Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
@@ -27,569 +27,133 @@ import io
|
|
27 |
import requests
|
28 |
import numpy as np
|
29 |
from urllib.parse import quote
|
|
|
30 |
|
31 |
-
#
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
{"title": "CosmosDB SQL API Client Library", "url": "https://learn.microsoft.com/en-us/python/api/overview/azure/cosmos-readme?view=azure-python", "emoji": "π»"},
|
37 |
-
{"title": "CosmosDB Index and Query Vectors", "url": "https://learn.microsoft.com/en-us/azure/cosmos-db/nosql/how-to-python-vector-index-query", "emoji": "π»"},
|
38 |
-
{"title": "CosmosDB NoSQL Materialized Views", "url": "https://learn.microsoft.com/en-us/azure/cosmos-db/nosql/materialized-views", "emoji": "π»"},
|
39 |
-
{"title": "LangChain Vector Store Guide", "url": "https://python.langchain.com/docs/integrations/vectorstores/azure_cosmos_db_no_sql/", "emoji": "π»"},
|
40 |
-
{"title": "Vector Database Prompt Engineering RAG for Python", "url": "https://learn.microsoft.com/en-us/azure/cosmos-db/vector-database?source=recommendations", "emoji": "π»"},
|
41 |
-
{"title": "MergeKit Official GitHub", "url": "https://github.com/arcee-ai/MergeKit", "emoji": "π»"},
|
42 |
-
{"title": "MergeKit Sample Usage", "url": "https://github.com/arcee-ai/MergeKit#examples", "emoji": "π"},
|
43 |
-
{"title": "DistillKit Official GitHub", "url": "https://github.com/arcee-ai/DistillKit", "emoji": "π»"},
|
44 |
-
{"title": "DistillKit Sample Usage", "url": "https://github.com/arcee-ai/DistillKit#usage", "emoji": "π"},
|
45 |
-
{"title": "arcee.ai Official Website", "url": "https://arcee.ai", "emoji": "π"},
|
46 |
-
]
|
47 |
|
48 |
# =============================================================================
|
49 |
# βββββββββββββ APP CONFIGURATION βββββββββββββ
|
50 |
# =============================================================================
|
51 |
-
|
52 |
-
title = "π GitCosmos"
|
53 |
-
helpURL = 'https://huggingface.co/awacke1'
|
54 |
-
bugURL = 'https://huggingface.co/spaces/awacke1/AzureCosmosDBUI/'
|
55 |
-
icons = 'πππ«'
|
56 |
-
st.set_page_config(
|
57 |
-
page_title=title,
|
58 |
-
page_icon=icons,
|
59 |
-
layout="wide",
|
60 |
-
initial_sidebar_state="auto",
|
61 |
-
menu_items={
|
62 |
-
'Get Help': helpURL,
|
63 |
-
'Report a bug': bugURL,
|
64 |
-
'About': title
|
65 |
-
}
|
66 |
-
)
|
67 |
-
|
68 |
-
ENDPOINT = "https://acae-afd.documents.azure.com:443/"
|
69 |
-
DATABASE_NAME = os.environ.get("COSMOS_DATABASE_NAME")
|
70 |
-
CONTAINER_NAME = os.environ.get("COSMOS_CONTAINER_NAME")
|
71 |
-
Key = os.environ.get("Key")
|
72 |
-
CosmosDBUrl = 'https://portal.azure.com/#@AaronCWackergmail.onmicrosoft.com/resource/subscriptions/003fba60-5b3f-48f4-ab36-3ed11bc40816/resourceGroups/datasets/providers/Microsoft.DocumentDB/databaseAccounts/acae-afd/dataExplorer'
|
73 |
|
74 |
# =============================================================================
|
75 |
# βββββββββββββ HELPER FUNCTIONS βββββββββββββ
|
76 |
# =============================================================================
|
77 |
-
|
78 |
-
with open(file_path, "rb") as file:
|
79 |
-
contents = file.read()
|
80 |
-
b64 = base64.b64encode(contents).decode()
|
81 |
-
file_name = os.path.basename(file_path)
|
82 |
-
return f'<a href="data:file/txt;base64,{b64}" download="{file_name}">Download {file_name} π</a>'
|
83 |
-
|
84 |
-
def generate_unique_id():
|
85 |
-
timestamp = datetime.utcnow().strftime('%Y%m%d%H%M%S%f')
|
86 |
-
unique_uuid = str(uuid.uuid4())
|
87 |
-
return f"{timestamp}-{unique_uuid}"
|
88 |
-
|
89 |
-
def generate_filename(prompt, file_type):
|
90 |
-
central = pytz.timezone('US/Central')
