count() medium_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=2).count() high_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=3).count() critical_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=4).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count message = 'Vulnerability scan has been completed for {} and discovered {} vulnerabilities.'.format( domain.name, vulnerability_count ) message += '\n\n*Vulnerability Stats:*' message += '\nCritical: {}'.format(critical_count) message += '\nHigh: {}'.format(high_count) message += '\nMedium: {}'.format(medium_count) message += '\nLow: {}'.format(low_count) message += '\nInfo: {}'.format(info_count) send_notification(message) def scan_failed(task): task.scan_status = 0 task.stop_scan_date = timezone.now() task.save() def create_scan_activity(task, message, status): scan_activity = ScanActivity() scan_activity.scan_of = task scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id def update_last_activity(id, activity_status): ScanActivity.objects.filter( id=id).update( status=activity_status, time=timezone.now()) def delete_scan_data(results_dir): # remove all txt,html,json files os.system('find {} -name "*.txt" -type f -delete'.format(results_dir)) os.system('find {} -name "*.html" -type f -delete'.format(results_dir)) os.system('find {} -name "*.json" -type f -delete'.format(results_dir)) def save_subdomain(subdomain_dict): subdomain = Subdomain() subdomain.discovered_date = timezone.now() subdomain.target_domain = subdomain_dict.get('target_domain') subdomain.scan_history = subdomain_dict.get('scan_history') subdomain.name = subdomain_dict.get('name') subdomain.http_url = subdomain_dict.get('http_url') subdomain.screenshot_path = subdomain_dict.get('screenshot_path') subdomain.http_header_path = subdomain_dict.get('http_header_path') subdomain.cname = subdomain_dict.get('cname') subdomain.is_cdn = subdomain_dict.get('is_cdn') subdomain.content_type = subdomain_dict.get('content_type') subdomain.webserver = subdomain_dict.get('webserver') subdomain.page_title = subdomain_dict.get('page_title') subdomain.is_imported_subdomain = subdomain_dict.get( 'is_imported_subdomain') if 'is_imported_subdomain' in subdomain_dict else False if 'http_status' in subdomain_dict: subdomain.http_status = subdomain_dict.get('http_status') if 'response_time' in subdomain_dict: subdomain.response_time = subdomain_dict.get('response_time') if 'content_length' in subdomain_dict: subdomain.content_length = subdomain_dict.get('content_length') subdomain.save() return subdomain def save_endpoint(endpoint_dict): endpoint = EndPoint() endpoint.discovered_date = timezone.now() endpoint.scan_history = endpoint_dict.get('scan_history') endpoint.target_domain = endpoint_dict.get('target_domain') if 'target_domain' in endpoint_dict else None endpoint.subdomain = endpoint_dict.get('subdomain') if 'target_domain' in endpoint_dict else None endpoint.http_url = endpoint_dict.get('http_url') endpoint.page_title = endpoint_dict.get('page_title') if 'page_title' in endpoint_dict else None endpoint.content_type = endpoint_dict.get('content_type') if 'content_type' in endpoint_dict else None endpoint.webserver = endpoint_dict.get('webserver') if 'webserver' in endpoint_dict else None endpoint.response_time = endpoint_dict.get('response_time') if 'response_time' in endpoint_dict else 0 endpoint.http_status = endpoint_dict.get('http_status') if 'http_status' in endpoint_dict else 0 endpoint.content_length = endpoint_dict.get('content_length') if 'content_length' in endpoint_dict else 0 endpoint.is_default = endpoint_dict.get('is_default') if 'is_default' in endpoint_dict else False endpoint.save() return endpoint def perform_osint(task, domain, yaml_configuration, results_dir): notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine has initiated OSINT on target {}'.format(domain.name)) if 'discover' in yaml_configuration[OSINT]: osint_discovery(task, domain, yaml_configuration, results_dir) if 'dork' in yaml_configuration[OSINT]: dorking(task, yaml_configuration) if notification and notification[0].send_scan_status_notif: send_notification('reNgine has completed performing OSINT on target {}'.format(domain.name)) def osint_discovery(task, domain, yaml_configuration, results_dir): if ALL in yaml_configuration[OSINT][OSINT_DISCOVER]: osint_lookup = 'emails metainfo employees' else: osint_lookup = ' '.join( str(lookup) for lookup in yaml_configuration[OSINT][OSINT_DISCOVER]) if 'metainfo' in osint_lookup: if INTENSITY in yaml_configuration[OSINT]: osint_intensity = yaml_configuration[OSINT][INTENSITY] else: osint_intensity = 'normal' if OSINT_DOCUMENTS_LIMIT in yaml_configuration[OSINT]: documents_limit = yaml_configuration[OSINT][OSINT_DOCUMENTS_LIMIT] else: documents_limit = 50 if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': domain.name, 'domain': domain, 'scan_id': task, 'documents_limit': documents_limit }) get_and_save_meta_info(meta_dict) elif osint_intensity == 'deep': # get all subdomains in scan_id subdomains = Subdomain.objects.filter(scan_history=task) for subdomain in subdomains: meta_dict = DottedDict({ 'osint_target': subdomain.name, 'domain': domain, 'scan_id': task, 'documents_limit': documents_limit }) get_and_save_meta_info(meta_dict) if 'emails' in osint_lookup: get_and_save_emails(task, results_dir) get_and_save_leaked_credentials(task, results_dir) if 'employees' in osint_lookup: get_and_save_employees(task, results_dir) def dorking(scan_history, yaml_configuration): # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh # look in stackoverflow if ALL in yaml_configuration[OSINT][OSINT_DORK]: dork_lookup = 'stackoverflow, 3rdparty, social_media, project_management, code_sharing, config_files, jenkins, cloud_buckets, php_error, exposed_documents, struts_rce, db_files, traefik, git_exposed' else: dork_lookup = ' '.join( str(lookup) for lookup in yaml_configuration[OSINT][OSINT_DORK]) if 'stackoverflow' in dork_lookup: dork = 'site:stackoverflow.com' dork_type = 'stackoverflow' get_and_save_dork_results( dork, dork_type, scan_history, in_target=False ) if '3rdparty' in dork_lookup: # look in 3rd party sitee dork_type = '3rdparty' lookup_websites = [ 'gitter.im', 'papaly.com', 'productforums.google.com', 'coggle.it', 'replt.it', 'ycombinator.com', 'libraries.io', 'npm.runkit.com', 'npmjs.com', 'scribd.com', 'gitter.im' ] dork = '' for website in lookup_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'social_media' in dork_lookup: dork_type = 'Social Media' social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'pinterest.com', 'tumblr.com', 'reddit.com' ] dork = '' for website in social_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'project_management' in dork_lookup: dork_type = 'Project Management' project_websites = [ 'trello.com', '*.atlassian.net' ] dork = '' for website in project_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'code_sharing' in dork_lookup: dork_type = 'Code Sharing Sites' code_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] dork = '' for website in code_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'config_files' in dork_lookup: dork_type = 'Config Files' config_file_ext = [ 'env', 'xml', 'conf', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] dork = '' for extension in config_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'jenkins' in dork_lookup: dork_type = 'Jenkins' dork = 'intitle:\"Dashboard [Jenkins]\"' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) if 'wordpress_files' in dork_lookup: dork_type = 'Wordpress Files' inurl_lookup = [ 'wp-content', 'wp-includes' ] dork = '' for lookup in inurl_lookup: dork = dork + ' | ' + 'inurl:' + lookup get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'cloud_buckets' in dork_lookup: dork_type = 'Cloud Buckets' cloud_websites = [ '.s3.amazonaws.com', 'storage.googleapis.com', 'amazonaws.com' ] dork = '' for website in cloud_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'php_error' in dork_lookup: dork_type = 'PHP Error' error_words = [ '\"PHP Parse error\"', '\"PHP Warning\"', '\"PHP Error\"' ] dork = '' for word in error_words: dork = dork + ' | ' + word get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'exposed_documents' in dork_lookup: dork_type = 'Exposed Documents' docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] dork = '' for extension in docs_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'struts_rce' in dork_lookup: dork_type = 'Apache Struts RCE' struts_file_ext = [ 'action', 'struts', 'do' ] dork = '' for extension in struts_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'db_files' in dork_lookup: dork_type = 'Database Files' db_file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] dork = '' for extension in db_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'traefik' in dork_lookup: dork = 'intitle:traefik inurl:8080/dashboard' dork_type = 'Traefik' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) if 'git_exposed' in dork_lookup: dork = 'inurl:\"/.git\"' dork_type = '.git Exposed' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) def get_and_save_dork_results(dork, type, scan_history, in_target=False): degoogle_obj = degoogle.dg() proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy if in_target: query = dork + " site:" + scan_history.domain.name else: query = dork + " \"{}\"".format(scan_history.domain.name) logger.info(query) degoogle_obj.query = query results = degoogle_obj.run() logger.info(results) for result in results: dork, _ = Dork.objects.get_or_create( type=type, description=result['desc'], url=result['url'] ) scan_history.dorks.add(dork) def get_and_save_employees(scan_history, results_dir): theHarvester_location = '/usr/src/github/theHarvester' # update proxies.yaml if Proxy.objects.all().exists(): proxy = Proxy.objects.all()[0] if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(theHarvester_location + '/proxies.yaml', 'w') as file: documents = yaml.dump(yaml_data, file) os.system('cd {} && python3 theHarvester.py -d {} -b all -f {}/theHarvester.html'.format( theHarvester_location, scan_history.domain.name, results_dir )) file_location = results_dir + '/theHarvester.html' print(file_location) # delete proxy environ var if os.environ.get(('https_proxy')): del os.environ['https_proxy'] if os.environ.get(('HTTPS_PROXY')): del os.environ['HTTPS_PROXY'] if os.path.isfile(file_location): logger.info('Parsing theHarvester results') options = FirefoxOptions() options.add_argument("--headless") driver = webdriver.Firefox(options=options) driver.get('file://'+file_location) tabledata = driver.execute_script('return tabledata') # save email addresses and linkedin employees for data in tabledata: if data['record'] == 'email': _email = data['result'] email, _ = Email.objects.get_or_create(address=_email) scan_history.emails.add(email) elif data['record'] == 'people': _employee = data['result'] split_val = _employee.split('-') name = split_val[0] if len(split_val) == 2: designation = split_val[1] else: designation = "" employee, _ = Employee.objects.get_or_create(name=name, designation=designation) scan_history.employees.add(employee) driver.quit() print(tabledata) def get_and_save_emails(scan_history, results_dir): leak_target_path = '{}/creds_target.txt'.format(results_dir) # get email address proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy emails = [] try: logger.info('OSINT: Getting emails from Google') email_from_google = get_emails_from_google(scan_history.domain.name) logger.info('OSINT: Getting emails from Bing') email_from_bing = get_emails_from_bing(scan_history.domain.name) logger.info('OSINT: Getting emails from Baidu') email_from_baidu = get_emails_from_baidu(scan_history.domain.name) emails = list(set(email_from_google + email_from_bing + email_from_baidu)) logger.info(emails) except Exception as e: logger.error(e) leak_target_file = open(leak_target_path, 'w') for _email in emails: email, _ = Email.objects.get_or_create(address=_email) scan_history.emails.add(email) leak_target_file.write('{}\n'.format(_email)) # fill leak_target_file with possible email address leak_target_file.write('%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.write('%.%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%.%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.write('%_%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%_%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.close() def get_and_save_leaked_credentials(scan_history, results_dir): logger.info('OSINT: Getting leaked credentials...') leak_target_file = '{}/creds_target.txt'.format(results_dir) leak_output_file = '{}/pwndb.json'.format(results_dir) pwndb_command = 'python3 /usr/src/github/pwndb/pwndb.py --proxy tor:9150 --output json --list {}'.format( leak_target_file ) try: pwndb_output = subprocess.getoutput(pwndb_command) creds = json.loads(pwndb_output) for cred in creds: if cred['username'] != 'donate': email_id = "{}@{}".format(cred['username'], cred['domain']) email_obj, _ = Email.objects.get_or_create( address=email_id, ) email_obj.password = cred['password'] email_obj.save() scan_history.emails.add(email_obj) except Exception as e: logger.error(e) pass def get_and_save_meta_info(meta_dict): logger.info('Getting METADATA for {}'.format(meta_dict.osint_target)) proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy result = metadata_extractor.extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if result: results = result.get_metadata() for meta in results: meta_finder_document = MetaFinderDocument() subdomain = Subdomain.objects.get(scan_history=meta_dict.scan_id, name=meta_dict.osint_target) meta_finder_document.subdomain = subdomain meta_finder_document.target_domain = meta_dict.domain meta_finder_document.scan_history = meta_dict.scan_id item = DottedDict(results[meta]) meta_finder_document.url = item.url meta_finder_document.doc_name = meta meta_finder_document.http_status = item.status_code metadata = results[meta]['metadata'] for data in metadata: if 'Producer' in metadata and metadata['Producer']: meta_finder_document.producer = metadata['Producer'].rstrip('\x00') if 'Creator' in metadata and metadata['Creator']: meta_finder_document.creator = metadata['Creator'].rstrip('\x00') if 'CreationDate' in metadata and metadata['CreationDate']: meta_finder_document.creation_date = metadata['CreationDate'].rstrip('\x00') if 'ModDate' in metadata and metadata['ModDate']: meta_finder_document.modified_date = metadata['ModDate'].rstrip('\x00') if 'Author' in metadata and metadata['Author']: meta_finder_document.author = metadata['Author'].rstrip('\x00') if 'Title' in metadata and metadata['Title']: meta_finder_document.title = metadata['Title'].rstrip('\x00') if 'OSInfo' in metadata and metadata['OSInfo']: meta_finder_document.os = metadata['OSInfo'].rstrip('\x00') meta_finder_document.save() @app.task(bind=True) def test_task(self): print('*' * 40) print('test task run') print('*' * 40)

radaram

43af3a6aecdece4923ee74b108853f7b9c51ed12

27d6ec5827a51fd74e3ab97a5cef38fc7f5d9168

Yes, of course, In version nuclei 2.5.2 passed key matched https://github.com/projectdiscovery/nuclei/blob/v2.5.2/v2/pkg/output/output.go#L78 `Matched string json:"matched,omitempty"` In nuclei 2.5.3(last version) passed key matched-at `Matched string json:"matched-at,omitempty"` https://github.com/projectdiscovery/nuclei/blob/v2.5.3/v2/pkg/output/output.go#L78

radaram

yogeshojha/rengine

Fix #529

Nuclei returns the response to stdout: `{"template-id":"tech-detect","info":{"name":"Wappalyzer Technology Detection","author":["hakluke"],"tags":["tech"],"reference":null,"severity":"info"},"matcher-name":"nginx","type":"http","host":"https://example.com:443","matched-at":"https://example.com:443","timestamp":"2021-10-31T09:39:47.1571248Z","curl-command":"curl -X 'GET' -d '' -H 'Accept: */*' -H 'Accept-Language: en' -H 'User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_9_2) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/36.0.1944.0 Safari/537.36' 'https://example.com'"}` It needs to read host_url from matched-at, not from matched.

2021-10-31 10:27:33+00:00

2021-11-01 16:58:16+00:00

web/reNgine/tasks.py

import os import traceback import yaml import json import csv import validators import random import requests import logging import metafinder.extractor as metadata_extractor import whatportis import subprocess from selenium.webdriver.firefox.options import Options as FirefoxOptions from selenium import webdriver from emailfinder.extractor import * from dotted_dict import DottedDict from celery import shared_task from discord_webhook import DiscordWebhook from reNgine.celery import app from startScan.models import * from targetApp.models import Domain from scanEngine.models import EngineType from django.conf import settings from django.shortcuts import get_object_or_404 from celery import shared_task from datetime import datetime from degoogle import degoogle from django.conf import settings from django.utils import timezone, dateformat from django.shortcuts import get_object_or_404 from django.core.exceptions import ObjectDoesNotExist from reNgine.celery import app from reNgine.definitions import * from startScan.models import * from targetApp.models import Domain from scanEngine.models import EngineType, Configuration, Wordlist from .common_func import * ''' task for background scan ''' @app.task def initiate_scan( domain_id, scan_history_id, scan_type, engine_type, imported_subdomains=None, out_of_scope_subdomains=[] ): ''' scan_type = 0 -> immediate scan, need not create scan object scan_type = 1 -> scheduled scan ''' engine_object = EngineType.objects.get(pk=engine_type) domain = Domain.objects.get(pk=domain_id) if scan_type == 1: task = ScanHistory() task.scan_status = -1 elif scan_type == 0: task = ScanHistory.objects.get(pk=scan_history_id) # save the last scan date for domain model domain.last_scan_date = timezone.now() domain.save() # once the celery task starts, change the task status to Started task.scan_type = engine_object task.celery_id = initiate_scan.request.id task.domain = domain task.scan_status = 1 task.start_scan_date = timezone.now() task.subdomain_discovery = True if engine_object.subdomain_discovery else False task.dir_file_search = True if engine_object.dir_file_search else False task.port_scan = True if engine_object.port_scan else False task.fetch_url = True if engine_object.fetch_url else False task.osint = True if engine_object.osint else False task.screenshot = True if engine_object.screenshot else False task.vulnerability_scan = True if engine_object.vulnerability_scan else False task.save() activity_id = create_scan_activity(task, "Scanning Started", 2) results_dir = '/usr/src/scan_results/' os.chdir(results_dir) notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine has initiated recon for target {} with engine type {}'.format(domain.name, engine_object.engine_name)) try: current_scan_dir = domain.name + '_' + str(random.randint(100000000000, 999999999999)) os.mkdir(current_scan_dir) task.results_dir = current_scan_dir task.save() except Exception as exception: logger.error(exception) scan_failed(task) yaml_configuration = None excluded_subdomains = '' try: yaml_configuration = yaml.load( task.scan_type.yaml_configuration, Loader=yaml.FullLoader) except Exception as exception: logger.error(exception) # TODO: Put failed reason on db ''' Add GF patterns name to db for dynamic URLs menu ''' if engine_object.fetch_url and GF_PATTERNS in yaml_configuration[FETCH_URL]: task.used_gf_patterns = ','.join( pattern for pattern in yaml_configuration[FETCH_URL][GF_PATTERNS]) task.save() results_dir = results_dir + current_scan_dir # put all imported subdomains into txt file and also in Subdomain model if imported_subdomains: extract_imported_subdomain( imported_subdomains, task, domain, results_dir) if yaml_configuration: ''' a target in itself is a subdomain, some tool give subdomains as www.yogeshojha.com but url and everything else resolves to yogeshojha.com In that case, we would already need to store target itself as subdomain ''' initial_subdomain_file = '/target_domain.txt' if task.subdomain_discovery else '/sorted_subdomain_collection.txt' subdomain_file = open(results_dir + initial_subdomain_file, "w") subdomain_file.write(domain.name + "\n") subdomain_file.close() if(task.subdomain_discovery): activity_id = create_scan_activity(task, "Subdomain Scanning", 1) subdomain_scan( task, domain, yaml_configuration, results_dir, activity_id, out_of_scope_subdomains ) else: skip_subdomain_scan(task, domain, results_dir) update_last_activity(activity_id, 2) activity_id = create_scan_activity(task, "HTTP Crawler", 1) http_crawler( task, domain, results_dir, activity_id) update_last_activity(activity_id, 2) try: if task.screenshot: activity_id = create_scan_activity( task, "Visual Recon - Screenshot", 1) grab_screenshot( task, domain, yaml_configuration, current_scan_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if(task.port_scan): activity_id = create_scan_activity(task, "Port Scanning", 1) port_scanning(task, domain, yaml_configuration, results_dir) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.osint: activity_id = create_scan_activity(task, "OSINT Running", 1) perform_osint(task, domain, yaml_configuration, results_dir) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.dir_file_search: activity_id = create_scan_activity(task, "Directory Search", 1) directory_brute( task, domain, yaml_configuration, results_dir, activity_id ) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.fetch_url: activity_id = create_scan_activity(task, "Fetching endpoints", 1) fetch_endpoints( task, domain, yaml_configuration, results_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.vulnerability_scan: activity_id = create_scan_activity(task, "Vulnerability Scan", 1) vulnerability_scan( task, domain, yaml_configuration, results_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) activity_id = create_scan_activity(task, "Scan Completed", 2) if notification and notification[0].send_scan_status_notif: send_notification('*Scan Completed*\nreNgine has finished performing recon on target {}.'.format(domain.name)) ''' Once the scan is completed, save the status to successful ''' if ScanActivity.objects.filter(scan_of=task).filter(status=0).all(): task.scan_status = 0 else: task.scan_status = 2 task.stop_scan_date = timezone.now() task.save() # cleanup results delete_scan_data(results_dir) return {"status": True} def skip_subdomain_scan(task, domain, results_dir): # store default target as subdomain ''' If the imported subdomain already has target domain saved, we can skip this ''' if not Subdomain.objects.filter( scan_history=task, name=domain.name).exists(): subdomain_dict = DottedDict({ 'name': domain.name, 'scan_history': task, 'target_domain': domain }) save_subdomain(subdomain_dict) # Save target into target_domain.txt with open('{}/target_domain.txt'.format(results_dir), 'w+') as file: file.write(domain.name + '\n') file.close() ''' We can have two conditions, either subdomain scan happens, or subdomain scan does not happen, in either cases, because we are using import subdomain, we need to collect and sort all the subdomains Write target domain into subdomain_collection ''' os.system( 'cat {0}/target_domain.txt > {0}/subdomain_collection.txt'.format(results_dir)) os.system( 'cat {0}/from_imported.txt > {0}/subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/from_imported.txt'.format(results_dir)) ''' Sort all Subdomains ''' os.system( 'sort -u {0}/subdomain_collection.txt -o {0}/sorted_subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/subdomain_collection.txt'.format(results_dir)) def extract_imported_subdomain(imported_subdomains, task, domain, results_dir): valid_imported_subdomains = [subdomain for subdomain in imported_subdomains if validators.domain( subdomain) and domain.name == get_domain_from_subdomain(subdomain)] # remove any duplicate valid_imported_subdomains = list(set(valid_imported_subdomains)) with open('{}/from_imported.txt'.format(results_dir), 'w+') as file: for subdomain_name in valid_imported_subdomains: # save _subdomain to Subdomain model db if not Subdomain.objects.filter( scan_history=task, name=subdomain_name).exists(): subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': subdomain_name, 'is_imported_subdomain': True }) save_subdomain(subdomain_dict) # save subdomain to file file.write('{}\n'.format(subdomain_name)) file.close() def subdomain_scan(task, domain, yaml_configuration, results_dir, activity_id, out_of_scope_subdomains=None): ''' This function is responsible for performing subdomain enumeration ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Subdomain Gathering for target {} has been started'.format(domain.name)) subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' # check for all the tools and add them into string # if tool selected is all then make string, no need for loop if ALL in yaml_configuration[SUBDOMAIN_DISCOVERY][USES_TOOLS]: tools = 'amass-active amass-passive assetfinder sublist3r subfinder oneforall' else: tools = ' '.join( str(tool) for tool in yaml_configuration[SUBDOMAIN_DISCOVERY][USES_TOOLS]) logging.info(tools) # check for THREADS, by default 10 threads = 10 if THREADS in yaml_configuration[SUBDOMAIN_DISCOVERY]: _threads = yaml_configuration[SUBDOMAIN_DISCOVERY][THREADS] if _threads > 0: threads = _threads if 'amass' in tools: if 'amass-passive' in tools: amass_command = 'amass enum -passive -d {} -o {}/from_amass.txt'.format( domain.name, results_dir) if USE_AMASS_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_AMASS_CONFIG]: amass_command += ' -config /root/.config/amass.ini' # Run Amass Passive logging.info(amass_command) os.system(amass_command) if 'amass-active' in tools: amass_command = 'amass enum -active -d {} -o {}/from_amass_active.txt'.format( domain.name, results_dir) if USE_AMASS_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_AMASS_CONFIG]: amass_command += ' -config /root/.config/amass.ini' if AMASS_WORDLIST in yaml_configuration[SUBDOMAIN_DISCOVERY]: wordlist = yaml_configuration[SUBDOMAIN_DISCOVERY][AMASS_WORDLIST] if wordlist == 'default': wordlist_path = '/usr/src/wordlist/deepmagic.com-prefixes-top50000.txt' else: wordlist_path = '/usr/src/wordlist/' + wordlist + '.txt' if not os.path.exists(wordlist_path): wordlist_path = '/usr/src/' + AMASS_WORDLIST amass_command = amass_command + \ ' -brute -w {}'.format(wordlist_path) if amass_config_path: amass_command = amass_command + \ ' -config {}'.format('/usr/src/scan_results/' + amass_config_path) # Run Amass Active logging.info(amass_command) os.system(amass_command) if 'assetfinder' in tools: assetfinder_command = 'assetfinder --subs-only {} > {}/from_assetfinder.txt'.format( domain.name, results_dir) # Run Assetfinder logging.info(assetfinder_command) os.system(assetfinder_command) if 'sublist3r' in tools: sublist3r_command = 'python3 /usr/src/github/Sublist3r/sublist3r.py -d {} -t {} -o {}/from_sublister.txt'.format( domain.name, threads, results_dir) # Run sublist3r logging.info(sublist3r_command) os.system(sublist3r_command) if 'subfinder' in tools: subfinder_command = 'subfinder -d {} -t {} -o {}/from_subfinder.txt'.format( domain.name, threads, results_dir) if USE_SUBFINDER_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_SUBFINDER_CONFIG]: subfinder_command += ' -config /root/.config/subfinder/config.yaml' # Run Subfinder logging.info(subfinder_command) os.system(subfinder_command) if 'oneforall' in tools: oneforall_command = 'python3 /usr/src/github/OneForAll/oneforall.py --target {} run'.format( domain.name, results_dir) # Run OneForAll logging.info(oneforall_command) os.system(oneforall_command) extract_subdomain = "cut -d',' -f6 /usr/src/github/OneForAll/results/{}.csv >> {}/from_oneforall.txt".format( domain.name, results_dir) os.system(extract_subdomain) # remove the results from oneforall directory os.system( 'rm -rf /usr/src/github/OneForAll/results/{}.*'.format(domain.name)) ''' All tools have gathered the list of subdomains with filename initials as from_* We will gather all the results in one single file, sort them and remove the older results from_* ''' os.system( 'cat {0}/*.txt > {0}/subdomain_collection.txt'.format(results_dir)) ''' Write target domain into subdomain_collection ''' os.system( 'cat {0}/target_domain.txt >> {0}/subdomain_collection.txt'.format(results_dir)) ''' Remove all the from_* files ''' os.system('rm -f {}/from*'.format(results_dir)) ''' Sort all Subdomains ''' os.system( 'sort -u {0}/subdomain_collection.txt -o {0}/sorted_subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/subdomain_collection.txt'.format(results_dir)) ''' The final results will be stored in sorted_subdomain_collection. ''' # parse the subdomain list file and store in db with open(subdomain_scan_results_file) as subdomain_list: for _subdomain in subdomain_list: __subdomain = _subdomain.rstrip('\n') if not Subdomain.objects.filter(scan_history=task, name=__subdomain).exists( ) and validators.domain(__subdomain) and __subdomain not in out_of_scope_subdomains: subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': __subdomain, }) save_subdomain(subdomain_dict) notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: subdomains_count = Subdomain.objects.filter(scan_history=task).count() send_notification('Subdomain Gathering for target {} has been completed and has discovered *{}* subdomains.'.format(domain.name, subdomains_count)) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/sorted_subdomain_collection.txt') # check for any subdomain changes and send notif if any if notification and notification[0].send_subdomain_changes_notif: newly_added_subdomain = get_new_added_subdomain(task.id, domain.id) if newly_added_subdomain: message = "**{} New Subdomains Discovered on domain {}**".format(newly_added_subdomain.count(), domain.name) for subdomain in newly_added_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) removed_subdomain = get_removed_subdomain(task.id, domain.id) if removed_subdomain: message = "**{} Subdomains are no longer available on domain {}**".format(removed_subdomain.count(), domain.name) for subdomain in removed_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) # check for interesting subdomains and send notif if any if notification and notification[0].send_interesting_notif: interesting_subdomain = get_interesting_subdomains(task.id, domain.id) print(interesting_subdomain) if interesting_subdomain: message = "**{} Interesting Subdomains Found on domain {}**".format(interesting_subdomain.count(), domain.name) for subdomain in interesting_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) def get_new_added_subdomain(scan_id, domain_id): scan_history = ScanHistory.objects.filter( domain=domain_id).filter( subdomain_discovery=True).filter( id__lte=scan_id) if scan_history.count() > 1: last_scan = scan_history.order_by('-start_scan_date')[1] scanned_host_q1 = Subdomain.objects.filter( scan_history__id=scan_id).values('name') scanned_host_q2 = Subdomain.objects.filter( scan_history__id=last_scan.id).values('name') added_subdomain = scanned_host_q1.difference(scanned_host_q2) return Subdomain.objects.filter( scan_history=scan_id).filter( name__in=added_subdomain) def get_removed_subdomain(scan_id, domain_id): scan_history = ScanHistory.objects.filter( domain=domain_id).filter( subdomain_discovery=True).filter( id__lte=scan_id) if scan_history.count() > 1: last_scan = scan_history.order_by('-start_scan_date')[1] scanned_host_q1 = Subdomain.objects.filter( scan_history__id=scan_id).values('name') scanned_host_q2 = Subdomain.objects.filter( scan_history__id=last_scan.id).values('name') removed_subdomains = scanned_host_q2.difference(scanned_host_q1) print() return Subdomain.objects.filter( scan_history=last_scan).filter( name__in=removed_subdomains) def http_crawler(task, domain, results_dir, activity_id): ''' This function is runs right after subdomain gathering, and gathers important like page title, http status, etc HTTP Crawler runs by default ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('HTTP Crawler for target {} has been initiated.'.format(domain.name)) alive_file_location = results_dir + '/alive.txt' httpx_results_file = results_dir + '/httpx.json' subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' httpx_command = 'httpx -status-code -content-length -title -tech-detect -cdn -ip -follow-host-redirects -random-agent' proxy = get_random_proxy() if proxy: httpx_command += " --http-proxy '{}'".format(proxy) httpx_command += ' -json -o {}'.format( httpx_results_file ) httpx_command = 'cat {} | {}'.format(subdomain_scan_results_file, httpx_command) print(httpx_command) os.system(httpx_command) # alive subdomains from httpx alive_file = open(alive_file_location, 'w') # writing httpx results if os.path.isfile(httpx_results_file): httpx_json_result = open(httpx_results_file, 'r') lines = httpx_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) try: # fallback for older versions of httpx if 'url' in json_st: subdomain = Subdomain.objects.get( scan_history=task, name=json_st['input']) else: subdomain = Subdomain.objects.get( scan_history=task, name=json_st['url'].split("//")[-1]) ''' Saving Default http urls to EndPoint ''' endpoint = EndPoint() endpoint.scan_history = task endpoint.target_domain = domain endpoint.subdomain = subdomain if 'url' in json_st: endpoint.http_url = json_st['url'] subdomain.http_url = json_st['url'] if 'status-code' in json_st: endpoint.http_status = json_st['status-code'] subdomain.http_status = json_st['status-code'] if 'title' in json_st: endpoint.page_title = json_st['title'] subdomain.page_title = json_st['title'] if 'content-length' in json_st: endpoint.content_length = json_st['content-length'] subdomain.content_length = json_st['content-length'] if 'content-type' in json_st: endpoint.content_type = json_st['content-type'] subdomain.content_type = json_st['content-type'] if 'webserver' in json_st: endpoint.webserver = json_st['webserver'] subdomain.webserver = json_st['webserver'] if 'response-time' in json_st: response_time = float( ''.join( ch for ch in json_st['response-time'] if not ch.isalpha())) if json_st['response-time'][-2:] == 'ms': response_time = response_time / 1000 endpoint.response_time = response_time subdomain.response_time = response_time if 'cnames' in json_st: cname_list = ','.join(json_st['cnames']) subdomain.cname = cname_list discovered_date = timezone.now() endpoint.discovered_date = discovered_date subdomain.discovered_date = discovered_date endpoint.is_default = True endpoint.save() subdomain.save() if 'technologies' in json_st: for _tech in json_st['technologies']: if Technology.objects.filter(name=_tech).exists(): tech = Technology.objects.get(name=_tech) else: tech = Technology(name=_tech) tech.save() subdomain.technologies.add(tech) endpoint.technologies.add(tech) if 'a' in json_st: for _ip in json_st['a']: if IpAddress.objects.filter(address=_ip).exists(): ip = IpAddress.objects.get(address=_ip) else: ip = IpAddress(address=_ip) if 'cdn' in json_st: ip.is_cdn = json_st['cdn'] ip.save() subdomain.ip_addresses.add(ip) # see if to ignore 404 or 5xx alive_file.write(json_st['url'] + '\n') subdomain.save() endpoint.save() except Exception as exception: logging.error(exception) alive_file.close() if notification and notification[0].send_scan_status_notif: alive_count = Subdomain.objects.filter( scan_history__id=task.id).values('name').distinct().filter( http_status__exact=200).count() send_notification('HTTP Crawler for target {} has been completed.\n\n {} subdomains were alive (http status 200).'.format(domain.name, alive_count)) def grab_screenshot(task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for taking screenshots ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine is currently gathering screenshots for {}'.format(domain.name)) output_screenshots_path = results_dir + '/screenshots' result_csv_path = results_dir + '/screenshots/Requests.csv' alive_subdomains_path = results_dir + '/alive.txt' eyewitness_command = 'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py' eyewitness_command += ' -f {} -d {} --no-prompt'.format( alive_subdomains_path, output_screenshots_path ) if EYEWITNESS in yaml_configuration \ and TIMEOUT in yaml_configuration[EYEWITNESS] \ and yaml_configuration[EYEWITNESS][TIMEOUT] > 0: eyewitness_command += ' --timeout {}'.format( yaml_configuration[EYEWITNESS][TIMEOUT] ) if EYEWITNESS in yaml_configuration \ and THREADS in yaml_configuration[EYEWITNESS] \ and yaml_configuration[EYEWITNESS][THREADS] > 0: eyewitness_command += ' --threads {}'.format( yaml_configuration[EYEWITNESS][THREADS] ) logger.info(eyewitness_command) os.system(eyewitness_command) if os.path.isfile(result_csv_path): logger.info('Gathering Eyewitness results') with open(result_csv_path, 'r') as file: reader = csv.reader(file) for row in reader: if row[3] == 'Successful' \ and Subdomain.objects.filter( scan_history__id=task.id).filter(name=row[2]).exists(): subdomain = Subdomain.objects.get( scan_history__id=task.id, name=row[2] ) subdomain.screenshot_path = row[4].replace( '/usr/src/scan_results/', '' ) subdomain.save() # remove all db, html extra files in screenshot results os.system('rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format( output_screenshots_path, )) os.system('rm -rf {0}/source'.format( output_screenshots_path, )) if notification and notification[0].send_scan_status_notif: send_notification('reNgine has finished gathering screenshots for {}'.format(domain.name)) def port_scanning(task, domain, yaml_configuration, results_dir): ''' This function is responsible for running the port scan ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Port Scan initiated for {}'.format(domain.name)) subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' port_results_file = results_dir + '/ports.json' # check the yaml_configuration and choose the ports to be scanned scan_ports = '-' # default port scan everything if PORTS in yaml_configuration[PORT_SCAN]: # TODO: legacy code, remove top-100 in future versions all_ports = yaml_configuration[PORT_SCAN][PORTS] if 'full' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -p {}'.format( subdomain_scan_results_file, port_results_file, '-') elif 'top-100' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -top-ports 100'.format( subdomain_scan_results_file, port_results_file) elif 'top-1000' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -top-ports 1000'.format( subdomain_scan_results_file, port_results_file) else: scan_ports = ','.join( str(port) for port in all_ports) naabu_command = 'cat {} | naabu -json -o {} -p {}'.format( subdomain_scan_results_file, port_results_file, scan_ports) # check for exclude ports if EXCLUDE_PORTS in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][EXCLUDE_PORTS]: exclude_ports = ','.join( str(port) for port in yaml_configuration['port_scan']['exclude_ports']) naabu_command = naabu_command + \ ' -exclude-ports {}'.format(exclude_ports) if NAABU_RATE in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][NAABU_RATE] > 0: naabu_command = naabu_command + \ ' -rate {}'.format( yaml_configuration[PORT_SCAN][NAABU_RATE]) if USE_NAABU_CONFIG in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][USE_NAABU_CONFIG]: naabu_command += ' -config /root/.config/naabu/naabu.conf' # run naabu os.system(naabu_command) # writing port results try: port_json_result = open(port_results_file, 'r') lines = port_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) port_number = json_st['port'] ip_address = json_st['ip'] # see if port already exists if Port.objects.filter(number__exact=port_number).exists(): port = Port.objects.get(number=port_number) else: port = Port() port.number = port_number if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port_detail = whatportis.get_ports(str(port_number)) if len(port_detail): port.service_name = port_detail[0].name port.description = port_detail[0].description port.save() if IpAddress.objects.filter(address=json_st['ip']).exists(): ip = IpAddress.objects.get(address=json_st['ip']) ip.ports.add(port) ip.save() except BaseException as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: port_count = Port.objects.filter( ports__in=IpAddress.objects.filter( ip_addresses__in=Subdomain.objects.filter( scan_history__id=task.id))).distinct().count() send_notification('reNgine has finished Port Scanning on {} and has identified {} ports.'.format(domain.name, port_count)) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/ports.json') def check_waf(): ''' This function will check for the WAF being used in subdomains using wafw00f ''' pass def directory_brute(task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for performing directory scan ''' # scan directories for all the alive subdomain with http status > # 200 notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Directory Bruteforce has been initiated for {}.'.format(domain.name)) alive_subdomains = Subdomain.objects.filter( scan_history__id=task.id).exclude(http_url__isnull=True) dirs_results = results_dir + '/dirs.json' # check the yaml settings if EXTENSIONS in yaml_configuration[DIR_FILE_SEARCH]: extensions = ','.join( str(ext) for ext in yaml_configuration[DIR_FILE_SEARCH][EXTENSIONS]) else: extensions = 'php,git,yaml,conf,db,mysql,bak,txt' # Threads if THREADS in yaml_configuration[DIR_FILE_SEARCH] \ and yaml_configuration[DIR_FILE_SEARCH][THREADS] > 0: threads = yaml_configuration[DIR_FILE_SEARCH][THREADS] else: threads = 10 for subdomain in alive_subdomains: # delete any existing dirs.json if os.path.isfile(dirs_results): os.system('rm -rf {}'.format(dirs_results)) dirsearch_command = 'python3 /usr/src/github/dirsearch/dirsearch.py' dirsearch_command += ' -u {}'.format(subdomain.http_url) if (WORDLIST not in yaml_configuration[DIR_FILE_SEARCH] or not yaml_configuration[DIR_FILE_SEARCH][WORDLIST] or 'default' in yaml_configuration[DIR_FILE_SEARCH][WORDLIST]): wordlist_location = '/usr/src/github/dirsearch/db/dicc.txt' else: wordlist_location = '/usr/src/wordlist/' + \ yaml_configuration[DIR_FILE_SEARCH][WORDLIST] + '.txt' dirsearch_command += ' -w {}'.format(wordlist_location) dirsearch_command += ' --format json -o {}'.format(dirs_results) dirsearch_command += ' -e {}'.format(extensions) dirsearch_command += ' -t {}'.format(threads) dirsearch_command += ' --random-agent --follow-redirects --exclude-status 403,401,404' if EXCLUDE_EXTENSIONS in yaml_configuration[DIR_FILE_SEARCH]: exclude_extensions = ','.join( str(ext) for ext in yaml_configuration[DIR_FILE_SEARCH][EXCLUDE_EXTENSIONS]) dirsearch_command += ' -X {}'.format(exclude_extensions) if EXCLUDE_TEXT in yaml_configuration[DIR_FILE_SEARCH]: exclude_text = ','.join( str(text) for text in yaml_configuration[DIR_FILE_SEARCH][EXCLUDE_TEXT]) dirsearch_command += ' -exclude-texts {}'.format(exclude_text) # check if recursive strategy is set to on if RECURSIVE_LEVEL in yaml_configuration[DIR_FILE_SEARCH]: dirsearch_command += ' --recursion-depth {}'.format(yaml_configuration[DIR_FILE_SEARCH][RECURSIVE_LEVEL]) if RECURSIVE_LEVEL in yaml_configuration[DIR_FILE_SEARCH]: dirsearch_command += ' --recursion-depth {}'.format(yaml_configuration[DIR_FILE_SEARCH][RECURSIVE_LEVEL]) # proxy proxy = get_random_proxy() if proxy: dirsearch_command += " --proxy '{}'".format(proxy) print(dirsearch_command) os.system(dirsearch_command) try: if os.path.isfile(dirs_results): with open(dirs_results, "r") as json_file: json_string = json_file.read() subdomain = Subdomain.objects.get( scan_history__id=task.id, http_url=subdomain.http_url) subdomain.directory_json = json_string subdomain.save() except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: send_notification('Directory Bruteforce has been completed for {}.'.format(domain.name)) def fetch_endpoints( task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for fetching all the urls associated with target and run HTTP probe It first runs gau to gather all urls from wayback, then we will use hakrawler to identify more urls ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine is currently gathering endpoints for {}.'.format(domain.name)) # check yaml settings if ALL in yaml_configuration[FETCH_URL][USES_TOOLS]: tools = 'gauplus hakrawler waybackurls gospider' else: tools = ' '.join( str(tool) for tool in yaml_configuration[FETCH_URL][USES_TOOLS]) if INTENSITY in yaml_configuration[FETCH_URL]: scan_type = yaml_configuration[FETCH_URL][INTENSITY] else: scan_type = 'normal' domain_regex = "\'https?://([a-z0-9]+[.])*{}.*\'".format(domain.name) if 'deep' in scan_type: # performs deep url gathering for all the subdomains present - # RECOMMENDED logger.info('Deep URLS Fetch') os.system(settings.TOOL_LOCATION + 'get_urls.sh %s %s %s %s %s' % ("None", results_dir, scan_type, domain_regex, tools)) else: # perform url gathering only for main domain - USE only for quick scan logger.info('Non Deep URLS Fetch') os.system( settings.TOOL_LOCATION + 'get_urls.sh %s %s %s %s %s' % ( domain.name, results_dir, scan_type, domain_regex, tools )) if IGNORE_FILE_EXTENSION in yaml_configuration[FETCH_URL]: ignore_extension = '|'.join( yaml_configuration[FETCH_URL][IGNORE_FILE_EXTENSION]) logger.info('Ignore extensions' + ignore_extension) os.system( 'cat {0}/all_urls.txt | grep -Eiv "\\.({1}).*" > {0}/temp_urls.txt'.format( results_dir, ignore_extension)) os.system( 'rm {0}/all_urls.txt && mv {0}/temp_urls.txt {0}/all_urls.txt'.format(results_dir)) ''' Store all the endpoints and then run the httpx ''' try: endpoint_final_url = results_dir + '/all_urls.txt' if os.path.isfile(endpoint_final_url): with open(endpoint_final_url) as endpoint_list: for url in endpoint_list: http_url = url.rstrip('\n') if not EndPoint.objects.filter(scan_history=task, http_url=http_url).exists(): _subdomain = get_subdomain_from_url(http_url) if Subdomain.objects.filter( scan_history=task).filter( name=_subdomain).exists(): subdomain = Subdomain.objects.get( scan_history=task, name=_subdomain) else: ''' gau or gosppider can gather interesting endpoints which when parsed can give subdomains that were not existent from subdomain scan. so storing them ''' logger.error( 'Subdomain {} not found, adding...'.format(_subdomain)) subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': _subdomain, }) subdomain = save_subdomain(subdomain_dict) endpoint_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'subdomain': subdomain, 'http_url': http_url, }) save_endpoint(endpoint_dict) except Exception as e: logger.error(e) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/all_urls.txt') ''' TODO: Go spider & waybackurls accumulates a lot of urls, which is good but nuclei takes forever to scan even a simple website, so we will do http probing and filter HTTP status 404, this way we can reduce the number of Non Existent URLS ''' logger.info('HTTP Probing on collected endpoints') httpx_command = 'httpx -l {0}/all_urls.txt -status-code -content-length -ip -cdn -title -tech-detect -json -follow-redirects -random-agent -o {0}/final_httpx_urls.json'.format(results_dir) proxy = get_random_proxy() if proxy: httpx_command += " --http-proxy '{}'".format(proxy) os.system(httpx_command) url_results_file = results_dir + '/final_httpx_urls.json' try: urls_json_result = open(url_results_file, 'r') lines = urls_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) http_url = json_st['url'] _subdomain = get_subdomain_from_url(http_url) if Subdomain.objects.filter( scan_history=task).filter( name=_subdomain).exists(): subdomain_obj = Subdomain.objects.get( scan_history=task, name=_subdomain) else: subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': _subdomain, }) subdomain_obj = save_subdomain(subdomain_dict) if EndPoint.objects.filter( scan_history=task).filter( http_url=http_url).exists(): endpoint = EndPoint.objects.get( scan_history=task, http_url=http_url) else: endpoint = EndPoint() endpoint_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'http_url': http_url, 'subdomain': subdomain_obj }) endpoint = save_endpoint(endpoint_dict) if 'title' in json_st: endpoint.page_title = json_st['title'] if 'webserver' in json_st: endpoint.webserver = json_st['webserver'] if 'content-length' in json_st: endpoint.content_length = json_st['content-length'] if 'content-type' in json_st: endpoint.content_type = json_st['content-type'] if 'status-code' in json_st: endpoint.http_status = json_st['status-code'] if 'response-time' in json_st: response_time = float(''.join(ch for ch in json_st['response-time'] if not ch.isalpha())) if json_st['response-time'][-2:] == 'ms': response_time = response_time / 1000 endpoint.response_time = response_time endpoint.save() if 'technologies' in json_st: for _tech in json_st['technologies']: if Technology.objects.filter(name=_tech).exists(): tech = Technology.objects.get(name=_tech) else: tech = Technology(name=_tech) tech.save() endpoint.technologies.add(tech) # get subdomain object subdomain = Subdomain.objects.get(scan_history=task, name=_subdomain) subdomain.technologies.add(tech) subdomain.save() except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: endpoint_count = EndPoint.objects.filter( scan_history__id=task.id).values('http_url').distinct().count() endpoint_alive_count = EndPoint.objects.filter( scan_history__id=task.id, http_status__exact=200).values('http_url').distinct().count() send_notification('reNgine has finished gathering endpoints for {} and has discovered *{}* unique endpoints.\n\n{} of those endpoints reported HTTP status 200.'.format( domain.name, endpoint_count, endpoint_alive_count )) # once endpoint is saved, run gf patterns TODO: run threads if GF_PATTERNS in yaml_configuration[FETCH_URL]: for pattern in yaml_configuration[FETCH_URL][GF_PATTERNS]: logger.info('Running GF for {}'.format(pattern)) gf_output_file_path = '{0}/gf_patterns_{1}.txt'.format( results_dir, pattern) gf_command = 'cat {0}/all_urls.txt | gf {1} >> {2}'.format( results_dir, pattern, gf_output_file_path) os.system(gf_command) if os.path.exists(gf_output_file_path): with open(gf_output_file_path) as gf_output: for line in gf_output: url = line.rstrip('\n') try: endpoint = EndPoint.objects.get( scan_history=task, http_url=url) earlier_pattern = endpoint.matched_gf_patterns new_pattern = earlier_pattern + ',' + pattern if earlier_pattern else pattern endpoint.matched_gf_patterns = new_pattern except Exception as e: # add the url in db logger.error(e) logger.info('Adding URL' + url) endpoint = EndPoint() endpoint.http_url = url endpoint.target_domain = domain endpoint.scan_history = task try: _subdomain = Subdomain.objects.get( scan_history=task, name=get_subdomain_from_url(url)) endpoint.subdomain = _subdomain except Exception as e: continue endpoint.matched_gf_patterns = pattern finally: endpoint.save() def vulnerability_scan( task, domain, yaml_configuration, results_dir, activity_id): notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Vulnerability scan has been initiated for {}.'.format(domain.name)) ''' This function will run nuclei as a vulnerability scanner ---- unfurl the urls to keep only domain and path, this will be sent to vuln scan ignore certain file extensions Thanks: https://github.com/six2dez/reconftw ''' urls_path = '/alive.txt' if task.scan_type.fetch_url: os.system('cat {0}/all_urls.txt | grep -Eiv "\\.(eot|jpg|jpeg|gif|css|tif|tiff|png|ttf|otf|woff|woff2|ico|pdf|svg|txt|js|doc|docx)$" | unfurl -u format %s://%d%p >> {0}/unfurl_urls.txt'.format(results_dir)) os.system( 'sort -u {0}/unfurl_urls.txt -o {0}/unfurl_urls.txt'.format(results_dir)) urls_path = '/unfurl_urls.txt' vulnerability_result_path = results_dir + '/vulnerability.json' vulnerability_scan_input_file = results_dir + urls_path nuclei_command = 'nuclei -json -l {} -o {}'.format( vulnerability_scan_input_file, vulnerability_result_path) # check nuclei config if USE_NUCLEI_CONFIG in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[VULNERABILITY_SCAN][USE_NUCLEI_CONFIG]: nuclei_command += ' -config /root/.config/nuclei/config.yaml' ''' Nuclei Templates Either custom template has to be supplied or default template, if neither has been supplied then use all templates including custom templates ''' if CUSTOM_NUCLEI_TEMPLATE in yaml_configuration[ VULNERABILITY_SCAN] or NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: # check yaml settings for templates if NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: if ALL in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_TEMPLATE]: template = NUCLEI_TEMPLATES_PATH else: _template = ','.join([NUCLEI_TEMPLATES_PATH + str(element) for element in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_TEMPLATE]]) template = _template.replace(',', ' -t ') # Update nuclei command with templates nuclei_command = nuclei_command + ' -t ' + template if CUSTOM_NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: # add .yaml to the custom template extensions _template = ','.join( [str(element) + '.yaml' for element in yaml_configuration[VULNERABILITY_SCAN][CUSTOM_NUCLEI_TEMPLATE]]) template = _template.replace(',', ' -t ') # Update nuclei command with templates nuclei_command = nuclei_command + ' -t ' + template else: nuclei_command = nuclei_command + ' -t /root/nuclei-templates' # check yaml settings for concurrency if NUCLEI_CONCURRENCY in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][NUCLEI_CONCURRENCY] > 0: concurrency = yaml_configuration[VULNERABILITY_SCAN][NUCLEI_CONCURRENCY] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -c ' + str(concurrency) if RATE_LIMIT in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][RATE_LIMIT] > 0: rate_limit = yaml_configuration[VULNERABILITY_SCAN][RATE_LIMIT] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -rl ' + str(rate_limit) if TIMEOUT in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][TIMEOUT] > 0: timeout = yaml_configuration[VULNERABILITY_SCAN][TIMEOUT] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -timeout ' + str(timeout) if RETRIES in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][RETRIES] > 0: retries = yaml_configuration[VULNERABILITY_SCAN][RETRIES] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -retries ' + str(retries) # for severity if NUCLEI_SEVERITY in yaml_configuration[VULNERABILITY_SCAN] and ALL not in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_SEVERITY]: _severity = ','.join( [str(element) for element in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_SEVERITY]]) severity = _severity.replace(" ", "") else: severity = "critical, high, medium, low, info" # update nuclei templates before running scan os.system('nuclei -update-templates') for _severity in severity.split(","): # delete any existing vulnerability.json file if os.path.isfile(vulnerability_result_path): os.system('rm {}'.format(vulnerability_result_path)) # run nuclei final_nuclei_command = nuclei_command + ' -severity ' + _severity proxy = get_random_proxy() if proxy: final_nuclei_command += " --proxy-url '{}'".format(proxy) logger.info(final_nuclei_command) os.system(final_nuclei_command) try: if os.path.isfile(vulnerability_result_path): urls_json_result = open(vulnerability_result_path, 'r') lines = urls_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) host = json_st['host'] _subdomain = get_subdomain_from_url(host) try: subdomain = Subdomain.objects.get( name=_subdomain, scan_history=task) vulnerability = Vulnerability() vulnerability.subdomain = subdomain vulnerability.scan_history = task vulnerability.target_domain = domain try: endpoint = EndPoint.objects.get( scan_history=task, target_domain=domain, http_url=host) vulnerability.endpoint = endpoint except Exception as exception: logger.error(exception) if 'name' in json_st['info']: vulnerability.name = json_st['info']['name'] if 'severity' in json_st['info']: if json_st['info']['severity'] == 'info': severity = 0 elif json_st['info']['severity'] == 'low': severity = 1 elif json_st['info']['severity'] == 'medium': severity = 2 elif json_st['info']['severity'] == 'high': severity = 3 elif json_st['info']['severity'] == 'critical': severity = 4 else: severity = 0 else: severity = 0 vulnerability.severity = severity if 'tags' in json_st['info']: vulnerability.tags = json_st['info']['tags'] if 'description' in json_st['info']: vulnerability.description = json_st['info']['description'] if 'reference' in json_st['info']: vulnerability.reference = json_st['info']['reference'] if 'matched' in json_st: vulnerability.http_url = json_st['matched'] if 'templateID' in json_st: vulnerability.template_used = json_st['templateID'] if 'description' in json_st: vulnerability.description = json_st['description'] if 'matcher_name' in json_st: vulnerability.matcher_name = json_st['matcher_name'] if 'extracted_results' in json_st: vulnerability.extracted_results = json_st['extracted_results'] vulnerability.discovered_date = timezone.now() vulnerability.open_status = True vulnerability.save() # send notification for all vulnerabilities except info if json_st['info']['severity'] != "info" and notification and notification[0].send_vuln_notif: message = "*Alert: Vulnerability Identified*" message += "\n\n" message += "A *{}* severity vulnerability has been identified.".format(json_st['info']['severity']) message += "\nVulnerability Name: {}".format(json_st['info']['name']) message += "\nVulnerable URL: {}".format(json_st['host']) send_notification(message) # send report to hackerone if Hackerone.objects.all().exists() and json_st['info']['severity'] != 'info' and json_st['info']['severity'] \ != 'low' and vulnerability.target_domain.h1_team_handle: hackerone = Hackerone.objects.all()[0] if hackerone.send_critical and json_st['info']['severity'] == 'critical': send_hackerone_report(vulnerability.id) elif hackerone.send_high and json_st['info']['severity'] == 'high': send_hackerone_report(vulnerability.id) elif hackerone.send_medium and json_st['info']['severity'] == 'medium': send_hackerone_report(vulnerability.id) except ObjectDoesNotExist: logger.error('Object not found') continue except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: info_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=0).count() low_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=1).count() medium_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=2).count() high_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=3).count() critical_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=4).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count message = 'Vulnerability scan has been completed for {} and discovered {} vulnerabilities.'.format( domain.name, vulnerability_count ) message += '\n\n*Vulnerability Stats:*' message += '\nCritical: {}'.format(critical_count) message += '\nHigh: {}'.format(high_count) message += '\nMedium: {}'.format(medium_count) message += '\nLow: {}'.format(low_count) message += '\nInfo: {}'.format(info_count) send_notification(message) def scan_failed(task): task.scan_status = 0 task.stop_scan_date = timezone.now() task.save() def create_scan_activity(task, message, status): scan_activity = ScanActivity() scan_activity.scan_of = task scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id def update_last_activity(id, activity_status): ScanActivity.objects.filter( id=id).update( status=activity_status, time=timezone.now()) def delete_scan_data(results_dir): # remove all txt,html,json files os.system('find {} -name "*.txt" -type f -delete'.format(results_dir)) os.system('find {} -name "*.html" -type f -delete'.format(results_dir)) os.system('find {} -name "*.json" -type f -delete'.format(results_dir)) def save_subdomain(subdomain_dict): subdomain = Subdomain() subdomain.discovered_date = timezone.now() subdomain.target_domain = subdomain_dict.get('target_domain') subdomain.scan_history = subdomain_dict.get('scan_history') subdomain.name = subdomain_dict.get('name') subdomain.http_url = subdomain_dict.get('http_url') subdomain.screenshot_path = subdomain_dict.get('screenshot_path') subdomain.http_header_path = subdomain_dict.get('http_header_path') subdomain.cname = subdomain_dict.get('cname') subdomain.is_cdn = subdomain_dict.get('is_cdn') subdomain.content_type = subdomain_dict.get('content_type') subdomain.webserver = subdomain_dict.get('webserver') subdomain.page_title = subdomain_dict.get('page_title') subdomain.is_imported_subdomain = subdomain_dict.get( 'is_imported_subdomain') if 'is_imported_subdomain' in subdomain_dict else False if 'http_status' in subdomain_dict: subdomain.http_status = subdomain_dict.get('http_status') if 'response_time' in subdomain_dict: subdomain.response_time = subdomain_dict.get('response_time') if 'content_length' in subdomain_dict: subdomain.content_length = subdomain_dict.get('content_length') subdomain.save() return subdomain def save_endpoint(endpoint_dict): endpoint = EndPoint() endpoint.discovered_date = timezone.now() endpoint.scan_history = endpoint_dict.get('scan_history') endpoint.target_domain = endpoint_dict.get('target_domain') if 'target_domain' in endpoint_dict else None endpoint.subdomain = endpoint_dict.get('subdomain') if 'target_domain' in endpoint_dict else None endpoint.http_url = endpoint_dict.get('http_url') endpoint.page_title = endpoint_dict.get('page_title') if 'page_title' in endpoint_dict else None endpoint.content_type = endpoint_dict.get('content_type') if 'content_type' in endpoint_dict else None endpoint.webserver = endpoint_dict.get('webserver') if 'webserver' in endpoint_dict else None endpoint.response_time = endpoint_dict.get('response_time') if 'response_time' in endpoint_dict else 0 endpoint.http_status = endpoint_dict.get('http_status') if 'http_status' in endpoint_dict else 0 endpoint.content_length = endpoint_dict.get('content_length') if 'content_length' in endpoint_dict else 0 endpoint.is_default = endpoint_dict.get('is_default') if 'is_default' in endpoint_dict else False endpoint.save() return endpoint def perform_osint(task, domain, yaml_configuration, results_dir): notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine has initiated OSINT on target {}'.format(domain.name)) if 'discover' in yaml_configuration[OSINT]: osint_discovery(task, domain, yaml_configuration, results_dir) if 'dork' in yaml_configuration[OSINT]: dorking(task, yaml_configuration) if notification and notification[0].send_scan_status_notif: send_notification('reNgine has completed performing OSINT on target {}'.format(domain.name)) def osint_discovery(task, domain, yaml_configuration, results_dir): if ALL in yaml_configuration[OSINT][OSINT_DISCOVER]: osint_lookup = 'emails metainfo employees' else: osint_lookup = ' '.join( str(lookup) for lookup in yaml_configuration[OSINT][OSINT_DISCOVER]) if 'metainfo' in osint_lookup: if INTENSITY in yaml_configuration[OSINT]: osint_intensity = yaml_configuration[OSINT][INTENSITY] else: osint_intensity = 'normal' if OSINT_DOCUMENTS_LIMIT in yaml_configuration[OSINT]: documents_limit = yaml_configuration[OSINT][OSINT_DOCUMENTS_LIMIT] else: documents_limit = 50 if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': domain.name, 'domain': domain, 'scan_id': task, 'documents_limit': documents_limit }) get_and_save_meta_info(meta_dict) elif osint_intensity == 'deep': # get all subdomains in scan_id subdomains = Subdomain.objects.filter(scan_history=task) for subdomain in subdomains: meta_dict = DottedDict({ 'osint_target': subdomain.name, 'domain': domain, 'scan_id': task, 'documents_limit': documents_limit }) get_and_save_meta_info(meta_dict) if 'emails' in osint_lookup: get_and_save_emails(task, results_dir) get_and_save_leaked_credentials(task, results_dir) if 'employees' in osint_lookup: get_and_save_employees(task, results_dir) def dorking(scan_history, yaml_configuration): # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh # look in stackoverflow if ALL in yaml_configuration[OSINT][OSINT_DORK]: dork_lookup = 'stackoverflow, 3rdparty, social_media, project_management, code_sharing, config_files, jenkins, cloud_buckets, php_error, exposed_documents, struts_rce, db_files, traefik, git_exposed' else: dork_lookup = ' '.join( str(lookup) for lookup in yaml_configuration[OSINT][OSINT_DORK]) if 'stackoverflow' in dork_lookup: dork = 'site:stackoverflow.com' dork_type = 'stackoverflow' get_and_save_dork_results( dork, dork_type, scan_history, in_target=False ) if '3rdparty' in dork_lookup: # look in 3rd party sitee dork_type = '3rdparty' lookup_websites = [ 'gitter.im', 'papaly.com', 'productforums.google.com', 'coggle.it', 'replt.it', 'ycombinator.com', 'libraries.io', 'npm.runkit.com', 'npmjs.com', 'scribd.com', 'gitter.im' ] dork = '' for website in lookup_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'social_media' in dork_lookup: dork_type = 'Social Media' social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'pinterest.com', 'tumblr.com', 'reddit.com' ] dork = '' for website in social_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'project_management' in dork_lookup: dork_type = 'Project Management' project_websites = [ 'trello.com', '*.atlassian.net' ] dork = '' for website in project_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'code_sharing' in dork_lookup: dork_type = 'Code Sharing Sites' code_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] dork = '' for website in code_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'config_files' in dork_lookup: dork_type = 'Config Files' config_file_ext = [ 'env', 'xml', 'conf', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] dork = '' for extension in config_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'jenkins' in dork_lookup: dork_type = 'Jenkins' dork = 'intitle:\"Dashboard [Jenkins]\"' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) if 'wordpress_files' in dork_lookup: dork_type = 'Wordpress Files' inurl_lookup = [ 'wp-content', 'wp-includes' ] dork = '' for lookup in inurl_lookup: dork = dork + ' | ' + 'inurl:' + lookup get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'cloud_buckets' in dork_lookup: dork_type = 'Cloud Buckets' cloud_websites = [ '.s3.amazonaws.com', 'storage.googleapis.com', 'amazonaws.com' ] dork = '' for website in cloud_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'php_error' in dork_lookup: dork_type = 'PHP Error' error_words = [ '\"PHP Parse error\"', '\"PHP Warning\"', '\"PHP Error\"' ] dork = '' for word in error_words: dork = dork + ' | ' + word get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'exposed_documents' in dork_lookup: dork_type = 'Exposed Documents' docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] dork = '' for extension in docs_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'struts_rce' in dork_lookup: dork_type = 'Apache Struts RCE' struts_file_ext = [ 'action', 'struts', 'do' ] dork = '' for extension in struts_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'db_files' in dork_lookup: dork_type = 'Database Files' db_file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] dork = '' for extension in db_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'traefik' in dork_lookup: dork = 'intitle:traefik inurl:8080/dashboard' dork_type = 'Traefik' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) if 'git_exposed' in dork_lookup: dork = 'inurl:\"/.git\"' dork_type = '.git Exposed' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) def get_and_save_dork_results(dork, type, scan_history, in_target=False): degoogle_obj = degoogle.dg() proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy if in_target: query = dork + " site:" + scan_history.domain.name else: query = dork + " \"{}\"".format(scan_history.domain.name) logger.info(query) degoogle_obj.query = query results = degoogle_obj.run() logger.info(results) for result in results: dork, _ = Dork.objects.get_or_create( type=type, description=result['desc'], url=result['url'] ) scan_history.dorks.add(dork) def get_and_save_employees(scan_history, results_dir): theHarvester_location = '/usr/src/github/theHarvester' # update proxies.yaml if Proxy.objects.all().exists(): proxy = Proxy.objects.all()[0] if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(theHarvester_location + '/proxies.yaml', 'w') as file: documents = yaml.dump(yaml_data, file) os.system('cd {} && python3 theHarvester.py -d {} -b all -f {}/theHarvester.html'.format( theHarvester_location, scan_history.domain.name, results_dir )) file_location = results_dir + '/theHarvester.html' print(file_location) # delete proxy environ var if os.environ.get(('https_proxy')): del os.environ['https_proxy'] if os.environ.get(('HTTPS_PROXY')): del os.environ['HTTPS_PROXY'] if os.path.isfile(file_location): logger.info('Parsing theHarvester results') options = FirefoxOptions() options.add_argument("--headless") driver = webdriver.Firefox(options=options) driver.get('file://'+file_location) tabledata = driver.execute_script('return tabledata') # save email addresses and linkedin employees for data in tabledata: if data['record'] == 'email': _email = data['result'] email, _ = Email.objects.get_or_create(address=_email) scan_history.emails.add(email) elif data['record'] == 'people': _employee = data['result'] split_val = _employee.split('-') name = split_val[0] if len(split_val) == 2: designation = split_val[1] else: designation = "" employee, _ = Employee.objects.get_or_create(name=name, designation=designation) scan_history.employees.add(employee) driver.quit() print(tabledata) def get_and_save_emails(scan_history, results_dir): leak_target_path = '{}/creds_target.txt'.format(results_dir) # get email address proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy emails = [] try: logger.info('OSINT: Getting emails from Google') email_from_google = get_emails_from_google(scan_history.domain.name) logger.info('OSINT: Getting emails from Bing') email_from_bing = get_emails_from_bing(scan_history.domain.name) logger.info('OSINT: Getting emails from Baidu') email_from_baidu = get_emails_from_baidu(scan_history.domain.name) emails = list(set(email_from_google + email_from_bing + email_from_baidu)) logger.info(emails) except Exception as e: logger.error(e) leak_target_file = open(leak_target_path, 'w') for _email in emails: email, _ = Email.objects.get_or_create(address=_email) scan_history.emails.add(email) leak_target_file.write('{}\n'.format(_email)) # fill leak_target_file with possible email address leak_target_file.write('%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.write('%.%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%.%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.write('%_%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%_%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.close() def get_and_save_leaked_credentials(scan_history, results_dir): logger.info('OSINT: Getting leaked credentials...') leak_target_file = '{}/creds_target.txt'.format(results_dir) leak_output_file = '{}/pwndb.json'.format(results_dir) pwndb_command = 'python3 /usr/src/github/pwndb/pwndb.py --proxy tor:9150 --output json --list {}'.format( leak_target_file ) try: pwndb_output = subprocess.getoutput(pwndb_command) creds = json.loads(pwndb_output) for cred in creds: if cred['username'] != 'donate': email_id = "{}@{}".format(cred['username'], cred['domain']) email_obj, _ = Email.objects.get_or_create( address=email_id, ) email_obj.password = cred['password'] email_obj.save() scan_history.emails.add(email_obj) except Exception as e: logger.error(e) pass def get_and_save_meta_info(meta_dict): logger.info('Getting METADATA for {}'.format(meta_dict.osint_target)) proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy result = metadata_extractor.extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if result: results = result.get_metadata() for meta in results: meta_finder_document = MetaFinderDocument() subdomain = Subdomain.objects.get(scan_history=meta_dict.scan_id, name=meta_dict.osint_target) meta_finder_document.subdomain = subdomain meta_finder_document.target_domain = meta_dict.domain meta_finder_document.scan_history = meta_dict.scan_id item = DottedDict(results[meta]) meta_finder_document.url = item.url meta_finder_document.doc_name = meta meta_finder_document.http_status = item.status_code metadata = results[meta]['metadata'] for data in metadata: if 'Producer' in metadata and metadata['Producer']: meta_finder_document.producer = metadata['Producer'].rstrip('\x00') if 'Creator' in metadata and metadata['Creator']: meta_finder_document.creator = metadata['Creator'].rstrip('\x00') if 'CreationDate' in metadata and metadata['CreationDate']: meta_finder_document.creation_date = metadata['CreationDate'].rstrip('\x00') if 'ModDate' in metadata and metadata['ModDate']: meta_finder_document.modified_date = metadata['ModDate'].rstrip('\x00') if 'Author' in metadata and metadata['Author']: meta_finder_document.author = metadata['Author'].rstrip('\x00') if 'Title' in metadata and metadata['Title']: meta_finder_document.title = metadata['Title'].rstrip('\x00') if 'OSInfo' in metadata and metadata['OSInfo']: meta_finder_document.os = metadata['OSInfo'].rstrip('\x00') meta_finder_document.save() @app.task(bind=True) def test_task(self): print('*' * 40) print('test task run') print('*' * 40)

import os import traceback import yaml import json import csv import validators import random import requests import logging import metafinder.extractor as metadata_extractor import whatportis import subprocess from selenium.webdriver.firefox.options import Options as FirefoxOptions from selenium import webdriver from emailfinder.extractor import * from dotted_dict import DottedDict from celery import shared_task from discord_webhook import DiscordWebhook from reNgine.celery import app from startScan.models import * from targetApp.models import Domain from scanEngine.models import EngineType from django.conf import settings from django.shortcuts import get_object_or_404 from celery import shared_task from datetime import datetime from degoogle import degoogle from django.conf import settings from django.utils import timezone, dateformat from django.shortcuts import get_object_or_404 from django.core.exceptions import ObjectDoesNotExist from reNgine.celery import app from reNgine.definitions import * from startScan.models import * from targetApp.models import Domain from scanEngine.models import EngineType, Configuration, Wordlist from .common_func import * ''' task for background scan ''' @app.task def initiate_scan( domain_id, scan_history_id, scan_type, engine_type, imported_subdomains=None, out_of_scope_subdomains=[] ): ''' scan_type = 0 -> immediate scan, need not create scan object scan_type = 1 -> scheduled scan ''' engine_object = EngineType.objects.get(pk=engine_type) domain = Domain.objects.get(pk=domain_id) if scan_type == 1: task = ScanHistory() task.scan_status = -1 elif scan_type == 0: task = ScanHistory.objects.get(pk=scan_history_id) # save the last scan date for domain model domain.last_scan_date = timezone.now() domain.save() # once the celery task starts, change the task status to Started task.scan_type = engine_object task.celery_id = initiate_scan.request.id task.domain = domain task.scan_status = 1 task.start_scan_date = timezone.now() task.subdomain_discovery = True if engine_object.subdomain_discovery else False task.dir_file_search = True if engine_object.dir_file_search else False task.port_scan = True if engine_object.port_scan else False task.fetch_url = True if engine_object.fetch_url else False task.osint = True if engine_object.osint else False task.screenshot = True if engine_object.screenshot else False task.vulnerability_scan = True if engine_object.vulnerability_scan else False task.save() activity_id = create_scan_activity(task, "Scanning Started", 2) results_dir = '/usr/src/scan_results/' os.chdir(results_dir) notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine has initiated recon for target {} with engine type {}'.format(domain.name, engine_object.engine_name)) try: current_scan_dir = domain.name + '_' + str(random.randint(100000000000, 999999999999)) os.mkdir(current_scan_dir) task.results_dir = current_scan_dir task.save() except Exception as exception: logger.error(exception) scan_failed(task) yaml_configuration = None excluded_subdomains = '' try: yaml_configuration = yaml.load( task.scan_type.yaml_configuration, Loader=yaml.FullLoader) except Exception as exception: logger.error(exception) # TODO: Put failed reason on db ''' Add GF patterns name to db for dynamic URLs menu ''' if engine_object.fetch_url and GF_PATTERNS in yaml_configuration[FETCH_URL]: task.used_gf_patterns = ','.join( pattern for pattern in yaml_configuration[FETCH_URL][GF_PATTERNS]) task.save() results_dir = results_dir + current_scan_dir # put all imported subdomains into txt file and also in Subdomain model if imported_subdomains: extract_imported_subdomain( imported_subdomains, task, domain, results_dir) if yaml_configuration: ''' a target in itself is a subdomain, some tool give subdomains as www.yogeshojha.com but url and everything else resolves to yogeshojha.com In that case, we would already need to store target itself as subdomain ''' initial_subdomain_file = '/target_domain.txt' if task.subdomain_discovery else '/sorted_subdomain_collection.txt' subdomain_file = open(results_dir + initial_subdomain_file, "w") subdomain_file.write(domain.name + "\n") subdomain_file.close() if(task.subdomain_discovery): activity_id = create_scan_activity(task, "Subdomain Scanning", 1) subdomain_scan( task, domain, yaml_configuration, results_dir, activity_id, out_of_scope_subdomains ) else: skip_subdomain_scan(task, domain, results_dir) update_last_activity(activity_id, 2) activity_id = create_scan_activity(task, "HTTP Crawler", 1) http_crawler( task, domain, results_dir, activity_id) update_last_activity(activity_id, 2) try: if task.screenshot: activity_id = create_scan_activity( task, "Visual Recon - Screenshot", 1) grab_screenshot( task, domain, yaml_configuration, current_scan_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if(task.port_scan): activity_id = create_scan_activity(task, "Port Scanning", 1) port_scanning(task, domain, yaml_configuration, results_dir) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.osint: activity_id = create_scan_activity(task, "OSINT Running", 1) perform_osint(task, domain, yaml_configuration, results_dir) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.dir_file_search: activity_id = create_scan_activity(task, "Directory Search", 1) directory_brute( task, domain, yaml_configuration, results_dir, activity_id ) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.fetch_url: activity_id = create_scan_activity(task, "Fetching endpoints", 1) fetch_endpoints( task, domain, yaml_configuration, results_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) try: if task.vulnerability_scan: activity_id = create_scan_activity(task, "Vulnerability Scan", 1) vulnerability_scan( task, domain, yaml_configuration, results_dir, activity_id) update_last_activity(activity_id, 2) except Exception as e: logger.error(e) update_last_activity(activity_id, 0) activity_id = create_scan_activity(task, "Scan Completed", 2) if notification and notification[0].send_scan_status_notif: send_notification('*Scan Completed*\nreNgine has finished performing recon on target {}.'.format(domain.name)) ''' Once the scan is completed, save the status to successful ''' if ScanActivity.objects.filter(scan_of=task).filter(status=0).all(): task.scan_status = 0 else: task.scan_status = 2 task.stop_scan_date = timezone.now() task.save() # cleanup results delete_scan_data(results_dir) return {"status": True} def skip_subdomain_scan(task, domain, results_dir): # store default target as subdomain ''' If the imported subdomain already has target domain saved, we can skip this ''' if not Subdomain.objects.filter( scan_history=task, name=domain.name).exists(): subdomain_dict = DottedDict({ 'name': domain.name, 'scan_history': task, 'target_domain': domain }) save_subdomain(subdomain_dict) # Save target into target_domain.txt with open('{}/target_domain.txt'.format(results_dir), 'w+') as file: file.write(domain.name + '\n') file.close() ''' We can have two conditions, either subdomain scan happens, or subdomain scan does not happen, in either cases, because we are using import subdomain, we need to collect and sort all the subdomains Write target domain into subdomain_collection ''' os.system( 'cat {0}/target_domain.txt > {0}/subdomain_collection.txt'.format(results_dir)) os.system( 'cat {0}/from_imported.txt > {0}/subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/from_imported.txt'.format(results_dir)) ''' Sort all Subdomains ''' os.system( 'sort -u {0}/subdomain_collection.txt -o {0}/sorted_subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/subdomain_collection.txt'.format(results_dir)) def extract_imported_subdomain(imported_subdomains, task, domain, results_dir): valid_imported_subdomains = [subdomain for subdomain in imported_subdomains if validators.domain( subdomain) and domain.name == get_domain_from_subdomain(subdomain)] # remove any duplicate valid_imported_subdomains = list(set(valid_imported_subdomains)) with open('{}/from_imported.txt'.format(results_dir), 'w+') as file: for subdomain_name in valid_imported_subdomains: # save _subdomain to Subdomain model db if not Subdomain.objects.filter( scan_history=task, name=subdomain_name).exists(): subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': subdomain_name, 'is_imported_subdomain': True }) save_subdomain(subdomain_dict) # save subdomain to file file.write('{}\n'.format(subdomain_name)) file.close() def subdomain_scan(task, domain, yaml_configuration, results_dir, activity_id, out_of_scope_subdomains=None): ''' This function is responsible for performing subdomain enumeration ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Subdomain Gathering for target {} has been started'.format(domain.name)) subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' # check for all the tools and add them into string # if tool selected is all then make string, no need for loop if ALL in yaml_configuration[SUBDOMAIN_DISCOVERY][USES_TOOLS]: tools = 'amass-active amass-passive assetfinder sublist3r subfinder oneforall' else: tools = ' '.join( str(tool) for tool in yaml_configuration[SUBDOMAIN_DISCOVERY][USES_TOOLS]) logging.info(tools) # check for THREADS, by default 10 threads = 10 if THREADS in yaml_configuration[SUBDOMAIN_DISCOVERY]: _threads = yaml_configuration[SUBDOMAIN_DISCOVERY][THREADS] if _threads > 0: threads = _threads if 'amass' in tools: if 'amass-passive' in tools: amass_command = 'amass enum -passive -d {} -o {}/from_amass.txt'.format( domain.name, results_dir) if USE_AMASS_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_AMASS_CONFIG]: amass_command += ' -config /root/.config/amass.ini' # Run Amass Passive logging.info(amass_command) os.system(amass_command) if 'amass-active' in tools: amass_command = 'amass enum -active -d {} -o {}/from_amass_active.txt'.format( domain.name, results_dir) if USE_AMASS_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_AMASS_CONFIG]: amass_command += ' -config /root/.config/amass.ini' if AMASS_WORDLIST in yaml_configuration[SUBDOMAIN_DISCOVERY]: wordlist = yaml_configuration[SUBDOMAIN_DISCOVERY][AMASS_WORDLIST] if wordlist == 'default': wordlist_path = '/usr/src/wordlist/deepmagic.com-prefixes-top50000.txt' else: wordlist_path = '/usr/src/wordlist/' + wordlist + '.txt' if not os.path.exists(wordlist_path): wordlist_path = '/usr/src/' + AMASS_WORDLIST amass_command = amass_command + \ ' -brute -w {}'.format(wordlist_path) if amass_config_path: amass_command = amass_command + \ ' -config {}'.format('/usr/src/scan_results/' + amass_config_path) # Run Amass Active logging.info(amass_command) os.system(amass_command) if 'assetfinder' in tools: assetfinder_command = 'assetfinder --subs-only {} > {}/from_assetfinder.txt'.format( domain.name, results_dir) # Run Assetfinder logging.info(assetfinder_command) os.system(assetfinder_command) if 'sublist3r' in tools: sublist3r_command = 'python3 /usr/src/github/Sublist3r/sublist3r.py -d {} -t {} -o {}/from_sublister.txt'.format( domain.name, threads, results_dir) # Run sublist3r logging.info(sublist3r_command) os.system(sublist3r_command) if 'subfinder' in tools: subfinder_command = 'subfinder -d {} -t {} -o {}/from_subfinder.txt'.format( domain.name, threads, results_dir) if USE_SUBFINDER_CONFIG in yaml_configuration[SUBDOMAIN_DISCOVERY] and yaml_configuration[SUBDOMAIN_DISCOVERY][USE_SUBFINDER_CONFIG]: subfinder_command += ' -config /root/.config/subfinder/config.yaml' # Run Subfinder logging.info(subfinder_command) os.system(subfinder_command) if 'oneforall' in tools: oneforall_command = 'python3 /usr/src/github/OneForAll/oneforall.py --target {} run'.format( domain.name, results_dir) # Run OneForAll logging.info(oneforall_command) os.system(oneforall_command) extract_subdomain = "cut -d',' -f6 /usr/src/github/OneForAll/results/{}.csv >> {}/from_oneforall.txt".format( domain.name, results_dir) os.system(extract_subdomain) # remove the results from oneforall directory os.system( 'rm -rf /usr/src/github/OneForAll/results/{}.*'.format(domain.name)) ''' All tools have gathered the list of subdomains with filename initials as from_* We will gather all the results in one single file, sort them and remove the older results from_* ''' os.system( 'cat {0}/*.txt > {0}/subdomain_collection.txt'.format(results_dir)) ''' Write target domain into subdomain_collection ''' os.system( 'cat {0}/target_domain.txt >> {0}/subdomain_collection.txt'.format(results_dir)) ''' Remove all the from_* files ''' os.system('rm -f {}/from*'.format(results_dir)) ''' Sort all Subdomains ''' os.system( 'sort -u {0}/subdomain_collection.txt -o {0}/sorted_subdomain_collection.txt'.format(results_dir)) os.system('rm -f {}/subdomain_collection.txt'.format(results_dir)) ''' The final results will be stored in sorted_subdomain_collection. ''' # parse the subdomain list file and store in db with open(subdomain_scan_results_file) as subdomain_list: for _subdomain in subdomain_list: __subdomain = _subdomain.rstrip('\n') if not Subdomain.objects.filter(scan_history=task, name=__subdomain).exists( ) and validators.domain(__subdomain) and __subdomain not in out_of_scope_subdomains: subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': __subdomain, }) save_subdomain(subdomain_dict) notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: subdomains_count = Subdomain.objects.filter(scan_history=task).count() send_notification('Subdomain Gathering for target {} has been completed and has discovered *{}* subdomains.'.format(domain.name, subdomains_count)) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/sorted_subdomain_collection.txt') # check for any subdomain changes and send notif if any if notification and notification[0].send_subdomain_changes_notif: newly_added_subdomain = get_new_added_subdomain(task.id, domain.id) if newly_added_subdomain: message = "**{} New Subdomains Discovered on domain {}**".format(newly_added_subdomain.count(), domain.name) for subdomain in newly_added_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) removed_subdomain = get_removed_subdomain(task.id, domain.id) if removed_subdomain: message = "**{} Subdomains are no longer available on domain {}**".format(removed_subdomain.count(), domain.name) for subdomain in removed_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) # check for interesting subdomains and send notif if any if notification and notification[0].send_interesting_notif: interesting_subdomain = get_interesting_subdomains(task.id, domain.id) print(interesting_subdomain) if interesting_subdomain: message = "**{} Interesting Subdomains Found on domain {}**".format(interesting_subdomain.count(), domain.name) for subdomain in interesting_subdomain: message += "\n• {}".format(subdomain.name) send_notification(message) def get_new_added_subdomain(scan_id, domain_id): scan_history = ScanHistory.objects.filter( domain=domain_id).filter( subdomain_discovery=True).filter( id__lte=scan_id) if scan_history.count() > 1: last_scan = scan_history.order_by('-start_scan_date')[1] scanned_host_q1 = Subdomain.objects.filter( scan_history__id=scan_id).values('name') scanned_host_q2 = Subdomain.objects.filter( scan_history__id=last_scan.id).values('name') added_subdomain = scanned_host_q1.difference(scanned_host_q2) return Subdomain.objects.filter( scan_history=scan_id).filter( name__in=added_subdomain) def get_removed_subdomain(scan_id, domain_id): scan_history = ScanHistory.objects.filter( domain=domain_id).filter( subdomain_discovery=True).filter( id__lte=scan_id) if scan_history.count() > 1: last_scan = scan_history.order_by('-start_scan_date')[1] scanned_host_q1 = Subdomain.objects.filter( scan_history__id=scan_id).values('name') scanned_host_q2 = Subdomain.objects.filter( scan_history__id=last_scan.id).values('name') removed_subdomains = scanned_host_q2.difference(scanned_host_q1) print() return Subdomain.objects.filter( scan_history=last_scan).filter( name__in=removed_subdomains) def http_crawler(task, domain, results_dir, activity_id): ''' This function is runs right after subdomain gathering, and gathers important like page title, http status, etc HTTP Crawler runs by default ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('HTTP Crawler for target {} has been initiated.'.format(domain.name)) alive_file_location = results_dir + '/alive.txt' httpx_results_file = results_dir + '/httpx.json' subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' httpx_command = 'httpx -status-code -content-length -title -tech-detect -cdn -ip -follow-host-redirects -random-agent' proxy = get_random_proxy() if proxy: httpx_command += " --http-proxy '{}'".format(proxy) httpx_command += ' -json -o {}'.format( httpx_results_file ) httpx_command = 'cat {} | {}'.format(subdomain_scan_results_file, httpx_command) print(httpx_command) os.system(httpx_command) # alive subdomains from httpx alive_file = open(alive_file_location, 'w') # writing httpx results if os.path.isfile(httpx_results_file): httpx_json_result = open(httpx_results_file, 'r') lines = httpx_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) try: # fallback for older versions of httpx if 'url' in json_st: subdomain = Subdomain.objects.get( scan_history=task, name=json_st['input']) else: subdomain = Subdomain.objects.get( scan_history=task, name=json_st['url'].split("//")[-1]) ''' Saving Default http urls to EndPoint ''' endpoint = EndPoint() endpoint.scan_history = task endpoint.target_domain = domain endpoint.subdomain = subdomain if 'url' in json_st: endpoint.http_url = json_st['url'] subdomain.http_url = json_st['url'] if 'status-code' in json_st: endpoint.http_status = json_st['status-code'] subdomain.http_status = json_st['status-code'] if 'title' in json_st: endpoint.page_title = json_st['title'] subdomain.page_title = json_st['title'] if 'content-length' in json_st: endpoint.content_length = json_st['content-length'] subdomain.content_length = json_st['content-length'] if 'content-type' in json_st: endpoint.content_type = json_st['content-type'] subdomain.content_type = json_st['content-type'] if 'webserver' in json_st: endpoint.webserver = json_st['webserver'] subdomain.webserver = json_st['webserver'] if 'response-time' in json_st: response_time = float( ''.join( ch for ch in json_st['response-time'] if not ch.isalpha())) if json_st['response-time'][-2:] == 'ms': response_time = response_time / 1000 endpoint.response_time = response_time subdomain.response_time = response_time if 'cnames' in json_st: cname_list = ','.join(json_st['cnames']) subdomain.cname = cname_list discovered_date = timezone.now() endpoint.discovered_date = discovered_date subdomain.discovered_date = discovered_date endpoint.is_default = True endpoint.save() subdomain.save() if 'technologies' in json_st: for _tech in json_st['technologies']: if Technology.objects.filter(name=_tech).exists(): tech = Technology.objects.get(name=_tech) else: tech = Technology(name=_tech) tech.save() subdomain.technologies.add(tech) endpoint.technologies.add(tech) if 'a' in json_st: for _ip in json_st['a']: if IpAddress.objects.filter(address=_ip).exists(): ip = IpAddress.objects.get(address=_ip) else: ip = IpAddress(address=_ip) if 'cdn' in json_st: ip.is_cdn = json_st['cdn'] ip.save() subdomain.ip_addresses.add(ip) # see if to ignore 404 or 5xx alive_file.write(json_st['url'] + '\n') subdomain.save() endpoint.save() except Exception as exception: logging.error(exception) alive_file.close() if notification and notification[0].send_scan_status_notif: alive_count = Subdomain.objects.filter( scan_history__id=task.id).values('name').distinct().filter( http_status__exact=200).count() send_notification('HTTP Crawler for target {} has been completed.\n\n {} subdomains were alive (http status 200).'.format(domain.name, alive_count)) def grab_screenshot(task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for taking screenshots ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine is currently gathering screenshots for {}'.format(domain.name)) output_screenshots_path = results_dir + '/screenshots' result_csv_path = results_dir + '/screenshots/Requests.csv' alive_subdomains_path = results_dir + '/alive.txt' eyewitness_command = 'python3 /usr/src/github/EyeWitness/Python/EyeWitness.py' eyewitness_command += ' -f {} -d {} --no-prompt'.format( alive_subdomains_path, output_screenshots_path ) if EYEWITNESS in yaml_configuration \ and TIMEOUT in yaml_configuration[EYEWITNESS] \ and yaml_configuration[EYEWITNESS][TIMEOUT] > 0: eyewitness_command += ' --timeout {}'.format( yaml_configuration[EYEWITNESS][TIMEOUT] ) if EYEWITNESS in yaml_configuration \ and THREADS in yaml_configuration[EYEWITNESS] \ and yaml_configuration[EYEWITNESS][THREADS] > 0: eyewitness_command += ' --threads {}'.format( yaml_configuration[EYEWITNESS][THREADS] ) logger.info(eyewitness_command) os.system(eyewitness_command) if os.path.isfile(result_csv_path): logger.info('Gathering Eyewitness results') with open(result_csv_path, 'r') as file: reader = csv.reader(file) for row in reader: if row[3] == 'Successful' \ and Subdomain.objects.filter( scan_history__id=task.id).filter(name=row[2]).exists(): subdomain = Subdomain.objects.get( scan_history__id=task.id, name=row[2] ) subdomain.screenshot_path = row[4].replace( '/usr/src/scan_results/', '' ) subdomain.save() # remove all db, html extra files in screenshot results os.system('rm -rf {0}/*.csv {0}/*.db {0}/*.js {0}/*.html {0}/*.css'.format( output_screenshots_path, )) os.system('rm -rf {0}/source'.format( output_screenshots_path, )) if notification and notification[0].send_scan_status_notif: send_notification('reNgine has finished gathering screenshots for {}'.format(domain.name)) def port_scanning(task, domain, yaml_configuration, results_dir): ''' This function is responsible for running the port scan ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Port Scan initiated for {}'.format(domain.name)) subdomain_scan_results_file = results_dir + '/sorted_subdomain_collection.txt' port_results_file = results_dir + '/ports.json' # check the yaml_configuration and choose the ports to be scanned scan_ports = '-' # default port scan everything if PORTS in yaml_configuration[PORT_SCAN]: # TODO: legacy code, remove top-100 in future versions all_ports = yaml_configuration[PORT_SCAN][PORTS] if 'full' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -p {}'.format( subdomain_scan_results_file, port_results_file, '-') elif 'top-100' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -top-ports 100'.format( subdomain_scan_results_file, port_results_file) elif 'top-1000' in all_ports: naabu_command = 'cat {} | naabu -json -o {} -top-ports 1000'.format( subdomain_scan_results_file, port_results_file) else: scan_ports = ','.join( str(port) for port in all_ports) naabu_command = 'cat {} | naabu -json -o {} -p {}'.format( subdomain_scan_results_file, port_results_file, scan_ports) # check for exclude ports if EXCLUDE_PORTS in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][EXCLUDE_PORTS]: exclude_ports = ','.join( str(port) for port in yaml_configuration['port_scan']['exclude_ports']) naabu_command = naabu_command + \ ' -exclude-ports {}'.format(exclude_ports) if NAABU_RATE in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][NAABU_RATE] > 0: naabu_command = naabu_command + \ ' -rate {}'.format( yaml_configuration[PORT_SCAN][NAABU_RATE]) if USE_NAABU_CONFIG in yaml_configuration[PORT_SCAN] and yaml_configuration[PORT_SCAN][USE_NAABU_CONFIG]: naabu_command += ' -config /root/.config/naabu/naabu.conf' # run naabu os.system(naabu_command) # writing port results try: port_json_result = open(port_results_file, 'r') lines = port_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) port_number = json_st['port'] ip_address = json_st['ip'] # see if port already exists if Port.objects.filter(number__exact=port_number).exists(): port = Port.objects.get(number=port_number) else: port = Port() port.number = port_number if port_number in UNCOMMON_WEB_PORTS: port.is_uncommon = True port_detail = whatportis.get_ports(str(port_number)) if len(port_detail): port.service_name = port_detail[0].name port.description = port_detail[0].description port.save() if IpAddress.objects.filter(address=json_st['ip']).exists(): ip = IpAddress.objects.get(address=json_st['ip']) ip.ports.add(port) ip.save() except BaseException as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: port_count = Port.objects.filter( ports__in=IpAddress.objects.filter( ip_addresses__in=Subdomain.objects.filter( scan_history__id=task.id))).distinct().count() send_notification('reNgine has finished Port Scanning on {} and has identified {} ports.'.format(domain.name, port_count)) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/ports.json') def check_waf(): ''' This function will check for the WAF being used in subdomains using wafw00f ''' pass def directory_brute(task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for performing directory scan ''' # scan directories for all the alive subdomain with http status > # 200 notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Directory Bruteforce has been initiated for {}.'.format(domain.name)) alive_subdomains = Subdomain.objects.filter( scan_history__id=task.id).exclude(http_url__isnull=True) dirs_results = results_dir + '/dirs.json' # check the yaml settings if EXTENSIONS in yaml_configuration[DIR_FILE_SEARCH]: extensions = ','.join( str(ext) for ext in yaml_configuration[DIR_FILE_SEARCH][EXTENSIONS]) else: extensions = 'php,git,yaml,conf,db,mysql,bak,txt' # Threads if THREADS in yaml_configuration[DIR_FILE_SEARCH] \ and yaml_configuration[DIR_FILE_SEARCH][THREADS] > 0: threads = yaml_configuration[DIR_FILE_SEARCH][THREADS] else: threads = 10 for subdomain in alive_subdomains: # delete any existing dirs.json if os.path.isfile(dirs_results): os.system('rm -rf {}'.format(dirs_results)) dirsearch_command = 'python3 /usr/src/github/dirsearch/dirsearch.py' dirsearch_command += ' -u {}'.format(subdomain.http_url) if (WORDLIST not in yaml_configuration[DIR_FILE_SEARCH] or not yaml_configuration[DIR_FILE_SEARCH][WORDLIST] or 'default' in yaml_configuration[DIR_FILE_SEARCH][WORDLIST]): wordlist_location = '/usr/src/github/dirsearch/db/dicc.txt' else: wordlist_location = '/usr/src/wordlist/' + \ yaml_configuration[DIR_FILE_SEARCH][WORDLIST] + '.txt' dirsearch_command += ' -w {}'.format(wordlist_location) dirsearch_command += ' --format json -o {}'.format(dirs_results) dirsearch_command += ' -e {}'.format(extensions) dirsearch_command += ' -t {}'.format(threads) dirsearch_command += ' --random-agent --follow-redirects --exclude-status 403,401,404' if EXCLUDE_EXTENSIONS in yaml_configuration[DIR_FILE_SEARCH]: exclude_extensions = ','.join( str(ext) for ext in yaml_configuration[DIR_FILE_SEARCH][EXCLUDE_EXTENSIONS]) dirsearch_command += ' -X {}'.format(exclude_extensions) if EXCLUDE_TEXT in yaml_configuration[DIR_FILE_SEARCH]: exclude_text = ','.join( str(text) for text in yaml_configuration[DIR_FILE_SEARCH][EXCLUDE_TEXT]) dirsearch_command += ' -exclude-texts {}'.format(exclude_text) # check if recursive strategy is set to on if RECURSIVE_LEVEL in yaml_configuration[DIR_FILE_SEARCH]: dirsearch_command += ' --recursion-depth {}'.format(yaml_configuration[DIR_FILE_SEARCH][RECURSIVE_LEVEL]) if RECURSIVE_LEVEL in yaml_configuration[DIR_FILE_SEARCH]: dirsearch_command += ' --recursion-depth {}'.format(yaml_configuration[DIR_FILE_SEARCH][RECURSIVE_LEVEL]) # proxy proxy = get_random_proxy() if proxy: dirsearch_command += " --proxy '{}'".format(proxy) print(dirsearch_command) os.system(dirsearch_command) try: if os.path.isfile(dirs_results): with open(dirs_results, "r") as json_file: json_string = json_file.read() subdomain = Subdomain.objects.get( scan_history__id=task.id, http_url=subdomain.http_url) subdomain.directory_json = json_string subdomain.save() except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: send_notification('Directory Bruteforce has been completed for {}.'.format(domain.name)) def fetch_endpoints( task, domain, yaml_configuration, results_dir, activity_id): ''' This function is responsible for fetching all the urls associated with target and run HTTP probe It first runs gau to gather all urls from wayback, then we will use hakrawler to identify more urls ''' notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine is currently gathering endpoints for {}.'.format(domain.name)) # check yaml settings if ALL in yaml_configuration[FETCH_URL][USES_TOOLS]: tools = 'gauplus hakrawler waybackurls gospider' else: tools = ' '.join( str(tool) for tool in yaml_configuration[FETCH_URL][USES_TOOLS]) if INTENSITY in yaml_configuration[FETCH_URL]: scan_type = yaml_configuration[FETCH_URL][INTENSITY] else: scan_type = 'normal' domain_regex = "\'https?://([a-z0-9]+[.])*{}.*\'".format(domain.name) if 'deep' in scan_type: # performs deep url gathering for all the subdomains present - # RECOMMENDED logger.info('Deep URLS Fetch') os.system(settings.TOOL_LOCATION + 'get_urls.sh %s %s %s %s %s' % ("None", results_dir, scan_type, domain_regex, tools)) else: # perform url gathering only for main domain - USE only for quick scan logger.info('Non Deep URLS Fetch') os.system( settings.TOOL_LOCATION + 'get_urls.sh %s %s %s %s %s' % ( domain.name, results_dir, scan_type, domain_regex, tools )) if IGNORE_FILE_EXTENSION in yaml_configuration[FETCH_URL]: ignore_extension = '|'.join( yaml_configuration[FETCH_URL][IGNORE_FILE_EXTENSION]) logger.info('Ignore extensions' + ignore_extension) os.system( 'cat {0}/all_urls.txt | grep -Eiv "\\.({1}).*" > {0}/temp_urls.txt'.format( results_dir, ignore_extension)) os.system( 'rm {0}/all_urls.txt && mv {0}/temp_urls.txt {0}/all_urls.txt'.format(results_dir)) ''' Store all the endpoints and then run the httpx ''' try: endpoint_final_url = results_dir + '/all_urls.txt' if os.path.isfile(endpoint_final_url): with open(endpoint_final_url) as endpoint_list: for url in endpoint_list: http_url = url.rstrip('\n') if not EndPoint.objects.filter(scan_history=task, http_url=http_url).exists(): _subdomain = get_subdomain_from_url(http_url) if Subdomain.objects.filter( scan_history=task).filter( name=_subdomain).exists(): subdomain = Subdomain.objects.get( scan_history=task, name=_subdomain) else: ''' gau or gosppider can gather interesting endpoints which when parsed can give subdomains that were not existent from subdomain scan. so storing them ''' logger.error( 'Subdomain {} not found, adding...'.format(_subdomain)) subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': _subdomain, }) subdomain = save_subdomain(subdomain_dict) endpoint_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'subdomain': subdomain, 'http_url': http_url, }) save_endpoint(endpoint_dict) except Exception as e: logger.error(e) if notification and notification[0].send_scan_output_file: send_files_to_discord(results_dir + '/all_urls.txt') ''' TODO: Go spider & waybackurls accumulates a lot of urls, which is good but nuclei takes forever to scan even a simple website, so we will do http probing and filter HTTP status 404, this way we can reduce the number of Non Existent URLS ''' logger.info('HTTP Probing on collected endpoints') httpx_command = 'httpx -l {0}/all_urls.txt -status-code -content-length -ip -cdn -title -tech-detect -json -follow-redirects -random-agent -o {0}/final_httpx_urls.json'.format(results_dir) proxy = get_random_proxy() if proxy: httpx_command += " --http-proxy '{}'".format(proxy) os.system(httpx_command) url_results_file = results_dir + '/final_httpx_urls.json' try: urls_json_result = open(url_results_file, 'r') lines = urls_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) http_url = json_st['url'] _subdomain = get_subdomain_from_url(http_url) if Subdomain.objects.filter( scan_history=task).filter( name=_subdomain).exists(): subdomain_obj = Subdomain.objects.get( scan_history=task, name=_subdomain) else: subdomain_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'name': _subdomain, }) subdomain_obj = save_subdomain(subdomain_dict) if EndPoint.objects.filter( scan_history=task).filter( http_url=http_url).exists(): endpoint = EndPoint.objects.get( scan_history=task, http_url=http_url) else: endpoint = EndPoint() endpoint_dict = DottedDict({ 'scan_history': task, 'target_domain': domain, 'http_url': http_url, 'subdomain': subdomain_obj }) endpoint = save_endpoint(endpoint_dict) if 'title' in json_st: endpoint.page_title = json_st['title'] if 'webserver' in json_st: endpoint.webserver = json_st['webserver'] if 'content-length' in json_st: endpoint.content_length = json_st['content-length'] if 'content-type' in json_st: endpoint.content_type = json_st['content-type'] if 'status-code' in json_st: endpoint.http_status = json_st['status-code'] if 'response-time' in json_st: response_time = float(''.join(ch for ch in json_st['response-time'] if not ch.isalpha())) if json_st['response-time'][-2:] == 'ms': response_time = response_time / 1000 endpoint.response_time = response_time endpoint.save() if 'technologies' in json_st: for _tech in json_st['technologies']: if Technology.objects.filter(name=_tech).exists(): tech = Technology.objects.get(name=_tech) else: tech = Technology(name=_tech) tech.save() endpoint.technologies.add(tech) # get subdomain object subdomain = Subdomain.objects.get(scan_history=task, name=_subdomain) subdomain.technologies.add(tech) subdomain.save() except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: endpoint_count = EndPoint.objects.filter( scan_history__id=task.id).values('http_url').distinct().count() endpoint_alive_count = EndPoint.objects.filter( scan_history__id=task.id, http_status__exact=200).values('http_url').distinct().count() send_notification('reNgine has finished gathering endpoints for {} and has discovered *{}* unique endpoints.\n\n{} of those endpoints reported HTTP status 200.'.format( domain.name, endpoint_count, endpoint_alive_count )) # once endpoint is saved, run gf patterns TODO: run threads if GF_PATTERNS in yaml_configuration[FETCH_URL]: for pattern in yaml_configuration[FETCH_URL][GF_PATTERNS]: logger.info('Running GF for {}'.format(pattern)) gf_output_file_path = '{0}/gf_patterns_{1}.txt'.format( results_dir, pattern) gf_command = 'cat {0}/all_urls.txt | gf {1} >> {2}'.format( results_dir, pattern, gf_output_file_path) os.system(gf_command) if os.path.exists(gf_output_file_path): with open(gf_output_file_path) as gf_output: for line in gf_output: url = line.rstrip('\n') try: endpoint = EndPoint.objects.get( scan_history=task, http_url=url) earlier_pattern = endpoint.matched_gf_patterns new_pattern = earlier_pattern + ',' + pattern if earlier_pattern else pattern endpoint.matched_gf_patterns = new_pattern except Exception as e: # add the url in db logger.error(e) logger.info('Adding URL' + url) endpoint = EndPoint() endpoint.http_url = url endpoint.target_domain = domain endpoint.scan_history = task try: _subdomain = Subdomain.objects.get( scan_history=task, name=get_subdomain_from_url(url)) endpoint.subdomain = _subdomain except Exception as e: continue endpoint.matched_gf_patterns = pattern finally: endpoint.save() def vulnerability_scan( task, domain, yaml_configuration, results_dir, activity_id): notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('Vulnerability scan has been initiated for {}.'.format(domain.name)) ''' This function will run nuclei as a vulnerability scanner ---- unfurl the urls to keep only domain and path, this will be sent to vuln scan ignore certain file extensions Thanks: https://github.com/six2dez/reconftw ''' urls_path = '/alive.txt' if task.scan_type.fetch_url: os.system('cat {0}/all_urls.txt | grep -Eiv "\\.(eot|jpg|jpeg|gif|css|tif|tiff|png|ttf|otf|woff|woff2|ico|pdf|svg|txt|js|doc|docx)$" | unfurl -u format %s://%d%p >> {0}/unfurl_urls.txt'.format(results_dir)) os.system( 'sort -u {0}/unfurl_urls.txt -o {0}/unfurl_urls.txt'.format(results_dir)) urls_path = '/unfurl_urls.txt' vulnerability_result_path = results_dir + '/vulnerability.json' vulnerability_scan_input_file = results_dir + urls_path nuclei_command = 'nuclei -json -l {} -o {}'.format( vulnerability_scan_input_file, vulnerability_result_path) # check nuclei config if USE_NUCLEI_CONFIG in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[VULNERABILITY_SCAN][USE_NUCLEI_CONFIG]: nuclei_command += ' -config /root/.config/nuclei/config.yaml' ''' Nuclei Templates Either custom template has to be supplied or default template, if neither has been supplied then use all templates including custom templates ''' if CUSTOM_NUCLEI_TEMPLATE in yaml_configuration[ VULNERABILITY_SCAN] or NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: # check yaml settings for templates if NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: if ALL in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_TEMPLATE]: template = NUCLEI_TEMPLATES_PATH else: _template = ','.join([NUCLEI_TEMPLATES_PATH + str(element) for element in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_TEMPLATE]]) template = _template.replace(',', ' -t ') # Update nuclei command with templates nuclei_command = nuclei_command + ' -t ' + template if CUSTOM_NUCLEI_TEMPLATE in yaml_configuration[VULNERABILITY_SCAN]: # add .yaml to the custom template extensions _template = ','.join( [str(element) + '.yaml' for element in yaml_configuration[VULNERABILITY_SCAN][CUSTOM_NUCLEI_TEMPLATE]]) template = _template.replace(',', ' -t ') # Update nuclei command with templates nuclei_command = nuclei_command + ' -t ' + template else: nuclei_command = nuclei_command + ' -t /root/nuclei-templates' # check yaml settings for concurrency if NUCLEI_CONCURRENCY in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][NUCLEI_CONCURRENCY] > 0: concurrency = yaml_configuration[VULNERABILITY_SCAN][NUCLEI_CONCURRENCY] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -c ' + str(concurrency) if RATE_LIMIT in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][RATE_LIMIT] > 0: rate_limit = yaml_configuration[VULNERABILITY_SCAN][RATE_LIMIT] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -rl ' + str(rate_limit) if TIMEOUT in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][TIMEOUT] > 0: timeout = yaml_configuration[VULNERABILITY_SCAN][TIMEOUT] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -timeout ' + str(timeout) if RETRIES in yaml_configuration[VULNERABILITY_SCAN] and yaml_configuration[ VULNERABILITY_SCAN][RETRIES] > 0: retries = yaml_configuration[VULNERABILITY_SCAN][RETRIES] # Update nuclei command with concurrent nuclei_command = nuclei_command + ' -retries ' + str(retries) # for severity if NUCLEI_SEVERITY in yaml_configuration[VULNERABILITY_SCAN] and ALL not in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_SEVERITY]: _severity = ','.join( [str(element) for element in yaml_configuration[VULNERABILITY_SCAN][NUCLEI_SEVERITY]]) severity = _severity.replace(" ", "") else: severity = "critical, high, medium, low, info" # update nuclei templates before running scan os.system('nuclei -update-templates') for _severity in severity.split(","): # delete any existing vulnerability.json file if os.path.isfile(vulnerability_result_path): os.system('rm {}'.format(vulnerability_result_path)) # run nuclei final_nuclei_command = nuclei_command + ' -severity ' + _severity proxy = get_random_proxy() if proxy: final_nuclei_command += " --proxy-url '{}'".format(proxy) logger.info(final_nuclei_command) os.system(final_nuclei_command) try: if os.path.isfile(vulnerability_result_path): urls_json_result = open(vulnerability_result_path, 'r') lines = urls_json_result.readlines() for line in lines: json_st = json.loads(line.strip()) host = json_st['host'] _subdomain = get_subdomain_from_url(host) try: subdomain = Subdomain.objects.get( name=_subdomain, scan_history=task) vulnerability = Vulnerability() vulnerability.subdomain = subdomain vulnerability.scan_history = task vulnerability.target_domain = domain try: endpoint = EndPoint.objects.get( scan_history=task, target_domain=domain, http_url=host) vulnerability.endpoint = endpoint except Exception as exception: logger.error(exception) if 'name' in json_st['info']: vulnerability.name = json_st['info']['name'] if 'severity' in json_st['info']: if json_st['info']['severity'] == 'info': severity = 0 elif json_st['info']['severity'] == 'low': severity = 1 elif json_st['info']['severity'] == 'medium': severity = 2 elif json_st['info']['severity'] == 'high': severity = 3 elif json_st['info']['severity'] == 'critical': severity = 4 else: severity = 0 else: severity = 0 vulnerability.severity = severity if 'tags' in json_st['info']: vulnerability.tags = json_st['info']['tags'] if 'description' in json_st['info']: vulnerability.description = json_st['info']['description'] if 'reference' in json_st['info']: vulnerability.reference = json_st['info']['reference'] if 'matched' in json_st: # TODO remove in rengine 1.1. 'matched' isn't used in nuclei 2.5.3 vulnerability.http_url = json_st['matched'] if 'matched-at' in json_st: vulnerability.http_url = json_st['matched-at'] if 'templateID' in json_st: vulnerability.template_used = json_st['templateID'] if 'description' in json_st: vulnerability.description = json_st['description'] if 'matcher_name' in json_st: vulnerability.matcher_name = json_st['matcher_name'] if 'extracted_results' in json_st: vulnerability.extracted_results = json_st['extracted_results'] vulnerability.discovered_date = timezone.now() vulnerability.open_status = True vulnerability.save() # send notification for all vulnerabilities except info if json_st['info']['severity'] != "info" and notification and notification[0].send_vuln_notif: message = "*Alert: Vulnerability Identified*" message += "\n\n" message += "A *{}* severity vulnerability has been identified.".format(json_st['info']['severity']) message += "\nVulnerability Name: {}".format(json_st['info']['name']) message += "\nVulnerable URL: {}".format(json_st['host']) send_notification(message) # send report to hackerone if Hackerone.objects.all().exists() and json_st['info']['severity'] != 'info' and json_st['info']['severity'] \ != 'low' and vulnerability.target_domain.h1_team_handle: hackerone = Hackerone.objects.all()[0] if hackerone.send_critical and json_st['info']['severity'] == 'critical': send_hackerone_report(vulnerability.id) elif hackerone.send_high and json_st['info']['severity'] == 'high': send_hackerone_report(vulnerability.id) elif hackerone.send_medium and json_st['info']['severity'] == 'medium': send_hackerone_report(vulnerability.id) except ObjectDoesNotExist: logger.error('Object not found') continue except Exception as exception: logging.error(exception) update_last_activity(activity_id, 0) if notification and notification[0].send_scan_status_notif: info_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=0).count() low_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=1).count() medium_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=2).count() high_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=3).count() critical_count = Vulnerability.objects.filter( scan_history__id=task.id, severity=4).count() vulnerability_count = info_count + low_count + medium_count + high_count + critical_count message = 'Vulnerability scan has been completed for {} and discovered {} vulnerabilities.'.format( domain.name, vulnerability_count ) message += '\n\n*Vulnerability Stats:*' message += '\nCritical: {}'.format(critical_count) message += '\nHigh: {}'.format(high_count) message += '\nMedium: {}'.format(medium_count) message += '\nLow: {}'.format(low_count) message += '\nInfo: {}'.format(info_count) send_notification(message) def scan_failed(task): task.scan_status = 0 task.stop_scan_date = timezone.now() task.save() def create_scan_activity(task, message, status): scan_activity = ScanActivity() scan_activity.scan_of = task scan_activity.title = message scan_activity.time = timezone.now() scan_activity.status = status scan_activity.save() return scan_activity.id def update_last_activity(id, activity_status): ScanActivity.objects.filter( id=id).update( status=activity_status, time=timezone.now()) def delete_scan_data(results_dir): # remove all txt,html,json files os.system('find {} -name "*.txt" -type f -delete'.format(results_dir)) os.system('find {} -name "*.html" -type f -delete'.format(results_dir)) os.system('find {} -name "*.json" -type f -delete'.format(results_dir)) def save_subdomain(subdomain_dict): subdomain = Subdomain() subdomain.discovered_date = timezone.now() subdomain.target_domain = subdomain_dict.get('target_domain') subdomain.scan_history = subdomain_dict.get('scan_history') subdomain.name = subdomain_dict.get('name') subdomain.http_url = subdomain_dict.get('http_url') subdomain.screenshot_path = subdomain_dict.get('screenshot_path') subdomain.http_header_path = subdomain_dict.get('http_header_path') subdomain.cname = subdomain_dict.get('cname') subdomain.is_cdn = subdomain_dict.get('is_cdn') subdomain.content_type = subdomain_dict.get('content_type') subdomain.webserver = subdomain_dict.get('webserver') subdomain.page_title = subdomain_dict.get('page_title') subdomain.is_imported_subdomain = subdomain_dict.get( 'is_imported_subdomain') if 'is_imported_subdomain' in subdomain_dict else False if 'http_status' in subdomain_dict: subdomain.http_status = subdomain_dict.get('http_status') if 'response_time' in subdomain_dict: subdomain.response_time = subdomain_dict.get('response_time') if 'content_length' in subdomain_dict: subdomain.content_length = subdomain_dict.get('content_length') subdomain.save() return subdomain def save_endpoint(endpoint_dict): endpoint = EndPoint() endpoint.discovered_date = timezone.now() endpoint.scan_history = endpoint_dict.get('scan_history') endpoint.target_domain = endpoint_dict.get('target_domain') if 'target_domain' in endpoint_dict else None endpoint.subdomain = endpoint_dict.get('subdomain') if 'target_domain' in endpoint_dict else None endpoint.http_url = endpoint_dict.get('http_url') endpoint.page_title = endpoint_dict.get('page_title') if 'page_title' in endpoint_dict else None endpoint.content_type = endpoint_dict.get('content_type') if 'content_type' in endpoint_dict else None endpoint.webserver = endpoint_dict.get('webserver') if 'webserver' in endpoint_dict else None endpoint.response_time = endpoint_dict.get('response_time') if 'response_time' in endpoint_dict else 0 endpoint.http_status = endpoint_dict.get('http_status') if 'http_status' in endpoint_dict else 0 endpoint.content_length = endpoint_dict.get('content_length') if 'content_length' in endpoint_dict else 0 endpoint.is_default = endpoint_dict.get('is_default') if 'is_default' in endpoint_dict else False endpoint.save() return endpoint def perform_osint(task, domain, yaml_configuration, results_dir): notification = Notification.objects.all() if notification and notification[0].send_scan_status_notif: send_notification('reNgine has initiated OSINT on target {}'.format(domain.name)) if 'discover' in yaml_configuration[OSINT]: osint_discovery(task, domain, yaml_configuration, results_dir) if 'dork' in yaml_configuration[OSINT]: dorking(task, yaml_configuration) if notification and notification[0].send_scan_status_notif: send_notification('reNgine has completed performing OSINT on target {}'.format(domain.name)) def osint_discovery(task, domain, yaml_configuration, results_dir): if ALL in yaml_configuration[OSINT][OSINT_DISCOVER]: osint_lookup = 'emails metainfo employees' else: osint_lookup = ' '.join( str(lookup) for lookup in yaml_configuration[OSINT][OSINT_DISCOVER]) if 'metainfo' in osint_lookup: if INTENSITY in yaml_configuration[OSINT]: osint_intensity = yaml_configuration[OSINT][INTENSITY] else: osint_intensity = 'normal' if OSINT_DOCUMENTS_LIMIT in yaml_configuration[OSINT]: documents_limit = yaml_configuration[OSINT][OSINT_DOCUMENTS_LIMIT] else: documents_limit = 50 if osint_intensity == 'normal': meta_dict = DottedDict({ 'osint_target': domain.name, 'domain': domain, 'scan_id': task, 'documents_limit': documents_limit }) get_and_save_meta_info(meta_dict) elif osint_intensity == 'deep': # get all subdomains in scan_id subdomains = Subdomain.objects.filter(scan_history=task) for subdomain in subdomains: meta_dict = DottedDict({ 'osint_target': subdomain.name, 'domain': domain, 'scan_id': task, 'documents_limit': documents_limit }) get_and_save_meta_info(meta_dict) if 'emails' in osint_lookup: get_and_save_emails(task, results_dir) get_and_save_leaked_credentials(task, results_dir) if 'employees' in osint_lookup: get_and_save_employees(task, results_dir) def dorking(scan_history, yaml_configuration): # Some dork sources: https://github.com/six2dez/degoogle_hunter/blob/master/degoogle_hunter.sh # look in stackoverflow if ALL in yaml_configuration[OSINT][OSINT_DORK]: dork_lookup = 'stackoverflow, 3rdparty, social_media, project_management, code_sharing, config_files, jenkins, cloud_buckets, php_error, exposed_documents, struts_rce, db_files, traefik, git_exposed' else: dork_lookup = ' '.join( str(lookup) for lookup in yaml_configuration[OSINT][OSINT_DORK]) if 'stackoverflow' in dork_lookup: dork = 'site:stackoverflow.com' dork_type = 'stackoverflow' get_and_save_dork_results( dork, dork_type, scan_history, in_target=False ) if '3rdparty' in dork_lookup: # look in 3rd party sitee dork_type = '3rdparty' lookup_websites = [ 'gitter.im', 'papaly.com', 'productforums.google.com', 'coggle.it', 'replt.it', 'ycombinator.com', 'libraries.io', 'npm.runkit.com', 'npmjs.com', 'scribd.com', 'gitter.im' ] dork = '' for website in lookup_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'social_media' in dork_lookup: dork_type = 'Social Media' social_websites = [ 'tiktok.com', 'facebook.com', 'twitter.com', 'youtube.com', 'pinterest.com', 'tumblr.com', 'reddit.com' ] dork = '' for website in social_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'project_management' in dork_lookup: dork_type = 'Project Management' project_websites = [ 'trello.com', '*.atlassian.net' ] dork = '' for website in project_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'code_sharing' in dork_lookup: dork_type = 'Code Sharing Sites' code_websites = [ 'github.com', 'gitlab.com', 'bitbucket.org' ] dork = '' for website in code_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'config_files' in dork_lookup: dork_type = 'Config Files' config_file_ext = [ 'env', 'xml', 'conf', 'cnf', 'inf', 'rdp', 'ora', 'txt', 'cfg', 'ini' ] dork = '' for extension in config_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'jenkins' in dork_lookup: dork_type = 'Jenkins' dork = 'intitle:\"Dashboard [Jenkins]\"' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) if 'wordpress_files' in dork_lookup: dork_type = 'Wordpress Files' inurl_lookup = [ 'wp-content', 'wp-includes' ] dork = '' for lookup in inurl_lookup: dork = dork + ' | ' + 'inurl:' + lookup get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'cloud_buckets' in dork_lookup: dork_type = 'Cloud Buckets' cloud_websites = [ '.s3.amazonaws.com', 'storage.googleapis.com', 'amazonaws.com' ] dork = '' for website in cloud_websites: dork = dork + ' | ' + 'site:' + website get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=False ) if 'php_error' in dork_lookup: dork_type = 'PHP Error' error_words = [ '\"PHP Parse error\"', '\"PHP Warning\"', '\"PHP Error\"' ] dork = '' for word in error_words: dork = dork + ' | ' + word get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'exposed_documents' in dork_lookup: dork_type = 'Exposed Documents' docs_file_ext = [ 'doc', 'docx', 'odt', 'pdf', 'rtf', 'sxw', 'psw', 'ppt', 'pptx', 'pps', 'csv' ] dork = '' for extension in docs_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'struts_rce' in dork_lookup: dork_type = 'Apache Struts RCE' struts_file_ext = [ 'action', 'struts', 'do' ] dork = '' for extension in struts_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'db_files' in dork_lookup: dork_type = 'Database Files' db_file_ext = [ 'sql', 'db', 'dbf', 'mdb' ] dork = '' for extension in db_file_ext: dork = dork + ' | ' + 'ext:' + extension get_and_save_dork_results( dork[3:], dork_type, scan_history, in_target=True ) if 'traefik' in dork_lookup: dork = 'intitle:traefik inurl:8080/dashboard' dork_type = 'Traefik' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) if 'git_exposed' in dork_lookup: dork = 'inurl:\"/.git\"' dork_type = '.git Exposed' get_and_save_dork_results( dork, dork_type, scan_history, in_target=True ) def get_and_save_dork_results(dork, type, scan_history, in_target=False): degoogle_obj = degoogle.dg() proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy if in_target: query = dork + " site:" + scan_history.domain.name else: query = dork + " \"{}\"".format(scan_history.domain.name) logger.info(query) degoogle_obj.query = query results = degoogle_obj.run() logger.info(results) for result in results: dork, _ = Dork.objects.get_or_create( type=type, description=result['desc'], url=result['url'] ) scan_history.dorks.add(dork) def get_and_save_employees(scan_history, results_dir): theHarvester_location = '/usr/src/github/theHarvester' # update proxies.yaml if Proxy.objects.all().exists(): proxy = Proxy.objects.all()[0] if proxy.use_proxy: proxy_list = proxy.proxies.splitlines() yaml_data = {'http' : proxy_list} with open(theHarvester_location + '/proxies.yaml', 'w') as file: documents = yaml.dump(yaml_data, file) os.system('cd {} && python3 theHarvester.py -d {} -b all -f {}/theHarvester.html'.format( theHarvester_location, scan_history.domain.name, results_dir )) file_location = results_dir + '/theHarvester.html' print(file_location) # delete proxy environ var if os.environ.get(('https_proxy')): del os.environ['https_proxy'] if os.environ.get(('HTTPS_PROXY')): del os.environ['HTTPS_PROXY'] if os.path.isfile(file_location): logger.info('Parsing theHarvester results') options = FirefoxOptions() options.add_argument("--headless") driver = webdriver.Firefox(options=options) driver.get('file://'+file_location) tabledata = driver.execute_script('return tabledata') # save email addresses and linkedin employees for data in tabledata: if data['record'] == 'email': _email = data['result'] email, _ = Email.objects.get_or_create(address=_email) scan_history.emails.add(email) elif data['record'] == 'people': _employee = data['result'] split_val = _employee.split('-') name = split_val[0] if len(split_val) == 2: designation = split_val[1] else: designation = "" employee, _ = Employee.objects.get_or_create(name=name, designation=designation) scan_history.employees.add(employee) driver.quit() print(tabledata) def get_and_save_emails(scan_history, results_dir): leak_target_path = '{}/creds_target.txt'.format(results_dir) # get email address proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy emails = [] try: logger.info('OSINT: Getting emails from Google') email_from_google = get_emails_from_google(scan_history.domain.name) logger.info('OSINT: Getting emails from Bing') email_from_bing = get_emails_from_bing(scan_history.domain.name) logger.info('OSINT: Getting emails from Baidu') email_from_baidu = get_emails_from_baidu(scan_history.domain.name) emails = list(set(email_from_google + email_from_bing + email_from_baidu)) logger.info(emails) except Exception as e: logger.error(e) leak_target_file = open(leak_target_path, 'w') for _email in emails: email, _ = Email.objects.get_or_create(address=_email) scan_history.emails.add(email) leak_target_file.write('{}\n'.format(_email)) # fill leak_target_file with possible email address leak_target_file.write('%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.write('%.%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%.%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.write('%_%@{}\n'.format(scan_history.domain.name)) leak_target_file.write('%_%@%.{}\n'.format(scan_history.domain.name)) leak_target_file.close() def get_and_save_leaked_credentials(scan_history, results_dir): logger.info('OSINT: Getting leaked credentials...') leak_target_file = '{}/creds_target.txt'.format(results_dir) leak_output_file = '{}/pwndb.json'.format(results_dir) pwndb_command = 'python3 /usr/src/github/pwndb/pwndb.py --proxy tor:9150 --output json --list {}'.format( leak_target_file ) try: pwndb_output = subprocess.getoutput(pwndb_command) creds = json.loads(pwndb_output) for cred in creds: if cred['username'] != 'donate': email_id = "{}@{}".format(cred['username'], cred['domain']) email_obj, _ = Email.objects.get_or_create( address=email_id, ) email_obj.password = cred['password'] email_obj.save() scan_history.emails.add(email_obj) except Exception as e: logger.error(e) pass def get_and_save_meta_info(meta_dict): logger.info('Getting METADATA for {}'.format(meta_dict.osint_target)) proxy = get_random_proxy() if proxy: os.environ['https_proxy'] = proxy os.environ['HTTPS_PROXY'] = proxy result = metadata_extractor.extract_metadata_from_google_search(meta_dict.osint_target, meta_dict.documents_limit) if result: results = result.get_metadata() for meta in results: meta_finder_document = MetaFinderDocument() subdomain = Subdomain.objects.get(scan_history=meta_dict.scan_id, name=meta_dict.osint_target) meta_finder_document.subdomain = subdomain meta_finder_document.target_domain = meta_dict.domain meta_finder_document.scan_history = meta_dict.scan_id item = DottedDict(results[meta]) meta_finder_document.url = item.url meta_finder_document.doc_name = meta meta_finder_document.http_status = item.status_code metadata = results[meta]['metadata'] for data in metadata: if 'Producer' in metadata and metadata['Producer']: meta_finder_document.producer = metadata['Producer'].rstrip('\x00') if 'Creator' in metadata and metadata['Creator']: meta_finder_document.creator = metadata['Creator'].rstrip('\x00') if 'CreationDate' in metadata and metadata['CreationDate']: meta_finder_document.creation_date = metadata['CreationDate'].rstrip('\x00') if 'ModDate' in metadata and metadata['ModDate']: meta_finder_document.modified_date = metadata['ModDate'].rstrip('\x00') if 'Author' in metadata and metadata['Author']: meta_finder_document.author = metadata['Author'].rstrip('\x00') if 'Title' in metadata and metadata['Title']: meta_finder_document.title = metadata['Title'].rstrip('\x00') if 'OSInfo' in metadata and metadata['OSInfo']: meta_finder_document.os = metadata['OSInfo'].rstrip('\x00') meta_finder_document.save() @app.task(bind=True) def test_task(self): print('*' * 40) print('test task run') print('*' * 40)