|
91 |
-
safe_date_time = datetime.now(central).strftime("%m%d_%H%M")
|
92 |
-
safe_prompt = re.sub(r'\W+', '', prompt)[:90]
|
93 |
-
return f"{safe_date_time}{safe_prompt}.{file_type}"
|
94 |
-
|
95 |
-
def create_file(filename, prompt, response, should_save=True):
|
96 |
-
if should_save:
|
97 |
-
with open(filename, 'w', encoding='utf-8') as file:
|
98 |
-
file.write(prompt + "\n\n" + response)
|
99 |
-
|
100 |
-
def load_file(file_name):
|
101 |
-
with open(file_name, "r", encoding='utf-8') as file:
|
102 |
-
return file.read()
|
103 |
-
|
104 |
-
def create_zip_of_files(files):
|
105 |
-
zip_name = "all_files.zip"
|
106 |
-
with zipfile.ZipFile(zip_name, 'w') as zipf:
|
107 |
-
for file in files:
|
108 |
-
zipf.write(file)
|
109 |
-
return zip_name
|
110 |
-
|
111 |
-
def preprocess_text(text):
|
112 |
-
text = text.replace('\r\n', '\\n').replace('\r', '\\n').replace('\n', '\\n')
|
113 |
-
text = text.replace('"', '\\"')
|
114 |
-
text = re.sub(r'[\t]', ' ', text)
|
115 |
-
text = re.sub(r'[^\x00-\x7F]+', '', text)
|
116 |
-
return text.strip()
|
117 |
-
|
118 |
-
def sanitize_json_text(text):
|
119 |
-
text = re.sub(r'[\x00-\x08\x0B\x0C\x0E-\x1F]', '', text)
|
120 |
-
return text.strip()
|
121 |
|
122 |
# =============================================================================
|
123 |
# βββββββββββββ COSMOS DB FUNCTIONS βββββββββββββ
|
124 |
# =============================================================================
|
125 |
-
def
|
126 |
-
|
127 |
-
|
128 |
-
def get_containers(database):
|
129 |
-
return [container['id'] for container in database.list_containers()]
|
130 |
-
|
131 |
-
def get_documents(container, limit=None):
|
132 |
-
query = "SELECT * FROM c ORDER BY c._ts DESC"
|
133 |
-
items = list(container.query_items(query=query, enable_cross_partition_query=True, max_item_count=limit))
|
134 |
-
return items
|
135 |
-
|
136 |
-
def insert_record(container, record):
|
137 |
-
try:
|
138 |
-
container.create_item(body=record)
|
139 |
-
return True, "Inserted! π"
|
140 |
-
except exceptions.CosmosHttpResponseError as e:
|
141 |
-
return False, f"HTTP error: {str(e)} π¨"
|
142 |
-
except Exception as e:
|
143 |
-
return False, f"Error: {str(e)} π±"
|
144 |
-
|
145 |
-
def update_record(container, updated_record):
|
146 |
try:
|
147 |
-
|
148 |
-
|
149 |
-
|
150 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
151 |
except Exception as e:
|
152 |
-
|
|
|
153 |
|
154 |
def delete_record(container, record):
|
|
|
155 |
try:
|
156 |
doc_id = record["id"]
|
157 |
partition_key_value = record.get("pk", doc_id)
|
158 |
-
|
159 |
container.delete_item(item=doc_id, partition_key=partition_key_value)
|
|
|
160 |
return True, f"Record {doc_id} deleted. ποΈ"
|
161 |
-
except exceptions.CosmosResourceNotFoundError:
|
|
|
162 |
return True, f"Record {doc_id} not found (already deleted). ποΈ"
|
163 |
except exceptions.CosmosHttpResponseError as e:
|
|
|
164 |
return False, f"HTTP error deleting {doc_id}: {str(e)} π¨"
|
165 |
except Exception as e:
|
|
|
166 |
return False, f"Unexpected error deleting {doc_id}: {str(e)} π±"
|
167 |
|
168 |
-
def
|
|
|
169 |
try:
|
170 |
-
|
171 |
-
|
172 |
-
|
173 |
-
|
174 |
-
|
175 |
-
|
176 |
-
|
177 |
-
|
178 |
-
|
179 |
-
|
180 |
-
|
181 |
-
|
182 |
-
|
183 |
-
|
184 |
-
|
185 |
-
|
186 |
-
|
187 |
-
|
188 |
-
|
189 |
-
def archive_current_container(database_name, container_name, client):
|
190 |
-
try:
|
191 |
-
base_dir = "./cosmos_archive"
|
192 |
-
if os.path.exists(base_dir):
|
193 |
-
shutil.rmtree(base_dir)
|
194 |
-
os.makedirs(base_dir)
|
195 |
-
db_client = client.get_database_client(database_name)
|
196 |
-