radaram

43af3a6aecdece4923ee74b108853f7b9c51ed12

27d6ec5827a51fd74e3ab97a5cef38fc7f5d9168

Сan I delete the old condition `if 'matched' in json_st`?

radaram

yogeshojha/rengine

web/Dockerfile: Update Go to v1.17 and add command to update Nuclei & Nuclei Templates

- Starting in Go 1.17, installing executables with go get is deprecated. go install may be used instead. [Deprecation of 'go get' for installing executables](https://golang.org/doc/go-get-install-deprecation). - Install and update Go package with `go install -v example.com/cmd@latest` or `GO111MODULE=on go install -v example.com/cmd@latest` - Add command to update Nuclei and Nuclei Templates

2021-10-25 09:33:01+00:00

2021-12-14 03:04:58+00:00

web/Dockerfile

# Base image FROM ubuntu:20.04 # Labels and Credits LABEL \ name="reNgine" \ author="Yogesh Ojha <yogesh.ojha11@gmail.com>" \ description="reNgine is a automated pipeline of recon process, useful for information gathering during web application penetration testing." # Environment Variables ENV DEBIAN_FRONTEND="noninteractive" \ DATABASE="postgres" ENV PYTHONDONTWRITEBYTECODE 1 ENV PYTHONUNBUFFERED 1 # Install essentials RUN apt update -y && apt install -y --no-install-recommends \ build-essential \ cmake \ firefox \ gcc \ git \ libpq-dev \ libpq-dev \ libpcap-dev \ netcat \ postgresql \ python3 \ python3-dev \ python3-pip \ python3-netaddr \ wget \ x11-utils \ xvfb # Download and install go 1.14 RUN wget https://dl.google.com/go/go1.16.5.linux-amd64.tar.gz RUN tar -xvf go1.16.5.linux-amd64.tar.gz RUN rm go1.16.5.linux-amd64.tar.gz RUN mv go /usr/local # Download geckodriver RUN wget https://github.com/mozilla/geckodriver/releases/download/v0.26.0/geckodriver-v0.26.0-linux64.tar.gz RUN tar -xvf geckodriver-v0.26.0-linux64.tar.gz RUN rm geckodriver-v0.26.0-linux64.tar.gz RUN mv geckodriver /usr/bin # ENV for Go ENV GOROOT="/usr/local/go" ENV PATH="${PATH}:${GOROOT}/bin" ENV PATH="${PATH}:${GOPATH}/bin" ENV GOPATH=$HOME/go ENV PATH="${PATH}:${GOROOT}/bin:${GOPATH}/bin" # Make directory for app WORKDIR /usr/src/app # set environment variables ENV PYTHONDONTWRITEBYTECODE 1 ENV PYTHONUNBUFFERED 1 # Download Go packages RUN go get -u github.com/tomnomnom/assetfinder github.com/hakluke/hakrawler RUN GO111MODULE=on go get -v github.com/projectdiscovery/httpx/cmd/httpx RUN GO111MODULE=on go get -v github.com/projectdiscovery/subfinder/v2/cmd/subfinder RUN GO111MODULE=on go get -v github.com/projectdiscovery/nuclei/v2/cmd/nuclei RUN GO111MODULE=on go get -v github.com/projectdiscovery/naabu/v2/cmd/naabu RUN GO111MODULE=on go get -u github.com/tomnomnom/unfurl RUN GO111MODULE=on go get -u -v github.com/bp0lr/gauplus RUN GO111MODULE=on go get github.com/tomnomnom/waybackurls RUN GO111MODULE=on go get -u github.com/jaeles-project/gospider RUN GO111MODULE=on go get -u github.com/tomnomnom/gf RUN go get -v github.com/OWASP/Amass/v3/... RUN go get -u github.com/tomnomnom/gf # Copy requirements COPY ./requirements.txt /tmp/requirements.txt RUN pip3 install --upgrade setuptools pip && \ pip3 install -r /tmp/requirements.txt # install eyewitness RUN python3 -m pip install fuzzywuzzy \ selenium \ python-Levenshtein \ pyvirtualdisplay \ netaddr # Copy source code COPY . /usr/src/app/ RUN chmod +x /usr/src/app/tools/get_urls.sh

# Base image FROM ubuntu:20.04 # Labels and Credits LABEL \ name="reNgine" \ author="Yogesh Ojha <yogesh.ojha11@gmail.com>" \ description="reNgine is a automated pipeline of recon process, useful for information gathering during web application penetration testing." # Environment Variables ENV DEBIAN_FRONTEND="noninteractive" \ DATABASE="postgres" ENV PYTHONDONTWRITEBYTECODE 1 ENV PYTHONUNBUFFERED 1 # Install essentials RUN apt update -y && apt install -y --no-install-recommends \ build-essential \ cmake \ firefox \ gcc \ git \ libpq-dev \ libpq-dev \ libpcap-dev \ netcat \ postgresql \ python3 \ python3-dev \ python3-pip \ python3-netaddr \ wget \ x11-utils \ xvfb # Download and install go 1.17 RUN wget https://golang.org/dl/go1.17.2.linux-amd64.tar.gz RUN tar -xvf go1.17.2.linux-amd64.tar.gz RUN rm go1.17.2.linux-amd64.tar.gz RUN mv go /usr/local # Download geckodriver RUN wget https://github.com/mozilla/geckodriver/releases/download/v0.26.0/geckodriver-v0.26.0-linux64.tar.gz RUN tar -xvf geckodriver-v0.26.0-linux64.tar.gz RUN rm geckodriver-v0.26.0-linux64.tar.gz RUN mv geckodriver /usr/bin # ENV for Go ENV GOROOT="/usr/local/go" ENV PATH="${PATH}:${GOROOT}/bin" ENV PATH="${PATH}:${GOPATH}/bin" ENV GOPATH=$HOME/go ENV PATH="${PATH}:${GOROOT}/bin:${GOPATH}/bin" # Make directory for app WORKDIR /usr/src/app # set environment variables ENV PYTHONDONTWRITEBYTECODE 1 ENV PYTHONUNBUFFERED 1 # Download Go packages RUN go install -v github.com/tomnomnom/assetfinder@latest RUN go install -v github.com/hakluke/hakrawler@latest RUN GO111MODULE=on go install -v github.com/projectdiscovery/httpx/cmd/httpx@latest RUN GO111MODULE=on go install -v github.com/projectdiscovery/subfinder/v2/cmd/subfinder@latest RUN GO111MODULE=on go install -v github.com/projectdiscovery/nuclei/v2/cmd/nuclei@latest RUN GO111MODULE=on go install -v github.com/projectdiscovery/naabu/v2/cmd/naabu@latest RUN GO111MODULE=on go install -v github.com/tomnomnom/unfurl@latest RUN GO111MODULE=on go install -v -v github.com/bp0lr/gauplus@latest RUN GO111MODULE=on go install -v github.com/tomnomnom/waybackurls@latest RUN GO111MODULE=on go install -v github.com/jaeles-project/gospider@latest RUN GO111MODULE=on go install -v github.com/tomnomnom/gf@latest RUN go install -v github.com/OWASP/Amass/v3/...@latest # Update Nuclei and Nuclei-Templates RUN nuclei -update RUN nuclei -update-templates # Copy requirements COPY ./requirements.txt /tmp/requirements.txt RUN pip3 install --upgrade setuptools pip && \ pip3 install -r /tmp/requirements.txt # install eyewitness RUN python3 -m pip install fuzzywuzzy \ selenium \ python-Levenshtein \ pyvirtualdisplay \ netaddr # Copy source code COPY . /usr/src/app/ RUN chmod +x /usr/src/app/tools/get_urls.sh

0x71rex

c46402b6381c17252e27baf0b1961849c51439a0

cf30e98e0440424019cb2cad600892ce405f850e

Sorry, my bad. What a silly mistake, there was extra '-v' on line 74 :grin:

0x71rex

yogeshojha/rengine

web/Dockerfile: Update Go to v1.17 and add command to update Nuclei & Nuclei Templates

- Starting in Go 1.17, installing executables with go get is deprecated. go install may be used instead. [Deprecation of 'go get' for installing executables](https://golang.org/doc/go-get-install-deprecation). - Install and update Go package with `go install -v example.com/cmd@latest` or `GO111MODULE=on go install -v example.com/cmd@latest` - Add command to update Nuclei and Nuclei Templates

2021-10-25 09:33:01+00:00

2021-12-14 03:04:58+00:00

web/Dockerfile

# Base image FROM ubuntu:20.04 # Labels and Credits LABEL \ name="reNgine" \ author="Yogesh Ojha <yogesh.ojha11@gmail.com>" \ description="reNgine is a automated pipeline of recon process, useful for information gathering during web application penetration testing." # Environment Variables ENV DEBIAN_FRONTEND="noninteractive" \ DATABASE="postgres" ENV PYTHONDONTWRITEBYTECODE 1 ENV PYTHONUNBUFFERED 1 # Install essentials RUN apt update -y && apt install -y --no-install-recommends \ build-essential \ cmake \ firefox \ gcc \ git \ libpq-dev \ libpq-dev \ libpcap-dev \ netcat \ postgresql \ python3 \ python3-dev \ python3-pip \ python3-netaddr \ wget \ x11-utils \ xvfb # Download and install go 1.14 RUN wget https://dl.google.com/go/go1.16.5.linux-amd64.tar.gz RUN tar -xvf go1.16.5.linux-amd64.tar.gz RUN rm go1.16.5.linux-amd64.tar.gz RUN mv go /usr/local # Download geckodriver RUN wget https://github.com/mozilla/geckodriver/releases/download/v0.26.0/geckodriver-v0.26.0-linux64.tar.gz RUN tar -xvf geckodriver-v0.26.0-linux64.tar.gz RUN rm geckodriver-v0.26.0-linux64.tar.gz RUN mv geckodriver /usr/bin # ENV for Go ENV GOROOT="/usr/local/go" ENV PATH="${PATH}:${GOROOT}/bin" ENV PATH="${PATH}:${GOPATH}/bin" ENV GOPATH=$HOME/go ENV PATH="${PATH}:${GOROOT}/bin:${GOPATH}/bin" # Make directory for app WORKDIR /usr/src/app # set environment variables ENV PYTHONDONTWRITEBYTECODE 1 ENV PYTHONUNBUFFERED 1 # Download Go packages RUN go get -u github.com/tomnomnom/assetfinder github.com/hakluke/hakrawler RUN GO111MODULE=on go get -v github.com/projectdiscovery/httpx/cmd/httpx RUN GO111MODULE=on go get -v github.com/projectdiscovery/subfinder/v2/cmd/subfinder RUN GO111MODULE=on go get -v github.com/projectdiscovery/nuclei/v2/cmd/nuclei RUN GO111MODULE=on go get -v github.com/projectdiscovery/naabu/v2/cmd/naabu RUN GO111MODULE=on go get -u github.com/tomnomnom/unfurl RUN GO111MODULE=on go get -u -v github.com/bp0lr/gauplus RUN GO111MODULE=on go get github.com/tomnomnom/waybackurls RUN GO111MODULE=on go get -u github.com/jaeles-project/gospider RUN GO111MODULE=on go get -u github.com/tomnomnom/gf RUN go get -v github.com/OWASP/Amass/v3/... RUN go get -u github.com/tomnomnom/gf # Copy requirements COPY ./requirements.txt /tmp/requirements.txt RUN pip3 install --upgrade setuptools pip && \ pip3 install -r /tmp/requirements.txt # install eyewitness RUN python3 -m pip install fuzzywuzzy \ selenium \ python-Levenshtein \ pyvirtualdisplay \ netaddr # Copy source code COPY . /usr/src/app/ RUN chmod +x /usr/src/app/tools/get_urls.sh

# Base image FROM ubuntu:20.04 # Labels and Credits LABEL \ name="reNgine" \ author="Yogesh Ojha <yogesh.ojha11@gmail.com>" \ description="reNgine is a automated pipeline of recon process, useful for information gathering during web application penetration testing." # Environment Variables ENV DEBIAN_FRONTEND="noninteractive" \ DATABASE="postgres" ENV PYTHONDONTWRITEBYTECODE 1 ENV PYTHONUNBUFFERED 1 # Install essentials RUN apt update -y && apt install -y --no-install-recommends \ build-essential \ cmake \ firefox \ gcc \ git \ libpq-dev \ libpq-dev \ libpcap-dev \ netcat \ postgresql \ python3 \ python3-dev \ python3-pip \ python3-netaddr \ wget \ x11-utils \ xvfb # Download and install go 1.17 RUN wget https://golang.org/dl/go1.17.2.linux-amd64.tar.gz RUN tar -xvf go1.17.2.linux-amd64.tar.gz RUN rm go1.17.2.linux-amd64.tar.gz RUN mv go /usr/local # Download geckodriver RUN wget https://github.com/mozilla/geckodriver/releases/download/v0.26.0/geckodriver-v0.26.0-linux64.tar.gz RUN tar -xvf geckodriver-v0.26.0-linux64.tar.gz RUN rm geckodriver-v0.26.0-linux64.tar.gz RUN mv geckodriver /usr/bin # ENV for Go ENV GOROOT="/usr/local/go" ENV PATH="${PATH}:${GOROOT}/bin" ENV PATH="${PATH}:${GOPATH}/bin" ENV GOPATH=$HOME/go ENV PATH="${PATH}:${GOROOT}/bin:${GOPATH}/bin" # Make directory for app WORKDIR /usr/src/app # set environment variables ENV PYTHONDONTWRITEBYTECODE 1 ENV PYTHONUNBUFFERED 1 # Download Go packages RUN go install -v github.com/tomnomnom/assetfinder@latest RUN go install -v github.com/hakluke/hakrawler@latest RUN GO111MODULE=on go install -v github.com/projectdiscovery/httpx/cmd/httpx@latest RUN GO111MODULE=on go install -v github.com/projectdiscovery/subfinder/v2/cmd/subfinder@latest RUN GO111MODULE=on go install -v github.com/projectdiscovery/nuclei/v2/cmd/nuclei@latest RUN GO111MODULE=on go install -v github.com/projectdiscovery/naabu/v2/cmd/naabu@latest RUN GO111MODULE=on go install -v github.com/tomnomnom/unfurl@latest RUN GO111MODULE=on go install -v -v github.com/bp0lr/gauplus@latest RUN GO111MODULE=on go install -v github.com/tomnomnom/waybackurls@latest RUN GO111MODULE=on go install -v github.com/jaeles-project/gospider@latest RUN GO111MODULE=on go install -v github.com/tomnomnom/gf@latest RUN go install -v github.com/OWASP/Amass/v3/...@latest # Update Nuclei and Nuclei-Templates RUN nuclei -update RUN nuclei -update-templates # Copy requirements COPY ./requirements.txt /tmp/requirements.txt RUN pip3 install --upgrade setuptools pip && \ pip3 install -r /tmp/requirements.txt # install eyewitness RUN python3 -m pip install fuzzywuzzy \ selenium \ python-Levenshtein \ pyvirtualdisplay \ netaddr # Copy source code COPY . /usr/src/app/ RUN chmod +x /usr/src/app/tools/get_urls.sh

0x71rex

c46402b6381c17252e27baf0b1961849c51439a0

cf30e98e0440424019cb2cad600892ce405f850e

Haven't remove the second "-v" in line 74 btw. 😁

0x71rex

yogeshojha/rengine

Installation: Check docker running status before installing reNgine.

### Changes - checks docker run status. Before it used to execute all `make build` commands even if docker was not running. - Early error shown to the user - Terminal text color changed - made this sentence "Before running this script, please make sure Docker is running and you have made changes to .env file." color RED because it can throw errors - "Changing the postgres username & password from .env is highly recommended." to Green since it's a suggestion

2021-08-24 16:52:53+00:00

2021-08-27 02:54:16+00:00

install.sh

#!/bin/bash tput setaf 2; cat web/art/1.0.txt tput setaf 3; echo "Before running this script, please make sure you have made changes to .env file." tput setaf 1; echo "Changing the postgres username & password from .env is highly recommended." tput setaf 4; read -p "Are you sure, you made changes to .env file (y/n)? " answer case ${answer:0:1} in y|Y|yes|YES|Yes ) echo "Continiuing Installation!" ;; * ) nano .env ;; esac echo " " tput setaf 3; echo "#########################################################################" echo "Please note that, this installation script is only intended for Linux" echo "For Mac and Windows, refer to the official guide https://rengine.wiki" echo "#########################################################################" echo " " tput setaf 4; echo "Installing reNgine and it's dependencies" echo " " if [ "$EUID" -ne 0 ] then tput setaf 1; echo "Error installing reNgine, Please run this script as root!" tput setaf 1; echo "Example: sudo ./install.sh" exit fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing Docker..." echo "#########################################################################" if [ -x "$(command -v docker)" ]; then tput setaf 2; echo "Docker already installed, skipping." else curl -fsSL https://get.docker.com -o get-docker.sh && sh get-docker.sh tput setaf 2; echo "Docker installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing docker-compose" echo "#########################################################################" if [ -x "$(command -v docker-compose)" ]; then tput setaf 2; echo "docker-compose already installed, skipping." else curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose tput setaf 2; echo "docker-compose installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing make" echo "#########################################################################" if [ -x "$(command -v make)" ]; then tput setaf 2; echo "make already installed, skipping." else apt install make fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing reNgine" echo "#########################################################################" make certs && make build && make up tput setaf 2; echo "reNgine is installed!!!" sleep 3 echo " " tput setaf 4; echo "#########################################################################" echo "Creating an account" echo "#########################################################################" make username tput setaf 2; echo "Thank you for installing reNgine, happy recon!!"

#!/bin/bash tput setaf 2; cat web/art/1.0.txt tput setaf 1; echo "Before running this script, please make sure Docker is running and you have made changes to .env file." tput setaf 2; echo "Changing the postgres username & password from .env is highly recommended." tput setaf 4; read -p "Are you sure, you made changes to .env file (y/n)? " answer case ${answer:0:1} in y|Y|yes|YES|Yes ) echo "Continiuing Installation!" ;; * ) nano .env ;; esac echo " " tput setaf 3; echo "#########################################################################" echo "Please note that, this installation script is only intended for Linux" echo "For Mac and Windows, refer to the official guide https://rengine.wiki" echo "#########################################################################" echo " " tput setaf 4; echo "Installing reNgine and it's dependencies" echo " " if [ "$EUID" -ne 0 ] then tput setaf 1; echo "Error installing reNgine, Please run this script as root!" tput setaf 1; echo "Example: sudo ./install.sh" exit fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing Docker..." echo "#########################################################################" if [ -x "$(command -v docker)" ]; then tput setaf 2; echo "Docker already installed, skipping." else curl -fsSL https://get.docker.com -o get-docker.sh && sh get-docker.sh tput setaf 2; echo "Docker installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing docker-compose" echo "#########################################################################" if [ -x "$(command -v docker-compose)" ]; then tput setaf 2; echo "docker-compose already installed, skipping." else curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose tput setaf 2; echo "docker-compose installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing make" echo "#########################################################################" if [ -x "$(command -v make)" ]; then tput setaf 2; echo "make already installed, skipping." else apt install make fi echo " " tput setaf 4; echo "#########################################################################" echo "Checking Docker status" echo "#########################################################################" if systemctl is-active docker >/dev/null 2>&1; then tput setaf 4; echo "Docker is running." else tput setaf 1; echo "Docker is not running. Please run docker and try again." echo "You can run docker service using sudo systemctl start docker" exit 1 fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing reNgine" echo "#########################################################################" make certs && make build && make up tput setaf 2; echo "reNgine is installed!!!" sleep 3 echo " " tput setaf 4; echo "#########################################################################" echo "Creating an account" echo "#########################################################################" make username tput setaf 2; echo "Thank you for installing reNgine, happy recon!!"

sbimochan

2bd2219659fcf0f0541fc4879bd69bfa79a500c7

e98433517e4a6198e6e2208fdf1b324f41be5bcb

It won't run in macOS or windows.

sbimochan

yogeshojha/rengine

Installation: Check docker running status before installing reNgine.

### Changes - checks docker run status. Before it used to execute all `make build` commands even if docker was not running. - Early error shown to the user - Terminal text color changed - made this sentence "Before running this script, please make sure Docker is running and you have made changes to .env file." color RED because it can throw errors - "Changing the postgres username & password from .env is highly recommended." to Green since it's a suggestion

2021-08-24 16:52:53+00:00

2021-08-27 02:54:16+00:00

install.sh

#!/bin/bash tput setaf 2; cat web/art/1.0.txt tput setaf 3; echo "Before running this script, please make sure you have made changes to .env file." tput setaf 1; echo "Changing the postgres username & password from .env is highly recommended." tput setaf 4; read -p "Are you sure, you made changes to .env file (y/n)? " answer case ${answer:0:1} in y|Y|yes|YES|Yes ) echo "Continiuing Installation!" ;; * ) nano .env ;; esac echo " " tput setaf 3; echo "#########################################################################" echo "Please note that, this installation script is only intended for Linux" echo "For Mac and Windows, refer to the official guide https://rengine.wiki" echo "#########################################################################" echo " " tput setaf 4; echo "Installing reNgine and it's dependencies" echo " " if [ "$EUID" -ne 0 ] then tput setaf 1; echo "Error installing reNgine, Please run this script as root!" tput setaf 1; echo "Example: sudo ./install.sh" exit fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing Docker..." echo "#########################################################################" if [ -x "$(command -v docker)" ]; then tput setaf 2; echo "Docker already installed, skipping." else curl -fsSL https://get.docker.com -o get-docker.sh && sh get-docker.sh tput setaf 2; echo "Docker installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing docker-compose" echo "#########################################################################" if [ -x "$(command -v docker-compose)" ]; then tput setaf 2; echo "docker-compose already installed, skipping." else curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose tput setaf 2; echo "docker-compose installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing make" echo "#########################################################################" if [ -x "$(command -v make)" ]; then tput setaf 2; echo "make already installed, skipping." else apt install make fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing reNgine" echo "#########################################################################" make certs && make build && make up tput setaf 2; echo "reNgine is installed!!!" sleep 3 echo " " tput setaf 4; echo "#########################################################################" echo "Creating an account" echo "#########################################################################" make username tput setaf 2; echo "Thank you for installing reNgine, happy recon!!"

#!/bin/bash tput setaf 2; cat web/art/1.0.txt tput setaf 1; echo "Before running this script, please make sure Docker is running and you have made changes to .env file." tput setaf 2; echo "Changing the postgres username & password from .env is highly recommended." tput setaf 4; read -p "Are you sure, you made changes to .env file (y/n)? " answer case ${answer:0:1} in y|Y|yes|YES|Yes ) echo "Continiuing Installation!" ;; * ) nano .env ;; esac echo " " tput setaf 3; echo "#########################################################################" echo "Please note that, this installation script is only intended for Linux" echo "For Mac and Windows, refer to the official guide https://rengine.wiki" echo "#########################################################################" echo " " tput setaf 4; echo "Installing reNgine and it's dependencies" echo " " if [ "$EUID" -ne 0 ] then tput setaf 1; echo "Error installing reNgine, Please run this script as root!" tput setaf 1; echo "Example: sudo ./install.sh" exit fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing Docker..." echo "#########################################################################" if [ -x "$(command -v docker)" ]; then tput setaf 2; echo "Docker already installed, skipping." else curl -fsSL https://get.docker.com -o get-docker.sh && sh get-docker.sh tput setaf 2; echo "Docker installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing docker-compose" echo "#########################################################################" if [ -x "$(command -v docker-compose)" ]; then tput setaf 2; echo "docker-compose already installed, skipping." else curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose tput setaf 2; echo "docker-compose installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing make" echo "#########################################################################" if [ -x "$(command -v make)" ]; then tput setaf 2; echo "make already installed, skipping." else apt install make fi echo " " tput setaf 4; echo "#########################################################################" echo "Checking Docker status" echo "#########################################################################" if systemctl is-active docker >/dev/null 2>&1; then tput setaf 4; echo "Docker is running." else tput setaf 1; echo "Docker is not running. Please run docker and try again." echo "You can run docker service using sudo systemctl start docker" exit 1 fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing reNgine" echo "#########################################################################" make certs && make build && make up tput setaf 2; echo "reNgine is installed!!!" sleep 3 echo " " tput setaf 4; echo "#########################################################################" echo "Creating an account" echo "#########################################################################" make username tput setaf 2; echo "Thank you for installing reNgine, happy recon!!"

sbimochan

2bd2219659fcf0f0541fc4879bd69bfa79a500c7

e98433517e4a6198e6e2208fdf1b324f41be5bcb

This quick install script is only for Ubuntu/Debian bases OS. https://rengine.wiki/install/quick-install/ Other OS will continue to use https://rengine.wiki/install/install/

yogeshojha

yogeshojha/rengine

Installation: Check docker running status before installing reNgine.

### Changes - checks docker run status. Before it used to execute all `make build` commands even if docker was not running. - Early error shown to the user - Terminal text color changed - made this sentence "Before running this script, please make sure Docker is running and you have made changes to .env file." color RED because it can throw errors - "Changing the postgres username & password from .env is highly recommended." to Green since it's a suggestion

2021-08-24 16:52:53+00:00

2021-08-27 02:54:16+00:00

install.sh

#!/bin/bash tput setaf 2; cat web/art/1.0.txt tput setaf 3; echo "Before running this script, please make sure you have made changes to .env file." tput setaf 1; echo "Changing the postgres username & password from .env is highly recommended." tput setaf 4; read -p "Are you sure, you made changes to .env file (y/n)? " answer case ${answer:0:1} in y|Y|yes|YES|Yes ) echo "Continiuing Installation!" ;; * ) nano .env ;; esac echo " " tput setaf 3; echo "#########################################################################" echo "Please note that, this installation script is only intended for Linux" echo "For Mac and Windows, refer to the official guide https://rengine.wiki" echo "#########################################################################" echo " " tput setaf 4; echo "Installing reNgine and it's dependencies" echo " " if [ "$EUID" -ne 0 ] then tput setaf 1; echo "Error installing reNgine, Please run this script as root!" tput setaf 1; echo "Example: sudo ./install.sh" exit fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing Docker..." echo "#########################################################################" if [ -x "$(command -v docker)" ]; then tput setaf 2; echo "Docker already installed, skipping." else curl -fsSL https://get.docker.com -o get-docker.sh && sh get-docker.sh tput setaf 2; echo "Docker installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing docker-compose" echo "#########################################################################" if [ -x "$(command -v docker-compose)" ]; then tput setaf 2; echo "docker-compose already installed, skipping." else curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose tput setaf 2; echo "docker-compose installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing make" echo "#########################################################################" if [ -x "$(command -v make)" ]; then tput setaf 2; echo "make already installed, skipping." else apt install make fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing reNgine" echo "#########################################################################" make certs && make build && make up tput setaf 2; echo "reNgine is installed!!!" sleep 3 echo " " tput setaf 4; echo "#########################################################################" echo "Creating an account" echo "#########################################################################" make username tput setaf 2; echo "Thank you for installing reNgine, happy recon!!"

#!/bin/bash tput setaf 2; cat web/art/1.0.txt tput setaf 1; echo "Before running this script, please make sure Docker is running and you have made changes to .env file." tput setaf 2; echo "Changing the postgres username & password from .env is highly recommended." tput setaf 4; read -p "Are you sure, you made changes to .env file (y/n)? " answer case ${answer:0:1} in y|Y|yes|YES|Yes ) echo "Continiuing Installation!" ;; * ) nano .env ;; esac echo " " tput setaf 3; echo "#########################################################################" echo "Please note that, this installation script is only intended for Linux" echo "For Mac and Windows, refer to the official guide https://rengine.wiki" echo "#########################################################################" echo " " tput setaf 4; echo "Installing reNgine and it's dependencies" echo " " if [ "$EUID" -ne 0 ] then tput setaf 1; echo "Error installing reNgine, Please run this script as root!" tput setaf 1; echo "Example: sudo ./install.sh" exit fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing Docker..." echo "#########################################################################" if [ -x "$(command -v docker)" ]; then tput setaf 2; echo "Docker already installed, skipping." else curl -fsSL https://get.docker.com -o get-docker.sh && sh get-docker.sh tput setaf 2; echo "Docker installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing docker-compose" echo "#########################################################################" if [ -x "$(command -v docker-compose)" ]; then tput setaf 2; echo "docker-compose already installed, skipping." else curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose tput setaf 2; echo "docker-compose installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing make" echo "#########################################################################" if [ -x "$(command -v make)" ]; then tput setaf 2; echo "make already installed, skipping." else apt install make fi echo " " tput setaf 4; echo "#########################################################################" echo "Checking Docker status" echo "#########################################################################" if systemctl is-active docker >/dev/null 2>&1; then tput setaf 4; echo "Docker is running." else tput setaf 1; echo "Docker is not running. Please run docker and try again." echo "You can run docker service using sudo systemctl start docker" exit 1 fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing reNgine" echo "#########################################################################" make certs && make build && make up tput setaf 2; echo "reNgine is installed!!!" sleep 3 echo " " tput setaf 4; echo "#########################################################################" echo "Creating an account" echo "#########################################################################" make username tput setaf 2; echo "Thank you for installing reNgine, happy recon!!"

sbimochan

2bd2219659fcf0f0541fc4879bd69bfa79a500c7

e98433517e4a6198e6e2208fdf1b324f41be5bcb

but docker info would run in all OS right? sorry why did it fail the docker test before?

sbimochan

yogeshojha/rengine

Installation: Check docker running status before installing reNgine.

### Changes - checks docker run status. Before it used to execute all `make build` commands even if docker was not running. - Early error shown to the user - Terminal text color changed - made this sentence "Before running this script, please make sure Docker is running and you have made changes to .env file." color RED because it can throw errors - "Changing the postgres username & password from .env is highly recommended." to Green since it's a suggestion

2021-08-24 16:52:53+00:00

2021-08-27 02:54:16+00:00

install.sh

#!/bin/bash tput setaf 2; cat web/art/1.0.txt tput setaf 3; echo "Before running this script, please make sure you have made changes to .env file." tput setaf 1; echo "Changing the postgres username & password from .env is highly recommended." tput setaf 4; read -p "Are you sure, you made changes to .env file (y/n)? " answer case ${answer:0:1} in y|Y|yes|YES|Yes ) echo "Continiuing Installation!" ;; * ) nano .env ;; esac echo " " tput setaf 3; echo "#########################################################################" echo "Please note that, this installation script is only intended for Linux" echo "For Mac and Windows, refer to the official guide https://rengine.wiki" echo "#########################################################################" echo " " tput setaf 4; echo "Installing reNgine and it's dependencies" echo " " if [ "$EUID" -ne 0 ] then tput setaf 1; echo "Error installing reNgine, Please run this script as root!" tput setaf 1; echo "Example: sudo ./install.sh" exit fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing Docker..." echo "#########################################################################" if [ -x "$(command -v docker)" ]; then tput setaf 2; echo "Docker already installed, skipping." else curl -fsSL https://get.docker.com -o get-docker.sh && sh get-docker.sh tput setaf 2; echo "Docker installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing docker-compose" echo "#########################################################################" if [ -x "$(command -v docker-compose)" ]; then tput setaf 2; echo "docker-compose already installed, skipping." else curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose tput setaf 2; echo "docker-compose installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing make" echo "#########################################################################" if [ -x "$(command -v make)" ]; then tput setaf 2; echo "make already installed, skipping." else apt install make fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing reNgine" echo "#########################################################################" make certs && make build && make up tput setaf 2; echo "reNgine is installed!!!" sleep 3 echo " " tput setaf 4; echo "#########################################################################" echo "Creating an account" echo "#########################################################################" make username tput setaf 2; echo "Thank you for installing reNgine, happy recon!!"

#!/bin/bash tput setaf 2; cat web/art/1.0.txt tput setaf 1; echo "Before running this script, please make sure Docker is running and you have made changes to .env file." tput setaf 2; echo "Changing the postgres username & password from .env is highly recommended." tput setaf 4; read -p "Are you sure, you made changes to .env file (y/n)? " answer case ${answer:0:1} in y|Y|yes|YES|Yes ) echo "Continiuing Installation!" ;; * ) nano .env ;; esac echo " " tput setaf 3; echo "#########################################################################" echo "Please note that, this installation script is only intended for Linux" echo "For Mac and Windows, refer to the official guide https://rengine.wiki" echo "#########################################################################" echo " " tput setaf 4; echo "Installing reNgine and it's dependencies" echo " " if [ "$EUID" -ne 0 ] then tput setaf 1; echo "Error installing reNgine, Please run this script as root!" tput setaf 1; echo "Example: sudo ./install.sh" exit fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing Docker..." echo "#########################################################################" if [ -x "$(command -v docker)" ]; then tput setaf 2; echo "Docker already installed, skipping." else curl -fsSL https://get.docker.com -o get-docker.sh && sh get-docker.sh tput setaf 2; echo "Docker installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing docker-compose" echo "#########################################################################" if [ -x "$(command -v docker-compose)" ]; then tput setaf 2; echo "docker-compose already installed, skipping." else curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose tput setaf 2; echo "docker-compose installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing make" echo "#########################################################################" if [ -x "$(command -v make)" ]; then tput setaf 2; echo "make already installed, skipping." else apt install make fi echo " " tput setaf 4; echo "#########################################################################" echo "Checking Docker status" echo "#########################################################################" if systemctl is-active docker >/dev/null 2>&1; then tput setaf 4; echo "Docker is running." else tput setaf 1; echo "Docker is not running. Please run docker and try again." echo "You can run docker service using sudo systemctl start docker" exit 1 fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing reNgine" echo "#########################################################################" make certs && make build && make up tput setaf 2; echo "reNgine is installed!!!" sleep 3 echo " " tput setaf 4; echo "#########################################################################" echo "Creating an account" echo "#########################################################################" make username tput setaf 2; echo "Thank you for installing reNgine, happy recon!!"

sbimochan

2bd2219659fcf0f0541fc4879bd69bfa79a500c7

e98433517e4a6198e6e2208fdf1b324f41be5bcb

Correct me if I am wrong, `docker info` will return True if docker is installed, and it has nothing to do with whether docker is running or not. You can give it a try Stop docker service `sudo systemctl stop docker` Now try with docker info `sudo docker info` On the other hand, try with `sudo systemctl is-active docker` You'll get active/inactive based on if docker service is running.

yogeshojha

yogeshojha/rengine

Installation: Check docker running status before installing reNgine.

### Changes - checks docker run status. Before it used to execute all `make build` commands even if docker was not running. - Early error shown to the user - Terminal text color changed - made this sentence "Before running this script, please make sure Docker is running and you have made changes to .env file." color RED because it can throw errors - "Changing the postgres username & password from .env is highly recommended." to Green since it's a suggestion

2021-08-24 16:52:53+00:00

2021-08-27 02:54:16+00:00

install.sh

#!/bin/bash tput setaf 2; cat web/art/1.0.txt tput setaf 3; echo "Before running this script, please make sure you have made changes to .env file." tput setaf 1; echo "Changing the postgres username & password from .env is highly recommended." tput setaf 4; read -p "Are you sure, you made changes to .env file (y/n)? " answer case ${answer:0:1} in y|Y|yes|YES|Yes ) echo "Continiuing Installation!" ;; * ) nano .env ;; esac echo " " tput setaf 3; echo "#########################################################################" echo "Please note that, this installation script is only intended for Linux" echo "For Mac and Windows, refer to the official guide https://rengine.wiki" echo "#########################################################################" echo " " tput setaf 4; echo "Installing reNgine and it's dependencies" echo " " if [ "$EUID" -ne 0 ] then tput setaf 1; echo "Error installing reNgine, Please run this script as root!" tput setaf 1; echo "Example: sudo ./install.sh" exit fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing Docker..." echo "#########################################################################" if [ -x "$(command -v docker)" ]; then tput setaf 2; echo "Docker already installed, skipping." else curl -fsSL https://get.docker.com -o get-docker.sh && sh get-docker.sh tput setaf 2; echo "Docker installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing docker-compose" echo "#########################################################################" if [ -x "$(command -v docker-compose)" ]; then tput setaf 2; echo "docker-compose already installed, skipping." else curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose tput setaf 2; echo "docker-compose installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing make" echo "#########################################################################" if [ -x "$(command -v make)" ]; then tput setaf 2; echo "make already installed, skipping." else apt install make fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing reNgine" echo "#########################################################################" make certs && make build && make up tput setaf 2; echo "reNgine is installed!!!" sleep 3 echo " " tput setaf 4; echo "#########################################################################" echo "Creating an account" echo "#########################################################################" make username tput setaf 2; echo "Thank you for installing reNgine, happy recon!!"

#!/bin/bash tput setaf 2; cat web/art/1.0.txt tput setaf 1; echo "Before running this script, please make sure Docker is running and you have made changes to .env file." tput setaf 2; echo "Changing the postgres username & password from .env is highly recommended." tput setaf 4; read -p "Are you sure, you made changes to .env file (y/n)? " answer case ${answer:0:1} in y|Y|yes|YES|Yes ) echo "Continiuing Installation!" ;; * ) nano .env ;; esac echo " " tput setaf 3; echo "#########################################################################" echo "Please note that, this installation script is only intended for Linux" echo "For Mac and Windows, refer to the official guide https://rengine.wiki" echo "#########################################################################" echo " " tput setaf 4; echo "Installing reNgine and it's dependencies" echo " " if [ "$EUID" -ne 0 ] then tput setaf 1; echo "Error installing reNgine, Please run this script as root!" tput setaf 1; echo "Example: sudo ./install.sh" exit fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing Docker..." echo "#########################################################################" if [ -x "$(command -v docker)" ]; then tput setaf 2; echo "Docker already installed, skipping." else curl -fsSL https://get.docker.com -o get-docker.sh && sh get-docker.sh tput setaf 2; echo "Docker installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing docker-compose" echo "#########################################################################" if [ -x "$(command -v docker-compose)" ]; then tput setaf 2; echo "docker-compose already installed, skipping." else curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose tput setaf 2; echo "docker-compose installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing make" echo "#########################################################################" if [ -x "$(command -v make)" ]; then tput setaf 2; echo "make already installed, skipping." else apt install make fi echo " " tput setaf 4; echo "#########################################################################" echo "Checking Docker status" echo "#########################################################################" if systemctl is-active docker >/dev/null 2>&1; then tput setaf 4; echo "Docker is running." else tput setaf 1; echo "Docker is not running. Please run docker and try again." echo "You can run docker service using sudo systemctl start docker" exit 1 fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing reNgine" echo "#########################################################################" make certs && make build && make up tput setaf 2; echo "reNgine is installed!!!" sleep 3 echo " " tput setaf 4; echo "#########################################################################" echo "Creating an account" echo "#########################################################################" make username tput setaf 2; echo "Thank you for installing reNgine, happy recon!!"

sbimochan

2bd2219659fcf0f0541fc4879bd69bfa79a500c7

e98433517e4a6198e6e2208fdf1b324f41be5bcb

let me test in my friend's linux.

sbimochan

yogeshojha/rengine

Installation: Check docker running status before installing reNgine.

### Changes - checks docker run status. Before it used to execute all `make build` commands even if docker was not running. - Early error shown to the user - Terminal text color changed - made this sentence "Before running this script, please make sure Docker is running and you have made changes to .env file." color RED because it can throw errors - "Changing the postgres username & password from .env is highly recommended." to Green since it's a suggestion

2021-08-24 16:52:53+00:00

2021-08-27 02:54:16+00:00

install.sh

#!/bin/bash tput setaf 2; cat web/art/1.0.txt tput setaf 3; echo "Before running this script, please make sure you have made changes to .env file." tput setaf 1; echo "Changing the postgres username & password from .env is highly recommended." tput setaf 4; read -p "Are you sure, you made changes to .env file (y/n)? " answer case ${answer:0:1} in y|Y|yes|YES|Yes ) echo "Continiuing Installation!" ;; * ) nano .env ;; esac echo " " tput setaf 3; echo "#########################################################################" echo "Please note that, this installation script is only intended for Linux" echo "For Mac and Windows, refer to the official guide https://rengine.wiki" echo "#########################################################################" echo " " tput setaf 4; echo "Installing reNgine and it's dependencies" echo " " if [ "$EUID" -ne 0 ] then tput setaf 1; echo "Error installing reNgine, Please run this script as root!" tput setaf 1; echo "Example: sudo ./install.sh" exit fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing Docker..." echo "#########################################################################" if [ -x "$(command -v docker)" ]; then tput setaf 2; echo "Docker already installed, skipping." else curl -fsSL https://get.docker.com -o get-docker.sh && sh get-docker.sh tput setaf 2; echo "Docker installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing docker-compose" echo "#########################################################################" if [ -x "$(command -v docker-compose)" ]; then tput setaf 2; echo "docker-compose already installed, skipping." else curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose tput setaf 2; echo "docker-compose installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing make" echo "#########################################################################" if [ -x "$(command -v make)" ]; then tput setaf 2; echo "make already installed, skipping." else apt install make fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing reNgine" echo "#########################################################################" make certs && make build && make up tput setaf 2; echo "reNgine is installed!!!" sleep 3 echo " " tput setaf 4; echo "#########################################################################" echo "Creating an account" echo "#########################################################################" make username tput setaf 2; echo "Thank you for installing reNgine, happy recon!!"

#!/bin/bash tput setaf 2; cat web/art/1.0.txt tput setaf 1; echo "Before running this script, please make sure Docker is running and you have made changes to .env file." tput setaf 2; echo "Changing the postgres username & password from .env is highly recommended." tput setaf 4; read -p "Are you sure, you made changes to .env file (y/n)? " answer case ${answer:0:1} in y|Y|yes|YES|Yes ) echo "Continiuing Installation!" ;; * ) nano .env ;; esac echo " " tput setaf 3; echo "#########################################################################" echo "Please note that, this installation script is only intended for Linux" echo "For Mac and Windows, refer to the official guide https://rengine.wiki" echo "#########################################################################" echo " " tput setaf 4; echo "Installing reNgine and it's dependencies" echo " " if [ "$EUID" -ne 0 ] then tput setaf 1; echo "Error installing reNgine, Please run this script as root!" tput setaf 1; echo "Example: sudo ./install.sh" exit fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing Docker..." echo "#########################################################################" if [ -x "$(command -v docker)" ]; then tput setaf 2; echo "Docker already installed, skipping." else curl -fsSL https://get.docker.com -o get-docker.sh && sh get-docker.sh tput setaf 2; echo "Docker installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing docker-compose" echo "#########################################################################" if [ -x "$(command -v docker-compose)" ]; then tput setaf 2; echo "docker-compose already installed, skipping." else curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose tput setaf 2; echo "docker-compose installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing make" echo "#########################################################################" if [ -x "$(command -v make)" ]; then tput setaf 2; echo "make already installed, skipping." else apt install make fi echo " " tput setaf 4; echo "#########################################################################" echo "Checking Docker status" echo "#########################################################################" if systemctl is-active docker >/dev/null 2>&1; then tput setaf 4; echo "Docker is running." else tput setaf 1; echo "Docker is not running. Please run docker and try again." echo "You can run docker service using sudo systemctl start docker" exit 1 fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing reNgine" echo "#########################################################################" make certs && make build && make up tput setaf 2; echo "reNgine is installed!!!" sleep 3 echo " " tput setaf 4; echo "#########################################################################" echo "Creating an account" echo "#########################################################################" make username tput setaf 2; echo "Thank you for installing reNgine, happy recon!!"

sbimochan

2bd2219659fcf0f0541fc4879bd69bfa79a500c7

e98433517e4a6198e6e2208fdf1b324f41be5bcb

confirmed. your way is right for linux. my way was right for 'macOS'

sbimochan

yogeshojha/rengine

Installation: Check docker running status before installing reNgine.

### Changes - checks docker run status. Before it used to execute all `make build` commands even if docker was not running. - Early error shown to the user - Terminal text color changed - made this sentence "Before running this script, please make sure Docker is running and you have made changes to .env file." color RED because it can throw errors - "Changing the postgres username & password from .env is highly recommended." to Green since it's a suggestion

2021-08-24 16:52:53+00:00

2021-08-27 02:54:16+00:00

install.sh

#!/bin/bash tput setaf 2; cat web/art/1.0.txt tput setaf 3; echo "Before running this script, please make sure you have made changes to .env file." tput setaf 1; echo "Changing the postgres username & password from .env is highly recommended." tput setaf 4; read -p "Are you sure, you made changes to .env file (y/n)? " answer case ${answer:0:1} in y|Y|yes|YES|Yes ) echo "Continiuing Installation!" ;; * ) nano .env ;; esac echo " " tput setaf 3; echo "#########################################################################" echo "Please note that, this installation script is only intended for Linux" echo "For Mac and Windows, refer to the official guide https://rengine.wiki" echo "#########################################################################" echo " " tput setaf 4; echo "Installing reNgine and it's dependencies" echo " " if [ "$EUID" -ne 0 ] then tput setaf 1; echo "Error installing reNgine, Please run this script as root!" tput setaf 1; echo "Example: sudo ./install.sh" exit fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing Docker..." echo "#########################################################################" if [ -x "$(command -v docker)" ]; then tput setaf 2; echo "Docker already installed, skipping." else curl -fsSL https://get.docker.com -o get-docker.sh && sh get-docker.sh tput setaf 2; echo "Docker installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing docker-compose" echo "#########################################################################" if [ -x "$(command -v docker-compose)" ]; then tput setaf 2; echo "docker-compose already installed, skipping." else curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose tput setaf 2; echo "docker-compose installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing make" echo "#########################################################################" if [ -x "$(command -v make)" ]; then tput setaf 2; echo "make already installed, skipping." else apt install make fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing reNgine" echo "#########################################################################" make certs && make build && make up tput setaf 2; echo "reNgine is installed!!!" sleep 3 echo " " tput setaf 4; echo "#########################################################################" echo "Creating an account" echo "#########################################################################" make username tput setaf 2; echo "Thank you for installing reNgine, happy recon!!"

#!/bin/bash tput setaf 2; cat web/art/1.0.txt tput setaf 1; echo "Before running this script, please make sure Docker is running and you have made changes to .env file." tput setaf 2; echo "Changing the postgres username & password from .env is highly recommended." tput setaf 4; read -p "Are you sure, you made changes to .env file (y/n)? " answer case ${answer:0:1} in y|Y|yes|YES|Yes ) echo "Continiuing Installation!" ;; * ) nano .env ;; esac echo " " tput setaf 3; echo "#########################################################################" echo "Please note that, this installation script is only intended for Linux" echo "For Mac and Windows, refer to the official guide https://rengine.wiki" echo "#########################################################################" echo " " tput setaf 4; echo "Installing reNgine and it's dependencies" echo " " if [ "$EUID" -ne 0 ] then tput setaf 1; echo "Error installing reNgine, Please run this script as root!" tput setaf 1; echo "Example: sudo ./install.sh" exit fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing Docker..." echo "#########################################################################" if [ -x "$(command -v docker)" ]; then tput setaf 2; echo "Docker already installed, skipping." else curl -fsSL https://get.docker.com -o get-docker.sh && sh get-docker.sh tput setaf 2; echo "Docker installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing docker-compose" echo "#########################################################################" if [ -x "$(command -v docker-compose)" ]; then tput setaf 2; echo "docker-compose already installed, skipping." else curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose tput setaf 2; echo "docker-compose installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing make" echo "#########################################################################" if [ -x "$(command -v make)" ]; then tput setaf 2; echo "make already installed, skipping." else apt install make fi echo " " tput setaf 4; echo "#########################################################################" echo "Checking Docker status" echo "#########################################################################" if systemctl is-active docker >/dev/null 2>&1; then tput setaf 4; echo "Docker is running." else tput setaf 1; echo "Docker is not running. Please run docker and try again." echo "You can run docker service using sudo systemctl start docker" exit 1 fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing reNgine" echo "#########################################################################" make certs && make build && make up tput setaf 2; echo "reNgine is installed!!!" sleep 3 echo " " tput setaf 4; echo "#########################################################################" echo "Creating an account" echo "#########################################################################" make username tput setaf 2; echo "Thank you for installing reNgine, happy recon!!"

sbimochan

2bd2219659fcf0f0541fc4879bd69bfa79a500c7

e98433517e4a6198e6e2208fdf1b324f41be5bcb

Sure take your time, for mac and windows users will continue to go through detailed installation steps. we also need to install make, and as I said, this script is only intended for Ubuntu/debian, that's why I've done `apt install make` this would anyways fail on macos

yogeshojha

yogeshojha/rengine

Installation: Check docker running status before installing reNgine.

### Changes - checks docker run status. Before it used to execute all `make build` commands even if docker was not running. - Early error shown to the user - Terminal text color changed - made this sentence "Before running this script, please make sure Docker is running and you have made changes to .env file." color RED because it can throw errors - "Changing the postgres username & password from .env is highly recommended." to Green since it's a suggestion

2021-08-24 16:52:53+00:00

2021-08-27 02:54:16+00:00

install.sh

#!/bin/bash tput setaf 2; cat web/art/1.0.txt tput setaf 3; echo "Before running this script, please make sure you have made changes to .env file." tput setaf 1; echo "Changing the postgres username & password from .env is highly recommended." tput setaf 4; read -p "Are you sure, you made changes to .env file (y/n)? " answer case ${answer:0:1} in y|Y|yes|YES|Yes ) echo "Continiuing Installation!" ;; * ) nano .env ;; esac echo " " tput setaf 3; echo "#########################################################################" echo "Please note that, this installation script is only intended for Linux" echo "For Mac and Windows, refer to the official guide https://rengine.wiki" echo "#########################################################################" echo " " tput setaf 4; echo "Installing reNgine and it's dependencies" echo " " if [ "$EUID" -ne 0 ] then tput setaf 1; echo "Error installing reNgine, Please run this script as root!" tput setaf 1; echo "Example: sudo ./install.sh" exit fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing Docker..." echo "#########################################################################" if [ -x "$(command -v docker)" ]; then tput setaf 2; echo "Docker already installed, skipping." else curl -fsSL https://get.docker.com -o get-docker.sh && sh get-docker.sh tput setaf 2; echo "Docker installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing docker-compose" echo "#########################################################################" if [ -x "$(command -v docker-compose)" ]; then tput setaf 2; echo "docker-compose already installed, skipping." else curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose tput setaf 2; echo "docker-compose installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing make" echo "#########################################################################" if [ -x "$(command -v make)" ]; then tput setaf 2; echo "make already installed, skipping." else apt install make fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing reNgine" echo "#########################################################################" make certs && make build && make up tput setaf 2; echo "reNgine is installed!!!" sleep 3 echo " " tput setaf 4; echo "#########################################################################" echo "Creating an account" echo "#########################################################################" make username tput setaf 2; echo "Thank you for installing reNgine, happy recon!!"

#!/bin/bash tput setaf 2; cat web/art/1.0.txt tput setaf 1; echo "Before running this script, please make sure Docker is running and you have made changes to .env file." tput setaf 2; echo "Changing the postgres username & password from .env is highly recommended." tput setaf 4; read -p "Are you sure, you made changes to .env file (y/n)? " answer case ${answer:0:1} in y|Y|yes|YES|Yes ) echo "Continiuing Installation!" ;; * ) nano .env ;; esac echo " " tput setaf 3; echo "#########################################################################" echo "Please note that, this installation script is only intended for Linux" echo "For Mac and Windows, refer to the official guide https://rengine.wiki" echo "#########################################################################" echo " " tput setaf 4; echo "Installing reNgine and it's dependencies" echo " " if [ "$EUID" -ne 0 ] then tput setaf 1; echo "Error installing reNgine, Please run this script as root!" tput setaf 1; echo "Example: sudo ./install.sh" exit fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing Docker..." echo "#########################################################################" if [ -x "$(command -v docker)" ]; then tput setaf 2; echo "Docker already installed, skipping." else curl -fsSL https://get.docker.com -o get-docker.sh && sh get-docker.sh tput setaf 2; echo "Docker installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing docker-compose" echo "#########################################################################" if [ -x "$(command -v docker-compose)" ]; then tput setaf 2; echo "docker-compose already installed, skipping." else curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose tput setaf 2; echo "docker-compose installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing make" echo "#########################################################################" if [ -x "$(command -v make)" ]; then tput setaf 2; echo "make already installed, skipping." else apt install make fi echo " " tput setaf 4; echo "#########################################################################" echo "Checking Docker status" echo "#########################################################################" if systemctl is-active docker >/dev/null 2>&1; then tput setaf 4; echo "Docker is running." else tput setaf 1; echo "Docker is not running. Please run docker and try again." echo "You can run docker service using sudo systemctl start docker" exit 1 fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing reNgine" echo "#########################################################################" make certs && make build && make up tput setaf 2; echo "reNgine is installed!!!" sleep 3 echo " " tput setaf 4; echo "#########################################################################" echo "Creating an account" echo "#########################################################################" make username tput setaf 2; echo "Thank you for installing reNgine, happy recon!!"

sbimochan

2bd2219659fcf0f0541fc4879bd69bfa79a500c7

e98433517e4a6198e6e2208fdf1b324f41be5bcb

Yep, so what we can do is, I will accept and merge this PR, and maybe if you want to work on a install script that works across different OS, that would be awesome.

yogeshojha

yogeshojha/rengine

Installation: Check docker running status before installing reNgine.

### Changes - checks docker run status. Before it used to execute all `make build` commands even if docker was not running. - Early error shown to the user - Terminal text color changed - made this sentence "Before running this script, please make sure Docker is running and you have made changes to .env file." color RED because it can throw errors - "Changing the postgres username & password from .env is highly recommended." to Green since it's a suggestion

2021-08-24 16:52:53+00:00

2021-08-27 02:54:16+00:00

install.sh

#!/bin/bash tput setaf 2; cat web/art/1.0.txt tput setaf 3; echo "Before running this script, please make sure you have made changes to .env file." tput setaf 1; echo "Changing the postgres username & password from .env is highly recommended." tput setaf 4; read -p "Are you sure, you made changes to .env file (y/n)? " answer case ${answer:0:1} in y|Y|yes|YES|Yes ) echo "Continiuing Installation!" ;; * ) nano .env ;; esac echo " " tput setaf 3; echo "#########################################################################" echo "Please note that, this installation script is only intended for Linux" echo "For Mac and Windows, refer to the official guide https://rengine.wiki" echo "#########################################################################" echo " " tput setaf 4; echo "Installing reNgine and it's dependencies" echo " " if [ "$EUID" -ne 0 ] then tput setaf 1; echo "Error installing reNgine, Please run this script as root!" tput setaf 1; echo "Example: sudo ./install.sh" exit fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing Docker..." echo "#########################################################################" if [ -x "$(command -v docker)" ]; then tput setaf 2; echo "Docker already installed, skipping." else curl -fsSL https://get.docker.com -o get-docker.sh && sh get-docker.sh tput setaf 2; echo "Docker installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing docker-compose" echo "#########################################################################" if [ -x "$(command -v docker-compose)" ]; then tput setaf 2; echo "docker-compose already installed, skipping." else curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose tput setaf 2; echo "docker-compose installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing make" echo "#########################################################################" if [ -x "$(command -v make)" ]; then tput setaf 2; echo "make already installed, skipping." else apt install make fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing reNgine" echo "#########################################################################" make certs && make build && make up tput setaf 2; echo "reNgine is installed!!!" sleep 3 echo " " tput setaf 4; echo "#########################################################################" echo "Creating an account" echo "#########################################################################" make username tput setaf 2; echo "Thank you for installing reNgine, happy recon!!"

#!/bin/bash tput setaf 2; cat web/art/1.0.txt tput setaf 1; echo "Before running this script, please make sure Docker is running and you have made changes to .env file." tput setaf 2; echo "Changing the postgres username & password from .env is highly recommended." tput setaf 4; read -p "Are you sure, you made changes to .env file (y/n)? " answer case ${answer:0:1} in y|Y|yes|YES|Yes ) echo "Continiuing Installation!" ;; * ) nano .env ;; esac echo " " tput setaf 3; echo "#########################################################################" echo "Please note that, this installation script is only intended for Linux" echo "For Mac and Windows, refer to the official guide https://rengine.wiki" echo "#########################################################################" echo " " tput setaf 4; echo "Installing reNgine and it's dependencies" echo " " if [ "$EUID" -ne 0 ] then tput setaf 1; echo "Error installing reNgine, Please run this script as root!" tput setaf 1; echo "Example: sudo ./install.sh" exit fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing Docker..." echo "#########################################################################" if [ -x "$(command -v docker)" ]; then tput setaf 2; echo "Docker already installed, skipping." else curl -fsSL https://get.docker.com -o get-docker.sh && sh get-docker.sh tput setaf 2; echo "Docker installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing docker-compose" echo "#########################################################################" if [ -x "$(command -v docker-compose)" ]; then tput setaf 2; echo "docker-compose already installed, skipping." else curl -L "https://github.com/docker/compose/releases/download/1.29.2/docker-compose-$(uname -s)-$(uname -m)" -o /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose ln -s /usr/local/bin/docker-compose /usr/bin/docker-compose tput setaf 2; echo "docker-compose installed!!!" fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing make" echo "#########################################################################" if [ -x "$(command -v make)" ]; then tput setaf 2; echo "make already installed, skipping." else apt install make fi echo " " tput setaf 4; echo "#########################################################################" echo "Checking Docker status" echo "#########################################################################" if systemctl is-active docker >/dev/null 2>&1; then tput setaf 4; echo "Docker is running." else tput setaf 1; echo "Docker is not running. Please run docker and try again." echo "You can run docker service using sudo systemctl start docker" exit 1 fi echo " " tput setaf 4; echo "#########################################################################" echo "Installing reNgine" echo "#########################################################################" make certs && make build && make up tput setaf 2; echo "reNgine is installed!!!" sleep 3 echo " " tput setaf 4; echo "#########################################################################" echo "Creating an account" echo "#########################################################################" make username tput setaf 2; echo "Thank you for installing reNgine, happy recon!!"

sbimochan

2bd2219659fcf0f0541fc4879bd69bfa79a500c7

e98433517e4a6198e6e2208fdf1b324f41be5bcb

Sounds perfect.

sbimochan

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

this now hard codes md5 hash. In the current version you can choose the hash function

PaulWestenthanner

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

do you know how much benefit the change from int(hexdigest) to int.from_bytes alone brings? I saw you mentioned 40%-60% for three changes combined. I think this is less readable and could use a comment. Also the byteorder might depend on the machine (c.f. https://docs.python.org/3/library/stdtypes.html#int.from_bytes and https://stackoverflow.com/questions/50509017/how-is-int-from-bytes-calculated)

PaulWestenthanner

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

you don't need the auto sample attribute anymore

PaulWestenthanner

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

`n_process` sounds like an integer. Better call it `process_list` or something more telling. You only copied the name but probably it's time to change it now

PaulWestenthanner

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

this ignores the `max_samples` parameter and might lead to the process crashing in case of too much data / too few CPUs

PaulWestenthanner

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

please add the `process_creation_method` parameter in the documentation and explain the options

PaulWestenthanner

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

how is this the same as the old code? wasn't the old code just doing `shm_result[column_index] += 1`?

PaulWestenthanner

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

why is this 2x2?

PaulWestenthanner

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

splitting this way is not very elegant if the last chunk has to be treated separately. using numpy array split could be helpful: https://stackoverflow.com/a/75981560. Wouldn't it be easier if the `hash_chunk` function would hash a chunk and return an array. Then it wouldn't need the `shm_result` and `shm_offset` parameters (what does shm stand for btw?). Then you'd just concatenate all the chunks in the end?

PaulWestenthanner

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

do you need to re-assign this to `np_result` or will this be updated in place?

PaulWestenthanner

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

Removing it!

bkhant1

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

I will add a comment! The `digest` version is about 30% faster than the `hexdigest` version. On my machine: ```python > import hashlib > hasher = hashlib.md5() > hasher.update(b"abdcde1234") > %timeit int(hasher.hexdigest(), 16) 659 ns ± 29.5 ns per loop (mean ± std. dev. of 7 runs, 1000000 loops each) > %timeit int.from_bytes(hasher.digest(), byteorder='big') 518 ns ± 11.5 ns per loop (mean ± std. dev. of 7 runs, 1000000 loops each) ``` The byte order is machine dependant only if the data comes directly from the machine. In that case, the `digest` should not be machine dependant. I tried it on - my system (MacOS Monterey little endian) ```python > import sys > import hashlib > sys.byteorder 'little' > hasher = hashlib.md5() > hasher.update(b"abc") > int.from_bytes(hasher.digest(), endian='big') 191415658344158766168031473277922803570 > int(hasher.hexdigest(), 16) 191415658344158766168031473277922803570 ``` - a docker system (`s390x/ubuntu` image, big endian): ```python > import sys > import hashlib > sys.byteorder 'big' > hasher = hashlib.md5() > hasher.update(b"abc") > int.from_bytes(hasher.digest(), endian='big') 191415658344158766168031473277922803570 > int(hasher.hexdigest(), 16) 191415658344158766168031473277922803570 ``` The result is the same.

bkhant1

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

That is actually another advantage of using a shared memory instead of queues. ### In the previous implementation, there were 3 mains things that could take up memory: 1. the input dataframe, stored in `self.X` - it is copied accross all processes wether we fork or spawn because the process target is `self.require_data`. 2. the output dataframe, which is the concatenation of all `hashing_parts`. There is only one instance of it in the main process. It takes space of exactly the output dataframe when all processes are done computing their chunk of the data. 3. the `data_part` that's temporary to each subprocess, and that's created to be put on the queue after the part is calculated. In the worst case scenario of the queue implementation, the total memory usage is: `input_size * number_of_processes + output_size + (ouput_size/max_sample)*number_of_processes` ### In the new implementation, only two things, because we write directly to the output shared memory instead of having the data transiting through the queue: 1. the input dataframe, stored in `self.X` - it is copied accross all processes wether we fork or spawn because the process has in its args `np_df`. 2. the output dataframe, which is the shared memory and is allocated only accross all processes. The total memory usage in the shared memory implementation is: `imput_size * number_of_processes + output_size` ### We don't need max samples anymore The memory usage is always smaller with the shared memory implementation, and does not depend on `max_samples`

bkhant1

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

It was, but it was in the context of writting to a subset of the output dataframe. In that case we write to the "global" output dataframe, so we need to add the offset specific to the chunk we're processing. I was worried about introducing a regression so I added `test_simple_example` where the expected result is hardcoded to the what the old code produced. In that case at least it produces the same result as the old code.

bkhant1

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

`shm` stands for "shared_memory" - will update the variable name 👍 > Wouldn't it be easier if the hash_chunk function would hash a chunk and return an array. Then it wouldn't need the shm_result and shm_offset parameters (what does shm stand for btw?). Then you'd just concatenate all the chunks in the end? That's a very good point, I will try that

bkhant1

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

My bad thanks for catching that! I'm replacing it with `hash_ctor = getattr(hashlib, hash_key)` outside the loop and `hash_ctor()` in the loop which is even faster! ```python > ctor = getattr(hashlib, 'md5') > %timeit ctor() 190 ns ± 0.927 ns per loop (mean ± std. dev. of 7 runs, 10000000 loops each) > %timeit hashlib.md5() 231 ns ± 4.86 ns per loop (mean ± std. dev. of 7 runs, 1000000 loops each)

bkhant1

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

Ok that makes sense. Which python/hashlib version are you using btw? Running `hasher.update("abc")` results in a `TypeError` on my machine since it requires (as per documentation) bytes not strings. With `hasher.update("abc".encode("utf-8"))` I'm getting the same results as you

PaulWestenthanner

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

I copied the wrong line from my shell! I updated my comment above.

bkhant1

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

Im not sure 😅 I am just using it to get the type of the default int array name in [that line](https://github.com/scikit-learn-contrib/category_encoders/pull/428/files/206de2d8327489c4ff4d2c7c4f566c0fc06210c1#diff-5871b042f65ccab77377b2e9a92ea2c9651cc039b020835b6d77bfcb01ffe475R187) so it could be 1 by 1!

bkhant1

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

It gets updated in place!

bkhant1

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

nice catch! without calling the `system()` function this was always `False`, was it?

PaulWestenthanner

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

I"ve seen your implementation of the mulitprocess pool (here https://github.com/bkhant1/category_encoders/compare/all_optis...bkhant1:category_encoders:multiproc_pool?expand=1) and like it a lot. I think it is very clean and you should add it to the PR

PaulWestenthanner

scikit-learn-contrib/category_encoders

Optimise `HashingEncoder` for both large and small dataframes

I used the HashingEncoder recently and found weird that any call to `fit` or `transform`, even for a dataframe with only 10s of rows and a couple of columns took at least 2s... I also had quite a large amount of data to encode, and that took a long time. That got me started on improving the performance of HashingEncoder, and here's the result! There are quite a few changes in there, each individual change should be in it's own commit, and here's a summary of the performance gain on my machine (macOS Monteray, i7 2.3ghz). | | Baseline | Numpy arrays instead of apply | Shared memory instead of queue | Fork instead of spawn | Faster hashlib usage | | --- | --- | --- | --- | --- | --- | | n_rows=30 n_features=3 n_components=10 n_process=4 | 3.55 s ± 150 ms per loop (mean ± std. dev. of ... | 3.62 s ± 140 ms per loop (mean ± std. dev. of ... | 2.2 s ± 41.6 ms per loop (mean ± std. dev. of ... | 56.6 ms ± 2.91 ms per loop (mean ± std. dev. o... | 47.3 ms ± 516 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=10 n_process=1 | 1.24 s ± 52.6 ms per loop (mean ± std. dev. of... | 1.42 s ± 170 ms per loop (mean ± std. dev. of ... | 1.74 ms ± 32.2 µs per loop (mean ± std. dev. o... | 2.08 ms ± 91.7 µs per loop (mean ± std. dev. o... | 1.86 ms ± 173 µs per loop (mean ± std. dev. of... | | n_rows=30 n_features=3 n_components=100 n_process=1 | 1.22 s ± 51.5 ms per loop (mean ± std. dev. of... | 1.33 s ± 60.7 ms per loop (mean ± std. dev. of... | 1.73 ms ± 29.7 µs per loop (mean ± std. dev. o... | 2.01 ms ± 148 µs per loop (mean ± std. dev. of... | 2.01 ms ± 225 µs per loop (mean ± std. dev. of... | | n_rows=10000 n_features=10 n_components=10 n_process=4 | 5.45 s ± 85.8 ms per loop (mean ± std. dev. of... | 5.36 s ± 57.5 ms per loop (mean ± std. dev. of... | 2.23 s ± 39.6 ms per loop (mean ± std. dev. of... | 120 ms ± 3.02 ms per loop (mean ± std. dev. of... | 96.4 ms ± 2.33 ms per loop (mean ± std. dev. o... | | n_rows=10000 n_features=10 n_components=10 n_process=1 | 1.61 s ± 30.1 ms per loop (mean ± std. dev. of... | 1.45 s ± 27.2 ms per loop (mean ± std. dev. of... | 227 ms ± 6.03 ms per loop (mean ± std. dev. of... | 236 ms ± 3.06 ms per loop (mean ± std. dev. of... | 170 ms ± 1.35 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=4 | 5.99 s ± 215 ms per loop (mean ± std. dev. of ... | 5.71 s ± 148 ms per loop (mean ± std. dev. of ... | 4.8 s ± 25.4 ms per loop (mean ± std. dev. of ... | 836 ms ± 42.3 ms per loop (mean ± std. dev. of... | 622 ms ± 33.2 ms per loop (mean ± std. dev. of... | | n_rows=100000 n_features=10 n_components=10 n_process=1 | 5.38 s ± 53 ms per loop (mean ± std. dev. of 7... | 3.73 s ± 56.5 ms per loop (mean ± std. dev. of... | 2.25 s ± 57.4 ms per loop (mean ± std. dev. of... | 3.76 s ± 1.61 s per loop (mean ± std. dev. of ... | 1.68 s ± 19.9 ms per loop (mean ± std. dev. of... | | n_rows=1000000 n_features=50 n_components=10 n_process=4 | 50.8 s ± 1.17 s per loop (mean ± std. dev. of ... | 56.4 s ± 2.11 s per loop (mean ± std. dev. of ... | 37.1 s ± 576 ms per loop (mean ± std. dev. of ... | 36.9 s ± 2.19 s per loop (mean ± std. dev. of ... | 26.6 s ± 1.8 s per loop (mean ± std. dev. of 7... | | n_rows=1000000 n_features=50 n_components=10 n_process=1 | 2min 22s ± 2.05 s per loop (mean ± std. dev. o... | 2min 19s ± 3.08 s per loop (mean ± std. dev. o... | 1min 47s ± 1.15 s per loop (mean ± std. dev. o... | 2min 10s ± 18.4 s per loop (mean ± std. dev. o... | 1min 21s ± 1.67 s per loop (mean ± std. dev. o... | The notebook that produced that table can be found [here](https://gist.github.com/bkhant1/ae2b813817d53b19a81f6774234fcfe3) ## Proposed Changes The changes are listed by commit. ### [Add a simple non-regression HashEncoder test](https://github.com/scikit-learn-contrib/category_encoders/commit/0afe06586c71388b8fd4034d196de8a7df4ad56c) To make sure I am not breaking it. ### [In HashingEncoder process the df as a numpy array instead of using apply](https://github.com/scikit-learn-contrib/category_encoders/commit/de124410f29778487a2910c8dd7f15ed15785705) It has no direct impact on performance, however it allows accessing the memory layout of the dataframe directly. That allows using shared memory to communicate between processes instead of a data queue, which does improve performance. ### [In HashEncoder use shared memory instead of queue for multiproccessing](https://github.com/scikit-learn-contrib/category_encoders/commit/5235a6b85e787b3a384c0d43f314c0e3146d3daf) It is faster to write directly in memory that to have to data transit through a queue. The multiprocessing method is similar to what it was with queues: the dataframe is split into chunks, and each process applies the hashing trick to its chunk of the dataframe. Instead of writting the result to a queue, it writes it directly in a shared memory segment, that is also the underlying memory of a numpy array that is used to build the output dataframe. ### [Allow forking processes instead of spwaning them and make it default](https://github.com/scikit-learn-contrib/category_encoders/commit/12f8f242959314ed770750902c1e5ab8ca81263e) This makes the HashEncoder transform method a lot faster on small datasets. The spawn process creation method creates a new python interpreter from scratch, and re-import all required module. In a minimal case (pandas and category_encoders.hashing only are imported) this adds a ~2s overhead to any call to transform. Fork creates a copy of the current process, and that's it. It is unsafe to use with threads, locks, file descriptors, ... but in that case the only thing the forked process will do is process some data and write it to ITS OWN segment of a shared memory. It is a lot faster as pandas doesn't have to be re-imported (around 20ms?) It might take up more memory as more than the necessary variables (the largest one by far being the HashEncoder instance, which include the user dataframe) will be copied. Add the option to use spawn instead of fork to potentially save some memory. ### [Remove python 2 check code and faster use of hashlib](https://github.com/scikit-learn-contrib/category_encoders/commit/d2d535b4b8b2c54adcb9b13a6b06b5fc8c55286c) Python 2 is not supported on master, the check isn't useful. Create int indexes from hashlib bytes digest instead of hex digest as it's faster. Call the md5 hashlib constructor directly instead of new('md5'), which is also faster.

2023-10-08 15:09:46+00:00

2023-11-11 14:34:26+00:00

category_encoders/hashing.py

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import math import platform __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) self.auto_sample = max_sample <= 0 self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def require_data(self, data_lock, new_start, done_index, hashing_parts, process_index): is_finished = False while not is_finished: if data_lock.acquire(): if new_start.value: end_index = 0 new_start.value = False else: end_index = done_index.value if all([self.data_lines > 0, end_index < self.data_lines]): start_index = end_index if (self.data_lines - end_index) <= self.max_sample: end_index = self.data_lines else: end_index += self.max_sample done_index.value = end_index data_lock.release() data_part = self.X.iloc[start_index: end_index] # Always get df and check it after merge all data parts data_part = self.hashing_trick(X_in=data_part, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) part_index = int(math.ceil(end_index / self.max_sample)) hashing_parts.put({part_index: data_part}) is_finished = end_index >= self.data_lines if self.verbose == 5: print(f"Process - {process_index} done hashing data : {start_index} ~ {end_index}") else: data_lock.release() is_finished = True else: data_lock.release() def _transform(self, X): """ Call _transform_single_cpu() if you want to use single CPU with all samples """ self.X = X self.data_lines = len(self.X) data_lock = multiprocessing.Manager().Lock() new_start = multiprocessing.Manager().Value('d', True) done_index = multiprocessing.Manager().Value('d', int(0)) hashing_parts = multiprocessing.Manager().Queue() if self.auto_sample: self.max_sample = int(self.data_lines / self.max_process) if self.max_sample == 0: self.max_sample = 1 if self.max_process == 1: self.require_data(data_lock, new_start, done_index, hashing_parts, process_index=1) else: n_process = [] for thread_idx in range(self.max_process): process = multiprocessing.Process(target=self.require_data, args=(data_lock, new_start, done_index, hashing_parts, thread_idx + 1)) process.daemon = True n_process.append(process) for process in n_process: process.start() for process in n_process: process.join() data = self.X if self.max_sample == 0 or self.max_sample == self.data_lines: if hashing_parts: data = list(hashing_parts.get().values())[0] else: list_data = {} while not hashing_parts.empty(): list_data.update(hashing_parts.get()) sort_data = [] for part_index in sorted(list_data): sort_data.append(list_data[part_index]) if sort_data: data = pd.concat(sort_data) return data def _transform_single_cpu(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick(X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols) if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.to_numpy() @staticmethod def hashing_trick(X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns def hash_fn(x): tmp = [0 for _ in range(N)] for val in x.array: if val is not None: hasher = hashlib.new(hashing_method) if sys.version_info[0] == 2: hasher.update(str(val)) else: hasher.update(bytes(str(val), 'utf-8')) tmp[int(hasher.hexdigest(), 16) % N] += 1 return tmp new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] X_cat = X_cat.apply(hash_fn, axis=1, result_type='expand') X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

"""The hashing module contains all methods and classes related to the hashing trick.""" import sys import hashlib import category_encoders.utils as util import multiprocessing import pandas as pd import numpy as np import math import platform from concurrent.futures import ProcessPoolExecutor __author__ = 'willmcginnis', 'LiuShulun' class HashingEncoder(util.BaseEncoder, util.UnsupervisedTransformerMixin): """ A multivariate hashing implementation with configurable dimensionality/precision. The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design. It's important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding. Default value of 'max_process' is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikit-learn-contrib/categorical-encoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows Parameters ---------- verbose: int integer indicating verbosity of the output. 0 for none. cols: list a list of columns to encode, if None, all string columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array). hash_method: str which hashing method to use. Any method from hashlib works. max_process: int how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding. max_sample: int how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value. n_components: int how many bits to use to represent the feature. By default, we use 8 bits. For high-cardinality features, consider using up-to 32 bits. process_creation_method: string either "fork", "spawn" or "forkserver" (availability depends on your platform). See https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods for more details and tradeoffs. Defaults to "fork" on linux/macos as it is the fastest option and to "spawn" on windows as it is the only one available Example ------- >>> from category_encoders.hashing import HashingEncoder >>> import pandas as pd >>> from sklearn.datasets import fetch_openml >>> bunch = fetch_openml(name="house_prices", as_frame=True) >>> display_cols = ["Id", "MSSubClass", "MSZoning", "LotFrontage", "YearBuilt", "Heating", "CentralAir"] >>> X = pd.DataFrame(bunch.data, columns=bunch.feature_names)[display_cols] >>> y = bunch.target >>> he = HashingEncoder(cols=['CentralAir', 'Heating']).fit(X, y) >>> numeric_dataset = he.transform(X) >>> print(numeric_dataset.info()) <class 'pandas.core.frame.DataFrame'> RangeIndex: 1460 entries, 0 to 1459 Data columns (total 13 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 col_0 1460 non-null int64 1 col_1 1460 non-null int64 2 col_2 1460 non-null int64 3 col_3 1460 non-null int64 4 col_4 1460 non-null int64 5 col_5 1460 non-null int64 6 col_6 1460 non-null int64 7 col_7 1460 non-null int64 8 Id 1460 non-null float64 9 MSSubClass 1460 non-null float64 10 MSZoning 1460 non-null object 11 LotFrontage 1201 non-null float64 12 YearBuilt 1460 non-null float64 dtypes: float64(4), int64(8), object(1) memory usage: 148.4+ KB None References ---------- .. [1] Feature Hashing for Large Scale Multitask Learning, from https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [2] Don't be tricked by the Hashing Trick, from https://booking.ai/dont-be-tricked-by-the-hashing-trick-192a6aae3087 """ prefit_ordinal = False encoding_relation = util.EncodingRelation.ONE_TO_M def __init__(self, max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5', process_creation_method='fork'): super().__init__(verbose=verbose, cols=cols, drop_invariant=drop_invariant, return_df=return_df, handle_unknown="does not apply", handle_missing="does not apply") if max_process not in range(1, 128): if platform.system() == 'Windows': self.max_process = 1 else: self.max_process = int(math.ceil(multiprocessing.cpu_count() / 2)) if self.max_process < 1: self.max_process = 1 elif self.max_process > 128: self.max_process = 128 else: self.max_process = max_process self.max_sample = int(max_sample) if platform.system() == 'Windows': self.process_creation_method = "spawn" else: self.process_creation_method = process_creation_method self.data_lines = 0 self.X = None self.n_components = n_components self.hash_method = hash_method def _fit(self, X, y=None, **kwargs): pass def _transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] Returns ------- p : array, shape = [n_samples, n_numeric + N] Transformed values with encoding applied. """ if self._dim is None: raise ValueError('Must train encoder before it can be used to transform data.') # first check the type X = util.convert_input(X) # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') if not list(self.cols): return X X = self.hashing_trick( X, hashing_method=self.hash_method, N=self.n_components, cols=self.cols, ) return X @staticmethod def hash_chunk(args): hash_method, np_df, N = args # Calling getattr outside the loop saves some time in the loop hasher_constructor = getattr(hashlib, hash_method) # Same when the call to getattr is implicit int_from_bytes = int.from_bytes result = np.zeros((np_df.shape[0], N), dtype='int') for i, row in enumerate(np_df): for val in row: if val is not None: hasher = hasher_constructor() # Computes an integer index from the hasher digest. The endian is # "big" as the code use to read: # column_index = int(hasher.hexdigest(), 16) % N # which is implicitly considering the hexdigest to be big endian, # even if the system is little endian. # Building the index that way is about 30% faster than using the # hexdigest. hasher.update(bytes(str(val), 'utf-8')) column_index = int_from_bytes(hasher.digest(), byteorder='big') % N result[i, column_index] += 1 return result def hashing_trick_with_np_parallel(self, df, N: int): np_df = df.to_numpy() ctx = multiprocessing.get_context(self.process_creation_method) with ProcessPoolExecutor(max_workers=self.max_process, mp_context=ctx) as executor: result = np.concatenate(list( executor.map( self.hash_chunk, zip( [self.hash_method]*self.max_process, np.array_split(np_df, self.max_process), [N]*self.max_process ) ) )) return pd.DataFrame(result, index=df.index) def hashing_trick_with_np_no_parallel(self, df, N): np_df = df.to_numpy() result = HashingEncoder.hash_chunk((self.hash_method, np_df, N)) return pd.DataFrame(result, index=df.index) def hashing_trick(self, X_in, hashing_method='md5', N=2, cols=None, make_copy=False): """A basic hashing implementation with configurable dimensionality/precision Performs the hashing trick on a pandas dataframe, `X`, using the hashing method from hashlib identified by `hashing_method`. The number of output dimensions (`N`), and columns to hash (`cols`) are also configurable. Parameters ---------- X_in: pandas dataframe description text hashing_method: string, optional description text N: int, optional description text cols: list, optional description text make_copy: bool, optional description text Returns ------- out : dataframe A hashing encoded dataframe. References ---------- Cite the relevant literature, e.g. [1]_. You may also cite these references in the notes section above. .. [1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML. """ if hashing_method not in hashlib.algorithms_available: raise ValueError(f"Hashing Method: {hashing_method} not Available. " f"Please use one from: [{', '.join([str(x) for x in hashlib.algorithms_available])}]") if make_copy: X = X_in.copy(deep=True) else: X = X_in if cols is None: cols = X.columns new_cols = [f'col_{d}' for d in range(N)] X_cat = X.loc[:, cols] X_num = X.loc[:, [x for x in X.columns if x not in cols]] if self.max_process == 1: X_cat = self.hashing_trick_with_np_no_parallel(X_cat, N) else: X_cat = self.hashing_trick_with_np_parallel(X_cat, N) X_cat.columns = new_cols X = pd.concat([X_cat, X_num], axis=1) return X

bkhant1

26ef26106fcbadb281c162b76258955f66f2c741

5c94e27436a3cf837d7c84a71c566e8320ce512f

maybe it would make sense to chance it to 1 by 1 as it would be sort of a minimal example. Also probably add a comment that you only need it for the datatype so people won't wonder why it is there

PaulWestenthanner

scikit-learn-contrib/category_encoders

(WIP) Partial fix for getting feature names out

I think this is a partial fix for this opened issue: https://github.com/scikit-learn-contrib/category_encoders/issues/395 It remains to check the behaviour of other estimators that are not ONE_TO_ONE. Please, let me know if you like the work in progress and I will try to continue.

2023-02-23 13:33:41+00:00

2023-03-13 11:48:24+00:00

category_encoders/__init__.py

""" .. module:: category_encoders :synopsis: :platform: """ from category_encoders.backward_difference import BackwardDifferenceEncoder from category_encoders.binary import BinaryEncoder from category_encoders.gray import GrayEncoder from category_encoders.count import CountEncoder from category_encoders.hashing import HashingEncoder from category_encoders.helmert import HelmertEncoder from category_encoders.one_hot import OneHotEncoder from category_encoders.ordinal import OrdinalEncoder from category_encoders.sum_coding import SumEncoder from category_encoders.polynomial import PolynomialEncoder from category_encoders.basen import BaseNEncoder from category_encoders.leave_one_out import LeaveOneOutEncoder from category_encoders.target_encoder import TargetEncoder from category_encoders.woe import WOEEncoder from category_encoders.m_estimate import MEstimateEncoder from category_encoders.james_stein import JamesSteinEncoder from category_encoders.cat_boost import CatBoostEncoder from category_encoders.rankhot import RankHotEncoder from category_encoders.glmm import GLMMEncoder from category_encoders.quantile_encoder import QuantileEncoder, SummaryEncoder __version__ = '2.6.0' __author__ = "willmcginnis", "cmougan", "paulwestenthanner" __all__ = [ "BackwardDifferenceEncoder", "BinaryEncoder", "GrayEncoder", "CountEncoder", "HashingEncoder", "HelmertEncoder", "OneHotEncoder", "OrdinalEncoder", "SumEncoder", "PolynomialEncoder", "BaseNEncoder", "LeaveOneOutEncoder", "TargetEncoder", "WOEEncoder", "MEstimateEncoder", "JamesSteinEncoder", "CatBoostEncoder", "GLMMEncoder", "QuantileEncoder", "SummaryEncoder", 'RankHotEncoder', ]

""" .. module:: category_encoders :synopsis: :platform: """ from category_encoders.backward_difference import BackwardDifferenceEncoder from category_encoders.binary import BinaryEncoder from category_encoders.gray import GrayEncoder from category_encoders.count import CountEncoder from category_encoders.hashing import HashingEncoder from category_encoders.helmert import HelmertEncoder from category_encoders.one_hot import OneHotEncoder from category_encoders.ordinal import OrdinalEncoder from category_encoders.sum_coding import SumEncoder from category_encoders.polynomial import PolynomialEncoder from category_encoders.basen import BaseNEncoder from category_encoders.leave_one_out import LeaveOneOutEncoder from category_encoders.target_encoder import TargetEncoder from category_encoders.woe import WOEEncoder from category_encoders.m_estimate import MEstimateEncoder from category_encoders.james_stein import JamesSteinEncoder from category_encoders.cat_boost import CatBoostEncoder from category_encoders.rankhot import RankHotEncoder from category_encoders.glmm import GLMMEncoder from category_encoders.quantile_encoder import QuantileEncoder, SummaryEncoder import sklearn import warnings from textwrap import dedent __version__ = '2.6.0' __author__ = "willmcginnis", "cmougan", "paulwestenthanner" __all__ = [ "BackwardDifferenceEncoder", "BinaryEncoder", "GrayEncoder", "CountEncoder", "HashingEncoder", "HelmertEncoder", "OneHotEncoder", "OrdinalEncoder", "SumEncoder", "PolynomialEncoder", "BaseNEncoder", "LeaveOneOutEncoder", "TargetEncoder", "WOEEncoder", "MEstimateEncoder", "JamesSteinEncoder", "CatBoostEncoder", "GLMMEncoder", "QuantileEncoder", "SummaryEncoder", 'RankHotEncoder', ]

JaimeArboleda

5eb7a2d6359d680bdadd0534bdb983e712a47f9c

570827e6b48737d0c9aece8aca31edd6da02c1b2

I don't really like the input warning here. I know of some users who use this library in their project and they try to suppress the warning for their end-users. Also, at the moment it also does not work anyway. I'd be an favor of not issuing a warning here but have something on the `index.rst` (c.f. below comment)

PaulWestenthanner

scikit-learn-contrib/category_encoders

(WIP) Partial fix for getting feature names out

I think this is a partial fix for this opened issue: https://github.com/scikit-learn-contrib/category_encoders/issues/395 It remains to check the behaviour of other estimators that are not ONE_TO_ONE. Please, let me know if you like the work in progress and I will try to continue.

2023-02-23 13:33:41+00:00

2023-03-13 11:48:24+00:00

category_encoders/__init__.py

""" .. module:: category_encoders :synopsis: :platform: """ from category_encoders.backward_difference import BackwardDifferenceEncoder from category_encoders.binary import BinaryEncoder from category_encoders.gray import GrayEncoder from category_encoders.count import CountEncoder from category_encoders.hashing import HashingEncoder from category_encoders.helmert import HelmertEncoder from category_encoders.one_hot import OneHotEncoder from category_encoders.ordinal import OrdinalEncoder from category_encoders.sum_coding import SumEncoder from category_encoders.polynomial import PolynomialEncoder from category_encoders.basen import BaseNEncoder from category_encoders.leave_one_out import LeaveOneOutEncoder from category_encoders.target_encoder import TargetEncoder from category_encoders.woe import WOEEncoder from category_encoders.m_estimate import MEstimateEncoder from category_encoders.james_stein import JamesSteinEncoder from category_encoders.cat_boost import CatBoostEncoder from category_encoders.rankhot import RankHotEncoder from category_encoders.glmm import GLMMEncoder from category_encoders.quantile_encoder import QuantileEncoder, SummaryEncoder __version__ = '2.6.0' __author__ = "willmcginnis", "cmougan", "paulwestenthanner" __all__ = [ "BackwardDifferenceEncoder", "BinaryEncoder", "GrayEncoder", "CountEncoder", "HashingEncoder", "HelmertEncoder", "OneHotEncoder", "OrdinalEncoder", "SumEncoder", "PolynomialEncoder", "BaseNEncoder", "LeaveOneOutEncoder", "TargetEncoder", "WOEEncoder", "MEstimateEncoder", "JamesSteinEncoder", "CatBoostEncoder", "GLMMEncoder", "QuantileEncoder", "SummaryEncoder", 'RankHotEncoder', ]

""" .. module:: category_encoders :synopsis: :platform: """ from category_encoders.backward_difference import BackwardDifferenceEncoder from category_encoders.binary import BinaryEncoder from category_encoders.gray import GrayEncoder from category_encoders.count import CountEncoder from category_encoders.hashing import HashingEncoder from category_encoders.helmert import HelmertEncoder from category_encoders.one_hot import OneHotEncoder from category_encoders.ordinal import OrdinalEncoder from category_encoders.sum_coding import SumEncoder from category_encoders.polynomial import PolynomialEncoder from category_encoders.basen import BaseNEncoder from category_encoders.leave_one_out import LeaveOneOutEncoder from category_encoders.target_encoder import TargetEncoder from category_encoders.woe import WOEEncoder from category_encoders.m_estimate import MEstimateEncoder from category_encoders.james_stein import JamesSteinEncoder from category_encoders.cat_boost import CatBoostEncoder from category_encoders.rankhot import RankHotEncoder from category_encoders.glmm import GLMMEncoder from category_encoders.quantile_encoder import QuantileEncoder, SummaryEncoder import sklearn import warnings from textwrap import dedent __version__ = '2.6.0' __author__ = "willmcginnis", "cmougan", "paulwestenthanner" __all__ = [ "BackwardDifferenceEncoder", "BinaryEncoder", "GrayEncoder", "CountEncoder", "HashingEncoder", "HelmertEncoder", "OneHotEncoder", "OrdinalEncoder", "SumEncoder", "PolynomialEncoder", "BaseNEncoder", "LeaveOneOutEncoder", "TargetEncoder", "WOEEncoder", "MEstimateEncoder", "JamesSteinEncoder", "CatBoostEncoder", "GLMMEncoder", "QuantileEncoder", "SummaryEncoder", 'RankHotEncoder', ]

JaimeArboleda

5eb7a2d6359d680bdadd0534bdb983e712a47f9c

570827e6b48737d0c9aece8aca31edd6da02c1b2

I fully agree, in fact warnings are annoying. Let's remove this. Do you want me to add a new commit or you prefer to do it in the merge process?

JaimeArboleda

scikit-learn-contrib/category_encoders

(WIP) Partial fix for getting feature names out

I think this is a partial fix for this opened issue: https://github.com/scikit-learn-contrib/category_encoders/issues/395 It remains to check the behaviour of other estimators that are not ONE_TO_ONE. Please, let me know if you like the work in progress and I will try to continue.

2023-02-23 13:33:41+00:00

2023-03-13 11:48:24+00:00

category_encoders/utils.py

"""A collection of shared utilities for all encoders, not intended for external use.""" from abc import abstractmethod from enum import Enum, auto import warnings import pandas as pd import numpy as np import sklearn.base from sklearn.base import BaseEstimator, TransformerMixin from sklearn.exceptions import NotFittedError from typing import Dict, List, Optional, Union from scipy.sparse import csr_matrix __author__ = 'willmcginnis' def convert_cols_to_list(cols): if isinstance(cols, pd.Series): return cols.tolist() elif isinstance(cols, np.ndarray): return cols.tolist() elif np.isscalar(cols): return [cols] elif isinstance(cols, set): return list(cols) elif isinstance(cols, tuple): return list(cols) elif pd.api.types.is_categorical_dtype(cols): return cols.astype(object).tolist() return cols def get_obj_cols(df): """ Returns names of 'object' columns in the DataFrame. """ obj_cols = [] for idx, dt in enumerate(df.dtypes): if dt == 'object' or is_category(dt): obj_cols.append(df.columns.values[idx]) if not obj_cols: print("Warning: No categorical columns found. Calling 'transform' will only return input data.") return obj_cols def is_category(dtype): return pd.api.types.is_categorical_dtype(dtype) def convert_inputs(X, y, columns=None, index=None, deep=False): """ Unite arraylike `X` and vectorlike `y` into a DataFrame and Series. If both are pandas types already, raises an error if their indexes do not match. If one is pandas, the returns will share that index. If neither is pandas, a default index will be used, unless `index` is passed. Parameters ---------- X: arraylike y: listlike columns: listlike Specifies column names to use for `X`. Ignored if `X` is already a dataframe. If `None`, use the default pandas column names. index: listlike The index to use, if neither `X` nor `y` is a pandas type. (If one has an index, then this has no effect.) If `None`, use the default pandas index. deep: bool Whether to deep-copy `X`. """ X_alt_index = y.index if isinstance(y, pd.Series) else index X = convert_input(X, columns=columns, deep=deep, index=X_alt_index) if y is not None: y = convert_input_vector(y, index=X.index) # N.B.: If either was already pandas, it keeps its index. if any(X.index != y.index): msg = "`X` and `y` both have indexes, but they do not match. If you are shuffling your input data on " \ "purpose (e.g. via permutation_test_score) use np arrays instead of data frames / series" raise ValueError(msg) if X.shape[0] != y.shape[0]: raise ValueError("The length of X is " + str(X.shape[0]) + " but length of y is " + str(y.shape[0]) + ".") return X, y def convert_input(X, columns=None, deep=False, index=None): """ Unite data into a DataFrame. Objects that do not contain column names take the names from the argument. Optionally perform deep copy of the data. """ if not isinstance(X, pd.DataFrame): if isinstance(X, pd.Series): X = pd.DataFrame(X, copy=deep) else: if columns is not None and np.size(X,1) != len(columns): raise ValueError('The count of the column names does not correspond to the count of the columns') if isinstance(X, list): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) # lists are always copied, but for consistency, we still pass the argument elif isinstance(X, (np.generic, np.ndarray)): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) elif isinstance(X, csr_matrix): X = pd.DataFrame(X.todense(), columns=columns, copy=deep, index=index) else: raise ValueError(f'Unexpected input type: {type(X)}') elif deep: X = X.copy(deep=True) return X def convert_input_vector(y, index): """ Unite target data type into a Series. If the target is a Series or a DataFrame, we preserve its index. But if the target does not contain index attribute, we use the index from the argument. """ if y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') if isinstance(y, pd.Series): return y elif isinstance(y, np.ndarray): if len(np.shape(y))==1: # vector return pd.Series(y, name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[0]==1: # single row in a matrix return pd.Series(y[0, :], name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[1]==1: # single column in a matrix return pd.Series(y[:, 0], name='target', index=index) else: raise ValueError(f'Unexpected input shape: {np.shape(y)}') elif np.isscalar(y): return pd.Series([y], name='target', index=index) elif isinstance(y, list): if len(y)==0: # empty list return pd.Series(y, name='target', index=index, dtype=float) elif len(y)>0 and not isinstance(y[0], list): # vector return pd.Series(y, name='target', index=index) elif len(y)>0 and isinstance(y[0], list) and len(y[0])==1: # single row in a matrix flatten = lambda y: [item for sublist in y for item in sublist] return pd.Series(flatten(y), name='target', index=index) elif len(y)==1 and len(y[0])==0 and isinstance(y[0], list): # single empty column in a matrix return pd.Series(y[0], name='target', index=index, dtype=float) elif len(y)==1 and isinstance(y[0], list): # single column in a matrix return pd.Series(y[0], name='target', index=index, dtype=type(y[0][0])) else: raise ValueError('Unexpected input shape') elif isinstance(y, pd.DataFrame): if len(list(y))==0: # empty DataFrame return pd.Series(name='target', index=index, dtype=float) if len(list(y))==1: # a single column return y.iloc[:, 0] else: raise ValueError(f'Unexpected input shape: {y.shape}') else: return pd.Series(y, name='target', index=index) # this covers tuples and other directly convertible types def get_generated_cols(X_original, X_transformed, to_transform): """ Returns a list of the generated/transformed columns. Arguments: X_original: df the original (input) DataFrame. X_transformed: df the transformed (current) DataFrame. to_transform: [str] a list of columns that were transformed (as in the original DataFrame), commonly self.feature_names_in. Output: a list of columns that were transformed (as in the current DataFrame). """ original_cols = list(X_original.columns) if len(to_transform) > 0: [original_cols.remove(c) for c in to_transform] current_cols = list(X_transformed.columns) if len(original_cols) > 0: [current_cols.remove(c) for c in original_cols] return current_cols def flatten_reverse_dict(d): sep = "___" [flat_dict] = pd.json_normalize(d, sep=sep).to_dict(orient='records') reversed_flat_dict = {v: tuple(k.split(sep)) for k, v in flat_dict.items()} return reversed_flat_dict class EncodingRelation(Enum): # one input feature get encoded into one output feature ONE_TO_ONE = auto() # one input feature get encoded into as many output features as it has distinct values ONE_TO_N_UNIQUE = auto() # one input feature get encoded into m output features that are not the number of distinct values ONE_TO_M = auto() # all N input features are encoded into M output features. # The encoding is done globally on all the input not on a per-feature basis N_TO_M = auto() def get_docstring_output_shape(in_out_relation: EncodingRelation): if in_out_relation == EncodingRelation.ONE_TO_ONE: return "n_features" elif in_out_relation == EncodingRelation.ONE_TO_N_UNIQUE: return "n_features * respective cardinality" elif in_out_relation == EncodingRelation.ONE_TO_M: return "M features (n_features < M)" elif in_out_relation == EncodingRelation.N_TO_M: return "M features (M can be anything)" class BaseEncoder(BaseEstimator): _dim: Optional[int] cols: List[str] use_default_cols: bool handle_missing: str handle_unknown: str verbose: int drop_invariant: bool invariant_cols: List[str] = [] return_df: bool supervised: bool encoding_relation: EncodingRelation INVARIANCE_THRESHOLD = 10e-5 # columns with variance less than this will be considered constant / invariant def __init__(self, verbose=0, cols=None, drop_invariant=False, return_df=True, handle_unknown='value', handle_missing='value', **kwargs): """ Parameters ---------- verbose: int integer indicating verbosity of output. 0 for none. cols: list a list of columns to encode, if None, all string and categorical columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform and inverse transform (otherwise it will be a numpy array). handle_missing: str how to handle missing values at fit time. Options are 'error', 'return_nan', and 'value'. Default 'value', which treat NaNs as a countable category at fit time. handle_unknown: str, int or dict of {column : option, ...}. how to handle unknown labels at transform time. Options are 'error' 'return_nan', 'value' and int. Defaults to None which uses NaN behaviour specified at fit time. Passing an int will fill with this int value. kwargs: dict. additional encoder specific parameters like regularisation. """ self.return_df = return_df self.drop_invariant = drop_invariant self.invariant_cols = [] self.verbose = verbose self.use_default_cols = cols is None # if True, even a repeated call of fit() will select string columns from X self.cols = cols # note that cols are only the columns to be encoded, feature_names_in_ are all columns self.mapping = None self.handle_unknown = handle_unknown self.handle_missing = handle_missing self._dim = None def fit(self, X, y=None, **kwargs): """Fits the encoder according to X and y. Parameters ---------- X : array-like, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. y : array-like, shape = [n_samples] Target values. Returns ------- self : encoder Returns self. """ self._check_fit_inputs(X, y) X, y = convert_inputs(X, y) self.feature_names_in_ = X.columns.tolist() self.n_features_in_ = len(self.feature_names_in_) self._dim = X.shape[1] self._determine_fit_columns(X) if not set(self.cols).issubset(X.columns): raise ValueError('X does not contain the columns listed in cols') if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') self._fit(X, y, **kwargs) # for finding invariant columns transform without y (as is done on the test set) X_transformed = self.transform(X, override_return_df=True) self.feature_names_out_ = X_transformed.columns.tolist() # drop all output columns with 0 variance. if self.drop_invariant: generated_cols = get_generated_cols(X, X_transformed, self.cols) self.invariant_cols = [x for x in generated_cols if X_transformed[x].var() <= self.INVARIANCE_THRESHOLD] self.feature_names_out_ = [x for x in self.feature_names_out_ if x not in self.invariant_cols] return self def _check_fit_inputs(self, X, y): if self._get_tags().get('supervised_encoder') and y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') def _check_transform_inputs(self, X): if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') if self._dim is None: raise NotFittedError('Must train encoder before it can be used to transform data.') # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') def _drop_invariants(self, X: pd.DataFrame, override_return_df: bool) -> Union[np.ndarray, pd.DataFrame]: if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.values def _determine_fit_columns(self, X: pd.DataFrame) -> None: """ Determine columns used by encoder. Note that the implementation also deals with re-fitting the same encoder object with different columns. :param X: input data frame :return: none, sets self.cols as a side effect """ # if columns aren't passed, just use every string column if self.use_default_cols: self.cols = get_obj_cols(X) else: self.cols = convert_cols_to_list(self.cols) def get_feature_names(self) -> List[str]: warnings.warn("`get_feature_names` is deprecated in all of sklearn. Use `get_feature_names_out` instead.", category=FutureWarning) return self.get_feature_names_out() def get_feature_names_out(self) -> List[str]: """ Returns the names of all transformed / added columns. Note that in sklearn the get_feature_names_out function takes the feature_names_in as an argument and determines the output feature names using the input. A fit is usually not necessary and if so a NotFittedError is raised. We just require a fit all the time and return the fitted output columns. Returns ------- feature_names: list A list with all feature names transformed or added. Note: potentially dropped features (because the feature is constant/invariant) are not included! """ out_feats = getattr(self, "feature_names_out_", None) if not isinstance(out_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return out_feats def get_feature_names_in(self) -> List[str]: """ Returns the names of all input columns present when fitting. These columns are necessary for the transform step. """ in_feats = getattr(self, "feature_names_in_", None) if not isinstance(in_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return in_feats @abstractmethod def _fit(self, X: pd.DataFrame, y: Optional[pd.Series], **kwargs): ... class SupervisedTransformerMixin(sklearn.base.TransformerMixin): def _more_tags(self): return {'supervised_encoder': True} def transform(self, X, y=None, override_return_df=False): """Perform the transformation to new categorical data. Some encoders behave differently on whether y is given or not. This is mainly due to regularisation in order to avoid overfitting. On training data transform should be called with y, on test data without. Parameters ---------- X : array-like, shape = [n_samples, n_features] y : array-like, shape = [n_samples] or None override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X, y = convert_inputs(X, y, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X, y) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X: pd.DataFrame, y: pd.Series) -> pd.DataFrame: ... def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y) class UnsupervisedTransformerMixin(sklearn.base.TransformerMixin): def transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X = convert_input(X, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X) -> pd.DataFrame: ... class TransformerWithTargetMixin: def _more_tags(self): return {'supervised_encoder': True} def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y)