container_client = db_client.get_container_client(container_name)
|
197 |
-
items = list(container_client.read_all_items())
|
198 |
-
container_dir = os.path.join(base_dir, container_name)
|
199 |
-
os.makedirs(container_dir)
|
200 |
-
for item in items:
|
201 |
-
item_id = item.get('id', f"unknown_{datetime.now().strftime('%Y%m%d%H%M%S')}")
|
202 |
-
with open(os.path.join(container_dir, f"{item_id}.json"), 'w') as f:
|
203 |
-
json.dump(item, f, indent=2)
|
204 |
-
archive_name = f"{container_name}_archive_{datetime.now().strftime('%Y%m%d%H%M%S')}"
|
205 |
-
shutil.make_archive(archive_name, 'zip', base_dir)
|
206 |
-
return get_download_link(f"{archive_name}.zip")
|
207 |
except Exception as e:
|
208 |
-
|
209 |
-
|
210 |
-
def create_new_container(database, container_id, partition_key_path):
|
211 |
-
try:
|
212 |
-
container = database.create_container(
|
213 |
-
id=container_id,
|
214 |
-
partition_key=PartitionKey(path=partition_key_path)
|
215 |
-
)
|
216 |
-
return container
|
217 |
-
except exceptions.CosmosResourceExistsError:
|
218 |
-
return database.get_container_client(container_id)
|
219 |
-
except exceptions.CosmosHttpResponseError as e:
|
220 |
-
st.error(f"Error creating container: {str(e)}")
|
221 |
-
return None
|
222 |
-
|
223 |
-
# =============================================================================
|
224 |
-
# βββββββββββββ GITHUB FUNCTIONS βββββββββββββ
|
225 |
-
# =============================================================================
|
226 |
-
def download_github_repo(url, local_path):
|
227 |
-
if os.path.exists(local_path):
|
228 |
-
shutil.rmtree(local_path)
|
229 |
-
Repo.clone_from(url, local_path)
|
230 |
-
|
231 |
-
def create_zip_file(source_dir, output_filename):
|
232 |
-
shutil.make_archive(output_filename, 'zip', source_dir)
|
233 |
-
|
234 |
-
def create_repo(g, repo_name):
|
235 |
-
user = g.get_user()
|
236 |
-
return user.create_repo(repo_name)
|
237 |
-
|
238 |
-
def push_to_github(local_path, repo, github_token):
|
239 |
-
repo_url = f"https://{github_token}@github.com/{repo.full_name}.git"
|
240 |
-
local_repo = Repo(local_path)
|
241 |
-
if 'origin' in [remote.name for remote in local_repo.remotes]:
|
242 |
-
origin = local_repo.remote('origin')
|
243 |
-
origin.set_url(repo_url)
|
244 |
-
else:
|
245 |
-
origin = local_repo.create_remote('origin', repo_url)
|
246 |
-
if not local_repo.heads:
|
247 |
-
local_repo.git.checkout('-b', 'main')
|
248 |
-
current_branch = 'main'
|
249 |
-
else:
|
250 |
-
current_branch = local_repo.active_branch.name
|
251 |
-
local_repo.git.add(A=True)
|
252 |
-
if local_repo.is_dirty():
|
253 |
-
local_repo.git.commit('-m', 'Initial commit')
|
254 |
-
origin.push(refspec=f'{current_branch}:{current_branch}')
|
255 |
-
|
256 |
-
# =============================================================================
|
257 |
-
# βββββββββββββ FILE & MEDIA MANAGEMENT FUNCTIONS βββββββββββββ
|
258 |
-
# =============================================================================
|
259 |
-
def display_file_viewer(file_path):
|
260 |
-
content = load_file(file_path)
|
261 |
-
if content:
|
262 |
-
st.markdown("### π File Viewer")
|
263 |
-
st.markdown(f"**{file_path}**")
|
264 |
-
file_stats = os.stat(file_path)
|
265 |
-
st.markdown(f"**Mod:** {datetime.fromtimestamp(file_stats.st_mtime).strftime('%Y-%m-%d %H:%M:%S')} | **Size:** {file_stats.st_size} bytes")
|
266 |
-
st.markdown("---")
|
267 |
-
st.markdown(content)
|
268 |
-
st.download_button("β¬οΈ", data=content, file_name=os.path.basename(file_path), mime="text/markdown")
|
269 |
-
|
270 |
-
def display_file_editor(file_path):
|
271 |
-
if 'file_content' not in st.session_state:
|
272 |
-
st.session_state.file_content = {}
|
273 |
-
if file_path not in st.session_state.file_content:
|
274 |
-
content = load_file(file_path)
|
275 |
-
if content:
|
276 |
-
st.session_state.file_content[file_path] = content
|
277 |
-
st.markdown("### βοΈ Edit File")
|
278 |
-
st.markdown(f"**Editing:** {file_path}")
|
279 |
-
new_content = st.text_area("Edit:", value=st.session_state.file_content.get(file_path, ""), height=400, key=f"editor_{hash(file_path)}")
|
280 |
-
col1, col2 = st.columns([1, 5])
|
281 |
-
with col1:
|
282 |
-
if st.button("πΎ Save"):
|
283 |
-
sanitized = sanitize_json_text(new_content)
|
284 |
-
try:
|
285 |
-
json.loads(sanitized)
|
286 |
-
with open(file_path, 'w', encoding='utf-8') as file:
|
287 |
-
file.write(sanitized)
|
288 |
-
st.session_state.file_content[file_path] = sanitized
|
289 |
-
st.success("Saved! π")
|
290 |
-
time.sleep(1)
|
291 |
-
st.rerun()
|
292 |
-
except Exception as e:
|
293 |
-
st.error(f"Save error: {str(e)}")
|
294 |
-
with col2:
|
295 |
-
st.download_button("β¬οΈ", data=new_content, file_name=os.path.basename(file_path), mime="text/markdown")
|
296 |
|
297 |
-
|
298 |
-
if 'file_view_mode' not in st.session_state:
|
299 |
-
st.session_state.file_view_mode = None
|
300 |
-
if 'current_file' not in st.session_state:
|
301 |
-
st.session_state.current_file = None
|
302 |
-
all_files = sorted(glob.glob("*.md"), reverse=True)
|
303 |
-
st.sidebar.subheader("π Files")
|
304 |
-
if st.sidebar.button("π Delete All"):
|
305 |
-
for file in all_files:
|
306 |
-
os.remove(file)
|
307 |
-
st.session_state.file_content = {}
|
308 |
-
st.session_state.current_file = None
|
309 |
-
st.session_state.file_view_mode = None
|
310 |
-
st.rerun()
|
311 |
-
if st.sidebar.button("β¬οΈ Download All"):
|
312 |
-
zip_file = create_zip_of_files(all_files)
|
313 |
-
st.sidebar.markdown(get_download_link(zip_file), unsafe_allow_html=True)
|
314 |
-
for file in all_files:
|
315 |
-
col1, col2, col3, col4 = st.sidebar.columns([1, 3, 1, 1])
|
316 |
-
with col1:
|
317 |
-
if st.button("π", key=f"view_{file}"):
|
318 |
-
st.session_state.current_file = file
|
319 |
-
st.session_state.file_view_mode = 'view'
|
320 |
-
st.rerun()
|
321 |
-
with col2:
|
322 |
-
st.markdown(get_download_link(file), unsafe_allow_html=True)
|
323 |
-
with col3:
|
324 |
-
if st.button("π", key=f"edit_{file}"):
|
325 |
-
st.session_state.current_file = file
|
326 |
-
st.session_state.file_view_mode = 'edit'
|
327 |
-
st.rerun()
|
328 |
-
with col4:
|
329 |
-
if st.button("π", key=f"delete_{file}"):
|
330 |
-
os.remove(file)
|
331 |
-
if file in st.session_state.file_content:
|
332 |
-
del st.session_state.file_content[file]
|
333 |
-
if st.session_state.current_file == file:
|
334 |
-
st.session_state.current_file = None
|
335 |
-
st.session_state.file_view_mode = None
|
336 |
-
st.rerun()
|
337 |
-
st.sidebar.subheader("External Help Links")
|
338 |
-
for link in external_links:
|
339 |
-
st.sidebar.markdown(f"{link['emoji']} [{link['title']}]({link['url']})", unsafe_allow_html=True)
|
340 |
-
if st.session_state.current_file:
|
341 |
-
if st.session_state.file_view_mode == 'view':
|
342 |
-
display_file_viewer(st.session_state.current_file)
|
343 |
-
elif st.session_state.file_view_mode == 'edit':
|
344 |
-
display_file_editor(st.session_state.current_file)
|
345 |
|
346 |
# =============================================================================
|
347 |
# βββββββββββββ UI FUNCTIONS βββββββββββββ
|
348 |
# =============================================================================
|
349 |
-
def
|
350 |
-
st.markdown("###
|
351 |
-
documents
|
352 |
-
|
353 |
-
|
354 |
-
|
355 |
-
st.