"""A collection of shared utilities for all encoders, not intended for external use.""" from abc import abstractmethod from enum import Enum, auto import warnings import pandas as pd import numpy as np import sklearn.base from sklearn.base import BaseEstimator, TransformerMixin from sklearn.exceptions import NotFittedError from typing import Dict, List, Optional, Union from scipy.sparse import csr_matrix __author__ = 'willmcginnis' def convert_cols_to_list(cols): if isinstance(cols, pd.Series): return cols.tolist() elif isinstance(cols, np.ndarray): return cols.tolist() elif np.isscalar(cols): return [cols] elif isinstance(cols, set): return list(cols) elif isinstance(cols, tuple): return list(cols) elif pd.api.types.is_categorical_dtype(cols): return cols.astype(object).tolist() return cols def get_obj_cols(df): """ Returns names of 'object' columns in the DataFrame. """ obj_cols = [] for idx, dt in enumerate(df.dtypes): if dt == 'object' or is_category(dt): obj_cols.append(df.columns.values[idx]) if not obj_cols: print("Warning: No categorical columns found. Calling 'transform' will only return input data.") return obj_cols def is_category(dtype): return pd.api.types.is_categorical_dtype(dtype) def convert_inputs(X, y, columns=None, index=None, deep=False): """ Unite arraylike `X` and vectorlike `y` into a DataFrame and Series. If both are pandas types already, raises an error if their indexes do not match. If one is pandas, the returns will share that index. If neither is pandas, a default index will be used, unless `index` is passed. Parameters ---------- X: arraylike y: listlike columns: listlike Specifies column names to use for `X`. Ignored if `X` is already a dataframe. If `None`, use the default pandas column names. index: listlike The index to use, if neither `X` nor `y` is a pandas type. (If one has an index, then this has no effect.) If `None`, use the default pandas index. deep: bool Whether to deep-copy `X`. """ X_alt_index = y.index if isinstance(y, pd.Series) else index X = convert_input(X, columns=columns, deep=deep, index=X_alt_index) if y is not None: y = convert_input_vector(y, index=X.index) # N.B.: If either was already pandas, it keeps its index. if any(X.index != y.index): msg = "`X` and `y` both have indexes, but they do not match. If you are shuffling your input data on " \ "purpose (e.g. via permutation_test_score) use np arrays instead of data frames / series" raise ValueError(msg) if X.shape[0] != y.shape[0]: raise ValueError("The length of X is " + str(X.shape[0]) + " but length of y is " + str(y.shape[0]) + ".") return X, y def convert_input(X, columns=None, deep=False, index=None): """ Unite data into a DataFrame. Objects that do not contain column names take the names from the argument. Optionally perform deep copy of the data. """ if not isinstance(X, pd.DataFrame): if isinstance(X, pd.Series): X = pd.DataFrame(X, copy=deep) else: if columns is not None and np.size(X,1) != len(columns): raise ValueError('The count of the column names does not correspond to the count of the columns') if isinstance(X, list): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) # lists are always copied, but for consistency, we still pass the argument elif isinstance(X, (np.generic, np.ndarray)): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) elif isinstance(X, csr_matrix): X = pd.DataFrame(X.todense(), columns=columns, copy=deep, index=index) else: raise ValueError(f'Unexpected input type: {type(X)}') elif deep: X = X.copy(deep=True) return X def convert_input_vector(y, index): """ Unite target data type into a Series. If the target is a Series or a DataFrame, we preserve its index. But if the target does not contain index attribute, we use the index from the argument. """ if y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') if isinstance(y, pd.Series): return y elif isinstance(y, np.ndarray): if len(np.shape(y))==1: # vector return pd.Series(y, name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[0]==1: # single row in a matrix return pd.Series(y[0, :], name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[1]==1: # single column in a matrix return pd.Series(y[:, 0], name='target', index=index) else: raise ValueError(f'Unexpected input shape: {np.shape(y)}') elif np.isscalar(y): return pd.Series([y], name='target', index=index) elif isinstance(y, list): if len(y)==0: # empty list return pd.Series(y, name='target', index=index, dtype=float) elif len(y)>0 and not isinstance(y[0], list): # vector return pd.Series(y, name='target', index=index) elif len(y)>0 and isinstance(y[0], list) and len(y[0])==1: # single row in a matrix flatten = lambda y: [item for sublist in y for item in sublist] return pd.Series(flatten(y), name='target', index=index) elif len(y)==1 and len(y[0])==0 and isinstance(y[0], list): # single empty column in a matrix return pd.Series(y[0], name='target', index=index, dtype=float) elif len(y)==1 and isinstance(y[0], list): # single column in a matrix return pd.Series(y[0], name='target', index=index, dtype=type(y[0][0])) else: raise ValueError('Unexpected input shape') elif isinstance(y, pd.DataFrame): if len(list(y))==0: # empty DataFrame return pd.Series(name='target', index=index, dtype=float) if len(list(y))==1: # a single column return y.iloc[:, 0] else: raise ValueError(f'Unexpected input shape: {y.shape}') else: return pd.Series(y, name='target', index=index) # this covers tuples and other directly convertible types def get_generated_cols(X_original, X_transformed, to_transform): """ Returns a list of the generated/transformed columns. Arguments: X_original: df the original (input) DataFrame. X_transformed: df the transformed (current) DataFrame. to_transform: [str] a list of columns that were transformed (as in the original DataFrame), commonly self.feature_names_in. Output: a list of columns that were transformed (as in the current DataFrame). """ original_cols = list(X_original.columns) if len(to_transform) > 0: [original_cols.remove(c) for c in to_transform] current_cols = list(X_transformed.columns) if len(original_cols) > 0: [current_cols.remove(c) for c in original_cols] return current_cols def flatten_reverse_dict(d): sep = "___" [flat_dict] = pd.json_normalize(d, sep=sep).to_dict(orient='records') reversed_flat_dict = {v: tuple(k.split(sep)) for k, v in flat_dict.items()} return reversed_flat_dict class EncodingRelation(Enum): # one input feature get encoded into one output feature ONE_TO_ONE = auto() # one input feature get encoded into as many output features as it has distinct values ONE_TO_N_UNIQUE = auto() # one input feature get encoded into m output features that are not the number of distinct values ONE_TO_M = auto() # all N input features are encoded into M output features. # The encoding is done globally on all the input not on a per-feature basis N_TO_M = auto() def get_docstring_output_shape(in_out_relation: EncodingRelation): if in_out_relation == EncodingRelation.ONE_TO_ONE: return "n_features" elif in_out_relation == EncodingRelation.ONE_TO_N_UNIQUE: return "n_features * respective cardinality" elif in_out_relation == EncodingRelation.ONE_TO_M: return "M features (n_features < M)" elif in_out_relation == EncodingRelation.N_TO_M: return "M features (M can be anything)" class BaseEncoder(BaseEstimator): _dim: Optional[int] cols: List[str] use_default_cols: bool handle_missing: str handle_unknown: str verbose: int drop_invariant: bool invariant_cols: List[str] = [] return_df: bool supervised: bool encoding_relation: EncodingRelation INVARIANCE_THRESHOLD = 10e-5 # columns with variance less than this will be considered constant / invariant def __init__(self, verbose=0, cols=None, drop_invariant=False, return_df=True, handle_unknown='value', handle_missing='value', **kwargs): """ Parameters ---------- verbose: int integer indicating verbosity of output. 0 for none. cols: list a list of columns to encode, if None, all string and categorical columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform and inverse transform (otherwise it will be a numpy array). handle_missing: str how to handle missing values at fit time. Options are 'error', 'return_nan', and 'value'. Default 'value', which treat NaNs as a countable category at fit time. handle_unknown: str, int or dict of {column : option, ...}. how to handle unknown labels at transform time. Options are 'error' 'return_nan', 'value' and int. Defaults to None which uses NaN behaviour specified at fit time. Passing an int will fill with this int value. kwargs: dict. additional encoder specific parameters like regularisation. """ self.return_df = return_df self.drop_invariant = drop_invariant self.invariant_cols = [] self.verbose = verbose self.use_default_cols = cols is None # if True, even a repeated call of fit() will select string columns from X self.cols = cols # note that cols are only the columns to be encoded, feature_names_in_ are all columns self.mapping = None self.handle_unknown = handle_unknown self.handle_missing = handle_missing self._dim = None def fit(self, X, y=None, **kwargs): """Fits the encoder according to X and y. Parameters ---------- X : array-like, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. y : array-like, shape = [n_samples] Target values. Returns ------- self : encoder Returns self. """ self._check_fit_inputs(X, y) X, y = convert_inputs(X, y) self.feature_names_in_ = X.columns.tolist() self.n_features_in_ = len(self.feature_names_in_) self._dim = X.shape[1] self._determine_fit_columns(X) if not set(self.cols).issubset(X.columns): raise ValueError('X does not contain the columns listed in cols') if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') self._fit(X, y, **kwargs) # for finding invariant columns transform without y (as is done on the test set) X_transformed = self.transform(X, override_return_df=True) self.feature_names_out_ = X_transformed.columns.tolist() # drop all output columns with 0 variance. if self.drop_invariant: generated_cols = get_generated_cols(X, X_transformed, self.cols) self.invariant_cols = [x for x in generated_cols if X_transformed[x].var() <= self.INVARIANCE_THRESHOLD] self.feature_names_out_ = [x for x in self.feature_names_out_ if x not in self.invariant_cols] return self def _check_fit_inputs(self, X, y): if self._get_tags().get('supervised_encoder') and y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') def _check_transform_inputs(self, X): if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') if self._dim is None: raise NotFittedError('Must train encoder before it can be used to transform data.') # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') def _drop_invariants(self, X: pd.DataFrame, override_return_df: bool) -> Union[np.ndarray, pd.DataFrame]: if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.values def _determine_fit_columns(self, X: pd.DataFrame) -> None: """ Determine columns used by encoder. Note that the implementation also deals with re-fitting the same encoder object with different columns. :param X: input data frame :return: none, sets self.cols as a side effect """ # if columns aren't passed, just use every string column if self.use_default_cols: self.cols = get_obj_cols(X) else: self.cols = convert_cols_to_list(self.cols) def get_feature_names(self) -> List[str]: warnings.warn("`get_feature_names` is deprecated in all of sklearn. Use `get_feature_names_out` instead.", category=FutureWarning) return self.get_feature_names_out() def get_feature_names_out(self, input_features=None) -> np.ndarray: """ Returns the names of all transformed / added columns. Note that in sklearn the get_feature_names_out function takes the feature_names_in as an argument and determines the output feature names using the input. A fit is usually not necessary and if so a NotFittedError is raised. We just require a fit all the time and return the fitted output columns. Returns ------- feature_names: np.ndarray A numpy array with all feature names transformed or added. Note: potentially dropped features (because the feature is constant/invariant) are not included! """ out_feats = getattr(self, "feature_names_out_", None) if not isinstance(out_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return np.array(out_feats, dtype=object) def get_feature_names_in(self) -> List[str]: """ Returns the names of all input columns present when fitting. These columns are necessary for the transform step. """ in_feats = getattr(self, "feature_names_in_", None) if not isinstance(in_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return in_feats @abstractmethod def _fit(self, X: pd.DataFrame, y: Optional[pd.Series], **kwargs): ... class SupervisedTransformerMixin(sklearn.base.TransformerMixin): def _more_tags(self): return {'supervised_encoder': True} def transform(self, X, y=None, override_return_df=False): """Perform the transformation to new categorical data. Some encoders behave differently on whether y is given or not. This is mainly due to regularisation in order to avoid overfitting. On training data transform should be called with y, on test data without. Parameters ---------- X : array-like, shape = [n_samples, n_features] y : array-like, shape = [n_samples] or None override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X, y = convert_inputs(X, y, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X, y) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X: pd.DataFrame, y: pd.Series) -> pd.DataFrame: ... def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y) class UnsupervisedTransformerMixin(sklearn.base.TransformerMixin): def transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X = convert_input(X, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X) -> pd.DataFrame: ... class TransformerWithTargetMixin: def _more_tags(self): return {'supervised_encoder': True} def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y)

JaimeArboleda

5eb7a2d6359d680bdadd0534bdb983e712a47f9c

570827e6b48737d0c9aece8aca31edd6da02c1b2

so we're not actually doing anything with the `input_features` parameter? I thought the point of including it would be to explicitly tell every encoder how to calculate output features from the given input features (this is what the 1-1 mixin does, right?) For some encoder it give a not-fitted error, but other don't need fitting. I feel this is taking some shortcut that sklearn does not take

PaulWestenthanner

scikit-learn-contrib/category_encoders

(WIP) Partial fix for getting feature names out

I think this is a partial fix for this opened issue: https://github.com/scikit-learn-contrib/category_encoders/issues/395 It remains to check the behaviour of other estimators that are not ONE_TO_ONE. Please, let me know if you like the work in progress and I will try to continue.

2023-02-23 13:33:41+00:00

2023-03-13 11:48:24+00:00

category_encoders/utils.py

"""A collection of shared utilities for all encoders, not intended for external use.""" from abc import abstractmethod from enum import Enum, auto import warnings import pandas as pd import numpy as np import sklearn.base from sklearn.base import BaseEstimator, TransformerMixin from sklearn.exceptions import NotFittedError from typing import Dict, List, Optional, Union from scipy.sparse import csr_matrix __author__ = 'willmcginnis' def convert_cols_to_list(cols): if isinstance(cols, pd.Series): return cols.tolist() elif isinstance(cols, np.ndarray): return cols.tolist() elif np.isscalar(cols): return [cols] elif isinstance(cols, set): return list(cols) elif isinstance(cols, tuple): return list(cols) elif pd.api.types.is_categorical_dtype(cols): return cols.astype(object).tolist() return cols def get_obj_cols(df): """ Returns names of 'object' columns in the DataFrame. """ obj_cols = [] for idx, dt in enumerate(df.dtypes): if dt == 'object' or is_category(dt): obj_cols.append(df.columns.values[idx]) if not obj_cols: print("Warning: No categorical columns found. Calling 'transform' will only return input data.") return obj_cols def is_category(dtype): return pd.api.types.is_categorical_dtype(dtype) def convert_inputs(X, y, columns=None, index=None, deep=False): """ Unite arraylike `X` and vectorlike `y` into a DataFrame and Series. If both are pandas types already, raises an error if their indexes do not match. If one is pandas, the returns will share that index. If neither is pandas, a default index will be used, unless `index` is passed. Parameters ---------- X: arraylike y: listlike columns: listlike Specifies column names to use for `X`. Ignored if `X` is already a dataframe. If `None`, use the default pandas column names. index: listlike The index to use, if neither `X` nor `y` is a pandas type. (If one has an index, then this has no effect.) If `None`, use the default pandas index. deep: bool Whether to deep-copy `X`. """ X_alt_index = y.index if isinstance(y, pd.Series) else index X = convert_input(X, columns=columns, deep=deep, index=X_alt_index) if y is not None: y = convert_input_vector(y, index=X.index) # N.B.: If either was already pandas, it keeps its index. if any(X.index != y.index): msg = "`X` and `y` both have indexes, but they do not match. If you are shuffling your input data on " \ "purpose (e.g. via permutation_test_score) use np arrays instead of data frames / series" raise ValueError(msg) if X.shape[0] != y.shape[0]: raise ValueError("The length of X is " + str(X.shape[0]) + " but length of y is " + str(y.shape[0]) + ".") return X, y def convert_input(X, columns=None, deep=False, index=None): """ Unite data into a DataFrame. Objects that do not contain column names take the names from the argument. Optionally perform deep copy of the data. """ if not isinstance(X, pd.DataFrame): if isinstance(X, pd.Series): X = pd.DataFrame(X, copy=deep) else: if columns is not None and np.size(X,1) != len(columns): raise ValueError('The count of the column names does not correspond to the count of the columns') if isinstance(X, list): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) # lists are always copied, but for consistency, we still pass the argument elif isinstance(X, (np.generic, np.ndarray)): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) elif isinstance(X, csr_matrix): X = pd.DataFrame(X.todense(), columns=columns, copy=deep, index=index) else: raise ValueError(f'Unexpected input type: {type(X)}') elif deep: X = X.copy(deep=True) return X def convert_input_vector(y, index): """ Unite target data type into a Series. If the target is a Series or a DataFrame, we preserve its index. But if the target does not contain index attribute, we use the index from the argument. """ if y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') if isinstance(y, pd.Series): return y elif isinstance(y, np.ndarray): if len(np.shape(y))==1: # vector return pd.Series(y, name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[0]==1: # single row in a matrix return pd.Series(y[0, :], name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[1]==1: # single column in a matrix return pd.Series(y[:, 0], name='target', index=index) else: raise ValueError(f'Unexpected input shape: {np.shape(y)}') elif np.isscalar(y): return pd.Series([y], name='target', index=index) elif isinstance(y, list): if len(y)==0: # empty list return pd.Series(y, name='target', index=index, dtype=float) elif len(y)>0 and not isinstance(y[0], list): # vector return pd.Series(y, name='target', index=index) elif len(y)>0 and isinstance(y[0], list) and len(y[0])==1: # single row in a matrix flatten = lambda y: [item for sublist in y for item in sublist] return pd.Series(flatten(y), name='target', index=index) elif len(y)==1 and len(y[0])==0 and isinstance(y[0], list): # single empty column in a matrix return pd.Series(y[0], name='target', index=index, dtype=float) elif len(y)==1 and isinstance(y[0], list): # single column in a matrix return pd.Series(y[0], name='target', index=index, dtype=type(y[0][0])) else: raise ValueError('Unexpected input shape') elif isinstance(y, pd.DataFrame): if len(list(y))==0: # empty DataFrame return pd.Series(name='target', index=index, dtype=float) if len(list(y))==1: # a single column return y.iloc[:, 0] else: raise ValueError(f'Unexpected input shape: {y.shape}') else: return pd.Series(y, name='target', index=index) # this covers tuples and other directly convertible types def get_generated_cols(X_original, X_transformed, to_transform): """ Returns a list of the generated/transformed columns. Arguments: X_original: df the original (input) DataFrame. X_transformed: df the transformed (current) DataFrame. to_transform: [str] a list of columns that were transformed (as in the original DataFrame), commonly self.feature_names_in. Output: a list of columns that were transformed (as in the current DataFrame). """ original_cols = list(X_original.columns) if len(to_transform) > 0: [original_cols.remove(c) for c in to_transform] current_cols = list(X_transformed.columns) if len(original_cols) > 0: [current_cols.remove(c) for c in original_cols] return current_cols def flatten_reverse_dict(d): sep = "___" [flat_dict] = pd.json_normalize(d, sep=sep).to_dict(orient='records') reversed_flat_dict = {v: tuple(k.split(sep)) for k, v in flat_dict.items()} return reversed_flat_dict class EncodingRelation(Enum): # one input feature get encoded into one output feature ONE_TO_ONE = auto() # one input feature get encoded into as many output features as it has distinct values ONE_TO_N_UNIQUE = auto() # one input feature get encoded into m output features that are not the number of distinct values ONE_TO_M = auto() # all N input features are encoded into M output features. # The encoding is done globally on all the input not on a per-feature basis N_TO_M = auto() def get_docstring_output_shape(in_out_relation: EncodingRelation): if in_out_relation == EncodingRelation.ONE_TO_ONE: return "n_features" elif in_out_relation == EncodingRelation.ONE_TO_N_UNIQUE: return "n_features * respective cardinality" elif in_out_relation == EncodingRelation.ONE_TO_M: return "M features (n_features < M)" elif in_out_relation == EncodingRelation.N_TO_M: return "M features (M can be anything)" class BaseEncoder(BaseEstimator): _dim: Optional[int] cols: List[str] use_default_cols: bool handle_missing: str handle_unknown: str verbose: int drop_invariant: bool invariant_cols: List[str] = [] return_df: bool supervised: bool encoding_relation: EncodingRelation INVARIANCE_THRESHOLD = 10e-5 # columns with variance less than this will be considered constant / invariant def __init__(self, verbose=0, cols=None, drop_invariant=False, return_df=True, handle_unknown='value', handle_missing='value', **kwargs): """ Parameters ---------- verbose: int integer indicating verbosity of output. 0 for none. cols: list a list of columns to encode, if None, all string and categorical columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform and inverse transform (otherwise it will be a numpy array). handle_missing: str how to handle missing values at fit time. Options are 'error', 'return_nan', and 'value'. Default 'value', which treat NaNs as a countable category at fit time. handle_unknown: str, int or dict of {column : option, ...}. how to handle unknown labels at transform time. Options are 'error' 'return_nan', 'value' and int. Defaults to None which uses NaN behaviour specified at fit time. Passing an int will fill with this int value. kwargs: dict. additional encoder specific parameters like regularisation. """ self.return_df = return_df self.drop_invariant = drop_invariant self.invariant_cols = [] self.verbose = verbose self.use_default_cols = cols is None # if True, even a repeated call of fit() will select string columns from X self.cols = cols # note that cols are only the columns to be encoded, feature_names_in_ are all columns self.mapping = None self.handle_unknown = handle_unknown self.handle_missing = handle_missing self._dim = None def fit(self, X, y=None, **kwargs): """Fits the encoder according to X and y. Parameters ---------- X : array-like, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. y : array-like, shape = [n_samples] Target values. Returns ------- self : encoder Returns self. """ self._check_fit_inputs(X, y) X, y = convert_inputs(X, y) self.feature_names_in_ = X.columns.tolist() self.n_features_in_ = len(self.feature_names_in_) self._dim = X.shape[1] self._determine_fit_columns(X) if not set(self.cols).issubset(X.columns): raise ValueError('X does not contain the columns listed in cols') if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') self._fit(X, y, **kwargs) # for finding invariant columns transform without y (as is done on the test set) X_transformed = self.transform(X, override_return_df=True) self.feature_names_out_ = X_transformed.columns.tolist() # drop all output columns with 0 variance. if self.drop_invariant: generated_cols = get_generated_cols(X, X_transformed, self.cols) self.invariant_cols = [x for x in generated_cols if X_transformed[x].var() <= self.INVARIANCE_THRESHOLD] self.feature_names_out_ = [x for x in self.feature_names_out_ if x not in self.invariant_cols] return self def _check_fit_inputs(self, X, y): if self._get_tags().get('supervised_encoder') and y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') def _check_transform_inputs(self, X): if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') if self._dim is None: raise NotFittedError('Must train encoder before it can be used to transform data.') # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') def _drop_invariants(self, X: pd.DataFrame, override_return_df: bool) -> Union[np.ndarray, pd.DataFrame]: if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.values def _determine_fit_columns(self, X: pd.DataFrame) -> None: """ Determine columns used by encoder. Note that the implementation also deals with re-fitting the same encoder object with different columns. :param X: input data frame :return: none, sets self.cols as a side effect """ # if columns aren't passed, just use every string column if self.use_default_cols: self.cols = get_obj_cols(X) else: self.cols = convert_cols_to_list(self.cols) def get_feature_names(self) -> List[str]: warnings.warn("`get_feature_names` is deprecated in all of sklearn. Use `get_feature_names_out` instead.", category=FutureWarning) return self.get_feature_names_out() def get_feature_names_out(self) -> List[str]: """ Returns the names of all transformed / added columns. Note that in sklearn the get_feature_names_out function takes the feature_names_in as an argument and determines the output feature names using the input. A fit is usually not necessary and if so a NotFittedError is raised. We just require a fit all the time and return the fitted output columns. Returns ------- feature_names: list A list with all feature names transformed or added. Note: potentially dropped features (because the feature is constant/invariant) are not included! """ out_feats = getattr(self, "feature_names_out_", None) if not isinstance(out_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return out_feats def get_feature_names_in(self) -> List[str]: """ Returns the names of all input columns present when fitting. These columns are necessary for the transform step. """ in_feats = getattr(self, "feature_names_in_", None) if not isinstance(in_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return in_feats @abstractmethod def _fit(self, X: pd.DataFrame, y: Optional[pd.Series], **kwargs): ... class SupervisedTransformerMixin(sklearn.base.TransformerMixin): def _more_tags(self): return {'supervised_encoder': True} def transform(self, X, y=None, override_return_df=False): """Perform the transformation to new categorical data. Some encoders behave differently on whether y is given or not. This is mainly due to regularisation in order to avoid overfitting. On training data transform should be called with y, on test data without. Parameters ---------- X : array-like, shape = [n_samples, n_features] y : array-like, shape = [n_samples] or None override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X, y = convert_inputs(X, y, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X, y) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X: pd.DataFrame, y: pd.Series) -> pd.DataFrame: ... def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y) class UnsupervisedTransformerMixin(sklearn.base.TransformerMixin): def transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X = convert_input(X, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X) -> pd.DataFrame: ... class TransformerWithTargetMixin: def _more_tags(self): return {'supervised_encoder': True} def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y)

"""A collection of shared utilities for all encoders, not intended for external use.""" from abc import abstractmethod from enum import Enum, auto import warnings import pandas as pd import numpy as np import sklearn.base from sklearn.base import BaseEstimator, TransformerMixin from sklearn.exceptions import NotFittedError from typing import Dict, List, Optional, Union from scipy.sparse import csr_matrix __author__ = 'willmcginnis' def convert_cols_to_list(cols): if isinstance(cols, pd.Series): return cols.tolist() elif isinstance(cols, np.ndarray): return cols.tolist() elif np.isscalar(cols): return [cols] elif isinstance(cols, set): return list(cols) elif isinstance(cols, tuple): return list(cols) elif pd.api.types.is_categorical_dtype(cols): return cols.astype(object).tolist() return cols def get_obj_cols(df): """ Returns names of 'object' columns in the DataFrame. """ obj_cols = [] for idx, dt in enumerate(df.dtypes): if dt == 'object' or is_category(dt): obj_cols.append(df.columns.values[idx]) if not obj_cols: print("Warning: No categorical columns found. Calling 'transform' will only return input data.") return obj_cols def is_category(dtype): return pd.api.types.is_categorical_dtype(dtype) def convert_inputs(X, y, columns=None, index=None, deep=False): """ Unite arraylike `X` and vectorlike `y` into a DataFrame and Series. If both are pandas types already, raises an error if their indexes do not match. If one is pandas, the returns will share that index. If neither is pandas, a default index will be used, unless `index` is passed. Parameters ---------- X: arraylike y: listlike columns: listlike Specifies column names to use for `X`. Ignored if `X` is already a dataframe. If `None`, use the default pandas column names. index: listlike The index to use, if neither `X` nor `y` is a pandas type. (If one has an index, then this has no effect.) If `None`, use the default pandas index. deep: bool Whether to deep-copy `X`. """ X_alt_index = y.index if isinstance(y, pd.Series) else index X = convert_input(X, columns=columns, deep=deep, index=X_alt_index) if y is not None: y = convert_input_vector(y, index=X.index) # N.B.: If either was already pandas, it keeps its index. if any(X.index != y.index): msg = "`X` and `y` both have indexes, but they do not match. If you are shuffling your input data on " \ "purpose (e.g. via permutation_test_score) use np arrays instead of data frames / series" raise ValueError(msg) if X.shape[0] != y.shape[0]: raise ValueError("The length of X is " + str(X.shape[0]) + " but length of y is " + str(y.shape[0]) + ".") return X, y def convert_input(X, columns=None, deep=False, index=None): """ Unite data into a DataFrame. Objects that do not contain column names take the names from the argument. Optionally perform deep copy of the data. """ if not isinstance(X, pd.DataFrame): if isinstance(X, pd.Series): X = pd.DataFrame(X, copy=deep) else: if columns is not None and np.size(X,1) != len(columns): raise ValueError('The count of the column names does not correspond to the count of the columns') if isinstance(X, list): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) # lists are always copied, but for consistency, we still pass the argument elif isinstance(X, (np.generic, np.ndarray)): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) elif isinstance(X, csr_matrix): X = pd.DataFrame(X.todense(), columns=columns, copy=deep, index=index) else: raise ValueError(f'Unexpected input type: {type(X)}') elif deep: X = X.copy(deep=True) return X def convert_input_vector(y, index): """ Unite target data type into a Series. If the target is a Series or a DataFrame, we preserve its index. But if the target does not contain index attribute, we use the index from the argument. """ if y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') if isinstance(y, pd.Series): return y elif isinstance(y, np.ndarray): if len(np.shape(y))==1: # vector return pd.Series(y, name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[0]==1: # single row in a matrix return pd.Series(y[0, :], name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[1]==1: # single column in a matrix return pd.Series(y[:, 0], name='target', index=index) else: raise ValueError(f'Unexpected input shape: {np.shape(y)}') elif np.isscalar(y): return pd.Series([y], name='target', index=index) elif isinstance(y, list): if len(y)==0: # empty list return pd.Series(y, name='target', index=index, dtype=float) elif len(y)>0 and not isinstance(y[0], list): # vector return pd.Series(y, name='target', index=index) elif len(y)>0 and isinstance(y[0], list) and len(y[0])==1: # single row in a matrix flatten = lambda y: [item for sublist in y for item in sublist] return pd.Series(flatten(y), name='target', index=index) elif len(y)==1 and len(y[0])==0 and isinstance(y[0], list): # single empty column in a matrix return pd.Series(y[0], name='target', index=index, dtype=float) elif len(y)==1 and isinstance(y[0], list): # single column in a matrix return pd.Series(y[0], name='target', index=index, dtype=type(y[0][0])) else: raise ValueError('Unexpected input shape') elif isinstance(y, pd.DataFrame): if len(list(y))==0: # empty DataFrame return pd.Series(name='target', index=index, dtype=float) if len(list(y))==1: # a single column return y.iloc[:, 0] else: raise ValueError(f'Unexpected input shape: {y.shape}') else: return pd.Series(y, name='target', index=index) # this covers tuples and other directly convertible types def get_generated_cols(X_original, X_transformed, to_transform): """ Returns a list of the generated/transformed columns. Arguments: X_original: df the original (input) DataFrame. X_transformed: df the transformed (current) DataFrame. to_transform: [str] a list of columns that were transformed (as in the original DataFrame), commonly self.feature_names_in. Output: a list of columns that were transformed (as in the current DataFrame). """ original_cols = list(X_original.columns) if len(to_transform) > 0: [original_cols.remove(c) for c in to_transform] current_cols = list(X_transformed.columns) if len(original_cols) > 0: [current_cols.remove(c) for c in original_cols] return current_cols def flatten_reverse_dict(d): sep = "___" [flat_dict] = pd.json_normalize(d, sep=sep).to_dict(orient='records') reversed_flat_dict = {v: tuple(k.split(sep)) for k, v in flat_dict.items()} return reversed_flat_dict class EncodingRelation(Enum): # one input feature get encoded into one output feature ONE_TO_ONE = auto() # one input feature get encoded into as many output features as it has distinct values ONE_TO_N_UNIQUE = auto() # one input feature get encoded into m output features that are not the number of distinct values ONE_TO_M = auto() # all N input features are encoded into M output features. # The encoding is done globally on all the input not on a per-feature basis N_TO_M = auto() def get_docstring_output_shape(in_out_relation: EncodingRelation): if in_out_relation == EncodingRelation.ONE_TO_ONE: return "n_features" elif in_out_relation == EncodingRelation.ONE_TO_N_UNIQUE: return "n_features * respective cardinality" elif in_out_relation == EncodingRelation.ONE_TO_M: return "M features (n_features < M)" elif in_out_relation == EncodingRelation.N_TO_M: return "M features (M can be anything)" class BaseEncoder(BaseEstimator): _dim: Optional[int] cols: List[str] use_default_cols: bool handle_missing: str handle_unknown: str verbose: int drop_invariant: bool invariant_cols: List[str] = [] return_df: bool supervised: bool encoding_relation: EncodingRelation INVARIANCE_THRESHOLD = 10e-5 # columns with variance less than this will be considered constant / invariant def __init__(self, verbose=0, cols=None, drop_invariant=False, return_df=True, handle_unknown='value', handle_missing='value', **kwargs): """ Parameters ---------- verbose: int integer indicating verbosity of output. 0 for none. cols: list a list of columns to encode, if None, all string and categorical columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform and inverse transform (otherwise it will be a numpy array). handle_missing: str how to handle missing values at fit time. Options are 'error', 'return_nan', and 'value'. Default 'value', which treat NaNs as a countable category at fit time. handle_unknown: str, int or dict of {column : option, ...}. how to handle unknown labels at transform time. Options are 'error' 'return_nan', 'value' and int. Defaults to None which uses NaN behaviour specified at fit time. Passing an int will fill with this int value. kwargs: dict. additional encoder specific parameters like regularisation. """ self.return_df = return_df self.drop_invariant = drop_invariant self.invariant_cols = [] self.verbose = verbose self.use_default_cols = cols is None # if True, even a repeated call of fit() will select string columns from X self.cols = cols # note that cols are only the columns to be encoded, feature_names_in_ are all columns self.mapping = None self.handle_unknown = handle_unknown self.handle_missing = handle_missing self._dim = None def fit(self, X, y=None, **kwargs): """Fits the encoder according to X and y. Parameters ---------- X : array-like, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. y : array-like, shape = [n_samples] Target values. Returns ------- self : encoder Returns self. """ self._check_fit_inputs(X, y) X, y = convert_inputs(X, y) self.feature_names_in_ = X.columns.tolist() self.n_features_in_ = len(self.feature_names_in_) self._dim = X.shape[1] self._determine_fit_columns(X) if not set(self.cols).issubset(X.columns): raise ValueError('X does not contain the columns listed in cols') if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') self._fit(X, y, **kwargs) # for finding invariant columns transform without y (as is done on the test set) X_transformed = self.transform(X, override_return_df=True) self.feature_names_out_ = X_transformed.columns.tolist() # drop all output columns with 0 variance. if self.drop_invariant: generated_cols = get_generated_cols(X, X_transformed, self.cols) self.invariant_cols = [x for x in generated_cols if X_transformed[x].var() <= self.INVARIANCE_THRESHOLD] self.feature_names_out_ = [x for x in self.feature_names_out_ if x not in self.invariant_cols] return self def _check_fit_inputs(self, X, y): if self._get_tags().get('supervised_encoder') and y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') def _check_transform_inputs(self, X): if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') if self._dim is None: raise NotFittedError('Must train encoder before it can be used to transform data.') # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') def _drop_invariants(self, X: pd.DataFrame, override_return_df: bool) -> Union[np.ndarray, pd.DataFrame]: if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.values def _determine_fit_columns(self, X: pd.DataFrame) -> None: """ Determine columns used by encoder. Note that the implementation also deals with re-fitting the same encoder object with different columns. :param X: input data frame :return: none, sets self.cols as a side effect """ # if columns aren't passed, just use every string column if self.use_default_cols: self.cols = get_obj_cols(X) else: self.cols = convert_cols_to_list(self.cols) def get_feature_names(self) -> List[str]: warnings.warn("`get_feature_names` is deprecated in all of sklearn. Use `get_feature_names_out` instead.", category=FutureWarning) return self.get_feature_names_out() def get_feature_names_out(self, input_features=None) -> np.ndarray: """ Returns the names of all transformed / added columns. Note that in sklearn the get_feature_names_out function takes the feature_names_in as an argument and determines the output feature names using the input. A fit is usually not necessary and if so a NotFittedError is raised. We just require a fit all the time and return the fitted output columns. Returns ------- feature_names: np.ndarray A numpy array with all feature names transformed or added. Note: potentially dropped features (because the feature is constant/invariant) are not included! """ out_feats = getattr(self, "feature_names_out_", None) if not isinstance(out_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return np.array(out_feats, dtype=object) def get_feature_names_in(self) -> List[str]: """ Returns the names of all input columns present when fitting. These columns are necessary for the transform step. """ in_feats = getattr(self, "feature_names_in_", None) if not isinstance(in_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return in_feats @abstractmethod def _fit(self, X: pd.DataFrame, y: Optional[pd.Series], **kwargs): ... class SupervisedTransformerMixin(sklearn.base.TransformerMixin): def _more_tags(self): return {'supervised_encoder': True} def transform(self, X, y=None, override_return_df=False): """Perform the transformation to new categorical data. Some encoders behave differently on whether y is given or not. This is mainly due to regularisation in order to avoid overfitting. On training data transform should be called with y, on test data without. Parameters ---------- X : array-like, shape = [n_samples, n_features] y : array-like, shape = [n_samples] or None override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X, y = convert_inputs(X, y, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X, y) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X: pd.DataFrame, y: pd.Series) -> pd.DataFrame: ... def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y) class UnsupervisedTransformerMixin(sklearn.base.TransformerMixin): def transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X = convert_input(X, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X) -> pd.DataFrame: ... class TransformerWithTargetMixin: def _more_tags(self): return {'supervised_encoder': True} def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y)

JaimeArboleda

5eb7a2d6359d680bdadd0534bdb983e712a47f9c

570827e6b48737d0c9aece8aca31edd6da02c1b2

Well, let me explain (for simplicity, every time I say a `Pipeline` it could be that or a `ColumnTransformer` or a `FeatureUnion`...): * We need to have the parameter `input_features` because otherwise, when the Encoder is part of a `Pipeline`, there will be an `Exception: BaseEncoder.get_feature_names_out() takes 1 positional argument but 2 were given`, because the method will be called with the parameter `input_features`. * Now, what can we do with this parameter? There were two options: modifying all encoders to take into account how to generate the output features given the input_features, or just using the `feature_names_out_`, which is available and computed. Given that this library internally work with dataframes, I thought that the second option was a much better idea. There is only one caveat, which happens again when the encoder is part of a `Pipeline`: if the previous transformer outputted a numpy array, then the column names won't be good (they will be like col_0_A, col_0_B...). That's why I think we should inform that, to ensure full compatibility and 100% correctness of feature names, `sklearn` should be used with the `set_output="pandas"` option. In this case, everything works smoothly, as the tests prove. * Another way to put it, in `sklearn` it makes sense to have a function with this signature, because an estimator can be fitted with a numpy array and then it should be able to "compute" the feature names of the output given a list of input_features. But for `category_encoders` this use case does not have too much sense, as the input is expected to be a pandas dataframe. And this assumption is spread everywhere: even other parameters work with column names, like the `mapping` or the `cols`; you won't see parameters expecting column names in `sklearn` because they expect that the input can be a numpy array. * So, yes, I agree the behavior is different than in `sklearn`, but I think it makes sense as long as the library is based on this "pandas assumption". I think a big redesign will be a good idea but as we discussed in the issue it will take a lot of effort. To be honest, if you have already set `sklearn` to output `pandas`, it does not make too much sense to use this method anyway... But at least this modification ensures that you don't get an `Exception` and it does not break anything. It's a very minor change, much less than I would have liked, but the opposite will be a big redesign that makes sense if you fully drop the "pandas assumption" of the library.

JaimeArboleda

scikit-learn-contrib/category_encoders

(WIP) Partial fix for getting feature names out

I think this is a partial fix for this opened issue: https://github.com/scikit-learn-contrib/category_encoders/issues/395 It remains to check the behaviour of other estimators that are not ONE_TO_ONE. Please, let me know if you like the work in progress and I will try to continue.

2023-02-23 13:33:41+00:00

2023-03-13 11:48:24+00:00

category_encoders/utils.py

"""A collection of shared utilities for all encoders, not intended for external use.""" from abc import abstractmethod from enum import Enum, auto import warnings import pandas as pd import numpy as np import sklearn.base from sklearn.base import BaseEstimator, TransformerMixin from sklearn.exceptions import NotFittedError from typing import Dict, List, Optional, Union from scipy.sparse import csr_matrix __author__ = 'willmcginnis' def convert_cols_to_list(cols): if isinstance(cols, pd.Series): return cols.tolist() elif isinstance(cols, np.ndarray): return cols.tolist() elif np.isscalar(cols): return [cols] elif isinstance(cols, set): return list(cols) elif isinstance(cols, tuple): return list(cols) elif pd.api.types.is_categorical_dtype(cols): return cols.astype(object).tolist() return cols def get_obj_cols(df): """ Returns names of 'object' columns in the DataFrame. """ obj_cols = [] for idx, dt in enumerate(df.dtypes): if dt == 'object' or is_category(dt): obj_cols.append(df.columns.values[idx]) if not obj_cols: print("Warning: No categorical columns found. Calling 'transform' will only return input data.") return obj_cols def is_category(dtype): return pd.api.types.is_categorical_dtype(dtype) def convert_inputs(X, y, columns=None, index=None, deep=False): """ Unite arraylike `X` and vectorlike `y` into a DataFrame and Series. If both are pandas types already, raises an error if their indexes do not match. If one is pandas, the returns will share that index. If neither is pandas, a default index will be used, unless `index` is passed. Parameters ---------- X: arraylike y: listlike columns: listlike Specifies column names to use for `X`. Ignored if `X` is already a dataframe. If `None`, use the default pandas column names. index: listlike The index to use, if neither `X` nor `y` is a pandas type. (If one has an index, then this has no effect.) If `None`, use the default pandas index. deep: bool Whether to deep-copy `X`. """ X_alt_index = y.index if isinstance(y, pd.Series) else index X = convert_input(X, columns=columns, deep=deep, index=X_alt_index) if y is not None: y = convert_input_vector(y, index=X.index) # N.B.: If either was already pandas, it keeps its index. if any(X.index != y.index): msg = "`X` and `y` both have indexes, but they do not match. If you are shuffling your input data on " \ "purpose (e.g. via permutation_test_score) use np arrays instead of data frames / series" raise ValueError(msg) if X.shape[0] != y.shape[0]: raise ValueError("The length of X is " + str(X.shape[0]) + " but length of y is " + str(y.shape[0]) + ".") return X, y def convert_input(X, columns=None, deep=False, index=None): """ Unite data into a DataFrame. Objects that do not contain column names take the names from the argument. Optionally perform deep copy of the data. """ if not isinstance(X, pd.DataFrame): if isinstance(X, pd.Series): X = pd.DataFrame(X, copy=deep) else: if columns is not None and np.size(X,1) != len(columns): raise ValueError('The count of the column names does not correspond to the count of the columns') if isinstance(X, list): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) # lists are always copied, but for consistency, we still pass the argument elif isinstance(X, (np.generic, np.ndarray)): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) elif isinstance(X, csr_matrix): X = pd.DataFrame(X.todense(), columns=columns, copy=deep, index=index) else: raise ValueError(f'Unexpected input type: {type(X)}') elif deep: X = X.copy(deep=True) return X def convert_input_vector(y, index): """ Unite target data type into a Series. If the target is a Series or a DataFrame, we preserve its index. But if the target does not contain index attribute, we use the index from the argument. """ if y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') if isinstance(y, pd.Series): return y elif isinstance(y, np.ndarray): if len(np.shape(y))==1: # vector return pd.Series(y, name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[0]==1: # single row in a matrix return pd.Series(y[0, :], name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[1]==1: # single column in a matrix return pd.Series(y[:, 0], name='target', index=index) else: raise ValueError(f'Unexpected input shape: {np.shape(y)}') elif np.isscalar(y): return pd.Series([y], name='target', index=index) elif isinstance(y, list): if len(y)==0: # empty list return pd.Series(y, name='target', index=index, dtype=float) elif len(y)>0 and not isinstance(y[0], list): # vector return pd.Series(y, name='target', index=index) elif len(y)>0 and isinstance(y[0], list) and len(y[0])==1: # single row in a matrix flatten = lambda y: [item for sublist in y for item in sublist] return pd.Series(flatten(y), name='target', index=index) elif len(y)==1 and len(y[0])==0 and isinstance(y[0], list): # single empty column in a matrix return pd.Series(y[0], name='target', index=index, dtype=float) elif len(y)==1 and isinstance(y[0], list): # single column in a matrix return pd.Series(y[0], name='target', index=index, dtype=type(y[0][0])) else: raise ValueError('Unexpected input shape') elif isinstance(y, pd.DataFrame): if len(list(y))==0: # empty DataFrame return pd.Series(name='target', index=index, dtype=float) if len(list(y))==1: # a single column return y.iloc[:, 0] else: raise ValueError(f'Unexpected input shape: {y.shape}') else: return pd.Series(y, name='target', index=index) # this covers tuples and other directly convertible types def get_generated_cols(X_original, X_transformed, to_transform): """ Returns a list of the generated/transformed columns. Arguments: X_original: df the original (input) DataFrame. X_transformed: df the transformed (current) DataFrame. to_transform: [str] a list of columns that were transformed (as in the original DataFrame), commonly self.feature_names_in. Output: a list of columns that were transformed (as in the current DataFrame). """ original_cols = list(X_original.columns) if len(to_transform) > 0: [original_cols.remove(c) for c in to_transform] current_cols = list(X_transformed.columns) if len(original_cols) > 0: [current_cols.remove(c) for c in original_cols] return current_cols def flatten_reverse_dict(d): sep = "___" [flat_dict] = pd.json_normalize(d, sep=sep).to_dict(orient='records') reversed_flat_dict = {v: tuple(k.split(sep)) for k, v in flat_dict.items()} return reversed_flat_dict class EncodingRelation(Enum): # one input feature get encoded into one output feature ONE_TO_ONE = auto() # one input feature get encoded into as many output features as it has distinct values ONE_TO_N_UNIQUE = auto() # one input feature get encoded into m output features that are not the number of distinct values ONE_TO_M = auto() # all N input features are encoded into M output features. # The encoding is done globally on all the input not on a per-feature basis N_TO_M = auto() def get_docstring_output_shape(in_out_relation: EncodingRelation): if in_out_relation == EncodingRelation.ONE_TO_ONE: return "n_features" elif in_out_relation == EncodingRelation.ONE_TO_N_UNIQUE: return "n_features * respective cardinality" elif in_out_relation == EncodingRelation.ONE_TO_M: return "M features (n_features < M)" elif in_out_relation == EncodingRelation.N_TO_M: return "M features (M can be anything)" class BaseEncoder(BaseEstimator): _dim: Optional[int] cols: List[str] use_default_cols: bool handle_missing: str handle_unknown: str verbose: int drop_invariant: bool invariant_cols: List[str] = [] return_df: bool supervised: bool encoding_relation: EncodingRelation INVARIANCE_THRESHOLD = 10e-5 # columns with variance less than this will be considered constant / invariant def __init__(self, verbose=0, cols=None, drop_invariant=False, return_df=True, handle_unknown='value', handle_missing='value', **kwargs): """ Parameters ---------- verbose: int integer indicating verbosity of output. 0 for none. cols: list a list of columns to encode, if None, all string and categorical columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform and inverse transform (otherwise it will be a numpy array). handle_missing: str how to handle missing values at fit time. Options are 'error', 'return_nan', and 'value'. Default 'value', which treat NaNs as a countable category at fit time. handle_unknown: str, int or dict of {column : option, ...}. how to handle unknown labels at transform time. Options are 'error' 'return_nan', 'value' and int. Defaults to None which uses NaN behaviour specified at fit time. Passing an int will fill with this int value. kwargs: dict. additional encoder specific parameters like regularisation. """ self.return_df = return_df self.drop_invariant = drop_invariant self.invariant_cols = [] self.verbose = verbose self.use_default_cols = cols is None # if True, even a repeated call of fit() will select string columns from X self.cols = cols # note that cols are only the columns to be encoded, feature_names_in_ are all columns self.mapping = None self.handle_unknown = handle_unknown self.handle_missing = handle_missing self._dim = None def fit(self, X, y=None, **kwargs): """Fits the encoder according to X and y. Parameters ---------- X : array-like, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. y : array-like, shape = [n_samples] Target values. Returns ------- self : encoder Returns self. """ self._check_fit_inputs(X, y) X, y = convert_inputs(X, y) self.feature_names_in_ = X.columns.tolist() self.n_features_in_ = len(self.feature_names_in_) self._dim = X.shape[1] self._determine_fit_columns(X) if not set(self.cols).issubset(X.columns): raise ValueError('X does not contain the columns listed in cols') if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') self._fit(X, y, **kwargs) # for finding invariant columns transform without y (as is done on the test set) X_transformed = self.transform(X, override_return_df=True) self.feature_names_out_ = X_transformed.columns.tolist() # drop all output columns with 0 variance. if self.drop_invariant: generated_cols = get_generated_cols(X, X_transformed, self.cols) self.invariant_cols = [x for x in generated_cols if X_transformed[x].var() <= self.INVARIANCE_THRESHOLD] self.feature_names_out_ = [x for x in self.feature_names_out_ if x not in self.invariant_cols] return self def _check_fit_inputs(self, X, y): if self._get_tags().get('supervised_encoder') and y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') def _check_transform_inputs(self, X): if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') if self._dim is None: raise NotFittedError('Must train encoder before it can be used to transform data.') # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') def _drop_invariants(self, X: pd.DataFrame, override_return_df: bool) -> Union[np.ndarray, pd.DataFrame]: if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.values def _determine_fit_columns(self, X: pd.DataFrame) -> None: """ Determine columns used by encoder. Note that the implementation also deals with re-fitting the same encoder object with different columns. :param X: input data frame :return: none, sets self.cols as a side effect """ # if columns aren't passed, just use every string column if self.use_default_cols: self.cols = get_obj_cols(X) else: self.cols = convert_cols_to_list(self.cols) def get_feature_names(self) -> List[str]: warnings.warn("`get_feature_names` is deprecated in all of sklearn. Use `get_feature_names_out` instead.", category=FutureWarning) return self.get_feature_names_out() def get_feature_names_out(self) -> List[str]: """ Returns the names of all transformed / added columns. Note that in sklearn the get_feature_names_out function takes the feature_names_in as an argument and determines the output feature names using the input. A fit is usually not necessary and if so a NotFittedError is raised. We just require a fit all the time and return the fitted output columns. Returns ------- feature_names: list A list with all feature names transformed or added. Note: potentially dropped features (because the feature is constant/invariant) are not included! """ out_feats = getattr(self, "feature_names_out_", None) if not isinstance(out_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return out_feats def get_feature_names_in(self) -> List[str]: """ Returns the names of all input columns present when fitting. These columns are necessary for the transform step. """ in_feats = getattr(self, "feature_names_in_", None) if not isinstance(in_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return in_feats @abstractmethod def _fit(self, X: pd.DataFrame, y: Optional[pd.Series], **kwargs): ... class SupervisedTransformerMixin(sklearn.base.TransformerMixin): def _more_tags(self): return {'supervised_encoder': True} def transform(self, X, y=None, override_return_df=False): """Perform the transformation to new categorical data. Some encoders behave differently on whether y is given or not. This is mainly due to regularisation in order to avoid overfitting. On training data transform should be called with y, on test data without. Parameters ---------- X : array-like, shape = [n_samples, n_features] y : array-like, shape = [n_samples] or None override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X, y = convert_inputs(X, y, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X, y) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X: pd.DataFrame, y: pd.Series) -> pd.DataFrame: ... def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y) class UnsupervisedTransformerMixin(sklearn.base.TransformerMixin): def transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X = convert_input(X, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X) -> pd.DataFrame: ... class TransformerWithTargetMixin: def _more_tags(self): return {'supervised_encoder': True} def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y)

"""A collection of shared utilities for all encoders, not intended for external use.""" from abc import abstractmethod from enum import Enum, auto import warnings import pandas as pd import numpy as np import sklearn.base from sklearn.base import BaseEstimator, TransformerMixin from sklearn.exceptions import NotFittedError from typing import Dict, List, Optional, Union from scipy.sparse import csr_matrix __author__ = 'willmcginnis' def convert_cols_to_list(cols): if isinstance(cols, pd.Series): return cols.tolist() elif isinstance(cols, np.ndarray): return cols.tolist() elif np.isscalar(cols): return [cols] elif isinstance(cols, set): return list(cols) elif isinstance(cols, tuple): return list(cols) elif pd.api.types.is_categorical_dtype(cols): return cols.astype(object).tolist() return cols def get_obj_cols(df): """ Returns names of 'object' columns in the DataFrame. """ obj_cols = [] for idx, dt in enumerate(df.dtypes): if dt == 'object' or is_category(dt): obj_cols.append(df.columns.values[idx]) if not obj_cols: print("Warning: No categorical columns found. Calling 'transform' will only return input data.") return obj_cols def is_category(dtype): return pd.api.types.is_categorical_dtype(dtype) def convert_inputs(X, y, columns=None, index=None, deep=False): """ Unite arraylike `X` and vectorlike `y` into a DataFrame and Series. If both are pandas types already, raises an error if their indexes do not match. If one is pandas, the returns will share that index. If neither is pandas, a default index will be used, unless `index` is passed. Parameters ---------- X: arraylike y: listlike columns: listlike Specifies column names to use for `X`. Ignored if `X` is already a dataframe. If `None`, use the default pandas column names. index: listlike The index to use, if neither `X` nor `y` is a pandas type. (If one has an index, then this has no effect.) If `None`, use the default pandas index. deep: bool Whether to deep-copy `X`. """ X_alt_index = y.index if isinstance(y, pd.Series) else index X = convert_input(X, columns=columns, deep=deep, index=X_alt_index) if y is not None: y = convert_input_vector(y, index=X.index) # N.B.: If either was already pandas, it keeps its index. if any(X.index != y.index): msg = "`X` and `y` both have indexes, but they do not match. If you are shuffling your input data on " \ "purpose (e.g. via permutation_test_score) use np arrays instead of data frames / series" raise ValueError(msg) if X.shape[0] != y.shape[0]: raise ValueError("The length of X is " + str(X.shape[0]) + " but length of y is " + str(y.shape[0]) + ".") return X, y def convert_input(X, columns=None, deep=False, index=None): """ Unite data into a DataFrame. Objects that do not contain column names take the names from the argument. Optionally perform deep copy of the data. """ if not isinstance(X, pd.DataFrame): if isinstance(X, pd.Series): X = pd.DataFrame(X, copy=deep) else: if columns is not None and np.size(X,1) != len(columns): raise ValueError('The count of the column names does not correspond to the count of the columns') if isinstance(X, list): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) # lists are always copied, but for consistency, we still pass the argument elif isinstance(X, (np.generic, np.ndarray)): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) elif isinstance(X, csr_matrix): X = pd.DataFrame(X.todense(), columns=columns, copy=deep, index=index) else: raise ValueError(f'Unexpected input type: {type(X)}') elif deep: X = X.copy(deep=True) return X def convert_input_vector(y, index): """ Unite target data type into a Series. If the target is a Series or a DataFrame, we preserve its index. But if the target does not contain index attribute, we use the index from the argument. """ if y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') if isinstance(y, pd.Series): return y elif isinstance(y, np.ndarray): if len(np.shape(y))==1: # vector return pd.Series(y, name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[0]==1: # single row in a matrix return pd.Series(y[0, :], name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[1]==1: # single column in a matrix return pd.Series(y[:, 0], name='target', index=index) else: raise ValueError(f'Unexpected input shape: {np.shape(y)}') elif np.isscalar(y): return pd.Series([y], name='target', index=index) elif isinstance(y, list): if len(y)==0: # empty list return pd.Series(y, name='target', index=index, dtype=float) elif len(y)>0 and not isinstance(y[0], list): # vector return pd.Series(y, name='target', index=index) elif len(y)>0 and isinstance(y[0], list) and len(y[0])==1: # single row in a matrix flatten = lambda y: [item for sublist in y for item in sublist] return pd.Series(flatten(y), name='target', index=index) elif len(y)==1 and len(y[0])==0 and isinstance(y[0], list): # single empty column in a matrix return pd.Series(y[0], name='target', index=index, dtype=float) elif len(y)==1 and isinstance(y[0], list): # single column in a matrix return pd.Series(y[0], name='target', index=index, dtype=type(y[0][0])) else: raise ValueError('Unexpected input shape') elif isinstance(y, pd.DataFrame): if len(list(y))==0: # empty DataFrame return pd.Series(name='target', index=index, dtype=float) if len(list(y))==1: # a single column return y.iloc[:, 0] else: raise ValueError(f'Unexpected input shape: {y.shape}') else: return pd.Series(y, name='target', index=index) # this covers tuples and other directly convertible types def get_generated_cols(X_original, X_transformed, to_transform): """ Returns a list of the generated/transformed columns. Arguments: X_original: df the original (input) DataFrame. X_transformed: df the transformed (current) DataFrame. to_transform: [str] a list of columns that were transformed (as in the original DataFrame), commonly self.feature_names_in. Output: a list of columns that were transformed (as in the current DataFrame). """ original_cols = list(X_original.columns) if len(to_transform) > 0: [original_cols.remove(c) for c in to_transform] current_cols = list(X_transformed.columns) if len(original_cols) > 0: [current_cols.remove(c) for c in original_cols] return current_cols def flatten_reverse_dict(d): sep = "___" [flat_dict] = pd.json_normalize(d, sep=sep).to_dict(orient='records') reversed_flat_dict = {v: tuple(k.split(sep)) for k, v in flat_dict.items()} return reversed_flat_dict class EncodingRelation(Enum): # one input feature get encoded into one output feature ONE_TO_ONE = auto() # one input feature get encoded into as many output features as it has distinct values ONE_TO_N_UNIQUE = auto() # one input feature get encoded into m output features that are not the number of distinct values ONE_TO_M = auto() # all N input features are encoded into M output features. # The encoding is done globally on all the input not on a per-feature basis N_TO_M = auto() def get_docstring_output_shape(in_out_relation: EncodingRelation): if in_out_relation == EncodingRelation.ONE_TO_ONE: return "n_features" elif in_out_relation == EncodingRelation.ONE_TO_N_UNIQUE: return "n_features * respective cardinality" elif in_out_relation == EncodingRelation.ONE_TO_M: return "M features (n_features < M)" elif in_out_relation == EncodingRelation.N_TO_M: return "M features (M can be anything)" class BaseEncoder(BaseEstimator): _dim: Optional[int] cols: List[str] use_default_cols: bool handle_missing: str handle_unknown: str verbose: int drop_invariant: bool invariant_cols: List[str] = [] return_df: bool supervised: bool encoding_relation: EncodingRelation INVARIANCE_THRESHOLD = 10e-5 # columns with variance less than this will be considered constant / invariant def __init__(self, verbose=0, cols=None, drop_invariant=False, return_df=True, handle_unknown='value', handle_missing='value', **kwargs): """ Parameters ---------- verbose: int integer indicating verbosity of output. 0 for none. cols: list a list of columns to encode, if None, all string and categorical columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform and inverse transform (otherwise it will be a numpy array). handle_missing: str how to handle missing values at fit time. Options are 'error', 'return_nan', and 'value'. Default 'value', which treat NaNs as a countable category at fit time. handle_unknown: str, int or dict of {column : option, ...}. how to handle unknown labels at transform time. Options are 'error' 'return_nan', 'value' and int. Defaults to None which uses NaN behaviour specified at fit time. Passing an int will fill with this int value. kwargs: dict. additional encoder specific parameters like regularisation. """ self.return_df = return_df self.drop_invariant = drop_invariant self.invariant_cols = [] self.verbose = verbose self.use_default_cols = cols is None # if True, even a repeated call of fit() will select string columns from X self.cols = cols # note that cols are only the columns to be encoded, feature_names_in_ are all columns self.mapping = None self.handle_unknown = handle_unknown self.handle_missing = handle_missing self._dim = None def fit(self, X, y=None, **kwargs): """Fits the encoder according to X and y. Parameters ---------- X : array-like, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. y : array-like, shape = [n_samples] Target values. Returns ------- self : encoder Returns self. """ self._check_fit_inputs(X, y) X, y = convert_inputs(X, y) self.feature_names_in_ = X.columns.tolist() self.n_features_in_ = len(self.feature_names_in_) self._dim = X.shape[1] self._determine_fit_columns(X) if not set(self.cols).issubset(X.columns): raise ValueError('X does not contain the columns listed in cols') if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') self._fit(X, y, **kwargs) # for finding invariant columns transform without y (as is done on the test set) X_transformed = self.transform(X, override_return_df=True) self.feature_names_out_ = X_transformed.columns.tolist() # drop all output columns with 0 variance. if self.drop_invariant: generated_cols = get_generated_cols(X, X_transformed, self.cols) self.invariant_cols = [x for x in generated_cols if X_transformed[x].var() <= self.INVARIANCE_THRESHOLD] self.feature_names_out_ = [x for x in self.feature_names_out_ if x not in self.invariant_cols] return self def _check_fit_inputs(self, X, y): if self._get_tags().get('supervised_encoder') and y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') def _check_transform_inputs(self, X): if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') if self._dim is None: raise NotFittedError('Must train encoder before it can be used to transform data.') # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') def _drop_invariants(self, X: pd.DataFrame, override_return_df: bool) -> Union[np.ndarray, pd.DataFrame]: if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.values def _determine_fit_columns(self, X: pd.DataFrame) -> None: """ Determine columns used by encoder. Note that the implementation also deals with re-fitting the same encoder object with different columns. :param X: input data frame :return: none, sets self.cols as a side effect """ # if columns aren't passed, just use every string column if self.use_default_cols: self.cols = get_obj_cols(X) else: self.cols = convert_cols_to_list(self.cols) def get_feature_names(self) -> List[str]: warnings.warn("`get_feature_names` is deprecated in all of sklearn. Use `get_feature_names_out` instead.", category=FutureWarning) return self.get_feature_names_out() def get_feature_names_out(self, input_features=None) -> np.ndarray: """ Returns the names of all transformed / added columns. Note that in sklearn the get_feature_names_out function takes the feature_names_in as an argument and determines the output feature names using the input. A fit is usually not necessary and if so a NotFittedError is raised. We just require a fit all the time and return the fitted output columns. Returns ------- feature_names: np.ndarray A numpy array with all feature names transformed or added. Note: potentially dropped features (because the feature is constant/invariant) are not included! """ out_feats = getattr(self, "feature_names_out_", None) if not isinstance(out_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return np.array(out_feats, dtype=object) def get_feature_names_in(self) -> List[str]: """ Returns the names of all input columns present when fitting. These columns are necessary for the transform step. """ in_feats = getattr(self, "feature_names_in_", None) if not isinstance(in_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return in_feats @abstractmethod def _fit(self, X: pd.DataFrame, y: Optional[pd.Series], **kwargs): ... class SupervisedTransformerMixin(sklearn.base.TransformerMixin): def _more_tags(self): return {'supervised_encoder': True} def transform(self, X, y=None, override_return_df=False): """Perform the transformation to new categorical data. Some encoders behave differently on whether y is given or not. This is mainly due to regularisation in order to avoid overfitting. On training data transform should be called with y, on test data without. Parameters ---------- X : array-like, shape = [n_samples, n_features] y : array-like, shape = [n_samples] or None override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X, y = convert_inputs(X, y, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X, y) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X: pd.DataFrame, y: pd.Series) -> pd.DataFrame: ... def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y) class UnsupervisedTransformerMixin(sklearn.base.TransformerMixin): def transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X = convert_input(X, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X) -> pd.DataFrame: ... class TransformerWithTargetMixin: def _more_tags(self): return {'supervised_encoder': True} def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y)

JaimeArboleda

5eb7a2d6359d680bdadd0534bdb983e712a47f9c

570827e6b48737d0c9aece8aca31edd6da02c1b2

I agree that the proposed change is better than what we have at the moment since it fixes a bug. I've got one more question: If we change all encoders to properly work with the input features, will we get compatibility even without setting `set_output=pandas`? Even if we still do all the internal stuff in pandas? If so I think it would be worth it because we'd get compatibility even for sklearn versions < 1.2. This is in fact what I intended when I first raised the issue. In one of the comment I outlined what the changes for each encoder / family of encoders would be. This is far from a complete redesign of the library and also won't get rid of all dataframes. Does this make sense? I somehow really don't like introducing dependency versions in the code since this will make it a maintenance hell in the long run

PaulWestenthanner

scikit-learn-contrib/category_encoders

(WIP) Partial fix for getting feature names out

I think this is a partial fix for this opened issue: https://github.com/scikit-learn-contrib/category_encoders/issues/395 It remains to check the behaviour of other estimators that are not ONE_TO_ONE. Please, let me know if you like the work in progress and I will try to continue.

2023-02-23 13:33:41+00:00

2023-03-13 11:48:24+00:00

category_encoders/utils.py

"""A collection of shared utilities for all encoders, not intended for external use.""" from abc import abstractmethod from enum import Enum, auto import warnings import pandas as pd import numpy as np import sklearn.base from sklearn.base import BaseEstimator, TransformerMixin from sklearn.exceptions import NotFittedError from typing import Dict, List, Optional, Union from scipy.sparse import csr_matrix __author__ = 'willmcginnis' def convert_cols_to_list(cols): if isinstance(cols, pd.Series): return cols.tolist() elif isinstance(cols, np.ndarray): return cols.tolist() elif np.isscalar(cols): return [cols] elif isinstance(cols, set): return list(cols) elif isinstance(cols, tuple): return list(cols) elif pd.api.types.is_categorical_dtype(cols): return cols.astype(object).tolist() return cols def get_obj_cols(df): """ Returns names of 'object' columns in the DataFrame. """ obj_cols = [] for idx, dt in enumerate(df.dtypes): if dt == 'object' or is_category(dt): obj_cols.append(df.columns.values[idx]) if not obj_cols: print("Warning: No categorical columns found. Calling 'transform' will only return input data.") return obj_cols def is_category(dtype): return pd.api.types.is_categorical_dtype(dtype) def convert_inputs(X, y, columns=None, index=None, deep=False): """ Unite arraylike `X` and vectorlike `y` into a DataFrame and Series. If both are pandas types already, raises an error if their indexes do not match. If one is pandas, the returns will share that index. If neither is pandas, a default index will be used, unless `index` is passed. Parameters ---------- X: arraylike y: listlike columns: listlike Specifies column names to use for `X`. Ignored if `X` is already a dataframe. If `None`, use the default pandas column names. index: listlike The index to use, if neither `X` nor `y` is a pandas type. (If one has an index, then this has no effect.) If `None`, use the default pandas index. deep: bool Whether to deep-copy `X`. """ X_alt_index = y.index if isinstance(y, pd.Series) else index X = convert_input(X, columns=columns, deep=deep, index=X_alt_index) if y is not None: y = convert_input_vector(y, index=X.index) # N.B.: If either was already pandas, it keeps its index. if any(X.index != y.index): msg = "`X` and `y` both have indexes, but they do not match. If you are shuffling your input data on " \ "purpose (e.g. via permutation_test_score) use np arrays instead of data frames / series" raise ValueError(msg) if X.shape[0] != y.shape[0]: raise ValueError("The length of X is " + str(X.shape[0]) + " but length of y is " + str(y.shape[0]) + ".") return X, y def convert_input(X, columns=None, deep=False, index=None): """ Unite data into a DataFrame. Objects that do not contain column names take the names from the argument. Optionally perform deep copy of the data. """ if not isinstance(X, pd.DataFrame): if isinstance(X, pd.Series): X = pd.DataFrame(X, copy=deep) else: if columns is not None and np.size(X,1) != len(columns): raise ValueError('The count of the column names does not correspond to the count of the columns') if isinstance(X, list): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) # lists are always copied, but for consistency, we still pass the argument elif isinstance(X, (np.generic, np.ndarray)): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) elif isinstance(X, csr_matrix): X = pd.DataFrame(X.todense(), columns=columns, copy=deep, index=index) else: raise ValueError(f'Unexpected input type: {type(X)}') elif deep: X = X.copy(deep=True) return X def convert_input_vector(y, index): """ Unite target data type into a Series. If the target is a Series or a DataFrame, we preserve its index. But if the target does not contain index attribute, we use the index from the argument. """ if y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') if isinstance(y, pd.Series): return y elif isinstance(y, np.ndarray): if len(np.shape(y))==1: # vector return pd.Series(y, name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[0]==1: # single row in a matrix return pd.Series(y[0, :], name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[1]==1: # single column in a matrix return pd.Series(y[:, 0], name='target', index=index) else: raise ValueError(f'Unexpected input shape: {np.shape(y)}') elif np.isscalar(y): return pd.Series([y], name='target', index=index) elif isinstance(y, list): if len(y)==0: # empty list return pd.Series(y, name='target', index=index, dtype=float) elif len(y)>0 and not isinstance(y[0], list): # vector return pd.Series(y, name='target', index=index) elif len(y)>0 and isinstance(y[0], list) and len(y[0])==1: # single row in a matrix flatten = lambda y: [item for sublist in y for item in sublist] return pd.Series(flatten(y), name='target', index=index) elif len(y)==1 and len(y[0])==0 and isinstance(y[0], list): # single empty column in a matrix return pd.Series(y[0], name='target', index=index, dtype=float) elif len(y)==1 and isinstance(y[0], list): # single column in a matrix return pd.Series(y[0], name='target', index=index, dtype=type(y[0][0])) else: raise ValueError('Unexpected input shape') elif isinstance(y, pd.DataFrame): if len(list(y))==0: # empty DataFrame return pd.Series(name='target', index=index, dtype=float) if len(list(y))==1: # a single column return y.iloc[:, 0] else: raise ValueError(f'Unexpected input shape: {y.shape}') else: return pd.Series(y, name='target', index=index) # this covers tuples and other directly convertible types def get_generated_cols(X_original, X_transformed, to_transform): """ Returns a list of the generated/transformed columns. Arguments: X_original: df the original (input) DataFrame. X_transformed: df the transformed (current) DataFrame. to_transform: [str] a list of columns that were transformed (as in the original DataFrame), commonly self.feature_names_in. Output: a list of columns that were transformed (as in the current DataFrame). """ original_cols = list(X_original.columns) if len(to_transform) > 0: [original_cols.remove(c) for c in to_transform] current_cols = list(X_transformed.columns) if len(original_cols) > 0: [current_cols.remove(c) for c in original_cols] return current_cols def flatten_reverse_dict(d): sep = "___" [flat_dict] = pd.json_normalize(d, sep=sep).to_dict(orient='records') reversed_flat_dict = {v: tuple(k.split(sep)) for k, v in flat_dict.items()} return reversed_flat_dict class EncodingRelation(Enum): # one input feature get encoded into one output feature ONE_TO_ONE = auto() # one input feature get encoded into as many output features as it has distinct values ONE_TO_N_UNIQUE = auto() # one input feature get encoded into m output features that are not the number of distinct values ONE_TO_M = auto() # all N input features are encoded into M output features. # The encoding is done globally on all the input not on a per-feature basis N_TO_M = auto() def get_docstring_output_shape(in_out_relation: EncodingRelation): if in_out_relation == EncodingRelation.ONE_TO_ONE: return "n_features" elif in_out_relation == EncodingRelation.ONE_TO_N_UNIQUE: return "n_features * respective cardinality" elif in_out_relation == EncodingRelation.ONE_TO_M: return "M features (n_features < M)" elif in_out_relation == EncodingRelation.N_TO_M: return "M features (M can be anything)" class BaseEncoder(BaseEstimator): _dim: Optional[int] cols: List[str] use_default_cols: bool handle_missing: str handle_unknown: str verbose: int drop_invariant: bool invariant_cols: List[str] = [] return_df: bool supervised: bool encoding_relation: EncodingRelation INVARIANCE_THRESHOLD = 10e-5 # columns with variance less than this will be considered constant / invariant def __init__(self, verbose=0, cols=None, drop_invariant=False, return_df=True, handle_unknown='value', handle_missing='value', **kwargs): """ Parameters ---------- verbose: int integer indicating verbosity of output. 0 for none. cols: list a list of columns to encode, if None, all string and categorical columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform and inverse transform (otherwise it will be a numpy array). handle_missing: str how to handle missing values at fit time. Options are 'error', 'return_nan', and 'value'. Default 'value', which treat NaNs as a countable category at fit time. handle_unknown: str, int or dict of {column : option, ...}. how to handle unknown labels at transform time. Options are 'error' 'return_nan', 'value' and int. Defaults to None which uses NaN behaviour specified at fit time. Passing an int will fill with this int value. kwargs: dict. additional encoder specific parameters like regularisation. """ self.return_df = return_df self.drop_invariant = drop_invariant self.invariant_cols = [] self.verbose = verbose self.use_default_cols = cols is None # if True, even a repeated call of fit() will select string columns from X self.cols = cols # note that cols are only the columns to be encoded, feature_names_in_ are all columns self.mapping = None self.handle_unknown = handle_unknown self.handle_missing = handle_missing self._dim = None def fit(self, X, y=None, **kwargs): """Fits the encoder according to X and y. Parameters ---------- X : array-like, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. y : array-like, shape = [n_samples] Target values. Returns ------- self : encoder Returns self. """ self._check_fit_inputs(X, y) X, y = convert_inputs(X, y) self.feature_names_in_ = X.columns.tolist() self.n_features_in_ = len(self.feature_names_in_) self._dim = X.shape[1] self._determine_fit_columns(X) if not set(self.cols).issubset(X.columns): raise ValueError('X does not contain the columns listed in cols') if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') self._fit(X, y, **kwargs) # for finding invariant columns transform without y (as is done on the test set) X_transformed = self.transform(X, override_return_df=True) self.feature_names_out_ = X_transformed.columns.tolist() # drop all output columns with 0 variance. if self.drop_invariant: generated_cols = get_generated_cols(X, X_transformed, self.cols) self.invariant_cols = [x for x in generated_cols if X_transformed[x].var() <= self.INVARIANCE_THRESHOLD] self.feature_names_out_ = [x for x in self.feature_names_out_ if x not in self.invariant_cols] return self def _check_fit_inputs(self, X, y): if self._get_tags().get('supervised_encoder') and y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') def _check_transform_inputs(self, X): if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') if self._dim is None: raise NotFittedError('Must train encoder before it can be used to transform data.') # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') def _drop_invariants(self, X: pd.DataFrame, override_return_df: bool) -> Union[np.ndarray, pd.DataFrame]: if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.values def _determine_fit_columns(self, X: pd.DataFrame) -> None: """ Determine columns used by encoder. Note that the implementation also deals with re-fitting the same encoder object with different columns. :param X: input data frame :return: none, sets self.cols as a side effect """ # if columns aren't passed, just use every string column if self.use_default_cols: self.cols = get_obj_cols(X) else: self.cols = convert_cols_to_list(self.cols) def get_feature_names(self) -> List[str]: warnings.warn("`get_feature_names` is deprecated in all of sklearn. Use `get_feature_names_out` instead.", category=FutureWarning) return self.get_feature_names_out() def get_feature_names_out(self) -> List[str]: """ Returns the names of all transformed / added columns. Note that in sklearn the get_feature_names_out function takes the feature_names_in as an argument and determines the output feature names using the input. A fit is usually not necessary and if so a NotFittedError is raised. We just require a fit all the time and return the fitted output columns. Returns ------- feature_names: list A list with all feature names transformed or added. Note: potentially dropped features (because the feature is constant/invariant) are not included! """ out_feats = getattr(self, "feature_names_out_", None) if not isinstance(out_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return out_feats def get_feature_names_in(self) -> List[str]: """ Returns the names of all input columns present when fitting. These columns are necessary for the transform step. """ in_feats = getattr(self, "feature_names_in_", None) if not isinstance(in_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return in_feats @abstractmethod def _fit(self, X: pd.DataFrame, y: Optional[pd.Series], **kwargs): ... class SupervisedTransformerMixin(sklearn.base.TransformerMixin): def _more_tags(self): return {'supervised_encoder': True} def transform(self, X, y=None, override_return_df=False): """Perform the transformation to new categorical data. Some encoders behave differently on whether y is given or not. This is mainly due to regularisation in order to avoid overfitting. On training data transform should be called with y, on test data without. Parameters ---------- X : array-like, shape = [n_samples, n_features] y : array-like, shape = [n_samples] or None override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X, y = convert_inputs(X, y, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X, y) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X: pd.DataFrame, y: pd.Series) -> pd.DataFrame: ... def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y) class UnsupervisedTransformerMixin(sklearn.base.TransformerMixin): def transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X = convert_input(X, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X) -> pd.DataFrame: ... class TransformerWithTargetMixin: def _more_tags(self): return {'supervised_encoder': True} def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y)

"""A collection of shared utilities for all encoders, not intended for external use.""" from abc import abstractmethod from enum import Enum, auto import warnings import pandas as pd import numpy as np import sklearn.base from sklearn.base import BaseEstimator, TransformerMixin from sklearn.exceptions import NotFittedError from typing import Dict, List, Optional, Union from scipy.sparse import csr_matrix __author__ = 'willmcginnis' def convert_cols_to_list(cols): if isinstance(cols, pd.Series): return cols.tolist() elif isinstance(cols, np.ndarray): return cols.tolist() elif np.isscalar(cols): return [cols] elif isinstance(cols, set): return list(cols) elif isinstance(cols, tuple): return list(cols) elif pd.api.types.is_categorical_dtype(cols): return cols.astype(object).tolist() return cols def get_obj_cols(df): """ Returns names of 'object' columns in the DataFrame. """ obj_cols = [] for idx, dt in enumerate(df.dtypes): if dt == 'object' or is_category(dt): obj_cols.append(df.columns.values[idx]) if not obj_cols: print("Warning: No categorical columns found. Calling 'transform' will only return input data.") return obj_cols def is_category(dtype): return pd.api.types.is_categorical_dtype(dtype) def convert_inputs(X, y, columns=None, index=None, deep=False): """ Unite arraylike `X` and vectorlike `y` into a DataFrame and Series. If both are pandas types already, raises an error if their indexes do not match. If one is pandas, the returns will share that index. If neither is pandas, a default index will be used, unless `index` is passed. Parameters ---------- X: arraylike y: listlike columns: listlike Specifies column names to use for `X`. Ignored if `X` is already a dataframe. If `None`, use the default pandas column names. index: listlike The index to use, if neither `X` nor `y` is a pandas type. (If one has an index, then this has no effect.) If `None`, use the default pandas index. deep: bool Whether to deep-copy `X`. """ X_alt_index = y.index if isinstance(y, pd.Series) else index X = convert_input(X, columns=columns, deep=deep, index=X_alt_index) if y is not None: y = convert_input_vector(y, index=X.index) # N.B.: If either was already pandas, it keeps its index. if any(X.index != y.index): msg = "`X` and `y` both have indexes, but they do not match. If you are shuffling your input data on " \ "purpose (e.g. via permutation_test_score) use np arrays instead of data frames / series" raise ValueError(msg) if X.shape[0] != y.shape[0]: raise ValueError("The length of X is " + str(X.shape[0]) + " but length of y is " + str(y.shape[0]) + ".") return X, y def convert_input(X, columns=None, deep=False, index=None): """ Unite data into a DataFrame. Objects that do not contain column names take the names from the argument. Optionally perform deep copy of the data. """ if not isinstance(X, pd.DataFrame): if isinstance(X, pd.Series): X = pd.DataFrame(X, copy=deep) else: if columns is not None and np.size(X,1) != len(columns): raise ValueError('The count of the column names does not correspond to the count of the columns') if isinstance(X, list): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) # lists are always copied, but for consistency, we still pass the argument elif isinstance(X, (np.generic, np.ndarray)): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) elif isinstance(X, csr_matrix): X = pd.DataFrame(X.todense(), columns=columns, copy=deep, index=index) else: raise ValueError(f'Unexpected input type: {type(X)}') elif deep: X = X.copy(deep=True) return X def convert_input_vector(y, index): """ Unite target data type into a Series. If the target is a Series or a DataFrame, we preserve its index. But if the target does not contain index attribute, we use the index from the argument. """ if y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') if isinstance(y, pd.Series): return y elif isinstance(y, np.ndarray): if len(np.shape(y))==1: # vector return pd.Series(y, name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[0]==1: # single row in a matrix return pd.Series(y[0, :], name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[1]==1: # single column in a matrix return pd.Series(y[:, 0], name='target', index=index) else: raise ValueError(f'Unexpected input shape: {np.shape(y)}') elif np.isscalar(y): return pd.Series([y], name='target', index=index) elif isinstance(y, list): if len(y)==0: # empty list return pd.Series(y, name='target', index=index, dtype=float) elif len(y)>0 and not isinstance(y[0], list): # vector return pd.Series(y, name='target', index=index) elif len(y)>0 and isinstance(y[0], list) and len(y[0])==1: # single row in a matrix flatten = lambda y: [item for sublist in y for item in sublist] return pd.Series(flatten(y), name='target', index=index) elif len(y)==1 and len(y[0])==0 and isinstance(y[0], list): # single empty column in a matrix return pd.Series(y[0], name='target', index=index, dtype=float) elif len(y)==1 and isinstance(y[0], list): # single column in a matrix return pd.Series(y[0], name='target', index=index, dtype=type(y[0][0])) else: raise ValueError('Unexpected input shape') elif isinstance(y, pd.DataFrame): if len(list(y))==0: # empty DataFrame return pd.Series(name='target', index=index, dtype=float) if len(list(y))==1: # a single column return y.iloc[:, 0] else: raise ValueError(f'Unexpected input shape: {y.shape}') else: return pd.Series(y, name='target', index=index) # this covers tuples and other directly convertible types def get_generated_cols(X_original, X_transformed, to_transform): """ Returns a list of the generated/transformed columns. Arguments: X_original: df the original (input) DataFrame. X_transformed: df the transformed (current) DataFrame. to_transform: [str] a list of columns that were transformed (as in the original DataFrame), commonly self.feature_names_in. Output: a list of columns that were transformed (as in the current DataFrame). """ original_cols = list(X_original.columns) if len(to_transform) > 0: [original_cols.remove(c) for c in to_transform] current_cols = list(X_transformed.columns) if len(original_cols) > 0: [current_cols.remove(c) for c in original_cols] return current_cols def flatten_reverse_dict(d): sep = "___" [flat_dict] = pd.json_normalize(d, sep=sep).to_dict(orient='records') reversed_flat_dict = {v: tuple(k.split(sep)) for k, v in flat_dict.items()} return reversed_flat_dict class EncodingRelation(Enum): # one input feature get encoded into one output feature ONE_TO_ONE = auto() # one input feature get encoded into as many output features as it has distinct values ONE_TO_N_UNIQUE = auto() # one input feature get encoded into m output features that are not the number of distinct values ONE_TO_M = auto() # all N input features are encoded into M output features. # The encoding is done globally on all the input not on a per-feature basis N_TO_M = auto() def get_docstring_output_shape(in_out_relation: EncodingRelation): if in_out_relation == EncodingRelation.ONE_TO_ONE: return "n_features" elif in_out_relation == EncodingRelation.ONE_TO_N_UNIQUE: return "n_features * respective cardinality" elif in_out_relation == EncodingRelation.ONE_TO_M: return "M features (n_features < M)" elif in_out_relation == EncodingRelation.N_TO_M: return "M features (M can be anything)" class BaseEncoder(BaseEstimator): _dim: Optional[int] cols: List[str] use_default_cols: bool handle_missing: str handle_unknown: str verbose: int drop_invariant: bool invariant_cols: List[str] = [] return_df: bool supervised: bool encoding_relation: EncodingRelation INVARIANCE_THRESHOLD = 10e-5 # columns with variance less than this will be considered constant / invariant def __init__(self, verbose=0, cols=None, drop_invariant=False, return_df=True, handle_unknown='value', handle_missing='value', **kwargs): """ Parameters ---------- verbose: int integer indicating verbosity of output. 0 for none. cols: list a list of columns to encode, if None, all string and categorical columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform and inverse transform (otherwise it will be a numpy array). handle_missing: str how to handle missing values at fit time. Options are 'error', 'return_nan', and 'value'. Default 'value', which treat NaNs as a countable category at fit time. handle_unknown: str, int or dict of {column : option, ...}. how to handle unknown labels at transform time. Options are 'error' 'return_nan', 'value' and int. Defaults to None which uses NaN behaviour specified at fit time. Passing an int will fill with this int value. kwargs: dict. additional encoder specific parameters like regularisation. """ self.return_df = return_df self.drop_invariant = drop_invariant self.invariant_cols = [] self.verbose = verbose self.use_default_cols = cols is None # if True, even a repeated call of fit() will select string columns from X self.cols = cols # note that cols are only the columns to be encoded, feature_names_in_ are all columns self.mapping = None self.handle_unknown = handle_unknown self.handle_missing = handle_missing self._dim = None def fit(self, X, y=None, **kwargs): """Fits the encoder according to X and y. Parameters ---------- X : array-like, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. y : array-like, shape = [n_samples] Target values. Returns ------- self : encoder Returns self. """ self._check_fit_inputs(X, y) X, y = convert_inputs(X, y) self.feature_names_in_ = X.columns.tolist() self.n_features_in_ = len(self.feature_names_in_) self._dim = X.shape[1] self._determine_fit_columns(X) if not set(self.cols).issubset(X.columns): raise ValueError('X does not contain the columns listed in cols') if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') self._fit(X, y, **kwargs) # for finding invariant columns transform without y (as is done on the test set) X_transformed = self.transform(X, override_return_df=True) self.feature_names_out_ = X_transformed.columns.tolist() # drop all output columns with 0 variance. if self.drop_invariant: generated_cols = get_generated_cols(X, X_transformed, self.cols) self.invariant_cols = [x for x in generated_cols if X_transformed[x].var() <= self.INVARIANCE_THRESHOLD] self.feature_names_out_ = [x for x in self.feature_names_out_ if x not in self.invariant_cols] return self def _check_fit_inputs(self, X, y): if self._get_tags().get('supervised_encoder') and y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') def _check_transform_inputs(self, X): if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') if self._dim is None: raise NotFittedError('Must train encoder before it can be used to transform data.') # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') def _drop_invariants(self, X: pd.DataFrame, override_return_df: bool) -> Union[np.ndarray, pd.DataFrame]: if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.values def _determine_fit_columns(self, X: pd.DataFrame) -> None: """ Determine columns used by encoder. Note that the implementation also deals with re-fitting the same encoder object with different columns. :param X: input data frame :return: none, sets self.cols as a side effect """ # if columns aren't passed, just use every string column if self.use_default_cols: self.cols = get_obj_cols(X) else: self.cols = convert_cols_to_list(self.cols) def get_feature_names(self) -> List[str]: warnings.warn("`get_feature_names` is deprecated in all of sklearn. Use `get_feature_names_out` instead.", category=FutureWarning) return self.get_feature_names_out() def get_feature_names_out(self, input_features=None) -> np.ndarray: """ Returns the names of all transformed / added columns. Note that in sklearn the get_feature_names_out function takes the feature_names_in as an argument and determines the output feature names using the input. A fit is usually not necessary and if so a NotFittedError is raised. We just require a fit all the time and return the fitted output columns. Returns ------- feature_names: np.ndarray A numpy array with all feature names transformed or added. Note: potentially dropped features (because the feature is constant/invariant) are not included! """ out_feats = getattr(self, "feature_names_out_", None) if not isinstance(out_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return np.array(out_feats, dtype=object) def get_feature_names_in(self) -> List[str]: """ Returns the names of all input columns present when fitting. These columns are necessary for the transform step. """ in_feats = getattr(self, "feature_names_in_", None) if not isinstance(in_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return in_feats @abstractmethod def _fit(self, X: pd.DataFrame, y: Optional[pd.Series], **kwargs): ... class SupervisedTransformerMixin(sklearn.base.TransformerMixin): def _more_tags(self): return {'supervised_encoder': True} def transform(self, X, y=None, override_return_df=False): """Perform the transformation to new categorical data. Some encoders behave differently on whether y is given or not. This is mainly due to regularisation in order to avoid overfitting. On training data transform should be called with y, on test data without. Parameters ---------- X : array-like, shape = [n_samples, n_features] y : array-like, shape = [n_samples] or None override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X, y = convert_inputs(X, y, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X, y) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X: pd.DataFrame, y: pd.Series) -> pd.DataFrame: ... def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y) class UnsupervisedTransformerMixin(sklearn.base.TransformerMixin): def transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X = convert_input(X, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X) -> pd.DataFrame: ... class TransformerWithTargetMixin: def _more_tags(self): return {'supervised_encoder': True} def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y)

JaimeArboleda

5eb7a2d6359d680bdadd0534bdb983e712a47f9c

570827e6b48737d0c9aece8aca31edd6da02c1b2

Well, my gut feeling is that it's not worth it: it will still be a very big effort and it will only fix one subtle bug (which doesn't make anything crash) and there will still be some inconsistencies. Let me try to explain... * **Current Situation**: If the `category_encoder` is inside a composed `sklearn` transfomer, which delegates the `get_feature_names_out` to every inner estimator, there is an Exception when calling `get_feature_names_out`, because the signature is not valid. The error is `TypeError: get_feature_names_out() takes 1 positional argument but 2 were given`. * **Super simple Fix**: By changing the signature but not doing anything with the input parameter, we remove this `Exception` and, making using of the attribute `feature_names_out_`, we get a right answer in most situations. I mean, will be right whenever the `category_encoder` is fitted and transformed alone, or whenever is included in a composed transformer but in the first layer. However, if the `category_encoder` is not in the first layer of a composed transformer, and if `set_output` is the default, then the answer will be wrong, because when fitted the previous transformer of the row will output a numpy array, and the `category_encoder` will take ["0", "1", ...] as the original column names. This is the only *issue* that will remain with this super simple fix (and the solution is as easy as setting `set_output=pandas`). * **More thorough fix**: If we modify every encoder by adding a treatment of the input_features in order to generate the column names (which requieres some work for some of the encoders), the only added value will be removing the wrong answer in the mentioned case (when the encoder is part of a composed transformer and has another `sklearn` transfomer doing something before). My point is that, even in this situation, there will be some inconsistences that are related to the fact of assuming that the input is a pandas dataframe. One of those is the fact that some parameters of the encoders depend on column names. For example, you can specify the mapping for an `OrdinalEncoder` using this: ```python OrdinalEncoder(mapping=[{'col': 'col_a', 'mapping': {'a': 2, 'b': 1, 'c': 3}}]) ``` This code will crash in the same situation as before: if the `OrdinalEncoder` is part o a pipeline and there is another `sklearn` transformer before, the columns received during `fit` won't correspond to the columns specified in the `mapping` parameter (unless `set_output=pandas`). You can take a look at this issue [here](https://colab.research.google.com/drive/1lJJvXyAOoG9MBtNiH2qSTrqbzZtEPTeB#scrollTo=nCJtM_1keNTD&line=2&uniqifier=1). That's why I was saying that if we wish to be fully aligned with `sklearn`, there will be more things to do (in particular, for example, no column names should be used in any parameter of any encoder). So I feel more inclined to say "all or nothing": if we want full compatibility with `sklearn`, that's fine, but it will be a breaking change a lots of work, and otherwise, I think the *super simple fix* is good enough. Well, that's my opinion...

JaimeArboleda

scikit-learn-contrib/category_encoders

(WIP) Partial fix for getting feature names out

I think this is a partial fix for this opened issue: https://github.com/scikit-learn-contrib/category_encoders/issues/395 It remains to check the behaviour of other estimators that are not ONE_TO_ONE. Please, let me know if you like the work in progress and I will try to continue.

2023-02-23 13:33:41+00:00

2023-03-13 11:48:24+00:00

category_encoders/utils.py

"""A collection of shared utilities for all encoders, not intended for external use.""" from abc import abstractmethod from enum import Enum, auto import warnings import pandas as pd import numpy as np import sklearn.base from sklearn.base import BaseEstimator, TransformerMixin from sklearn.exceptions import NotFittedError from typing import Dict, List, Optional, Union from scipy.sparse import csr_matrix __author__ = 'willmcginnis' def convert_cols_to_list(cols): if isinstance(cols, pd.Series): return cols.tolist() elif isinstance(cols, np.ndarray): return cols.tolist() elif np.isscalar(cols): return [cols] elif isinstance(cols, set): return list(cols) elif isinstance(cols, tuple): return list(cols) elif pd.api.types.is_categorical_dtype(cols): return cols.astype(object).tolist() return cols def get_obj_cols(df): """ Returns names of 'object' columns in the DataFrame. """ obj_cols = [] for idx, dt in enumerate(df.dtypes): if dt == 'object' or is_category(dt): obj_cols.append(df.columns.values[idx]) if not obj_cols: print("Warning: No categorical columns found. Calling 'transform' will only return input data.") return obj_cols def is_category(dtype): return pd.api.types.is_categorical_dtype(dtype) def convert_inputs(X, y, columns=None, index=None, deep=False): """ Unite arraylike `X` and vectorlike `y` into a DataFrame and Series. If both are pandas types already, raises an error if their indexes do not match. If one is pandas, the returns will share that index. If neither is pandas, a default index will be used, unless `index` is passed. Parameters ---------- X: arraylike y: listlike columns: listlike Specifies column names to use for `X`. Ignored if `X` is already a dataframe. If `None`, use the default pandas column names. index: listlike The index to use, if neither `X` nor `y` is a pandas type. (If one has an index, then this has no effect.) If `None`, use the default pandas index. deep: bool Whether to deep-copy `X`. """ X_alt_index = y.index if isinstance(y, pd.Series) else index X = convert_input(X, columns=columns, deep=deep, index=X_alt_index) if y is not None: y = convert_input_vector(y, index=X.index) # N.B.: If either was already pandas, it keeps its index. if any(X.index != y.index): msg = "`X` and `y` both have indexes, but they do not match. If you are shuffling your input data on " \ "purpose (e.g. via permutation_test_score) use np arrays instead of data frames / series" raise ValueError(msg) if X.shape[0] != y.shape[0]: raise ValueError("The length of X is " + str(X.shape[0]) + " but length of y is " + str(y.shape[0]) + ".") return X, y def convert_input(X, columns=None, deep=False, index=None): """ Unite data into a DataFrame. Objects that do not contain column names take the names from the argument. Optionally perform deep copy of the data. """ if not isinstance(X, pd.DataFrame): if isinstance(X, pd.Series): X = pd.DataFrame(X, copy=deep) else: if columns is not None and np.size(X,1) != len(columns): raise ValueError('The count of the column names does not correspond to the count of the columns') if isinstance(X, list): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) # lists are always copied, but for consistency, we still pass the argument elif isinstance(X, (np.generic, np.ndarray)): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) elif isinstance(X, csr_matrix): X = pd.DataFrame(X.todense(), columns=columns, copy=deep, index=index) else: raise ValueError(f'Unexpected input type: {type(X)}') elif deep: X = X.copy(deep=True) return X def convert_input_vector(y, index): """ Unite target data type into a Series. If the target is a Series or a DataFrame, we preserve its index. But if the target does not contain index attribute, we use the index from the argument. """ if y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') if isinstance(y, pd.Series): return y elif isinstance(y, np.ndarray): if len(np.shape(y))==1: # vector return pd.Series(y, name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[0]==1: # single row in a matrix return pd.Series(y[0, :], name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[1]==1: # single column in a matrix return pd.Series(y[:, 0], name='target', index=index) else: raise ValueError(f'Unexpected input shape: {np.shape(y)}') elif np.isscalar(y): return pd.Series([y], name='target', index=index) elif isinstance(y, list): if len(y)==0: # empty list return pd.Series(y, name='target', index=index, dtype=float) elif len(y)>0 and not isinstance(y[0], list): # vector return pd.Series(y, name='target', index=index) elif len(y)>0 and isinstance(y[0], list) and len(y[0])==1: # single row in a matrix flatten = lambda y: [item for sublist in y for item in sublist] return pd.Series(flatten(y), name='target', index=index) elif len(y)==1 and len(y[0])==0 and isinstance(y[0], list): # single empty column in a matrix return pd.Series(y[0], name='target', index=index, dtype=float) elif len(y)==1 and isinstance(y[0], list): # single column in a matrix return pd.Series(y[0], name='target', index=index, dtype=type(y[0][0])) else: raise ValueError('Unexpected input shape') elif isinstance(y, pd.DataFrame): if len(list(y))==0: # empty DataFrame return pd.Series(name='target', index=index, dtype=float) if len(list(y))==1: # a single column return y.iloc[:, 0] else: raise ValueError(f'Unexpected input shape: {y.shape}') else: return pd.Series(y, name='target', index=index) # this covers tuples and other directly convertible types def get_generated_cols(X_original, X_transformed, to_transform): """ Returns a list of the generated/transformed columns. Arguments: X_original: df the original (input) DataFrame. X_transformed: df the transformed (current) DataFrame. to_transform: [str] a list of columns that were transformed (as in the original DataFrame), commonly self.feature_names_in. Output: a list of columns that were transformed (as in the current DataFrame). """ original_cols = list(X_original.columns) if len(to_transform) > 0: [original_cols.remove(c) for c in to_transform] current_cols = list(X_transformed.columns) if len(original_cols) > 0: [current_cols.remove(c) for c in original_cols] return current_cols def flatten_reverse_dict(d): sep = "___" [flat_dict] = pd.json_normalize(d, sep=sep).to_dict(orient='records') reversed_flat_dict = {v: tuple(k.split(sep)) for k, v in flat_dict.items()} return reversed_flat_dict class EncodingRelation(Enum): # one input feature get encoded into one output feature ONE_TO_ONE = auto() # one input feature get encoded into as many output features as it has distinct values ONE_TO_N_UNIQUE = auto() # one input feature get encoded into m output features that are not the number of distinct values ONE_TO_M = auto() # all N input features are encoded into M output features. # The encoding is done globally on all the input not on a per-feature basis N_TO_M = auto() def get_docstring_output_shape(in_out_relation: EncodingRelation): if in_out_relation == EncodingRelation.ONE_TO_ONE: return "n_features" elif in_out_relation == EncodingRelation.ONE_TO_N_UNIQUE: return "n_features * respective cardinality" elif in_out_relation == EncodingRelation.ONE_TO_M: return "M features (n_features < M)" elif in_out_relation == EncodingRelation.N_TO_M: return "M features (M can be anything)" class BaseEncoder(BaseEstimator): _dim: Optional[int] cols: List[str] use_default_cols: bool handle_missing: str handle_unknown: str verbose: int drop_invariant: bool invariant_cols: List[str] = [] return_df: bool supervised: bool encoding_relation: EncodingRelation INVARIANCE_THRESHOLD = 10e-5 # columns with variance less than this will be considered constant / invariant def __init__(self, verbose=0, cols=None, drop_invariant=False, return_df=True, handle_unknown='value', handle_missing='value', **kwargs): """ Parameters ---------- verbose: int integer indicating verbosity of output. 0 for none. cols: list a list of columns to encode, if None, all string and categorical columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform and inverse transform (otherwise it will be a numpy array). handle_missing: str how to handle missing values at fit time. Options are 'error', 'return_nan', and 'value'. Default 'value', which treat NaNs as a countable category at fit time. handle_unknown: str, int or dict of {column : option, ...}. how to handle unknown labels at transform time. Options are 'error' 'return_nan', 'value' and int. Defaults to None which uses NaN behaviour specified at fit time. Passing an int will fill with this int value. kwargs: dict. additional encoder specific parameters like regularisation. """ self.return_df = return_df self.drop_invariant = drop_invariant self.invariant_cols = [] self.verbose = verbose self.use_default_cols = cols is None # if True, even a repeated call of fit() will select string columns from X self.cols = cols # note that cols are only the columns to be encoded, feature_names_in_ are all columns self.mapping = None self.handle_unknown = handle_unknown self.handle_missing = handle_missing self._dim = None def fit(self, X, y=None, **kwargs): """Fits the encoder according to X and y. Parameters ---------- X : array-like, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. y : array-like, shape = [n_samples] Target values. Returns ------- self : encoder Returns self. """ self._check_fit_inputs(X, y) X, y = convert_inputs(X, y) self.feature_names_in_ = X.columns.tolist() self.n_features_in_ = len(self.feature_names_in_) self._dim = X.shape[1] self._determine_fit_columns(X) if not set(self.cols).issubset(X.columns): raise ValueError('X does not contain the columns listed in cols') if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') self._fit(X, y, **kwargs) # for finding invariant columns transform without y (as is done on the test set) X_transformed = self.transform(X, override_return_df=True) self.feature_names_out_ = X_transformed.columns.tolist() # drop all output columns with 0 variance. if self.drop_invariant: generated_cols = get_generated_cols(X, X_transformed, self.cols) self.invariant_cols = [x for x in generated_cols if X_transformed[x].var() <= self.INVARIANCE_THRESHOLD] self.feature_names_out_ = [x for x in self.feature_names_out_ if x not in self.invariant_cols] return self def _check_fit_inputs(self, X, y): if self._get_tags().get('supervised_encoder') and y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') def _check_transform_inputs(self, X): if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') if self._dim is None: raise NotFittedError('Must train encoder before it can be used to transform data.') # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') def _drop_invariants(self, X: pd.DataFrame, override_return_df: bool) -> Union[np.ndarray, pd.DataFrame]: if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.values def _determine_fit_columns(self, X: pd.DataFrame) -> None: """ Determine columns used by encoder. Note that the implementation also deals with re-fitting the same encoder object with different columns. :param X: input data frame :return: none, sets self.cols as a side effect """ # if columns aren't passed, just use every string column if self.use_default_cols: self.cols = get_obj_cols(X) else: self.cols = convert_cols_to_list(self.cols) def get_feature_names(self) -> List[str]: warnings.warn("`get_feature_names` is deprecated in all of sklearn. Use `get_feature_names_out` instead.", category=FutureWarning) return self.get_feature_names_out() def get_feature_names_out(self) -> List[str]: """ Returns the names of all transformed / added columns. Note that in sklearn the get_feature_names_out function takes the feature_names_in as an argument and determines the output feature names using the input. A fit is usually not necessary and if so a NotFittedError is raised. We just require a fit all the time and return the fitted output columns. Returns ------- feature_names: list A list with all feature names transformed or added. Note: potentially dropped features (because the feature is constant/invariant) are not included! """ out_feats = getattr(self, "feature_names_out_", None) if not isinstance(out_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return out_feats def get_feature_names_in(self) -> List[str]: """ Returns the names of all input columns present when fitting. These columns are necessary for the transform step. """ in_feats = getattr(self, "feature_names_in_", None) if not isinstance(in_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return in_feats @abstractmethod def _fit(self, X: pd.DataFrame, y: Optional[pd.Series], **kwargs): ... class SupervisedTransformerMixin(sklearn.base.TransformerMixin): def _more_tags(self): return {'supervised_encoder': True} def transform(self, X, y=None, override_return_df=False): """Perform the transformation to new categorical data. Some encoders behave differently on whether y is given or not. This is mainly due to regularisation in order to avoid overfitting. On training data transform should be called with y, on test data without. Parameters ---------- X : array-like, shape = [n_samples, n_features] y : array-like, shape = [n_samples] or None override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X, y = convert_inputs(X, y, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X, y) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X: pd.DataFrame, y: pd.Series) -> pd.DataFrame: ... def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y) class UnsupervisedTransformerMixin(sklearn.base.TransformerMixin): def transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X = convert_input(X, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X) -> pd.DataFrame: ... class TransformerWithTargetMixin: def _more_tags(self): return {'supervised_encoder': True} def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y)

"""A collection of shared utilities for all encoders, not intended for external use.""" from abc import abstractmethod from enum import Enum, auto import warnings import pandas as pd import numpy as np import sklearn.base from sklearn.base import BaseEstimator, TransformerMixin from sklearn.exceptions import NotFittedError from typing import Dict, List, Optional, Union from scipy.sparse import csr_matrix __author__ = 'willmcginnis' def convert_cols_to_list(cols): if isinstance(cols, pd.Series): return cols.tolist() elif isinstance(cols, np.ndarray): return cols.tolist() elif np.isscalar(cols): return [cols] elif isinstance(cols, set): return list(cols) elif isinstance(cols, tuple): return list(cols) elif pd.api.types.is_categorical_dtype(cols): return cols.astype(object).tolist() return cols def get_obj_cols(df): """ Returns names of 'object' columns in the DataFrame. """ obj_cols = [] for idx, dt in enumerate(df.dtypes): if dt == 'object' or is_category(dt): obj_cols.append(df.columns.values[idx]) if not obj_cols: print("Warning: No categorical columns found. Calling 'transform' will only return input data.") return obj_cols def is_category(dtype): return pd.api.types.is_categorical_dtype(dtype) def convert_inputs(X, y, columns=None, index=None, deep=False): """ Unite arraylike `X` and vectorlike `y` into a DataFrame and Series. If both are pandas types already, raises an error if their indexes do not match. If one is pandas, the returns will share that index. If neither is pandas, a default index will be used, unless `index` is passed. Parameters ---------- X: arraylike y: listlike columns: listlike Specifies column names to use for `X`. Ignored if `X` is already a dataframe. If `None`, use the default pandas column names. index: listlike The index to use, if neither `X` nor `y` is a pandas type. (If one has an index, then this has no effect.) If `None`, use the default pandas index. deep: bool Whether to deep-copy `X`. """ X_alt_index = y.index if isinstance(y, pd.Series) else index X = convert_input(X, columns=columns, deep=deep, index=X_alt_index) if y is not None: y = convert_input_vector(y, index=X.index) # N.B.: If either was already pandas, it keeps its index. if any(X.index != y.index): msg = "`X` and `y` both have indexes, but they do not match. If you are shuffling your input data on " \ "purpose (e.g. via permutation_test_score) use np arrays instead of data frames / series" raise ValueError(msg) if X.shape[0] != y.shape[0]: raise ValueError("The length of X is " + str(X.shape[0]) + " but length of y is " + str(y.shape[0]) + ".") return X, y def convert_input(X, columns=None, deep=False, index=None): """ Unite data into a DataFrame. Objects that do not contain column names take the names from the argument. Optionally perform deep copy of the data. """ if not isinstance(X, pd.DataFrame): if isinstance(X, pd.Series): X = pd.DataFrame(X, copy=deep) else: if columns is not None and np.size(X,1) != len(columns): raise ValueError('The count of the column names does not correspond to the count of the columns') if isinstance(X, list): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) # lists are always copied, but for consistency, we still pass the argument elif isinstance(X, (np.generic, np.ndarray)): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) elif isinstance(X, csr_matrix): X = pd.DataFrame(X.todense(), columns=columns, copy=deep, index=index) else: raise ValueError(f'Unexpected input type: {type(X)}') elif deep: X = X.copy(deep=True) return X def convert_input_vector(y, index): """ Unite target data type into a Series. If the target is a Series or a DataFrame, we preserve its index. But if the target does not contain index attribute, we use the index from the argument. """ if y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') if isinstance(y, pd.Series): return y elif isinstance(y, np.ndarray): if len(np.shape(y))==1: # vector return pd.Series(y, name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[0]==1: # single row in a matrix return pd.Series(y[0, :], name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[1]==1: # single column in a matrix return pd.Series(y[:, 0], name='target', index=index) else: raise ValueError(f'Unexpected input shape: {np.shape(y)}') elif np.isscalar(y): return pd.Series([y], name='target', index=index) elif isinstance(y, list): if len(y)==0: # empty list return pd.Series(y, name='target', index=index, dtype=float) elif len(y)>0 and not isinstance(y[0], list): # vector return pd.Series(y, name='target', index=index) elif len(y)>0 and isinstance(y[0], list) and len(y[0])==1: # single row in a matrix flatten = lambda y: [item for sublist in y for item in sublist] return pd.Series(flatten(y), name='target', index=index) elif len(y)==1 and len(y[0])==0 and isinstance(y[0], list): # single empty column in a matrix return pd.Series(y[0], name='target', index=index, dtype=float) elif len(y)==1 and isinstance(y[0], list): # single column in a matrix return pd.Series(y[0], name='target', index=index, dtype=type(y[0][0])) else: raise ValueError('Unexpected input shape') elif isinstance(y, pd.DataFrame): if len(list(y))==0: # empty DataFrame return pd.Series(name='target', index=index, dtype=float) if len(list(y))==1: # a single column return y.iloc[:, 0] else: raise ValueError(f'Unexpected input shape: {y.shape}') else: return pd.Series(y, name='target', index=index) # this covers tuples and other directly convertible types def get_generated_cols(X_original, X_transformed, to_transform): """ Returns a list of the generated/transformed columns. Arguments: X_original: df the original (input) DataFrame. X_transformed: df the transformed (current) DataFrame. to_transform: [str] a list of columns that were transformed (as in the original DataFrame), commonly self.feature_names_in. Output: a list of columns that were transformed (as in the current DataFrame). """ original_cols = list(X_original.columns) if len(to_transform) > 0: [original_cols.remove(c) for c in to_transform] current_cols = list(X_transformed.columns) if len(original_cols) > 0: [current_cols.remove(c) for c in original_cols] return current_cols def flatten_reverse_dict(d): sep = "___" [flat_dict] = pd.json_normalize(d, sep=sep).to_dict(orient='records') reversed_flat_dict = {v: tuple(k.split(sep)) for k, v in flat_dict.items()} return reversed_flat_dict class EncodingRelation(Enum): # one input feature get encoded into one output feature ONE_TO_ONE = auto() # one input feature get encoded into as many output features as it has distinct values ONE_TO_N_UNIQUE = auto() # one input feature get encoded into m output features that are not the number of distinct values ONE_TO_M = auto() # all N input features are encoded into M output features. # The encoding is done globally on all the input not on a per-feature basis N_TO_M = auto() def get_docstring_output_shape(in_out_relation: EncodingRelation): if in_out_relation == EncodingRelation.ONE_TO_ONE: return "n_features" elif in_out_relation == EncodingRelation.ONE_TO_N_UNIQUE: return "n_features * respective cardinality" elif in_out_relation == EncodingRelation.ONE_TO_M: return "M features (n_features < M)" elif in_out_relation == EncodingRelation.N_TO_M: return "M features (M can be anything)" class BaseEncoder(BaseEstimator): _dim: Optional[int] cols: List[str] use_default_cols: bool handle_missing: str handle_unknown: str verbose: int drop_invariant: bool invariant_cols: List[str] = [] return_df: bool supervised: bool encoding_relation: EncodingRelation INVARIANCE_THRESHOLD = 10e-5 # columns with variance less than this will be considered constant / invariant def __init__(self, verbose=0, cols=None, drop_invariant=False, return_df=True, handle_unknown='value', handle_missing='value', **kwargs): """ Parameters ---------- verbose: int integer indicating verbosity of output. 0 for none. cols: list a list of columns to encode, if None, all string and categorical columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform and inverse transform (otherwise it will be a numpy array). handle_missing: str how to handle missing values at fit time. Options are 'error', 'return_nan', and 'value'. Default 'value', which treat NaNs as a countable category at fit time. handle_unknown: str, int or dict of {column : option, ...}. how to handle unknown labels at transform time. Options are 'error' 'return_nan', 'value' and int. Defaults to None which uses NaN behaviour specified at fit time. Passing an int will fill with this int value. kwargs: dict. additional encoder specific parameters like regularisation. """ self.return_df = return_df self.drop_invariant = drop_invariant self.invariant_cols = [] self.verbose = verbose self.use_default_cols = cols is None # if True, even a repeated call of fit() will select string columns from X self.cols = cols # note that cols are only the columns to be encoded, feature_names_in_ are all columns self.mapping = None self.handle_unknown = handle_unknown self.handle_missing = handle_missing self._dim = None def fit(self, X, y=None, **kwargs): """Fits the encoder according to X and y. Parameters ---------- X : array-like, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. y : array-like, shape = [n_samples] Target values. Returns ------- self : encoder Returns self. """ self._check_fit_inputs(X, y) X, y = convert_inputs(X, y) self.feature_names_in_ = X.columns.tolist() self.n_features_in_ = len(self.feature_names_in_) self._dim = X.shape[1] self._determine_fit_columns(X) if not set(self.cols).issubset(X.columns): raise ValueError('X does not contain the columns listed in cols') if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') self._fit(X, y, **kwargs) # for finding invariant columns transform without y (as is done on the test set) X_transformed = self.transform(X, override_return_df=True) self.feature_names_out_ = X_transformed.columns.tolist() # drop all output columns with 0 variance. if self.drop_invariant: generated_cols = get_generated_cols(X, X_transformed, self.cols) self.invariant_cols = [x for x in generated_cols if X_transformed[x].var() <= self.INVARIANCE_THRESHOLD] self.feature_names_out_ = [x for x in self.feature_names_out_ if x not in self.invariant_cols] return self def _check_fit_inputs(self, X, y): if self._get_tags().get('supervised_encoder') and y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') def _check_transform_inputs(self, X): if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') if self._dim is None: raise NotFittedError('Must train encoder before it can be used to transform data.') # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') def _drop_invariants(self, X: pd.DataFrame, override_return_df: bool) -> Union[np.ndarray, pd.DataFrame]: if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.values def _determine_fit_columns(self, X: pd.DataFrame) -> None: """ Determine columns used by encoder. Note that the implementation also deals with re-fitting the same encoder object with different columns. :param X: input data frame :return: none, sets self.cols as a side effect """ # if columns aren't passed, just use every string column if self.use_default_cols: self.cols = get_obj_cols(X) else: self.cols = convert_cols_to_list(self.cols) def get_feature_names(self) -> List[str]: warnings.warn("`get_feature_names` is deprecated in all of sklearn. Use `get_feature_names_out` instead.", category=FutureWarning) return self.get_feature_names_out() def get_feature_names_out(self, input_features=None) -> np.ndarray: """ Returns the names of all transformed / added columns. Note that in sklearn the get_feature_names_out function takes the feature_names_in as an argument and determines the output feature names using the input. A fit is usually not necessary and if so a NotFittedError is raised. We just require a fit all the time and return the fitted output columns. Returns ------- feature_names: np.ndarray A numpy array with all feature names transformed or added. Note: potentially dropped features (because the feature is constant/invariant) are not included! """ out_feats = getattr(self, "feature_names_out_", None) if not isinstance(out_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return np.array(out_feats, dtype=object) def get_feature_names_in(self) -> List[str]: """ Returns the names of all input columns present when fitting. These columns are necessary for the transform step. """ in_feats = getattr(self, "feature_names_in_", None) if not isinstance(in_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return in_feats @abstractmethod def _fit(self, X: pd.DataFrame, y: Optional[pd.Series], **kwargs): ... class SupervisedTransformerMixin(sklearn.base.TransformerMixin): def _more_tags(self): return {'supervised_encoder': True} def transform(self, X, y=None, override_return_df=False): """Perform the transformation to new categorical data. Some encoders behave differently on whether y is given or not. This is mainly due to regularisation in order to avoid overfitting. On training data transform should be called with y, on test data without. Parameters ---------- X : array-like, shape = [n_samples, n_features] y : array-like, shape = [n_samples] or None override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X, y = convert_inputs(X, y, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X, y) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X: pd.DataFrame, y: pd.Series) -> pd.DataFrame: ... def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y) class UnsupervisedTransformerMixin(sklearn.base.TransformerMixin): def transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X = convert_input(X, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X) -> pd.DataFrame: ... class TransformerWithTargetMixin: def _more_tags(self): return {'supervised_encoder': True} def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y)

JaimeArboleda

5eb7a2d6359d680bdadd0534bdb983e712a47f9c

570827e6b48737d0c9aece8aca31edd6da02c1b2

Sorry that I did not answer you question properly: regarding the versions, I would not specify the dependency of 1.2 in the requirements or the setup, because with the simple fix there the remaining bugs are not, in my opinion, a big issue. If you want, I could refactor the tests so that I don't require the version for running all tests, but only for some of the asserts (the one that will fail because the column names won't match what `get_feature_names_out` outputs). This way it will be more clear what does not work if not setting the output as pandas.