|
356 |
-
return
|
357 |
-
|
358 |
-
if 'saved_docs' not in st.session_state:
|
359 |
-
st.session_state.saved_docs = {}
|
360 |
|
361 |
-
|
362 |
-
|
363 |
-
|
364 |
-
|
365 |
-
|
366 |
-
|
367 |
-
|
368 |
-
|
369 |
-
edited_content = st.text_area("Edit JSON", value=initial_value, height=300, key=doc_key)
|
370 |
-
col_save, col_delete = st.columns(2)
|
371 |
-
with col_save:
|
372 |
-
if st.button("πΎ Save", key=f"save_{doc['id']}"):
|
373 |
-
try:
|
374 |
-
cleaned_content = sanitize_json_text(edited_content)
|
375 |
-
updated_doc = json.loads(cleaned_content)
|
376 |
-
updated_doc['id'] = doc['id']
|
377 |
-
updated_doc['pk'] = doc.get('pk', doc['id'])
|
378 |
-
for field in ['_ts', '_rid', '_self', '_etag', '_attachments']:
|
379 |
-
updated_doc.pop(field, None)
|
380 |
-
success, message = update_record(container, updated_doc)
|
381 |
-
if success:
|
382 |
-
st.success(f"Saved {doc['id']}")
|
383 |
-
st.session_state.saved_docs[doc['id']] = json.dumps(updated_doc, indent=2)
|
384 |
-
st.rerun()
|
385 |
-
else:
|
386 |
-
st.error(message)
|
387 |
-
except json.JSONDecodeError as e:
|
388 |
-
st.error(f"Invalid JSON format: {str(e)}\nProblematic input:\n{cleaned_content}")
|
389 |
-
except Exception as e:
|
390 |
-
st.error(f"Save error: {str(e)}")
|
391 |
-
with col_delete:
|
392 |
-
if st.button("ποΈ Delete", key=f"delete_{doc['id']}"):
|
393 |
-
success, message = delete_record(container, doc)
|
394 |
-
if success:
|
395 |
-
st.success(message)
|
396 |
-
if doc['id'] in st.session_state.saved_docs:
|
397 |
-
del st.session_state.saved_docs[doc['id']]
|
398 |
-
st.rerun()
|
399 |
-
else:
|
400 |
-
st.error(message)
|
401 |
-
|
402 |
-
def new_item_default(container):
|
403 |
-
new_id = generate_unique_id()
|
404 |
-
default_doc = {
|
405 |
-
"id": new_id,
|
406 |
-
"pk": new_id,
|
407 |
-
"name": "New Document",
|
408 |
-
"content": "Start editing here...",
|
409 |
-
"timestamp": datetime.now().isoformat(),
|
410 |
-
"type": "sample"
|
411 |
-
}
|
412 |
-
success, message = insert_record(container, default_doc)
|
413 |
-
if success:
|
414 |
-
st.success("New document created! β¨")
|
415 |
-
st.rerun()
|
416 |
-
else:
|
417 |
-
st.error(f"Error creating new item: {message}")
|
418 |
-
|
419 |
-
def add_field_to_doc():
|
420 |
-
key = st.session_state.new_field_key
|
421 |
-
value = st.session_state.new_field_value
|
422 |
-
try:
|
423 |
-
doc = json.loads(st.session_state.doc_editor)
|
424 |
-
doc[key] = value
|
425 |
-
st.session_state.doc_editor = json.dumps(doc, indent=2)
|
426 |
-
container = st.session_state.current_container
|
427 |
-
success, message = update_record(container, doc)
|
428 |
-
if success:
|
429 |
-
st.success(f"Added field {key} π")
|
430 |
-
else:
|
431 |
-
st.error(message)
|
432 |
-
except Exception as e:
|
433 |
-
st.error(f"Error adding field: {str(e)}")
|
434 |
-
|
435 |
-
def new_ai_record(container):
|
436 |
-
new_id = generate_unique_id()