JaimeArboleda

scikit-learn-contrib/category_encoders

(WIP) Partial fix for getting feature names out

I think this is a partial fix for this opened issue: https://github.com/scikit-learn-contrib/category_encoders/issues/395 It remains to check the behaviour of other estimators that are not ONE_TO_ONE. Please, let me know if you like the work in progress and I will try to continue.

2023-02-23 13:33:41+00:00

2023-03-13 11:48:24+00:00

category_encoders/utils.py

"""A collection of shared utilities for all encoders, not intended for external use.""" from abc import abstractmethod from enum import Enum, auto import warnings import pandas as pd import numpy as np import sklearn.base from sklearn.base import BaseEstimator, TransformerMixin from sklearn.exceptions import NotFittedError from typing import Dict, List, Optional, Union from scipy.sparse import csr_matrix __author__ = 'willmcginnis' def convert_cols_to_list(cols): if isinstance(cols, pd.Series): return cols.tolist() elif isinstance(cols, np.ndarray): return cols.tolist() elif np.isscalar(cols): return [cols] elif isinstance(cols, set): return list(cols) elif isinstance(cols, tuple): return list(cols) elif pd.api.types.is_categorical_dtype(cols): return cols.astype(object).tolist() return cols def get_obj_cols(df): """ Returns names of 'object' columns in the DataFrame. """ obj_cols = [] for idx, dt in enumerate(df.dtypes): if dt == 'object' or is_category(dt): obj_cols.append(df.columns.values[idx]) if not obj_cols: print("Warning: No categorical columns found. Calling 'transform' will only return input data.") return obj_cols def is_category(dtype): return pd.api.types.is_categorical_dtype(dtype) def convert_inputs(X, y, columns=None, index=None, deep=False): """ Unite arraylike `X` and vectorlike `y` into a DataFrame and Series. If both are pandas types already, raises an error if their indexes do not match. If one is pandas, the returns will share that index. If neither is pandas, a default index will be used, unless `index` is passed. Parameters ---------- X: arraylike y: listlike columns: listlike Specifies column names to use for `X`. Ignored if `X` is already a dataframe. If `None`, use the default pandas column names. index: listlike The index to use, if neither `X` nor `y` is a pandas type. (If one has an index, then this has no effect.) If `None`, use the default pandas index. deep: bool Whether to deep-copy `X`. """ X_alt_index = y.index if isinstance(y, pd.Series) else index X = convert_input(X, columns=columns, deep=deep, index=X_alt_index) if y is not None: y = convert_input_vector(y, index=X.index) # N.B.: If either was already pandas, it keeps its index. if any(X.index != y.index): msg = "`X` and `y` both have indexes, but they do not match. If you are shuffling your input data on " \ "purpose (e.g. via permutation_test_score) use np arrays instead of data frames / series" raise ValueError(msg) if X.shape[0] != y.shape[0]: raise ValueError("The length of X is " + str(X.shape[0]) + " but length of y is " + str(y.shape[0]) + ".") return X, y def convert_input(X, columns=None, deep=False, index=None): """ Unite data into a DataFrame. Objects that do not contain column names take the names from the argument. Optionally perform deep copy of the data. """ if not isinstance(X, pd.DataFrame): if isinstance(X, pd.Series): X = pd.DataFrame(X, copy=deep) else: if columns is not None and np.size(X,1) != len(columns): raise ValueError('The count of the column names does not correspond to the count of the columns') if isinstance(X, list): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) # lists are always copied, but for consistency, we still pass the argument elif isinstance(X, (np.generic, np.ndarray)): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) elif isinstance(X, csr_matrix): X = pd.DataFrame(X.todense(), columns=columns, copy=deep, index=index) else: raise ValueError(f'Unexpected input type: {type(X)}') elif deep: X = X.copy(deep=True) return X def convert_input_vector(y, index): """ Unite target data type into a Series. If the target is a Series or a DataFrame, we preserve its index. But if the target does not contain index attribute, we use the index from the argument. """ if y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') if isinstance(y, pd.Series): return y elif isinstance(y, np.ndarray): if len(np.shape(y))==1: # vector return pd.Series(y, name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[0]==1: # single row in a matrix return pd.Series(y[0, :], name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[1]==1: # single column in a matrix return pd.Series(y[:, 0], name='target', index=index) else: raise ValueError(f'Unexpected input shape: {np.shape(y)}') elif np.isscalar(y): return pd.Series([y], name='target', index=index) elif isinstance(y, list): if len(y)==0: # empty list return pd.Series(y, name='target', index=index, dtype=float) elif len(y)>0 and not isinstance(y[0], list): # vector return pd.Series(y, name='target', index=index) elif len(y)>0 and isinstance(y[0], list) and len(y[0])==1: # single row in a matrix flatten = lambda y: [item for sublist in y for item in sublist] return pd.Series(flatten(y), name='target', index=index) elif len(y)==1 and len(y[0])==0 and isinstance(y[0], list): # single empty column in a matrix return pd.Series(y[0], name='target', index=index, dtype=float) elif len(y)==1 and isinstance(y[0], list): # single column in a matrix return pd.Series(y[0], name='target', index=index, dtype=type(y[0][0])) else: raise ValueError('Unexpected input shape') elif isinstance(y, pd.DataFrame): if len(list(y))==0: # empty DataFrame return pd.Series(name='target', index=index, dtype=float) if len(list(y))==1: # a single column return y.iloc[:, 0] else: raise ValueError(f'Unexpected input shape: {y.shape}') else: return pd.Series(y, name='target', index=index) # this covers tuples and other directly convertible types def get_generated_cols(X_original, X_transformed, to_transform): """ Returns a list of the generated/transformed columns. Arguments: X_original: df the original (input) DataFrame. X_transformed: df the transformed (current) DataFrame. to_transform: [str] a list of columns that were transformed (as in the original DataFrame), commonly self.feature_names_in. Output: a list of columns that were transformed (as in the current DataFrame). """ original_cols = list(X_original.columns) if len(to_transform) > 0: [original_cols.remove(c) for c in to_transform] current_cols = list(X_transformed.columns) if len(original_cols) > 0: [current_cols.remove(c) for c in original_cols] return current_cols def flatten_reverse_dict(d): sep = "___" [flat_dict] = pd.json_normalize(d, sep=sep).to_dict(orient='records') reversed_flat_dict = {v: tuple(k.split(sep)) for k, v in flat_dict.items()} return reversed_flat_dict class EncodingRelation(Enum): # one input feature get encoded into one output feature ONE_TO_ONE = auto() # one input feature get encoded into as many output features as it has distinct values ONE_TO_N_UNIQUE = auto() # one input feature get encoded into m output features that are not the number of distinct values ONE_TO_M = auto() # all N input features are encoded into M output features. # The encoding is done globally on all the input not on a per-feature basis N_TO_M = auto() def get_docstring_output_shape(in_out_relation: EncodingRelation): if in_out_relation == EncodingRelation.ONE_TO_ONE: return "n_features" elif in_out_relation == EncodingRelation.ONE_TO_N_UNIQUE: return "n_features * respective cardinality" elif in_out_relation == EncodingRelation.ONE_TO_M: return "M features (n_features < M)" elif in_out_relation == EncodingRelation.N_TO_M: return "M features (M can be anything)" class BaseEncoder(BaseEstimator): _dim: Optional[int] cols: List[str] use_default_cols: bool handle_missing: str handle_unknown: str verbose: int drop_invariant: bool invariant_cols: List[str] = [] return_df: bool supervised: bool encoding_relation: EncodingRelation INVARIANCE_THRESHOLD = 10e-5 # columns with variance less than this will be considered constant / invariant def __init__(self, verbose=0, cols=None, drop_invariant=False, return_df=True, handle_unknown='value', handle_missing='value', **kwargs): """ Parameters ---------- verbose: int integer indicating verbosity of output. 0 for none. cols: list a list of columns to encode, if None, all string and categorical columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform and inverse transform (otherwise it will be a numpy array). handle_missing: str how to handle missing values at fit time. Options are 'error', 'return_nan', and 'value'. Default 'value', which treat NaNs as a countable category at fit time. handle_unknown: str, int or dict of {column : option, ...}. how to handle unknown labels at transform time. Options are 'error' 'return_nan', 'value' and int. Defaults to None which uses NaN behaviour specified at fit time. Passing an int will fill with this int value. kwargs: dict. additional encoder specific parameters like regularisation. """ self.return_df = return_df self.drop_invariant = drop_invariant self.invariant_cols = [] self.verbose = verbose self.use_default_cols = cols is None # if True, even a repeated call of fit() will select string columns from X self.cols = cols # note that cols are only the columns to be encoded, feature_names_in_ are all columns self.mapping = None self.handle_unknown = handle_unknown self.handle_missing = handle_missing self._dim = None def fit(self, X, y=None, **kwargs): """Fits the encoder according to X and y. Parameters ---------- X : array-like, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. y : array-like, shape = [n_samples] Target values. Returns ------- self : encoder Returns self. """ self._check_fit_inputs(X, y) X, y = convert_inputs(X, y) self.feature_names_in_ = X.columns.tolist() self.n_features_in_ = len(self.feature_names_in_) self._dim = X.shape[1] self._determine_fit_columns(X) if not set(self.cols).issubset(X.columns): raise ValueError('X does not contain the columns listed in cols') if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') self._fit(X, y, **kwargs) # for finding invariant columns transform without y (as is done on the test set) X_transformed = self.transform(X, override_return_df=True) self.feature_names_out_ = X_transformed.columns.tolist() # drop all output columns with 0 variance. if self.drop_invariant: generated_cols = get_generated_cols(X, X_transformed, self.cols) self.invariant_cols = [x for x in generated_cols if X_transformed[x].var() <= self.INVARIANCE_THRESHOLD] self.feature_names_out_ = [x for x in self.feature_names_out_ if x not in self.invariant_cols] return self def _check_fit_inputs(self, X, y): if self._get_tags().get('supervised_encoder') and y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') def _check_transform_inputs(self, X): if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') if self._dim is None: raise NotFittedError('Must train encoder before it can be used to transform data.') # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') def _drop_invariants(self, X: pd.DataFrame, override_return_df: bool) -> Union[np.ndarray, pd.DataFrame]: if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.values def _determine_fit_columns(self, X: pd.DataFrame) -> None: """ Determine columns used by encoder. Note that the implementation also deals with re-fitting the same encoder object with different columns. :param X: input data frame :return: none, sets self.cols as a side effect """ # if columns aren't passed, just use every string column if self.use_default_cols: self.cols = get_obj_cols(X) else: self.cols = convert_cols_to_list(self.cols) def get_feature_names(self) -> List[str]: warnings.warn("`get_feature_names` is deprecated in all of sklearn. Use `get_feature_names_out` instead.", category=FutureWarning) return self.get_feature_names_out() def get_feature_names_out(self) -> List[str]: """ Returns the names of all transformed / added columns. Note that in sklearn the get_feature_names_out function takes the feature_names_in as an argument and determines the output feature names using the input. A fit is usually not necessary and if so a NotFittedError is raised. We just require a fit all the time and return the fitted output columns. Returns ------- feature_names: list A list with all feature names transformed or added. Note: potentially dropped features (because the feature is constant/invariant) are not included! """ out_feats = getattr(self, "feature_names_out_", None) if not isinstance(out_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return out_feats def get_feature_names_in(self) -> List[str]: """ Returns the names of all input columns present when fitting. These columns are necessary for the transform step. """ in_feats = getattr(self, "feature_names_in_", None) if not isinstance(in_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return in_feats @abstractmethod def _fit(self, X: pd.DataFrame, y: Optional[pd.Series], **kwargs): ... class SupervisedTransformerMixin(sklearn.base.TransformerMixin): def _more_tags(self): return {'supervised_encoder': True} def transform(self, X, y=None, override_return_df=False): """Perform the transformation to new categorical data. Some encoders behave differently on whether y is given or not. This is mainly due to regularisation in order to avoid overfitting. On training data transform should be called with y, on test data without. Parameters ---------- X : array-like, shape = [n_samples, n_features] y : array-like, shape = [n_samples] or None override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X, y = convert_inputs(X, y, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X, y) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X: pd.DataFrame, y: pd.Series) -> pd.DataFrame: ... def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y) class UnsupervisedTransformerMixin(sklearn.base.TransformerMixin): def transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X = convert_input(X, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X) -> pd.DataFrame: ... class TransformerWithTargetMixin: def _more_tags(self): return {'supervised_encoder': True} def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y)

"""A collection of shared utilities for all encoders, not intended for external use.""" from abc import abstractmethod from enum import Enum, auto import warnings import pandas as pd import numpy as np import sklearn.base from sklearn.base import BaseEstimator, TransformerMixin from sklearn.exceptions import NotFittedError from typing import Dict, List, Optional, Union from scipy.sparse import csr_matrix __author__ = 'willmcginnis' def convert_cols_to_list(cols): if isinstance(cols, pd.Series): return cols.tolist() elif isinstance(cols, np.ndarray): return cols.tolist() elif np.isscalar(cols): return [cols] elif isinstance(cols, set): return list(cols) elif isinstance(cols, tuple): return list(cols) elif pd.api.types.is_categorical_dtype(cols): return cols.astype(object).tolist() return cols def get_obj_cols(df): """ Returns names of 'object' columns in the DataFrame. """ obj_cols = [] for idx, dt in enumerate(df.dtypes): if dt == 'object' or is_category(dt): obj_cols.append(df.columns.values[idx]) if not obj_cols: print("Warning: No categorical columns found. Calling 'transform' will only return input data.") return obj_cols def is_category(dtype): return pd.api.types.is_categorical_dtype(dtype) def convert_inputs(X, y, columns=None, index=None, deep=False): """ Unite arraylike `X` and vectorlike `y` into a DataFrame and Series. If both are pandas types already, raises an error if their indexes do not match. If one is pandas, the returns will share that index. If neither is pandas, a default index will be used, unless `index` is passed. Parameters ---------- X: arraylike y: listlike columns: listlike Specifies column names to use for `X`. Ignored if `X` is already a dataframe. If `None`, use the default pandas column names. index: listlike The index to use, if neither `X` nor `y` is a pandas type. (If one has an index, then this has no effect.) If `None`, use the default pandas index. deep: bool Whether to deep-copy `X`. """ X_alt_index = y.index if isinstance(y, pd.Series) else index X = convert_input(X, columns=columns, deep=deep, index=X_alt_index) if y is not None: y = convert_input_vector(y, index=X.index) # N.B.: If either was already pandas, it keeps its index. if any(X.index != y.index): msg = "`X` and `y` both have indexes, but they do not match. If you are shuffling your input data on " \ "purpose (e.g. via permutation_test_score) use np arrays instead of data frames / series" raise ValueError(msg) if X.shape[0] != y.shape[0]: raise ValueError("The length of X is " + str(X.shape[0]) + " but length of y is " + str(y.shape[0]) + ".") return X, y def convert_input(X, columns=None, deep=False, index=None): """ Unite data into a DataFrame. Objects that do not contain column names take the names from the argument. Optionally perform deep copy of the data. """ if not isinstance(X, pd.DataFrame): if isinstance(X, pd.Series): X = pd.DataFrame(X, copy=deep) else: if columns is not None and np.size(X,1) != len(columns): raise ValueError('The count of the column names does not correspond to the count of the columns') if isinstance(X, list): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) # lists are always copied, but for consistency, we still pass the argument elif isinstance(X, (np.generic, np.ndarray)): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) elif isinstance(X, csr_matrix): X = pd.DataFrame(X.todense(), columns=columns, copy=deep, index=index) else: raise ValueError(f'Unexpected input type: {type(X)}') elif deep: X = X.copy(deep=True) return X def convert_input_vector(y, index): """ Unite target data type into a Series. If the target is a Series or a DataFrame, we preserve its index. But if the target does not contain index attribute, we use the index from the argument. """ if y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') if isinstance(y, pd.Series): return y elif isinstance(y, np.ndarray): if len(np.shape(y))==1: # vector return pd.Series(y, name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[0]==1: # single row in a matrix return pd.Series(y[0, :], name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[1]==1: # single column in a matrix return pd.Series(y[:, 0], name='target', index=index) else: raise ValueError(f'Unexpected input shape: {np.shape(y)}') elif np.isscalar(y): return pd.Series([y], name='target', index=index) elif isinstance(y, list): if len(y)==0: # empty list return pd.Series(y, name='target', index=index, dtype=float) elif len(y)>0 and not isinstance(y[0], list): # vector return pd.Series(y, name='target', index=index) elif len(y)>0 and isinstance(y[0], list) and len(y[0])==1: # single row in a matrix flatten = lambda y: [item for sublist in y for item in sublist] return pd.Series(flatten(y), name='target', index=index) elif len(y)==1 and len(y[0])==0 and isinstance(y[0], list): # single empty column in a matrix return pd.Series(y[0], name='target', index=index, dtype=float) elif len(y)==1 and isinstance(y[0], list): # single column in a matrix return pd.Series(y[0], name='target', index=index, dtype=type(y[0][0])) else: raise ValueError('Unexpected input shape') elif isinstance(y, pd.DataFrame): if len(list(y))==0: # empty DataFrame return pd.Series(name='target', index=index, dtype=float) if len(list(y))==1: # a single column return y.iloc[:, 0] else: raise ValueError(f'Unexpected input shape: {y.shape}') else: return pd.Series(y, name='target', index=index) # this covers tuples and other directly convertible types def get_generated_cols(X_original, X_transformed, to_transform): """ Returns a list of the generated/transformed columns. Arguments: X_original: df the original (input) DataFrame. X_transformed: df the transformed (current) DataFrame. to_transform: [str] a list of columns that were transformed (as in the original DataFrame), commonly self.feature_names_in. Output: a list of columns that were transformed (as in the current DataFrame). """ original_cols = list(X_original.columns) if len(to_transform) > 0: [original_cols.remove(c) for c in to_transform] current_cols = list(X_transformed.columns) if len(original_cols) > 0: [current_cols.remove(c) for c in original_cols] return current_cols def flatten_reverse_dict(d): sep = "___" [flat_dict] = pd.json_normalize(d, sep=sep).to_dict(orient='records') reversed_flat_dict = {v: tuple(k.split(sep)) for k, v in flat_dict.items()} return reversed_flat_dict class EncodingRelation(Enum): # one input feature get encoded into one output feature ONE_TO_ONE = auto() # one input feature get encoded into as many output features as it has distinct values ONE_TO_N_UNIQUE = auto() # one input feature get encoded into m output features that are not the number of distinct values ONE_TO_M = auto() # all N input features are encoded into M output features. # The encoding is done globally on all the input not on a per-feature basis N_TO_M = auto() def get_docstring_output_shape(in_out_relation: EncodingRelation): if in_out_relation == EncodingRelation.ONE_TO_ONE: return "n_features" elif in_out_relation == EncodingRelation.ONE_TO_N_UNIQUE: return "n_features * respective cardinality" elif in_out_relation == EncodingRelation.ONE_TO_M: return "M features (n_features < M)" elif in_out_relation == EncodingRelation.N_TO_M: return "M features (M can be anything)" class BaseEncoder(BaseEstimator): _dim: Optional[int] cols: List[str] use_default_cols: bool handle_missing: str handle_unknown: str verbose: int drop_invariant: bool invariant_cols: List[str] = [] return_df: bool supervised: bool encoding_relation: EncodingRelation INVARIANCE_THRESHOLD = 10e-5 # columns with variance less than this will be considered constant / invariant def __init__(self, verbose=0, cols=None, drop_invariant=False, return_df=True, handle_unknown='value', handle_missing='value', **kwargs): """ Parameters ---------- verbose: int integer indicating verbosity of output. 0 for none. cols: list a list of columns to encode, if None, all string and categorical columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform and inverse transform (otherwise it will be a numpy array). handle_missing: str how to handle missing values at fit time. Options are 'error', 'return_nan', and 'value'. Default 'value', which treat NaNs as a countable category at fit time. handle_unknown: str, int or dict of {column : option, ...}. how to handle unknown labels at transform time. Options are 'error' 'return_nan', 'value' and int. Defaults to None which uses NaN behaviour specified at fit time. Passing an int will fill with this int value. kwargs: dict. additional encoder specific parameters like regularisation. """ self.return_df = return_df self.drop_invariant = drop_invariant self.invariant_cols = [] self.verbose = verbose self.use_default_cols = cols is None # if True, even a repeated call of fit() will select string columns from X self.cols = cols # note that cols are only the columns to be encoded, feature_names_in_ are all columns self.mapping = None self.handle_unknown = handle_unknown self.handle_missing = handle_missing self._dim = None def fit(self, X, y=None, **kwargs): """Fits the encoder according to X and y. Parameters ---------- X : array-like, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. y : array-like, shape = [n_samples] Target values. Returns ------- self : encoder Returns self. """ self._check_fit_inputs(X, y) X, y = convert_inputs(X, y) self.feature_names_in_ = X.columns.tolist() self.n_features_in_ = len(self.feature_names_in_) self._dim = X.shape[1] self._determine_fit_columns(X) if not set(self.cols).issubset(X.columns): raise ValueError('X does not contain the columns listed in cols') if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') self._fit(X, y, **kwargs) # for finding invariant columns transform without y (as is done on the test set) X_transformed = self.transform(X, override_return_df=True) self.feature_names_out_ = X_transformed.columns.tolist() # drop all output columns with 0 variance. if self.drop_invariant: generated_cols = get_generated_cols(X, X_transformed, self.cols) self.invariant_cols = [x for x in generated_cols if X_transformed[x].var() <= self.INVARIANCE_THRESHOLD] self.feature_names_out_ = [x for x in self.feature_names_out_ if x not in self.invariant_cols] return self def _check_fit_inputs(self, X, y): if self._get_tags().get('supervised_encoder') and y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') def _check_transform_inputs(self, X): if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') if self._dim is None: raise NotFittedError('Must train encoder before it can be used to transform data.') # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') def _drop_invariants(self, X: pd.DataFrame, override_return_df: bool) -> Union[np.ndarray, pd.DataFrame]: if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.values def _determine_fit_columns(self, X: pd.DataFrame) -> None: """ Determine columns used by encoder. Note that the implementation also deals with re-fitting the same encoder object with different columns. :param X: input data frame :return: none, sets self.cols as a side effect """ # if columns aren't passed, just use every string column if self.use_default_cols: self.cols = get_obj_cols(X) else: self.cols = convert_cols_to_list(self.cols) def get_feature_names(self) -> List[str]: warnings.warn("`get_feature_names` is deprecated in all of sklearn. Use `get_feature_names_out` instead.", category=FutureWarning) return self.get_feature_names_out() def get_feature_names_out(self, input_features=None) -> np.ndarray: """ Returns the names of all transformed / added columns. Note that in sklearn the get_feature_names_out function takes the feature_names_in as an argument and determines the output feature names using the input. A fit is usually not necessary and if so a NotFittedError is raised. We just require a fit all the time and return the fitted output columns. Returns ------- feature_names: np.ndarray A numpy array with all feature names transformed or added. Note: potentially dropped features (because the feature is constant/invariant) are not included! """ out_feats = getattr(self, "feature_names_out_", None) if not isinstance(out_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return np.array(out_feats, dtype=object) def get_feature_names_in(self) -> List[str]: """ Returns the names of all input columns present when fitting. These columns are necessary for the transform step. """ in_feats = getattr(self, "feature_names_in_", None) if not isinstance(in_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return in_feats @abstractmethod def _fit(self, X: pd.DataFrame, y: Optional[pd.Series], **kwargs): ... class SupervisedTransformerMixin(sklearn.base.TransformerMixin): def _more_tags(self): return {'supervised_encoder': True} def transform(self, X, y=None, override_return_df=False): """Perform the transformation to new categorical data. Some encoders behave differently on whether y is given or not. This is mainly due to regularisation in order to avoid overfitting. On training data transform should be called with y, on test data without. Parameters ---------- X : array-like, shape = [n_samples, n_features] y : array-like, shape = [n_samples] or None override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X, y = convert_inputs(X, y, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X, y) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X: pd.DataFrame, y: pd.Series) -> pd.DataFrame: ... def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y) class UnsupervisedTransformerMixin(sklearn.base.TransformerMixin): def transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X = convert_input(X, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X) -> pd.DataFrame: ... class TransformerWithTargetMixin: def _more_tags(self): return {'supervised_encoder': True} def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y)

JaimeArboleda

5eb7a2d6359d680bdadd0534bdb983e712a47f9c

570827e6b48737d0c9aece8aca31edd6da02c1b2

> I've got one more question: If we change all encoders to properly work with the input features, will we get compatibility even without setting `set_output=pandas`? Ok I see, this is not true since the columns to encode are always referenced by name in the `cols` parameter. Then I agree with you that changing `get_feature_names_out` to actually do something with the `input_features` is pretty much useless since it won't fix the issue.

PaulWestenthanner

scikit-learn-contrib/category_encoders

(WIP) Partial fix for getting feature names out

I think this is a partial fix for this opened issue: https://github.com/scikit-learn-contrib/category_encoders/issues/395 It remains to check the behaviour of other estimators that are not ONE_TO_ONE. Please, let me know if you like the work in progress and I will try to continue.

2023-02-23 13:33:41+00:00

2023-03-13 11:48:24+00:00

category_encoders/utils.py

"""A collection of shared utilities for all encoders, not intended for external use.""" from abc import abstractmethod from enum import Enum, auto import warnings import pandas as pd import numpy as np import sklearn.base from sklearn.base import BaseEstimator, TransformerMixin from sklearn.exceptions import NotFittedError from typing import Dict, List, Optional, Union from scipy.sparse import csr_matrix __author__ = 'willmcginnis' def convert_cols_to_list(cols): if isinstance(cols, pd.Series): return cols.tolist() elif isinstance(cols, np.ndarray): return cols.tolist() elif np.isscalar(cols): return [cols] elif isinstance(cols, set): return list(cols) elif isinstance(cols, tuple): return list(cols) elif pd.api.types.is_categorical_dtype(cols): return cols.astype(object).tolist() return cols def get_obj_cols(df): """ Returns names of 'object' columns in the DataFrame. """ obj_cols = [] for idx, dt in enumerate(df.dtypes): if dt == 'object' or is_category(dt): obj_cols.append(df.columns.values[idx]) if not obj_cols: print("Warning: No categorical columns found. Calling 'transform' will only return input data.") return obj_cols def is_category(dtype): return pd.api.types.is_categorical_dtype(dtype) def convert_inputs(X, y, columns=None, index=None, deep=False): """ Unite arraylike `X` and vectorlike `y` into a DataFrame and Series. If both are pandas types already, raises an error if their indexes do not match. If one is pandas, the returns will share that index. If neither is pandas, a default index will be used, unless `index` is passed. Parameters ---------- X: arraylike y: listlike columns: listlike Specifies column names to use for `X`. Ignored if `X` is already a dataframe. If `None`, use the default pandas column names. index: listlike The index to use, if neither `X` nor `y` is a pandas type. (If one has an index, then this has no effect.) If `None`, use the default pandas index. deep: bool Whether to deep-copy `X`. """ X_alt_index = y.index if isinstance(y, pd.Series) else index X = convert_input(X, columns=columns, deep=deep, index=X_alt_index) if y is not None: y = convert_input_vector(y, index=X.index) # N.B.: If either was already pandas, it keeps its index. if any(X.index != y.index): msg = "`X` and `y` both have indexes, but they do not match. If you are shuffling your input data on " \ "purpose (e.g. via permutation_test_score) use np arrays instead of data frames / series" raise ValueError(msg) if X.shape[0] != y.shape[0]: raise ValueError("The length of X is " + str(X.shape[0]) + " but length of y is " + str(y.shape[0]) + ".") return X, y def convert_input(X, columns=None, deep=False, index=None): """ Unite data into a DataFrame. Objects that do not contain column names take the names from the argument. Optionally perform deep copy of the data. """ if not isinstance(X, pd.DataFrame): if isinstance(X, pd.Series): X = pd.DataFrame(X, copy=deep) else: if columns is not None and np.size(X,1) != len(columns): raise ValueError('The count of the column names does not correspond to the count of the columns') if isinstance(X, list): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) # lists are always copied, but for consistency, we still pass the argument elif isinstance(X, (np.generic, np.ndarray)): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) elif isinstance(X, csr_matrix): X = pd.DataFrame(X.todense(), columns=columns, copy=deep, index=index) else: raise ValueError(f'Unexpected input type: {type(X)}') elif deep: X = X.copy(deep=True) return X def convert_input_vector(y, index): """ Unite target data type into a Series. If the target is a Series or a DataFrame, we preserve its index. But if the target does not contain index attribute, we use the index from the argument. """ if y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') if isinstance(y, pd.Series): return y elif isinstance(y, np.ndarray): if len(np.shape(y))==1: # vector return pd.Series(y, name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[0]==1: # single row in a matrix return pd.Series(y[0, :], name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[1]==1: # single column in a matrix return pd.Series(y[:, 0], name='target', index=index) else: raise ValueError(f'Unexpected input shape: {np.shape(y)}') elif np.isscalar(y): return pd.Series([y], name='target', index=index) elif isinstance(y, list): if len(y)==0: # empty list return pd.Series(y, name='target', index=index, dtype=float) elif len(y)>0 and not isinstance(y[0], list): # vector return pd.Series(y, name='target', index=index) elif len(y)>0 and isinstance(y[0], list) and len(y[0])==1: # single row in a matrix flatten = lambda y: [item for sublist in y for item in sublist] return pd.Series(flatten(y), name='target', index=index) elif len(y)==1 and len(y[0])==0 and isinstance(y[0], list): # single empty column in a matrix return pd.Series(y[0], name='target', index=index, dtype=float) elif len(y)==1 and isinstance(y[0], list): # single column in a matrix return pd.Series(y[0], name='target', index=index, dtype=type(y[0][0])) else: raise ValueError('Unexpected input shape') elif isinstance(y, pd.DataFrame): if len(list(y))==0: # empty DataFrame return pd.Series(name='target', index=index, dtype=float) if len(list(y))==1: # a single column return y.iloc[:, 0] else: raise ValueError(f'Unexpected input shape: {y.shape}') else: return pd.Series(y, name='target', index=index) # this covers tuples and other directly convertible types def get_generated_cols(X_original, X_transformed, to_transform): """ Returns a list of the generated/transformed columns. Arguments: X_original: df the original (input) DataFrame. X_transformed: df the transformed (current) DataFrame. to_transform: [str] a list of columns that were transformed (as in the original DataFrame), commonly self.feature_names_in. Output: a list of columns that were transformed (as in the current DataFrame). """ original_cols = list(X_original.columns) if len(to_transform) > 0: [original_cols.remove(c) for c in to_transform] current_cols = list(X_transformed.columns) if len(original_cols) > 0: [current_cols.remove(c) for c in original_cols] return current_cols def flatten_reverse_dict(d): sep = "___" [flat_dict] = pd.json_normalize(d, sep=sep).to_dict(orient='records') reversed_flat_dict = {v: tuple(k.split(sep)) for k, v in flat_dict.items()} return reversed_flat_dict class EncodingRelation(Enum): # one input feature get encoded into one output feature ONE_TO_ONE = auto() # one input feature get encoded into as many output features as it has distinct values ONE_TO_N_UNIQUE = auto() # one input feature get encoded into m output features that are not the number of distinct values ONE_TO_M = auto() # all N input features are encoded into M output features. # The encoding is done globally on all the input not on a per-feature basis N_TO_M = auto() def get_docstring_output_shape(in_out_relation: EncodingRelation): if in_out_relation == EncodingRelation.ONE_TO_ONE: return "n_features" elif in_out_relation == EncodingRelation.ONE_TO_N_UNIQUE: return "n_features * respective cardinality" elif in_out_relation == EncodingRelation.ONE_TO_M: return "M features (n_features < M)" elif in_out_relation == EncodingRelation.N_TO_M: return "M features (M can be anything)" class BaseEncoder(BaseEstimator): _dim: Optional[int] cols: List[str] use_default_cols: bool handle_missing: str handle_unknown: str verbose: int drop_invariant: bool invariant_cols: List[str] = [] return_df: bool supervised: bool encoding_relation: EncodingRelation INVARIANCE_THRESHOLD = 10e-5 # columns with variance less than this will be considered constant / invariant def __init__(self, verbose=0, cols=None, drop_invariant=False, return_df=True, handle_unknown='value', handle_missing='value', **kwargs): """ Parameters ---------- verbose: int integer indicating verbosity of output. 0 for none. cols: list a list of columns to encode, if None, all string and categorical columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform and inverse transform (otherwise it will be a numpy array). handle_missing: str how to handle missing values at fit time. Options are 'error', 'return_nan', and 'value'. Default 'value', which treat NaNs as a countable category at fit time. handle_unknown: str, int or dict of {column : option, ...}. how to handle unknown labels at transform time. Options are 'error' 'return_nan', 'value' and int. Defaults to None which uses NaN behaviour specified at fit time. Passing an int will fill with this int value. kwargs: dict. additional encoder specific parameters like regularisation. """ self.return_df = return_df self.drop_invariant = drop_invariant self.invariant_cols = [] self.verbose = verbose self.use_default_cols = cols is None # if True, even a repeated call of fit() will select string columns from X self.cols = cols # note that cols are only the columns to be encoded, feature_names_in_ are all columns self.mapping = None self.handle_unknown = handle_unknown self.handle_missing = handle_missing self._dim = None def fit(self, X, y=None, **kwargs): """Fits the encoder according to X and y. Parameters ---------- X : array-like, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. y : array-like, shape = [n_samples] Target values. Returns ------- self : encoder Returns self. """ self._check_fit_inputs(X, y) X, y = convert_inputs(X, y) self.feature_names_in_ = X.columns.tolist() self.n_features_in_ = len(self.feature_names_in_) self._dim = X.shape[1] self._determine_fit_columns(X) if not set(self.cols).issubset(X.columns): raise ValueError('X does not contain the columns listed in cols') if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') self._fit(X, y, **kwargs) # for finding invariant columns transform without y (as is done on the test set) X_transformed = self.transform(X, override_return_df=True) self.feature_names_out_ = X_transformed.columns.tolist() # drop all output columns with 0 variance. if self.drop_invariant: generated_cols = get_generated_cols(X, X_transformed, self.cols) self.invariant_cols = [x for x in generated_cols if X_transformed[x].var() <= self.INVARIANCE_THRESHOLD] self.feature_names_out_ = [x for x in self.feature_names_out_ if x not in self.invariant_cols] return self def _check_fit_inputs(self, X, y): if self._get_tags().get('supervised_encoder') and y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') def _check_transform_inputs(self, X): if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') if self._dim is None: raise NotFittedError('Must train encoder before it can be used to transform data.') # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') def _drop_invariants(self, X: pd.DataFrame, override_return_df: bool) -> Union[np.ndarray, pd.DataFrame]: if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.values def _determine_fit_columns(self, X: pd.DataFrame) -> None: """ Determine columns used by encoder. Note that the implementation also deals with re-fitting the same encoder object with different columns. :param X: input data frame :return: none, sets self.cols as a side effect """ # if columns aren't passed, just use every string column if self.use_default_cols: self.cols = get_obj_cols(X) else: self.cols = convert_cols_to_list(self.cols) def get_feature_names(self) -> List[str]: warnings.warn("`get_feature_names` is deprecated in all of sklearn. Use `get_feature_names_out` instead.", category=FutureWarning) return self.get_feature_names_out() def get_feature_names_out(self) -> List[str]: """ Returns the names of all transformed / added columns. Note that in sklearn the get_feature_names_out function takes the feature_names_in as an argument and determines the output feature names using the input. A fit is usually not necessary and if so a NotFittedError is raised. We just require a fit all the time and return the fitted output columns. Returns ------- feature_names: list A list with all feature names transformed or added. Note: potentially dropped features (because the feature is constant/invariant) are not included! """ out_feats = getattr(self, "feature_names_out_", None) if not isinstance(out_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return out_feats def get_feature_names_in(self) -> List[str]: """ Returns the names of all input columns present when fitting. These columns are necessary for the transform step. """ in_feats = getattr(self, "feature_names_in_", None) if not isinstance(in_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return in_feats @abstractmethod def _fit(self, X: pd.DataFrame, y: Optional[pd.Series], **kwargs): ... class SupervisedTransformerMixin(sklearn.base.TransformerMixin): def _more_tags(self): return {'supervised_encoder': True} def transform(self, X, y=None, override_return_df=False): """Perform the transformation to new categorical data. Some encoders behave differently on whether y is given or not. This is mainly due to regularisation in order to avoid overfitting. On training data transform should be called with y, on test data without. Parameters ---------- X : array-like, shape = [n_samples, n_features] y : array-like, shape = [n_samples] or None override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X, y = convert_inputs(X, y, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X, y) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X: pd.DataFrame, y: pd.Series) -> pd.DataFrame: ... def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y) class UnsupervisedTransformerMixin(sklearn.base.TransformerMixin): def transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X = convert_input(X, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X) -> pd.DataFrame: ... class TransformerWithTargetMixin: def _more_tags(self): return {'supervised_encoder': True} def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y)

"""A collection of shared utilities for all encoders, not intended for external use.""" from abc import abstractmethod from enum import Enum, auto import warnings import pandas as pd import numpy as np import sklearn.base from sklearn.base import BaseEstimator, TransformerMixin from sklearn.exceptions import NotFittedError from typing import Dict, List, Optional, Union from scipy.sparse import csr_matrix __author__ = 'willmcginnis' def convert_cols_to_list(cols): if isinstance(cols, pd.Series): return cols.tolist() elif isinstance(cols, np.ndarray): return cols.tolist() elif np.isscalar(cols): return [cols] elif isinstance(cols, set): return list(cols) elif isinstance(cols, tuple): return list(cols) elif pd.api.types.is_categorical_dtype(cols): return cols.astype(object).tolist() return cols def get_obj_cols(df): """ Returns names of 'object' columns in the DataFrame. """ obj_cols = [] for idx, dt in enumerate(df.dtypes): if dt == 'object' or is_category(dt): obj_cols.append(df.columns.values[idx]) if not obj_cols: print("Warning: No categorical columns found. Calling 'transform' will only return input data.") return obj_cols def is_category(dtype): return pd.api.types.is_categorical_dtype(dtype) def convert_inputs(X, y, columns=None, index=None, deep=False): """ Unite arraylike `X` and vectorlike `y` into a DataFrame and Series. If both are pandas types already, raises an error if their indexes do not match. If one is pandas, the returns will share that index. If neither is pandas, a default index will be used, unless `index` is passed. Parameters ---------- X: arraylike y: listlike columns: listlike Specifies column names to use for `X`. Ignored if `X` is already a dataframe. If `None`, use the default pandas column names. index: listlike The index to use, if neither `X` nor `y` is a pandas type. (If one has an index, then this has no effect.) If `None`, use the default pandas index. deep: bool Whether to deep-copy `X`. """ X_alt_index = y.index if isinstance(y, pd.Series) else index X = convert_input(X, columns=columns, deep=deep, index=X_alt_index) if y is not None: y = convert_input_vector(y, index=X.index) # N.B.: If either was already pandas, it keeps its index. if any(X.index != y.index): msg = "`X` and `y` both have indexes, but they do not match. If you are shuffling your input data on " \ "purpose (e.g. via permutation_test_score) use np arrays instead of data frames / series" raise ValueError(msg) if X.shape[0] != y.shape[0]: raise ValueError("The length of X is " + str(X.shape[0]) + " but length of y is " + str(y.shape[0]) + ".") return X, y def convert_input(X, columns=None, deep=False, index=None): """ Unite data into a DataFrame. Objects that do not contain column names take the names from the argument. Optionally perform deep copy of the data. """ if not isinstance(X, pd.DataFrame): if isinstance(X, pd.Series): X = pd.DataFrame(X, copy=deep) else: if columns is not None and np.size(X,1) != len(columns): raise ValueError('The count of the column names does not correspond to the count of the columns') if isinstance(X, list): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) # lists are always copied, but for consistency, we still pass the argument elif isinstance(X, (np.generic, np.ndarray)): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) elif isinstance(X, csr_matrix): X = pd.DataFrame(X.todense(), columns=columns, copy=deep, index=index) else: raise ValueError(f'Unexpected input type: {type(X)}') elif deep: X = X.copy(deep=True) return X def convert_input_vector(y, index): """ Unite target data type into a Series. If the target is a Series or a DataFrame, we preserve its index. But if the target does not contain index attribute, we use the index from the argument. """ if y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') if isinstance(y, pd.Series): return y elif isinstance(y, np.ndarray): if len(np.shape(y))==1: # vector return pd.Series(y, name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[0]==1: # single row in a matrix return pd.Series(y[0, :], name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[1]==1: # single column in a matrix return pd.Series(y[:, 0], name='target', index=index) else: raise ValueError(f'Unexpected input shape: {np.shape(y)}') elif np.isscalar(y): return pd.Series([y], name='target', index=index) elif isinstance(y, list): if len(y)==0: # empty list return pd.Series(y, name='target', index=index, dtype=float) elif len(y)>0 and not isinstance(y[0], list): # vector return pd.Series(y, name='target', index=index) elif len(y)>0 and isinstance(y[0], list) and len(y[0])==1: # single row in a matrix flatten = lambda y: [item for sublist in y for item in sublist] return pd.Series(flatten(y), name='target', index=index) elif len(y)==1 and len(y[0])==0 and isinstance(y[0], list): # single empty column in a matrix return pd.Series(y[0], name='target', index=index, dtype=float) elif len(y)==1 and isinstance(y[0], list): # single column in a matrix return pd.Series(y[0], name='target', index=index, dtype=type(y[0][0])) else: raise ValueError('Unexpected input shape') elif isinstance(y, pd.DataFrame): if len(list(y))==0: # empty DataFrame return pd.Series(name='target', index=index, dtype=float) if len(list(y))==1: # a single column return y.iloc[:, 0] else: raise ValueError(f'Unexpected input shape: {y.shape}') else: return pd.Series(y, name='target', index=index) # this covers tuples and other directly convertible types def get_generated_cols(X_original, X_transformed, to_transform): """ Returns a list of the generated/transformed columns. Arguments: X_original: df the original (input) DataFrame. X_transformed: df the transformed (current) DataFrame. to_transform: [str] a list of columns that were transformed (as in the original DataFrame), commonly self.feature_names_in. Output: a list of columns that were transformed (as in the current DataFrame). """ original_cols = list(X_original.columns) if len(to_transform) > 0: [original_cols.remove(c) for c in to_transform] current_cols = list(X_transformed.columns) if len(original_cols) > 0: [current_cols.remove(c) for c in original_cols] return current_cols def flatten_reverse_dict(d): sep = "___" [flat_dict] = pd.json_normalize(d, sep=sep).to_dict(orient='records') reversed_flat_dict = {v: tuple(k.split(sep)) for k, v in flat_dict.items()} return reversed_flat_dict class EncodingRelation(Enum): # one input feature get encoded into one output feature ONE_TO_ONE = auto() # one input feature get encoded into as many output features as it has distinct values ONE_TO_N_UNIQUE = auto() # one input feature get encoded into m output features that are not the number of distinct values ONE_TO_M = auto() # all N input features are encoded into M output features. # The encoding is done globally on all the input not on a per-feature basis N_TO_M = auto() def get_docstring_output_shape(in_out_relation: EncodingRelation): if in_out_relation == EncodingRelation.ONE_TO_ONE: return "n_features" elif in_out_relation == EncodingRelation.ONE_TO_N_UNIQUE: return "n_features * respective cardinality" elif in_out_relation == EncodingRelation.ONE_TO_M: return "M features (n_features < M)" elif in_out_relation == EncodingRelation.N_TO_M: return "M features (M can be anything)" class BaseEncoder(BaseEstimator): _dim: Optional[int] cols: List[str] use_default_cols: bool handle_missing: str handle_unknown: str verbose: int drop_invariant: bool invariant_cols: List[str] = [] return_df: bool supervised: bool encoding_relation: EncodingRelation INVARIANCE_THRESHOLD = 10e-5 # columns with variance less than this will be considered constant / invariant def __init__(self, verbose=0, cols=None, drop_invariant=False, return_df=True, handle_unknown='value', handle_missing='value', **kwargs): """ Parameters ---------- verbose: int integer indicating verbosity of output. 0 for none. cols: list a list of columns to encode, if None, all string and categorical columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform and inverse transform (otherwise it will be a numpy array). handle_missing: str how to handle missing values at fit time. Options are 'error', 'return_nan', and 'value'. Default 'value', which treat NaNs as a countable category at fit time. handle_unknown: str, int or dict of {column : option, ...}. how to handle unknown labels at transform time. Options are 'error' 'return_nan', 'value' and int. Defaults to None which uses NaN behaviour specified at fit time. Passing an int will fill with this int value. kwargs: dict. additional encoder specific parameters like regularisation. """ self.return_df = return_df self.drop_invariant = drop_invariant self.invariant_cols = [] self.verbose = verbose self.use_default_cols = cols is None # if True, even a repeated call of fit() will select string columns from X self.cols = cols # note that cols are only the columns to be encoded, feature_names_in_ are all columns self.mapping = None self.handle_unknown = handle_unknown self.handle_missing = handle_missing self._dim = None def fit(self, X, y=None, **kwargs): """Fits the encoder according to X and y. Parameters ---------- X : array-like, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. y : array-like, shape = [n_samples] Target values. Returns ------- self : encoder Returns self. """ self._check_fit_inputs(X, y) X, y = convert_inputs(X, y) self.feature_names_in_ = X.columns.tolist() self.n_features_in_ = len(self.feature_names_in_) self._dim = X.shape[1] self._determine_fit_columns(X) if not set(self.cols).issubset(X.columns): raise ValueError('X does not contain the columns listed in cols') if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') self._fit(X, y, **kwargs) # for finding invariant columns transform without y (as is done on the test set) X_transformed = self.transform(X, override_return_df=True) self.feature_names_out_ = X_transformed.columns.tolist() # drop all output columns with 0 variance. if self.drop_invariant: generated_cols = get_generated_cols(X, X_transformed, self.cols) self.invariant_cols = [x for x in generated_cols if X_transformed[x].var() <= self.INVARIANCE_THRESHOLD] self.feature_names_out_ = [x for x in self.feature_names_out_ if x not in self.invariant_cols] return self def _check_fit_inputs(self, X, y): if self._get_tags().get('supervised_encoder') and y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') def _check_transform_inputs(self, X): if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') if self._dim is None: raise NotFittedError('Must train encoder before it can be used to transform data.') # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') def _drop_invariants(self, X: pd.DataFrame, override_return_df: bool) -> Union[np.ndarray, pd.DataFrame]: if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.values def _determine_fit_columns(self, X: pd.DataFrame) -> None: """ Determine columns used by encoder. Note that the implementation also deals with re-fitting the same encoder object with different columns. :param X: input data frame :return: none, sets self.cols as a side effect """ # if columns aren't passed, just use every string column if self.use_default_cols: self.cols = get_obj_cols(X) else: self.cols = convert_cols_to_list(self.cols) def get_feature_names(self) -> List[str]: warnings.warn("`get_feature_names` is deprecated in all of sklearn. Use `get_feature_names_out` instead.", category=FutureWarning) return self.get_feature_names_out() def get_feature_names_out(self, input_features=None) -> np.ndarray: """ Returns the names of all transformed / added columns. Note that in sklearn the get_feature_names_out function takes the feature_names_in as an argument and determines the output feature names using the input. A fit is usually not necessary and if so a NotFittedError is raised. We just require a fit all the time and return the fitted output columns. Returns ------- feature_names: np.ndarray A numpy array with all feature names transformed or added. Note: potentially dropped features (because the feature is constant/invariant) are not included! """ out_feats = getattr(self, "feature_names_out_", None) if not isinstance(out_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return np.array(out_feats, dtype=object) def get_feature_names_in(self) -> List[str]: """ Returns the names of all input columns present when fitting. These columns are necessary for the transform step. """ in_feats = getattr(self, "feature_names_in_", None) if not isinstance(in_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return in_feats @abstractmethod def _fit(self, X: pd.DataFrame, y: Optional[pd.Series], **kwargs): ... class SupervisedTransformerMixin(sklearn.base.TransformerMixin): def _more_tags(self): return {'supervised_encoder': True} def transform(self, X, y=None, override_return_df=False): """Perform the transformation to new categorical data. Some encoders behave differently on whether y is given or not. This is mainly due to regularisation in order to avoid overfitting. On training data transform should be called with y, on test data without. Parameters ---------- X : array-like, shape = [n_samples, n_features] y : array-like, shape = [n_samples] or None override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X, y = convert_inputs(X, y, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X, y) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X: pd.DataFrame, y: pd.Series) -> pd.DataFrame: ... def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y) class UnsupervisedTransformerMixin(sklearn.base.TransformerMixin): def transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X = convert_input(X, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X) -> pd.DataFrame: ... class TransformerWithTargetMixin: def _more_tags(self): return {'supervised_encoder': True} def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y)

JaimeArboleda

5eb7a2d6359d680bdadd0534bdb983e712a47f9c

570827e6b48737d0c9aece8aca31edd6da02c1b2

I pretty much agree with you that we should go for the super simple fix. Let me think a little about the version stuff, maybe there is another solution, but I think we're very close to getting this merged

PaulWestenthanner

scikit-learn-contrib/category_encoders

(WIP) Partial fix for getting feature names out

I think this is a partial fix for this opened issue: https://github.com/scikit-learn-contrib/category_encoders/issues/395 It remains to check the behaviour of other estimators that are not ONE_TO_ONE. Please, let me know if you like the work in progress and I will try to continue.