|
437 |
-
default_doc = {
|
438 |
-
"id": new_id,
|
439 |
-
"pk": new_id,
|
440 |
-
"name": "AI Modality Record",
|
441 |
-
"function_url": "https://example.com/function",
|
442 |
-
"input_text": "### Input (markdown)\n\nType your input here.",
|
443 |
-
"output_text": "### Output (markdown)\n\nResult will appear here.",
|
444 |
-
"timestamp": datetime.now().isoformat(),
|
445 |
-
"type": "ai_modality"
|
446 |
-
}
|
447 |
-
success, message = insert_record(container, default_doc)
|
448 |
-
if success:
|
449 |
-
st.success("New AI modality record created! π‘")
|
450 |
-
st.rerun()
|
451 |
-
else:
|
452 |
-
st.error(f"Error creating AI record: {message}")
|
453 |
-
|
454 |
-
def new_links_record(container):
|
455 |
-
new_id = generate_unique_id()
|
456 |
-
links_md = "\n".join([f"- {link['emoji']} [{link['title']}]({link['url']})" for link in external_links])
|
457 |
-
default_doc = {
|
458 |
-
"id": new_id,
|
459 |
-
"pk": new_id,
|
460 |
-
"name": "Portal Links Record",
|
461 |
-
"function_url": "",
|
462 |
-
"input_text": links_md,
|
463 |
-
"output_text": "",
|
464 |
-
"timestamp": datetime.now().isoformat(),
|
465 |
-
"type": "ai_modality"
|
466 |
-
}
|
467 |
-
success, message = insert_record(container, default_doc)
|
468 |
-
if success:
|
469 |
-
st.success("New Portal Links record created! π")
|
470 |
-
st.rerun()
|
471 |
-
else:
|
472 |
-
st.error(f"Error creating links record: {message}")
|
473 |
-
|
474 |
-
def vector_keyword_search(keyword, doc):
|
475 |
-
keyword = keyword.lower()
|
476 |
-
for key, value in doc.items():
|
477 |
-
if isinstance(value, str) and keyword in value.lower():
|
478 |
-
return True
|
479 |
-
return False
|
480 |
-
|
481 |
-
def search_documents_ui(container):
|
482 |
-
with st.sidebar.form("search_form"):
|
483 |
-
keyword = st.text_input("Search Keyword", key="search_keyword")
|
484 |
-
col1, col2 = st.columns(2)
|
485 |
-
with col1:
|
486 |
-
search_submitted = st.form_submit_button("π Search")
|
487 |
-
with col2:
|
488 |
-
clear_submitted = st.form_submit_button("ποΈ Clear")
|
489 |
-
if search_submitted and keyword:
|
490 |
-
st.session_state.active_search = keyword # Use a separate key
|
491 |
-
st.rerun()
|
492 |
-
if clear_submitted:
|
493 |
-
if 'active_search' in st.session_state:
|
494 |
-
del st.session_state.active_search
|
495 |
-
st.rerun()
|
496 |
-
|
497 |
-
def validate_and_preprocess_image(file_data, target_size=(576, 1024)):
|
498 |
-
try:
|
499 |
-
if isinstance(file_data, bytes):
|
500 |
-
img = Image.open(io.BytesIO(file_data))
|
501 |
-
elif hasattr(file_data, 'read'):
|
502 |
-
if hasattr(file_data, 'seek'):
|
503 |
-
file_data.seek(0)
|
504 |
-
img = Image.open(file_data)
|
505 |
else:
|
506 |
-
|
507 |
-
if img.mode != 'RGB':
|
508 |
-
img = img.convert('RGB')
|
509 |
-
aspect_ratio = img.size[0] / img.size[1]
|
510 |
-
if aspect_ratio > target_size[0] / target_size[1]:
|
511 |
-
new_width = target_size[0]