2023-02-23 13:33:41+00:00

2023-03-13 11:48:24+00:00

category_encoders/utils.py

"""A collection of shared utilities for all encoders, not intended for external use.""" from abc import abstractmethod from enum import Enum, auto import warnings import pandas as pd import numpy as np import sklearn.base from sklearn.base import BaseEstimator, TransformerMixin from sklearn.exceptions import NotFittedError from typing import Dict, List, Optional, Union from scipy.sparse import csr_matrix __author__ = 'willmcginnis' def convert_cols_to_list(cols): if isinstance(cols, pd.Series): return cols.tolist() elif isinstance(cols, np.ndarray): return cols.tolist() elif np.isscalar(cols): return [cols] elif isinstance(cols, set): return list(cols) elif isinstance(cols, tuple): return list(cols) elif pd.api.types.is_categorical_dtype(cols): return cols.astype(object).tolist() return cols def get_obj_cols(df): """ Returns names of 'object' columns in the DataFrame. """ obj_cols = [] for idx, dt in enumerate(df.dtypes): if dt == 'object' or is_category(dt): obj_cols.append(df.columns.values[idx]) if not obj_cols: print("Warning: No categorical columns found. Calling 'transform' will only return input data.") return obj_cols def is_category(dtype): return pd.api.types.is_categorical_dtype(dtype) def convert_inputs(X, y, columns=None, index=None, deep=False): """ Unite arraylike `X` and vectorlike `y` into a DataFrame and Series. If both are pandas types already, raises an error if their indexes do not match. If one is pandas, the returns will share that index. If neither is pandas, a default index will be used, unless `index` is passed. Parameters ---------- X: arraylike y: listlike columns: listlike Specifies column names to use for `X`. Ignored if `X` is already a dataframe. If `None`, use the default pandas column names. index: listlike The index to use, if neither `X` nor `y` is a pandas type. (If one has an index, then this has no effect.) If `None`, use the default pandas index. deep: bool Whether to deep-copy `X`. """ X_alt_index = y.index if isinstance(y, pd.Series) else index X = convert_input(X, columns=columns, deep=deep, index=X_alt_index) if y is not None: y = convert_input_vector(y, index=X.index) # N.B.: If either was already pandas, it keeps its index. if any(X.index != y.index): msg = "`X` and `y` both have indexes, but they do not match. If you are shuffling your input data on " \ "purpose (e.g. via permutation_test_score) use np arrays instead of data frames / series" raise ValueError(msg) if X.shape[0] != y.shape[0]: raise ValueError("The length of X is " + str(X.shape[0]) + " but length of y is " + str(y.shape[0]) + ".") return X, y def convert_input(X, columns=None, deep=False, index=None): """ Unite data into a DataFrame. Objects that do not contain column names take the names from the argument. Optionally perform deep copy of the data. """ if not isinstance(X, pd.DataFrame): if isinstance(X, pd.Series): X = pd.DataFrame(X, copy=deep) else: if columns is not None and np.size(X,1) != len(columns): raise ValueError('The count of the column names does not correspond to the count of the columns') if isinstance(X, list): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) # lists are always copied, but for consistency, we still pass the argument elif isinstance(X, (np.generic, np.ndarray)): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) elif isinstance(X, csr_matrix): X = pd.DataFrame(X.todense(), columns=columns, copy=deep, index=index) else: raise ValueError(f'Unexpected input type: {type(X)}') elif deep: X = X.copy(deep=True) return X def convert_input_vector(y, index): """ Unite target data type into a Series. If the target is a Series or a DataFrame, we preserve its index. But if the target does not contain index attribute, we use the index from the argument. """ if y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') if isinstance(y, pd.Series): return y elif isinstance(y, np.ndarray): if len(np.shape(y))==1: # vector return pd.Series(y, name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[0]==1: # single row in a matrix return pd.Series(y[0, :], name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[1]==1: # single column in a matrix return pd.Series(y[:, 0], name='target', index=index) else: raise ValueError(f'Unexpected input shape: {np.shape(y)}') elif np.isscalar(y): return pd.Series([y], name='target', index=index) elif isinstance(y, list): if len(y)==0: # empty list return pd.Series(y, name='target', index=index, dtype=float) elif len(y)>0 and not isinstance(y[0], list): # vector return pd.Series(y, name='target', index=index) elif len(y)>0 and isinstance(y[0], list) and len(y[0])==1: # single row in a matrix flatten = lambda y: [item for sublist in y for item in sublist] return pd.Series(flatten(y), name='target', index=index) elif len(y)==1 and len(y[0])==0 and isinstance(y[0], list): # single empty column in a matrix return pd.Series(y[0], name='target', index=index, dtype=float) elif len(y)==1 and isinstance(y[0], list): # single column in a matrix return pd.Series(y[0], name='target', index=index, dtype=type(y[0][0])) else: raise ValueError('Unexpected input shape') elif isinstance(y, pd.DataFrame): if len(list(y))==0: # empty DataFrame return pd.Series(name='target', index=index, dtype=float) if len(list(y))==1: # a single column return y.iloc[:, 0] else: raise ValueError(f'Unexpected input shape: {y.shape}') else: return pd.Series(y, name='target', index=index) # this covers tuples and other directly convertible types def get_generated_cols(X_original, X_transformed, to_transform): """ Returns a list of the generated/transformed columns. Arguments: X_original: df the original (input) DataFrame. X_transformed: df the transformed (current) DataFrame. to_transform: [str] a list of columns that were transformed (as in the original DataFrame), commonly self.feature_names_in. Output: a list of columns that were transformed (as in the current DataFrame). """ original_cols = list(X_original.columns) if len(to_transform) > 0: [original_cols.remove(c) for c in to_transform] current_cols = list(X_transformed.columns) if len(original_cols) > 0: [current_cols.remove(c) for c in original_cols] return current_cols def flatten_reverse_dict(d): sep = "___" [flat_dict] = pd.json_normalize(d, sep=sep).to_dict(orient='records') reversed_flat_dict = {v: tuple(k.split(sep)) for k, v in flat_dict.items()} return reversed_flat_dict class EncodingRelation(Enum): # one input feature get encoded into one output feature ONE_TO_ONE = auto() # one input feature get encoded into as many output features as it has distinct values ONE_TO_N_UNIQUE = auto() # one input feature get encoded into m output features that are not the number of distinct values ONE_TO_M = auto() # all N input features are encoded into M output features. # The encoding is done globally on all the input not on a per-feature basis N_TO_M = auto() def get_docstring_output_shape(in_out_relation: EncodingRelation): if in_out_relation == EncodingRelation.ONE_TO_ONE: return "n_features" elif in_out_relation == EncodingRelation.ONE_TO_N_UNIQUE: return "n_features * respective cardinality" elif in_out_relation == EncodingRelation.ONE_TO_M: return "M features (n_features < M)" elif in_out_relation == EncodingRelation.N_TO_M: return "M features (M can be anything)" class BaseEncoder(BaseEstimator): _dim: Optional[int] cols: List[str] use_default_cols: bool handle_missing: str handle_unknown: str verbose: int drop_invariant: bool invariant_cols: List[str] = [] return_df: bool supervised: bool encoding_relation: EncodingRelation INVARIANCE_THRESHOLD = 10e-5 # columns with variance less than this will be considered constant / invariant def __init__(self, verbose=0, cols=None, drop_invariant=False, return_df=True, handle_unknown='value', handle_missing='value', **kwargs): """ Parameters ---------- verbose: int integer indicating verbosity of output. 0 for none. cols: list a list of columns to encode, if None, all string and categorical columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform and inverse transform (otherwise it will be a numpy array). handle_missing: str how to handle missing values at fit time. Options are 'error', 'return_nan', and 'value'. Default 'value', which treat NaNs as a countable category at fit time. handle_unknown: str, int or dict of {column : option, ...}. how to handle unknown labels at transform time. Options are 'error' 'return_nan', 'value' and int. Defaults to None which uses NaN behaviour specified at fit time. Passing an int will fill with this int value. kwargs: dict. additional encoder specific parameters like regularisation. """ self.return_df = return_df self.drop_invariant = drop_invariant self.invariant_cols = [] self.verbose = verbose self.use_default_cols = cols is None # if True, even a repeated call of fit() will select string columns from X self.cols = cols # note that cols are only the columns to be encoded, feature_names_in_ are all columns self.mapping = None self.handle_unknown = handle_unknown self.handle_missing = handle_missing self._dim = None def fit(self, X, y=None, **kwargs): """Fits the encoder according to X and y. Parameters ---------- X : array-like, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. y : array-like, shape = [n_samples] Target values. Returns ------- self : encoder Returns self. """ self._check_fit_inputs(X, y) X, y = convert_inputs(X, y) self.feature_names_in_ = X.columns.tolist() self.n_features_in_ = len(self.feature_names_in_) self._dim = X.shape[1] self._determine_fit_columns(X) if not set(self.cols).issubset(X.columns): raise ValueError('X does not contain the columns listed in cols') if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') self._fit(X, y, **kwargs) # for finding invariant columns transform without y (as is done on the test set) X_transformed = self.transform(X, override_return_df=True) self.feature_names_out_ = X_transformed.columns.tolist() # drop all output columns with 0 variance. if self.drop_invariant: generated_cols = get_generated_cols(X, X_transformed, self.cols) self.invariant_cols = [x for x in generated_cols if X_transformed[x].var() <= self.INVARIANCE_THRESHOLD] self.feature_names_out_ = [x for x in self.feature_names_out_ if x not in self.invariant_cols] return self def _check_fit_inputs(self, X, y): if self._get_tags().get('supervised_encoder') and y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') def _check_transform_inputs(self, X): if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') if self._dim is None: raise NotFittedError('Must train encoder before it can be used to transform data.') # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') def _drop_invariants(self, X: pd.DataFrame, override_return_df: bool) -> Union[np.ndarray, pd.DataFrame]: if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.values def _determine_fit_columns(self, X: pd.DataFrame) -> None: """ Determine columns used by encoder. Note that the implementation also deals with re-fitting the same encoder object with different columns. :param X: input data frame :return: none, sets self.cols as a side effect """ # if columns aren't passed, just use every string column if self.use_default_cols: self.cols = get_obj_cols(X) else: self.cols = convert_cols_to_list(self.cols) def get_feature_names(self) -> List[str]: warnings.warn("`get_feature_names` is deprecated in all of sklearn. Use `get_feature_names_out` instead.", category=FutureWarning) return self.get_feature_names_out() def get_feature_names_out(self) -> List[str]: """ Returns the names of all transformed / added columns. Note that in sklearn the get_feature_names_out function takes the feature_names_in as an argument and determines the output feature names using the input. A fit is usually not necessary and if so a NotFittedError is raised. We just require a fit all the time and return the fitted output columns. Returns ------- feature_names: list A list with all feature names transformed or added. Note: potentially dropped features (because the feature is constant/invariant) are not included! """ out_feats = getattr(self, "feature_names_out_", None) if not isinstance(out_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return out_feats def get_feature_names_in(self) -> List[str]: """ Returns the names of all input columns present when fitting. These columns are necessary for the transform step. """ in_feats = getattr(self, "feature_names_in_", None) if not isinstance(in_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return in_feats @abstractmethod def _fit(self, X: pd.DataFrame, y: Optional[pd.Series], **kwargs): ... class SupervisedTransformerMixin(sklearn.base.TransformerMixin): def _more_tags(self): return {'supervised_encoder': True} def transform(self, X, y=None, override_return_df=False): """Perform the transformation to new categorical data. Some encoders behave differently on whether y is given or not. This is mainly due to regularisation in order to avoid overfitting. On training data transform should be called with y, on test data without. Parameters ---------- X : array-like, shape = [n_samples, n_features] y : array-like, shape = [n_samples] or None override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X, y = convert_inputs(X, y, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X, y) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X: pd.DataFrame, y: pd.Series) -> pd.DataFrame: ... def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y) class UnsupervisedTransformerMixin(sklearn.base.TransformerMixin): def transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X = convert_input(X, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X) -> pd.DataFrame: ... class TransformerWithTargetMixin: def _more_tags(self): return {'supervised_encoder': True} def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y)

"""A collection of shared utilities for all encoders, not intended for external use.""" from abc import abstractmethod from enum import Enum, auto import warnings import pandas as pd import numpy as np import sklearn.base from sklearn.base import BaseEstimator, TransformerMixin from sklearn.exceptions import NotFittedError from typing import Dict, List, Optional, Union from scipy.sparse import csr_matrix __author__ = 'willmcginnis' def convert_cols_to_list(cols): if isinstance(cols, pd.Series): return cols.tolist() elif isinstance(cols, np.ndarray): return cols.tolist() elif np.isscalar(cols): return [cols] elif isinstance(cols, set): return list(cols) elif isinstance(cols, tuple): return list(cols) elif pd.api.types.is_categorical_dtype(cols): return cols.astype(object).tolist() return cols def get_obj_cols(df): """ Returns names of 'object' columns in the DataFrame. """ obj_cols = [] for idx, dt in enumerate(df.dtypes): if dt == 'object' or is_category(dt): obj_cols.append(df.columns.values[idx]) if not obj_cols: print("Warning: No categorical columns found. Calling 'transform' will only return input data.") return obj_cols def is_category(dtype): return pd.api.types.is_categorical_dtype(dtype) def convert_inputs(X, y, columns=None, index=None, deep=False): """ Unite arraylike `X` and vectorlike `y` into a DataFrame and Series. If both are pandas types already, raises an error if their indexes do not match. If one is pandas, the returns will share that index. If neither is pandas, a default index will be used, unless `index` is passed. Parameters ---------- X: arraylike y: listlike columns: listlike Specifies column names to use for `X`. Ignored if `X` is already a dataframe. If `None`, use the default pandas column names. index: listlike The index to use, if neither `X` nor `y` is a pandas type. (If one has an index, then this has no effect.) If `None`, use the default pandas index. deep: bool Whether to deep-copy `X`. """ X_alt_index = y.index if isinstance(y, pd.Series) else index X = convert_input(X, columns=columns, deep=deep, index=X_alt_index) if y is not None: y = convert_input_vector(y, index=X.index) # N.B.: If either was already pandas, it keeps its index. if any(X.index != y.index): msg = "`X` and `y` both have indexes, but they do not match. If you are shuffling your input data on " \ "purpose (e.g. via permutation_test_score) use np arrays instead of data frames / series" raise ValueError(msg) if X.shape[0] != y.shape[0]: raise ValueError("The length of X is " + str(X.shape[0]) + " but length of y is " + str(y.shape[0]) + ".") return X, y def convert_input(X, columns=None, deep=False, index=None): """ Unite data into a DataFrame. Objects that do not contain column names take the names from the argument. Optionally perform deep copy of the data. """ if not isinstance(X, pd.DataFrame): if isinstance(X, pd.Series): X = pd.DataFrame(X, copy=deep) else: if columns is not None and np.size(X,1) != len(columns): raise ValueError('The count of the column names does not correspond to the count of the columns') if isinstance(X, list): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) # lists are always copied, but for consistency, we still pass the argument elif isinstance(X, (np.generic, np.ndarray)): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) elif isinstance(X, csr_matrix): X = pd.DataFrame(X.todense(), columns=columns, copy=deep, index=index) else: raise ValueError(f'Unexpected input type: {type(X)}') elif deep: X = X.copy(deep=True) return X def convert_input_vector(y, index): """ Unite target data type into a Series. If the target is a Series or a DataFrame, we preserve its index. But if the target does not contain index attribute, we use the index from the argument. """ if y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') if isinstance(y, pd.Series): return y elif isinstance(y, np.ndarray): if len(np.shape(y))==1: # vector return pd.Series(y, name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[0]==1: # single row in a matrix return pd.Series(y[0, :], name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[1]==1: # single column in a matrix return pd.Series(y[:, 0], name='target', index=index) else: raise ValueError(f'Unexpected input shape: {np.shape(y)}') elif np.isscalar(y): return pd.Series([y], name='target', index=index) elif isinstance(y, list): if len(y)==0: # empty list return pd.Series(y, name='target', index=index, dtype=float) elif len(y)>0 and not isinstance(y[0], list): # vector return pd.Series(y, name='target', index=index) elif len(y)>0 and isinstance(y[0], list) and len(y[0])==1: # single row in a matrix flatten = lambda y: [item for sublist in y for item in sublist] return pd.Series(flatten(y), name='target', index=index) elif len(y)==1 and len(y[0])==0 and isinstance(y[0], list): # single empty column in a matrix return pd.Series(y[0], name='target', index=index, dtype=float) elif len(y)==1 and isinstance(y[0], list): # single column in a matrix return pd.Series(y[0], name='target', index=index, dtype=type(y[0][0])) else: raise ValueError('Unexpected input shape') elif isinstance(y, pd.DataFrame): if len(list(y))==0: # empty DataFrame return pd.Series(name='target', index=index, dtype=float) if len(list(y))==1: # a single column return y.iloc[:, 0] else: raise ValueError(f'Unexpected input shape: {y.shape}') else: return pd.Series(y, name='target', index=index) # this covers tuples and other directly convertible types def get_generated_cols(X_original, X_transformed, to_transform): """ Returns a list of the generated/transformed columns. Arguments: X_original: df the original (input) DataFrame. X_transformed: df the transformed (current) DataFrame. to_transform: [str] a list of columns that were transformed (as in the original DataFrame), commonly self.feature_names_in. Output: a list of columns that were transformed (as in the current DataFrame). """ original_cols = list(X_original.columns) if len(to_transform) > 0: [original_cols.remove(c) for c in to_transform] current_cols = list(X_transformed.columns) if len(original_cols) > 0: [current_cols.remove(c) for c in original_cols] return current_cols def flatten_reverse_dict(d): sep = "___" [flat_dict] = pd.json_normalize(d, sep=sep).to_dict(orient='records') reversed_flat_dict = {v: tuple(k.split(sep)) for k, v in flat_dict.items()} return reversed_flat_dict class EncodingRelation(Enum): # one input feature get encoded into one output feature ONE_TO_ONE = auto() # one input feature get encoded into as many output features as it has distinct values ONE_TO_N_UNIQUE = auto() # one input feature get encoded into m output features that are not the number of distinct values ONE_TO_M = auto() # all N input features are encoded into M output features. # The encoding is done globally on all the input not on a per-feature basis N_TO_M = auto() def get_docstring_output_shape(in_out_relation: EncodingRelation): if in_out_relation == EncodingRelation.ONE_TO_ONE: return "n_features" elif in_out_relation == EncodingRelation.ONE_TO_N_UNIQUE: return "n_features * respective cardinality" elif in_out_relation == EncodingRelation.ONE_TO_M: return "M features (n_features < M)" elif in_out_relation == EncodingRelation.N_TO_M: return "M features (M can be anything)" class BaseEncoder(BaseEstimator): _dim: Optional[int] cols: List[str] use_default_cols: bool handle_missing: str handle_unknown: str verbose: int drop_invariant: bool invariant_cols: List[str] = [] return_df: bool supervised: bool encoding_relation: EncodingRelation INVARIANCE_THRESHOLD = 10e-5 # columns with variance less than this will be considered constant / invariant def __init__(self, verbose=0, cols=None, drop_invariant=False, return_df=True, handle_unknown='value', handle_missing='value', **kwargs): """ Parameters ---------- verbose: int integer indicating verbosity of output. 0 for none. cols: list a list of columns to encode, if None, all string and categorical columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform and inverse transform (otherwise it will be a numpy array). handle_missing: str how to handle missing values at fit time. Options are 'error', 'return_nan', and 'value'. Default 'value', which treat NaNs as a countable category at fit time. handle_unknown: str, int or dict of {column : option, ...}. how to handle unknown labels at transform time. Options are 'error' 'return_nan', 'value' and int. Defaults to None which uses NaN behaviour specified at fit time. Passing an int will fill with this int value. kwargs: dict. additional encoder specific parameters like regularisation. """ self.return_df = return_df self.drop_invariant = drop_invariant self.invariant_cols = [] self.verbose = verbose self.use_default_cols = cols is None # if True, even a repeated call of fit() will select string columns from X self.cols = cols # note that cols are only the columns to be encoded, feature_names_in_ are all columns self.mapping = None self.handle_unknown = handle_unknown self.handle_missing = handle_missing self._dim = None def fit(self, X, y=None, **kwargs): """Fits the encoder according to X and y. Parameters ---------- X : array-like, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. y : array-like, shape = [n_samples] Target values. Returns ------- self : encoder Returns self. """ self._check_fit_inputs(X, y) X, y = convert_inputs(X, y) self.feature_names_in_ = X.columns.tolist() self.n_features_in_ = len(self.feature_names_in_) self._dim = X.shape[1] self._determine_fit_columns(X) if not set(self.cols).issubset(X.columns): raise ValueError('X does not contain the columns listed in cols') if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') self._fit(X, y, **kwargs) # for finding invariant columns transform without y (as is done on the test set) X_transformed = self.transform(X, override_return_df=True) self.feature_names_out_ = X_transformed.columns.tolist() # drop all output columns with 0 variance. if self.drop_invariant: generated_cols = get_generated_cols(X, X_transformed, self.cols) self.invariant_cols = [x for x in generated_cols if X_transformed[x].var() <= self.INVARIANCE_THRESHOLD] self.feature_names_out_ = [x for x in self.feature_names_out_ if x not in self.invariant_cols] return self def _check_fit_inputs(self, X, y): if self._get_tags().get('supervised_encoder') and y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') def _check_transform_inputs(self, X): if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') if self._dim is None: raise NotFittedError('Must train encoder before it can be used to transform data.') # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') def _drop_invariants(self, X: pd.DataFrame, override_return_df: bool) -> Union[np.ndarray, pd.DataFrame]: if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.values def _determine_fit_columns(self, X: pd.DataFrame) -> None: """ Determine columns used by encoder. Note that the implementation also deals with re-fitting the same encoder object with different columns. :param X: input data frame :return: none, sets self.cols as a side effect """ # if columns aren't passed, just use every string column if self.use_default_cols: self.cols = get_obj_cols(X) else: self.cols = convert_cols_to_list(self.cols) def get_feature_names(self) -> List[str]: warnings.warn("`get_feature_names` is deprecated in all of sklearn. Use `get_feature_names_out` instead.", category=FutureWarning) return self.get_feature_names_out() def get_feature_names_out(self, input_features=None) -> np.ndarray: """ Returns the names of all transformed / added columns. Note that in sklearn the get_feature_names_out function takes the feature_names_in as an argument and determines the output feature names using the input. A fit is usually not necessary and if so a NotFittedError is raised. We just require a fit all the time and return the fitted output columns. Returns ------- feature_names: np.ndarray A numpy array with all feature names transformed or added. Note: potentially dropped features (because the feature is constant/invariant) are not included! """ out_feats = getattr(self, "feature_names_out_", None) if not isinstance(out_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return np.array(out_feats, dtype=object) def get_feature_names_in(self) -> List[str]: """ Returns the names of all input columns present when fitting. These columns are necessary for the transform step. """ in_feats = getattr(self, "feature_names_in_", None) if not isinstance(in_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return in_feats @abstractmethod def _fit(self, X: pd.DataFrame, y: Optional[pd.Series], **kwargs): ... class SupervisedTransformerMixin(sklearn.base.TransformerMixin): def _more_tags(self): return {'supervised_encoder': True} def transform(self, X, y=None, override_return_df=False): """Perform the transformation to new categorical data. Some encoders behave differently on whether y is given or not. This is mainly due to regularisation in order to avoid overfitting. On training data transform should be called with y, on test data without. Parameters ---------- X : array-like, shape = [n_samples, n_features] y : array-like, shape = [n_samples] or None override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X, y = convert_inputs(X, y, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X, y) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X: pd.DataFrame, y: pd.Series) -> pd.DataFrame: ... def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y) class UnsupervisedTransformerMixin(sklearn.base.TransformerMixin): def transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X = convert_input(X, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X) -> pd.DataFrame: ... class TransformerWithTargetMixin: def _more_tags(self): return {'supervised_encoder': True} def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y)

JaimeArboleda

5eb7a2d6359d680bdadd0534bdb983e712a47f9c

570827e6b48737d0c9aece8aca31edd6da02c1b2

Hi, I've added some thoughts on the version. I'll resolve the discussion here. Please have a look at the other comment I've made and if you're fine with it we're ready to merge!

PaulWestenthanner

scikit-learn-contrib/category_encoders

(WIP) Partial fix for getting feature names out

I think this is a partial fix for this opened issue: https://github.com/scikit-learn-contrib/category_encoders/issues/395 It remains to check the behaviour of other estimators that are not ONE_TO_ONE. Please, let me know if you like the work in progress and I will try to continue.

2023-02-23 13:33:41+00:00

2023-03-13 11:48:24+00:00

category_encoders/utils.py

"""A collection of shared utilities for all encoders, not intended for external use.""" from abc import abstractmethod from enum import Enum, auto import warnings import pandas as pd import numpy as np import sklearn.base from sklearn.base import BaseEstimator, TransformerMixin from sklearn.exceptions import NotFittedError from typing import Dict, List, Optional, Union from scipy.sparse import csr_matrix __author__ = 'willmcginnis' def convert_cols_to_list(cols): if isinstance(cols, pd.Series): return cols.tolist() elif isinstance(cols, np.ndarray): return cols.tolist() elif np.isscalar(cols): return [cols] elif isinstance(cols, set): return list(cols) elif isinstance(cols, tuple): return list(cols) elif pd.api.types.is_categorical_dtype(cols): return cols.astype(object).tolist() return cols def get_obj_cols(df): """ Returns names of 'object' columns in the DataFrame. """ obj_cols = [] for idx, dt in enumerate(df.dtypes): if dt == 'object' or is_category(dt): obj_cols.append(df.columns.values[idx]) if not obj_cols: print("Warning: No categorical columns found. Calling 'transform' will only return input data.") return obj_cols def is_category(dtype): return pd.api.types.is_categorical_dtype(dtype) def convert_inputs(X, y, columns=None, index=None, deep=False): """ Unite arraylike `X` and vectorlike `y` into a DataFrame and Series. If both are pandas types already, raises an error if their indexes do not match. If one is pandas, the returns will share that index. If neither is pandas, a default index will be used, unless `index` is passed. Parameters ---------- X: arraylike y: listlike columns: listlike Specifies column names to use for `X`. Ignored if `X` is already a dataframe. If `None`, use the default pandas column names. index: listlike The index to use, if neither `X` nor `y` is a pandas type. (If one has an index, then this has no effect.) If `None`, use the default pandas index. deep: bool Whether to deep-copy `X`. """ X_alt_index = y.index if isinstance(y, pd.Series) else index X = convert_input(X, columns=columns, deep=deep, index=X_alt_index) if y is not None: y = convert_input_vector(y, index=X.index) # N.B.: If either was already pandas, it keeps its index. if any(X.index != y.index): msg = "`X` and `y` both have indexes, but they do not match. If you are shuffling your input data on " \ "purpose (e.g. via permutation_test_score) use np arrays instead of data frames / series" raise ValueError(msg) if X.shape[0] != y.shape[0]: raise ValueError("The length of X is " + str(X.shape[0]) + " but length of y is " + str(y.shape[0]) + ".") return X, y def convert_input(X, columns=None, deep=False, index=None): """ Unite data into a DataFrame. Objects that do not contain column names take the names from the argument. Optionally perform deep copy of the data. """ if not isinstance(X, pd.DataFrame): if isinstance(X, pd.Series): X = pd.DataFrame(X, copy=deep) else: if columns is not None and np.size(X,1) != len(columns): raise ValueError('The count of the column names does not correspond to the count of the columns') if isinstance(X, list): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) # lists are always copied, but for consistency, we still pass the argument elif isinstance(X, (np.generic, np.ndarray)): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) elif isinstance(X, csr_matrix): X = pd.DataFrame(X.todense(), columns=columns, copy=deep, index=index) else: raise ValueError(f'Unexpected input type: {type(X)}') elif deep: X = X.copy(deep=True) return X def convert_input_vector(y, index): """ Unite target data type into a Series. If the target is a Series or a DataFrame, we preserve its index. But if the target does not contain index attribute, we use the index from the argument. """ if y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') if isinstance(y, pd.Series): return y elif isinstance(y, np.ndarray): if len(np.shape(y))==1: # vector return pd.Series(y, name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[0]==1: # single row in a matrix return pd.Series(y[0, :], name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[1]==1: # single column in a matrix return pd.Series(y[:, 0], name='target', index=index) else: raise ValueError(f'Unexpected input shape: {np.shape(y)}') elif np.isscalar(y): return pd.Series([y], name='target', index=index) elif isinstance(y, list): if len(y)==0: # empty list return pd.Series(y, name='target', index=index, dtype=float) elif len(y)>0 and not isinstance(y[0], list): # vector return pd.Series(y, name='target', index=index) elif len(y)>0 and isinstance(y[0], list) and len(y[0])==1: # single row in a matrix flatten = lambda y: [item for sublist in y for item in sublist] return pd.Series(flatten(y), name='target', index=index) elif len(y)==1 and len(y[0])==0 and isinstance(y[0], list): # single empty column in a matrix return pd.Series(y[0], name='target', index=index, dtype=float) elif len(y)==1 and isinstance(y[0], list): # single column in a matrix return pd.Series(y[0], name='target', index=index, dtype=type(y[0][0])) else: raise ValueError('Unexpected input shape') elif isinstance(y, pd.DataFrame): if len(list(y))==0: # empty DataFrame return pd.Series(name='target', index=index, dtype=float) if len(list(y))==1: # a single column return y.iloc[:, 0] else: raise ValueError(f'Unexpected input shape: {y.shape}') else: return pd.Series(y, name='target', index=index) # this covers tuples and other directly convertible types def get_generated_cols(X_original, X_transformed, to_transform): """ Returns a list of the generated/transformed columns. Arguments: X_original: df the original (input) DataFrame. X_transformed: df the transformed (current) DataFrame. to_transform: [str] a list of columns that were transformed (as in the original DataFrame), commonly self.feature_names_in. Output: a list of columns that were transformed (as in the current DataFrame). """ original_cols = list(X_original.columns) if len(to_transform) > 0: [original_cols.remove(c) for c in to_transform] current_cols = list(X_transformed.columns) if len(original_cols) > 0: [current_cols.remove(c) for c in original_cols] return current_cols def flatten_reverse_dict(d): sep = "___" [flat_dict] = pd.json_normalize(d, sep=sep).to_dict(orient='records') reversed_flat_dict = {v: tuple(k.split(sep)) for k, v in flat_dict.items()} return reversed_flat_dict class EncodingRelation(Enum): # one input feature get encoded into one output feature ONE_TO_ONE = auto() # one input feature get encoded into as many output features as it has distinct values ONE_TO_N_UNIQUE = auto() # one input feature get encoded into m output features that are not the number of distinct values ONE_TO_M = auto() # all N input features are encoded into M output features. # The encoding is done globally on all the input not on a per-feature basis N_TO_M = auto() def get_docstring_output_shape(in_out_relation: EncodingRelation): if in_out_relation == EncodingRelation.ONE_TO_ONE: return "n_features" elif in_out_relation == EncodingRelation.ONE_TO_N_UNIQUE: return "n_features * respective cardinality" elif in_out_relation == EncodingRelation.ONE_TO_M: return "M features (n_features < M)" elif in_out_relation == EncodingRelation.N_TO_M: return "M features (M can be anything)" class BaseEncoder(BaseEstimator): _dim: Optional[int] cols: List[str] use_default_cols: bool handle_missing: str handle_unknown: str verbose: int drop_invariant: bool invariant_cols: List[str] = [] return_df: bool supervised: bool encoding_relation: EncodingRelation INVARIANCE_THRESHOLD = 10e-5 # columns with variance less than this will be considered constant / invariant def __init__(self, verbose=0, cols=None, drop_invariant=False, return_df=True, handle_unknown='value', handle_missing='value', **kwargs): """ Parameters ---------- verbose: int integer indicating verbosity of output. 0 for none. cols: list a list of columns to encode, if None, all string and categorical columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform and inverse transform (otherwise it will be a numpy array). handle_missing: str how to handle missing values at fit time. Options are 'error', 'return_nan', and 'value'. Default 'value', which treat NaNs as a countable category at fit time. handle_unknown: str, int or dict of {column : option, ...}. how to handle unknown labels at transform time. Options are 'error' 'return_nan', 'value' and int. Defaults to None which uses NaN behaviour specified at fit time. Passing an int will fill with this int value. kwargs: dict. additional encoder specific parameters like regularisation. """ self.return_df = return_df self.drop_invariant = drop_invariant self.invariant_cols = [] self.verbose = verbose self.use_default_cols = cols is None # if True, even a repeated call of fit() will select string columns from X self.cols = cols # note that cols are only the columns to be encoded, feature_names_in_ are all columns self.mapping = None self.handle_unknown = handle_unknown self.handle_missing = handle_missing self._dim = None def fit(self, X, y=None, **kwargs): """Fits the encoder according to X and y. Parameters ---------- X : array-like, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. y : array-like, shape = [n_samples] Target values. Returns ------- self : encoder Returns self. """ self._check_fit_inputs(X, y) X, y = convert_inputs(X, y) self.feature_names_in_ = X.columns.tolist() self.n_features_in_ = len(self.feature_names_in_) self._dim = X.shape[1] self._determine_fit_columns(X) if not set(self.cols).issubset(X.columns): raise ValueError('X does not contain the columns listed in cols') if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') self._fit(X, y, **kwargs) # for finding invariant columns transform without y (as is done on the test set) X_transformed = self.transform(X, override_return_df=True) self.feature_names_out_ = X_transformed.columns.tolist() # drop all output columns with 0 variance. if self.drop_invariant: generated_cols = get_generated_cols(X, X_transformed, self.cols) self.invariant_cols = [x for x in generated_cols if X_transformed[x].var() <= self.INVARIANCE_THRESHOLD] self.feature_names_out_ = [x for x in self.feature_names_out_ if x not in self.invariant_cols] return self def _check_fit_inputs(self, X, y): if self._get_tags().get('supervised_encoder') and y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') def _check_transform_inputs(self, X): if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') if self._dim is None: raise NotFittedError('Must train encoder before it can be used to transform data.') # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') def _drop_invariants(self, X: pd.DataFrame, override_return_df: bool) -> Union[np.ndarray, pd.DataFrame]: if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.values def _determine_fit_columns(self, X: pd.DataFrame) -> None: """ Determine columns used by encoder. Note that the implementation also deals with re-fitting the same encoder object with different columns. :param X: input data frame :return: none, sets self.cols as a side effect """ # if columns aren't passed, just use every string column if self.use_default_cols: self.cols = get_obj_cols(X) else: self.cols = convert_cols_to_list(self.cols) def get_feature_names(self) -> List[str]: warnings.warn("`get_feature_names` is deprecated in all of sklearn. Use `get_feature_names_out` instead.", category=FutureWarning) return self.get_feature_names_out() def get_feature_names_out(self) -> List[str]: """ Returns the names of all transformed / added columns. Note that in sklearn the get_feature_names_out function takes the feature_names_in as an argument and determines the output feature names using the input. A fit is usually not necessary and if so a NotFittedError is raised. We just require a fit all the time and return the fitted output columns. Returns ------- feature_names: list A list with all feature names transformed or added. Note: potentially dropped features (because the feature is constant/invariant) are not included! """ out_feats = getattr(self, "feature_names_out_", None) if not isinstance(out_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return out_feats def get_feature_names_in(self) -> List[str]: """ Returns the names of all input columns present when fitting. These columns are necessary for the transform step. """ in_feats = getattr(self, "feature_names_in_", None) if not isinstance(in_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return in_feats @abstractmethod def _fit(self, X: pd.DataFrame, y: Optional[pd.Series], **kwargs): ... class SupervisedTransformerMixin(sklearn.base.TransformerMixin): def _more_tags(self): return {'supervised_encoder': True} def transform(self, X, y=None, override_return_df=False): """Perform the transformation to new categorical data. Some encoders behave differently on whether y is given or not. This is mainly due to regularisation in order to avoid overfitting. On training data transform should be called with y, on test data without. Parameters ---------- X : array-like, shape = [n_samples, n_features] y : array-like, shape = [n_samples] or None override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X, y = convert_inputs(X, y, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X, y) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X: pd.DataFrame, y: pd.Series) -> pd.DataFrame: ... def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y) class UnsupervisedTransformerMixin(sklearn.base.TransformerMixin): def transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X = convert_input(X, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X) -> pd.DataFrame: ... class TransformerWithTargetMixin: def _more_tags(self): return {'supervised_encoder': True} def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y)

"""A collection of shared utilities for all encoders, not intended for external use.""" from abc import abstractmethod from enum import Enum, auto import warnings import pandas as pd import numpy as np import sklearn.base from sklearn.base import BaseEstimator, TransformerMixin from sklearn.exceptions import NotFittedError from typing import Dict, List, Optional, Union from scipy.sparse import csr_matrix __author__ = 'willmcginnis' def convert_cols_to_list(cols): if isinstance(cols, pd.Series): return cols.tolist() elif isinstance(cols, np.ndarray): return cols.tolist() elif np.isscalar(cols): return [cols] elif isinstance(cols, set): return list(cols) elif isinstance(cols, tuple): return list(cols) elif pd.api.types.is_categorical_dtype(cols): return cols.astype(object).tolist() return cols def get_obj_cols(df): """ Returns names of 'object' columns in the DataFrame. """ obj_cols = [] for idx, dt in enumerate(df.dtypes): if dt == 'object' or is_category(dt): obj_cols.append(df.columns.values[idx]) if not obj_cols: print("Warning: No categorical columns found. Calling 'transform' will only return input data.") return obj_cols def is_category(dtype): return pd.api.types.is_categorical_dtype(dtype) def convert_inputs(X, y, columns=None, index=None, deep=False): """ Unite arraylike `X` and vectorlike `y` into a DataFrame and Series. If both are pandas types already, raises an error if their indexes do not match. If one is pandas, the returns will share that index. If neither is pandas, a default index will be used, unless `index` is passed. Parameters ---------- X: arraylike y: listlike columns: listlike Specifies column names to use for `X`. Ignored if `X` is already a dataframe. If `None`, use the default pandas column names. index: listlike The index to use, if neither `X` nor `y` is a pandas type. (If one has an index, then this has no effect.) If `None`, use the default pandas index. deep: bool Whether to deep-copy `X`. """ X_alt_index = y.index if isinstance(y, pd.Series) else index X = convert_input(X, columns=columns, deep=deep, index=X_alt_index) if y is not None: y = convert_input_vector(y, index=X.index) # N.B.: If either was already pandas, it keeps its index. if any(X.index != y.index): msg = "`X` and `y` both have indexes, but they do not match. If you are shuffling your input data on " \ "purpose (e.g. via permutation_test_score) use np arrays instead of data frames / series" raise ValueError(msg) if X.shape[0] != y.shape[0]: raise ValueError("The length of X is " + str(X.shape[0]) + " but length of y is " + str(y.shape[0]) + ".") return X, y def convert_input(X, columns=None, deep=False, index=None): """ Unite data into a DataFrame. Objects that do not contain column names take the names from the argument. Optionally perform deep copy of the data. """ if not isinstance(X, pd.DataFrame): if isinstance(X, pd.Series): X = pd.DataFrame(X, copy=deep) else: if columns is not None and np.size(X,1) != len(columns): raise ValueError('The count of the column names does not correspond to the count of the columns') if isinstance(X, list): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) # lists are always copied, but for consistency, we still pass the argument elif isinstance(X, (np.generic, np.ndarray)): X = pd.DataFrame(X, columns=columns, copy=deep, index=index) elif isinstance(X, csr_matrix): X = pd.DataFrame(X.todense(), columns=columns, copy=deep, index=index) else: raise ValueError(f'Unexpected input type: {type(X)}') elif deep: X = X.copy(deep=True) return X def convert_input_vector(y, index): """ Unite target data type into a Series. If the target is a Series or a DataFrame, we preserve its index. But if the target does not contain index attribute, we use the index from the argument. """ if y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') if isinstance(y, pd.Series): return y elif isinstance(y, np.ndarray): if len(np.shape(y))==1: # vector return pd.Series(y, name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[0]==1: # single row in a matrix return pd.Series(y[0, :], name='target', index=index) elif len(np.shape(y))==2 and np.shape(y)[1]==1: # single column in a matrix return pd.Series(y[:, 0], name='target', index=index) else: raise ValueError(f'Unexpected input shape: {np.shape(y)}') elif np.isscalar(y): return pd.Series([y], name='target', index=index) elif isinstance(y, list): if len(y)==0: # empty list return pd.Series(y, name='target', index=index, dtype=float) elif len(y)>0 and not isinstance(y[0], list): # vector return pd.Series(y, name='target', index=index) elif len(y)>0 and isinstance(y[0], list) and len(y[0])==1: # single row in a matrix flatten = lambda y: [item for sublist in y for item in sublist] return pd.Series(flatten(y), name='target', index=index) elif len(y)==1 and len(y[0])==0 and isinstance(y[0], list): # single empty column in a matrix return pd.Series(y[0], name='target', index=index, dtype=float) elif len(y)==1 and isinstance(y[0], list): # single column in a matrix return pd.Series(y[0], name='target', index=index, dtype=type(y[0][0])) else: raise ValueError('Unexpected input shape') elif isinstance(y, pd.DataFrame): if len(list(y))==0: # empty DataFrame return pd.Series(name='target', index=index, dtype=float) if len(list(y))==1: # a single column return y.iloc[:, 0] else: raise ValueError(f'Unexpected input shape: {y.shape}') else: return pd.Series(y, name='target', index=index) # this covers tuples and other directly convertible types def get_generated_cols(X_original, X_transformed, to_transform): """ Returns a list of the generated/transformed columns. Arguments: X_original: df the original (input) DataFrame. X_transformed: df the transformed (current) DataFrame. to_transform: [str] a list of columns that were transformed (as in the original DataFrame), commonly self.feature_names_in. Output: a list of columns that were transformed (as in the current DataFrame). """ original_cols = list(X_original.columns) if len(to_transform) > 0: [original_cols.remove(c) for c in to_transform] current_cols = list(X_transformed.columns) if len(original_cols) > 0: [current_cols.remove(c) for c in original_cols] return current_cols def flatten_reverse_dict(d): sep = "___" [flat_dict] = pd.json_normalize(d, sep=sep).to_dict(orient='records') reversed_flat_dict = {v: tuple(k.split(sep)) for k, v in flat_dict.items()} return reversed_flat_dict class EncodingRelation(Enum): # one input feature get encoded into one output feature ONE_TO_ONE = auto() # one input feature get encoded into as many output features as it has distinct values ONE_TO_N_UNIQUE = auto() # one input feature get encoded into m output features that are not the number of distinct values ONE_TO_M = auto() # all N input features are encoded into M output features. # The encoding is done globally on all the input not on a per-feature basis N_TO_M = auto() def get_docstring_output_shape(in_out_relation: EncodingRelation): if in_out_relation == EncodingRelation.ONE_TO_ONE: return "n_features" elif in_out_relation == EncodingRelation.ONE_TO_N_UNIQUE: return "n_features * respective cardinality" elif in_out_relation == EncodingRelation.ONE_TO_M: return "M features (n_features < M)" elif in_out_relation == EncodingRelation.N_TO_M: return "M features (M can be anything)" class BaseEncoder(BaseEstimator): _dim: Optional[int] cols: List[str] use_default_cols: bool handle_missing: str handle_unknown: str verbose: int drop_invariant: bool invariant_cols: List[str] = [] return_df: bool supervised: bool encoding_relation: EncodingRelation INVARIANCE_THRESHOLD = 10e-5 # columns with variance less than this will be considered constant / invariant def __init__(self, verbose=0, cols=None, drop_invariant=False, return_df=True, handle_unknown='value', handle_missing='value', **kwargs): """ Parameters ---------- verbose: int integer indicating verbosity of output. 0 for none. cols: list a list of columns to encode, if None, all string and categorical columns will be encoded. drop_invariant: bool boolean for whether or not to drop columns with 0 variance. return_df: bool boolean for whether to return a pandas DataFrame from transform and inverse transform (otherwise it will be a numpy array). handle_missing: str how to handle missing values at fit time. Options are 'error', 'return_nan', and 'value'. Default 'value', which treat NaNs as a countable category at fit time. handle_unknown: str, int or dict of {column : option, ...}. how to handle unknown labels at transform time. Options are 'error' 'return_nan', 'value' and int. Defaults to None which uses NaN behaviour specified at fit time. Passing an int will fill with this int value. kwargs: dict. additional encoder specific parameters like regularisation. """ self.return_df = return_df self.drop_invariant = drop_invariant self.invariant_cols = [] self.verbose = verbose self.use_default_cols = cols is None # if True, even a repeated call of fit() will select string columns from X self.cols = cols # note that cols are only the columns to be encoded, feature_names_in_ are all columns self.mapping = None self.handle_unknown = handle_unknown self.handle_missing = handle_missing self._dim = None def fit(self, X, y=None, **kwargs): """Fits the encoder according to X and y. Parameters ---------- X : array-like, shape = [n_samples, n_features] Training vectors, where n_samples is the number of samples and n_features is the number of features. y : array-like, shape = [n_samples] Target values. Returns ------- self : encoder Returns self. """ self._check_fit_inputs(X, y) X, y = convert_inputs(X, y) self.feature_names_in_ = X.columns.tolist() self.n_features_in_ = len(self.feature_names_in_) self._dim = X.shape[1] self._determine_fit_columns(X) if not set(self.cols).issubset(X.columns): raise ValueError('X does not contain the columns listed in cols') if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') self._fit(X, y, **kwargs) # for finding invariant columns transform without y (as is done on the test set) X_transformed = self.transform(X, override_return_df=True) self.feature_names_out_ = X_transformed.columns.tolist() # drop all output columns with 0 variance. if self.drop_invariant: generated_cols = get_generated_cols(X, X_transformed, self.cols) self.invariant_cols = [x for x in generated_cols if X_transformed[x].var() <= self.INVARIANCE_THRESHOLD] self.feature_names_out_ = [x for x in self.feature_names_out_ if x not in self.invariant_cols] return self def _check_fit_inputs(self, X, y): if self._get_tags().get('supervised_encoder') and y is None: raise ValueError('Supervised encoders need a target for the fitting. The target cannot be None') def _check_transform_inputs(self, X): if self.handle_missing == 'error': if X[self.cols].isnull().any().any(): raise ValueError('Columns to be encoded can not contain null') if self._dim is None: raise NotFittedError('Must train encoder before it can be used to transform data.') # then make sure that it is the right size if X.shape[1] != self._dim: raise ValueError(f'Unexpected input dimension {X.shape[1]}, expected {self._dim}') def _drop_invariants(self, X: pd.DataFrame, override_return_df: bool) -> Union[np.ndarray, pd.DataFrame]: if self.drop_invariant: X = X.drop(columns=self.invariant_cols) if self.return_df or override_return_df: return X else: return X.values def _determine_fit_columns(self, X: pd.DataFrame) -> None: """ Determine columns used by encoder. Note that the implementation also deals with re-fitting the same encoder object with different columns. :param X: input data frame :return: none, sets self.cols as a side effect """ # if columns aren't passed, just use every string column if self.use_default_cols: self.cols = get_obj_cols(X) else: self.cols = convert_cols_to_list(self.cols) def get_feature_names(self) -> List[str]: warnings.warn("`get_feature_names` is deprecated in all of sklearn. Use `get_feature_names_out` instead.", category=FutureWarning) return self.get_feature_names_out() def get_feature_names_out(self, input_features=None) -> np.ndarray: """ Returns the names of all transformed / added columns. Note that in sklearn the get_feature_names_out function takes the feature_names_in as an argument and determines the output feature names using the input. A fit is usually not necessary and if so a NotFittedError is raised. We just require a fit all the time and return the fitted output columns. Returns ------- feature_names: np.ndarray A numpy array with all feature names transformed or added. Note: potentially dropped features (because the feature is constant/invariant) are not included! """ out_feats = getattr(self, "feature_names_out_", None) if not isinstance(out_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return np.array(out_feats, dtype=object) def get_feature_names_in(self) -> List[str]: """ Returns the names of all input columns present when fitting. These columns are necessary for the transform step. """ in_feats = getattr(self, "feature_names_in_", None) if not isinstance(in_feats, list): raise NotFittedError("Estimator has to be fitted to return feature names.") else: return in_feats @abstractmethod def _fit(self, X: pd.DataFrame, y: Optional[pd.Series], **kwargs): ... class SupervisedTransformerMixin(sklearn.base.TransformerMixin): def _more_tags(self): return {'supervised_encoder': True} def transform(self, X, y=None, override_return_df=False): """Perform the transformation to new categorical data. Some encoders behave differently on whether y is given or not. This is mainly due to regularisation in order to avoid overfitting. On training data transform should be called with y, on test data without. Parameters ---------- X : array-like, shape = [n_samples, n_features] y : array-like, shape = [n_samples] or None override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X, y = convert_inputs(X, y, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X, y) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X: pd.DataFrame, y: pd.Series) -> pd.DataFrame: ... def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y) class UnsupervisedTransformerMixin(sklearn.base.TransformerMixin): def transform(self, X, override_return_df=False): """Perform the transformation to new categorical data. Parameters ---------- X : array-like, shape = [n_samples, n_features] override_return_df : bool override self.return_df to force to return a data frame Returns ------- p : array or DataFrame, shape = [n_samples, n_features_out] Transformed values with encoding applied. """ # first check the type X = convert_input(X, deep=True) self._check_transform_inputs(X) if not list(self.cols): return X X = self._transform(X) return self._drop_invariants(X, override_return_df) @abstractmethod def _transform(self, X) -> pd.DataFrame: ... class TransformerWithTargetMixin: def _more_tags(self): return {'supervised_encoder': True} def fit_transform(self, X, y=None, **fit_params): """ Encoders that utilize the target must make sure that the training data are transformed with: transform(X, y) and not with: transform(X) """ if y is None: raise TypeError('fit_transform() missing argument: ''y''') return self.fit(X, y, **fit_params).transform(X, y)

JaimeArboleda

5eb7a2d6359d680bdadd0534bdb983e712a47f9c

570827e6b48737d0c9aece8aca31edd6da02c1b2

Hi! Ok, I have added both suggestions!

JaimeArboleda

scikit-learn-contrib/category_encoders

(WIP) Partial fix for getting feature names out

I think this is a partial fix for this opened issue: https://github.com/scikit-learn-contrib/category_encoders/issues/395 It remains to check the behaviour of other estimators that are not ONE_TO_ONE. Please, let me know if you like the work in progress and I will try to continue.

2023-02-23 13:33:41+00:00

2023-03-13 11:48:24+00:00

docs/source/index.rst

.. Category Encoders documentation master file, created by sphinx-quickstart on Sat Jan 16 13:08:19 2016. You can adapt this file completely to your liking, but it should at least contain the root `toctree` directive. Category Encoders ================= A set of scikit-learn-style transformers for encoding categorical variables into numeric with different techniques. While ordinal, one-hot, and hashing encoders have similar equivalents in the existing scikit-learn version, the transformers in this library all share a few useful properties: * First-class support for pandas dataframes as an input (and optionally as output) * Can explicitly configure which columns in the data are encoded by name or index, or infer non-numeric columns regardless of input type * Can drop any columns with very low variance based on training set optionally * Portability: train a transformer on data, pickle it, reuse it later and get the same thing out. * Full compatibility with sklearn pipelines, input an array-like dataset like any other transformer Usage ----- install as: .. code-block:: python pip install category_encoders or .. code-block:: python conda install -c conda-forge category_encoders To use: .. code-block:: python import category_encoders as ce encoder = ce.BackwardDifferenceEncoder(cols=[...]) encoder = ce.BaseNEncoder(cols=[...]) encoder = ce.BinaryEncoder(cols=[...]) encoder = ce.CatBoostEncoder(cols=[...]) encoder = ce.CountEncoder(cols=[...]) encoder = ce.GLMMEncoder(cols=[...]) encoder = ce.GrayEncoder(cols=[...]) encoder = ce.HashingEncoder(cols=[...]) encoder = ce.HelmertEncoder(cols=[...]) encoder = ce.JamesSteinEncoder(cols=[...]) encoder = ce.LeaveOneOutEncoder(cols=[...]) encoder = ce.MEstimateEncoder(cols=[...]) encoder = ce.OneHotEncoder(cols=[...]) encoder = ce.OrdinalEncoder(cols=[...]) encoder = ce.PolynomialEncoder(cols=[...]) encoder = ce.QuantileEncoder(cols=[...]) encoder = ce.RankHotEncoder(cols=[...]) encoder = ce.SumEncoder(cols=[...]) encoder = ce.TargetEncoder(cols=[...]) encoder = ce.WOEEncoder(cols=[...]) encoder.fit(X, y) X_cleaned = encoder.transform(X_dirty) All of these are fully compatible sklearn transformers, so they can be used in pipelines or in your existing scripts. If the cols parameter isn't passed, every non-numeric column will be converted. See below for detailed documentation Contents: .. toctree:: :maxdepth: 3 backward_difference basen binary catboost count glmm gray hashing helmert jamesstein leaveoneout mestimate onehot ordinal polynomial quantile rankhot sum summary targetencoder woe wrapper Indices and tables ================== * :ref:`genindex` * :ref:`modindex` * :ref:`search`

.. Category Encoders documentation master file, created by sphinx-quickstart on Sat Jan 16 13:08:19 2016. You can adapt this file completely to your liking, but it should at least contain the root `toctree` directive. Category Encoders ================= A set of scikit-learn-style transformers for encoding categorical variables into numeric with different techniques. While ordinal, one-hot, and hashing encoders have similar equivalents in the existing scikit-learn version, the transformers in this library all share a few useful properties: * First-class support for pandas dataframes as an input (and optionally as output) * Can explicitly configure which columns in the data are encoded by name or index, or infer non-numeric columns regardless of input type * Can drop any columns with very low variance based on training set optionally * Portability: train a transformer on data, pickle it, reuse it later and get the same thing out. * Full compatibility with sklearn pipelines, input an array-like dataset like any other transformer (\*) (\*) For full compatibility with Pipelines and ColumnTransformers, and consistent behaviour of `get_feature_names_out`, it's recommended to upgrade `sklearn` to a version at least '1.2.0' and to set output as pandas: .. code-block:: python import sklearn sklearn.set_config(transform_output="pandas") Usage ----- install as: .. code-block:: python pip install category_encoders or .. code-block:: python conda install -c conda-forge category_encoders To use: .. code-block:: python import category_encoders as ce encoder = ce.BackwardDifferenceEncoder(cols=[...]) encoder = ce.BaseNEncoder(cols=[...]) encoder = ce.BinaryEncoder(cols=[...]) encoder = ce.CatBoostEncoder(cols=[...]) encoder = ce.CountEncoder(cols=[...]) encoder = ce.GLMMEncoder(cols=[...]) encoder = ce.GrayEncoder(cols=[...]) encoder = ce.HashingEncoder(cols=[...]) encoder = ce.HelmertEncoder(cols=[...]) encoder = ce.JamesSteinEncoder(cols=[...]) encoder = ce.LeaveOneOutEncoder(cols=[...]) encoder = ce.MEstimateEncoder(cols=[...]) encoder = ce.OneHotEncoder(cols=[...]) encoder = ce.OrdinalEncoder(cols=[...]) encoder = ce.PolynomialEncoder(cols=[...]) encoder = ce.QuantileEncoder(cols=[...]) encoder = ce.RankHotEncoder(cols=[...]) encoder = ce.SumEncoder(cols=[...]) encoder = ce.TargetEncoder(cols=[...]) encoder = ce.WOEEncoder(cols=[...]) encoder.fit(X, y) X_cleaned = encoder.transform(X_dirty) All of these are fully compatible sklearn transformers, so they can be used in pipelines or in your existing scripts. If the cols parameter isn't passed, every non-numeric column will be converted. See below for detailed documentation Known issues: ---- `CategoryEncoders` internally works with `pandas DataFrames` as apposed to `sklearn` which works with `numpy arrays`. This can cause problems in `sklearn` versions prior to 1.2.0. In order to ensure full compatibility with `sklearn` set `sklearn` to also output `DataFrames`. This can be done by .. code-block::python sklearn.set_config(transform_output="pandas") for a whole project or just for a single pipeline using .. code-block::python Pipeline( steps=[ ("preprocessor", SomePreprocessor().set_output("pandas"), ("encoder", SomeEncoder()), ] ) If you experience another bug, feel free to report it on [github](https://github.com/scikit-learn-contrib/category_encoders/issues) Contents: ---- .. toctree:: :maxdepth: 3 backward_difference basen binary catboost count glmm gray hashing helmert jamesstein leaveoneout mestimate onehot ordinal polynomial quantile rankhot sum summary targetencoder woe wrapper Indices and tables ================== * :ref:`genindex` * :ref:`modindex` * :ref:`search`

JaimeArboleda

5eb7a2d6359d680bdadd0534bdb983e712a47f9c

570827e6b48737d0c9aece8aca31edd6da02c1b2

could you please introduce a section `known issues` here after usage and before contents, that states something like: """ CategoryEncoders internally works with pandas `DataFrames` as apposed to sklearn which works with numpy arrays. This can cause problems in sklearn versions prior to `1.2.0`. In order to ensure full compatibility with sklearn set sklearn to also output `DataFrames`. This can be done by ```python sklearn.set_config(transform_output="pandas") ``` for a whole project or just for a single pipeline using ```python Pipeline( steps=[ ("preprocessor", SomePreprocessor().set_output("pandas"), ("encoder", SomeEncoder()), ] ) ``` If you experience another bug feel free to report it on github https://github.com/scikit-learn-contrib/category_encoders/issues """ I think this makes it sufficiently clear and is better than the status quo. Do you agree?

PaulWestenthanner

scikit-learn-contrib/category_encoders

(WIP) Partial fix for getting feature names out

I think this is a partial fix for this opened issue: https://github.com/scikit-learn-contrib/category_encoders/issues/395 It remains to check the behaviour of other estimators that are not ONE_TO_ONE. Please, let me know if you like the work in progress and I will try to continue.

2023-02-23 13:33:41+00:00

2023-03-13 11:48:24+00:00

docs/source/index.rst

.. Category Encoders documentation master file, created by sphinx-quickstart on Sat Jan 16 13:08:19 2016. You can adapt this file completely to your liking, but it should at least contain the root `toctree` directive. Category Encoders ================= A set of scikit-learn-style transformers for encoding categorical variables into numeric with different techniques. While ordinal, one-hot, and hashing encoders have similar equivalents in the existing scikit-learn version, the transformers in this library all share a few useful properties: * First-class support for pandas dataframes as an input (and optionally as output) * Can explicitly configure which columns in the data are encoded by name or index, or infer non-numeric columns regardless of input type * Can drop any columns with very low variance based on training set optionally * Portability: train a transformer on data, pickle it, reuse it later and get the same thing out. * Full compatibility with sklearn pipelines, input an array-like dataset like any other transformer Usage ----- install as: .. code-block:: python pip install category_encoders or .. code-block:: python conda install -c conda-forge category_encoders To use: .. code-block:: python import category_encoders as ce encoder = ce.BackwardDifferenceEncoder(cols=[...]) encoder = ce.BaseNEncoder(cols=[...]) encoder = ce.BinaryEncoder(cols=[...]) encoder = ce.CatBoostEncoder(cols=[...]) encoder = ce.CountEncoder(cols=[...]) encoder = ce.GLMMEncoder(cols=[...]) encoder = ce.GrayEncoder(cols=[...]) encoder = ce.HashingEncoder(cols=[...]) encoder = ce.HelmertEncoder(cols=[...]) encoder = ce.JamesSteinEncoder(cols=[...]) encoder = ce.LeaveOneOutEncoder(cols=[...]) encoder = ce.MEstimateEncoder(cols=[...]) encoder = ce.OneHotEncoder(cols=[...]) encoder = ce.OrdinalEncoder(cols=[...]) encoder = ce.PolynomialEncoder(cols=[...]) encoder = ce.QuantileEncoder(cols=[...]) encoder = ce.RankHotEncoder(cols=[...]) encoder = ce.SumEncoder(cols=[...]) encoder = ce.TargetEncoder(cols=[...]) encoder = ce.WOEEncoder(cols=[...]) encoder.fit(X, y) X_cleaned = encoder.transform(X_dirty) All of these are fully compatible sklearn transformers, so they can be used in pipelines or in your existing scripts. If the cols parameter isn't passed, every non-numeric column will be converted. See below for detailed documentation Contents: .. toctree:: :maxdepth: 3 backward_difference basen binary catboost count glmm gray hashing helmert jamesstein leaveoneout mestimate onehot ordinal polynomial quantile rankhot sum summary targetencoder woe wrapper Indices and tables ================== * :ref:`genindex` * :ref:`modindex` * :ref:`search`

.. Category Encoders documentation master file, created by sphinx-quickstart on Sat Jan 16 13:08:19 2016. You can adapt this file completely to your liking, but it should at least contain the root `toctree` directive. Category Encoders ================= A set of scikit-learn-style transformers for encoding categorical variables into numeric with different techniques. While ordinal, one-hot, and hashing encoders have similar equivalents in the existing scikit-learn version, the transformers in this library all share a few useful properties: * First-class support for pandas dataframes as an input (and optionally as output) * Can explicitly configure which columns in the data are encoded by name or index, or infer non-numeric columns regardless of input type * Can drop any columns with very low variance based on training set optionally * Portability: train a transformer on data, pickle it, reuse it later and get the same thing out. * Full compatibility with sklearn pipelines, input an array-like dataset like any other transformer (\*) (\*) For full compatibility with Pipelines and ColumnTransformers, and consistent behaviour of `get_feature_names_out`, it's recommended to upgrade `sklearn` to a version at least '1.2.0' and to set output as pandas: .. code-block:: python import sklearn sklearn.set_config(transform_output="pandas") Usage ----- install as: .. code-block:: python pip install category_encoders or .. code-block:: python conda install -c conda-forge category_encoders To use: .. code-block:: python import category_encoders as ce encoder = ce.BackwardDifferenceEncoder(cols=[...]) encoder = ce.BaseNEncoder(cols=[...]) encoder = ce.BinaryEncoder(cols=[...]) encoder = ce.CatBoostEncoder(cols=[...]) encoder = ce.CountEncoder(cols=[...]) encoder = ce.GLMMEncoder(cols=[...]) encoder = ce.GrayEncoder(cols=[...]) encoder = ce.HashingEncoder(cols=[...]) encoder = ce.HelmertEncoder(cols=[...]) encoder = ce.JamesSteinEncoder(cols=[...]) encoder = ce.LeaveOneOutEncoder(cols=[...]) encoder = ce.MEstimateEncoder(cols=[...]) encoder = ce.OneHotEncoder(cols=[...]) encoder = ce.OrdinalEncoder(cols=[...]) encoder = ce.PolynomialEncoder(cols=[...]) encoder = ce.QuantileEncoder(cols=[...]) encoder = ce.RankHotEncoder(cols=[...]) encoder = ce.SumEncoder(cols=[...]) encoder = ce.TargetEncoder(cols=[...]) encoder = ce.WOEEncoder(cols=[...]) encoder.fit(X, y) X_cleaned = encoder.transform(X_dirty) All of these are fully compatible sklearn transformers, so they can be used in pipelines or in your existing scripts. If the cols parameter isn't passed, every non-numeric column will be converted. See below for detailed documentation Known issues: ---- `CategoryEncoders` internally works with `pandas DataFrames` as apposed to `sklearn` which works with `numpy arrays`. This can cause problems in `sklearn` versions prior to 1.2.0. In order to ensure full compatibility with `sklearn` set `sklearn` to also output `DataFrames`. This can be done by .. code-block::python sklearn.set_config(transform_output="pandas") for a whole project or just for a single pipeline using .. code-block::python Pipeline( steps=[ ("preprocessor", SomePreprocessor().set_output("pandas"), ("encoder", SomeEncoder()), ] ) If you experience another bug, feel free to report it on [github](https://github.com/scikit-learn-contrib/category_encoders/issues) Contents: ---- .. toctree:: :maxdepth: 3 backward_difference basen binary catboost count glmm gray hashing helmert jamesstein leaveoneout mestimate onehot ordinal polynomial quantile rankhot sum summary targetencoder woe wrapper Indices and tables ================== * :ref:`genindex` * :ref:`modindex` * :ref:`search`

JaimeArboleda

5eb7a2d6359d680bdadd0534bdb983e712a47f9c

570827e6b48737d0c9aece8aca31edd6da02c1b2

I think it's very good, yes. Thanks! I don't see any need for more. Let me add this piece and remove the other one so that you can merge :)

JaimeArboleda