|
512 |
-
new_height = int(new_width / aspect_ratio)
|
513 |
-
else:
|
514 |
-
new_height = target_size[1]
|
515 |
-
new_width = int(new_height * aspect_ratio)
|
516 |
-
new_width = (new_width // 2) * 2
|
517 |
-
new_height = (new_height // 2) * 2
|
518 |
-
resized_img = img.resize((new_width, new_height), Image.Resampling.LANCZOS)
|
519 |
-
final_img = Image.new('RGB', target_size, (255, 255, 255))
|
520 |
-
paste_x = (target_size[0] - new_width) // 2
|
521 |
-
paste_y = (target_size[1] - new_height) // 2
|
522 |
-
final_img.paste(resized_img, (paste_x, paste_y))
|
523 |
-
return final_img
|
524 |
-
except Exception as e:
|
525 |
-
st.error(f"Image error: {str(e)}")
|
526 |
-
return None
|
527 |
|
528 |
-
def
|
529 |
-
st.markdown("###
|
530 |
-
|
|
|
531 |
with col1:
|
532 |
-
|
|
|
|
|
533 |
with col2:
|
534 |
-
st.
|
535 |
-
|
536 |
-
|
537 |
-
|
538 |
-
|
539 |
-
|
540 |
-
if uploaded_file:
|
541 |
-
file_data = uploaded_file.read()
|
542 |
-
preview1, preview2 = st.columns(2)
|
543 |
-
with preview1:
|
544 |
-
st.write("Original")
|
545 |
-
st.image(Image.open(io.BytesIO(file_data)), use_column_width=True)
|
546 |
-
with preview2:
|
547 |
-
proc_img = validate_and_preprocess_image(io.BytesIO(file_data))
|
548 |
-
if proc_img:
|
549 |
-
st.write("Processed")
|
550 |
-
st.image(proc_img, use_column_width=True)
|
551 |
else:
|
552 |
-
|
553 |
-
|
554 |
-
|
555 |
-
with tempfile.NamedTemporaryFile(suffix='.png', delete=False) as temp_file:
|
556 |
-
proc_img.save(temp_file.name, format='PNG')
|
557 |
-
try:
|
558 |
-
client = Client("awacke1/stable-video-diffusion", hf_token=os.environ.get("HUGGINGFACE_TOKEN"))
|
559 |
-
result = client.predict(
|
560 |
-
image=temp_file.name,
|
561 |
-
seed=seed if seed is not None else int(time.time() * 1000),
|
562 |
-
randomize_seed=seed is None,
|
563 |
-
motion_bucket_id=motion,
|
564 |
-
fps_id=fps,
|
565 |
-
api_name="/video"
|
566 |
-
)
|
567 |
-
video_path = result[0].get('video') if isinstance(result[0], dict) else None
|
568 |
-
if video_path and os.path.exists(video_path):
|
569 |
-
video_filename = f"generated_video_{datetime.now().strftime('%Y%m%d_%H%M%S')}.mp4"
|
570 |
-
shutil.copy(video_path, video_filename)
|
571 |
-
st.success(f"Video generated! π")
|
572 |
-
st.video(video_filename)
|
573 |
-
if container:
|
574 |
-
video_record = {
|
575 |
-
"id": generate_unique_id(),
|
576 |
-
"pk": generate_unique_id(),
|
577 |
-
"type": "generated_video",
|
578 |
-
"filename": video_filename,
|
579 |
-
"seed": seed if seed is not None else "random",
|
580 |
-
"motion": motion,
|
581 |
-
"fps": fps,
|
582 |
-
"timestamp": datetime.now().isoformat()
|
583 |
-
}
|
584 |
-
success, message = insert_record(container, video_record)
|
585 |
-
if success:
|
586 |
-
st.success("DB record saved!")
|
587 |
-
else:
|
588 |
-
st.error(f"DB error: {message}")
|
589 |
-
except Exception as e:
|
590 |
-
st.error(f"Video gen error: {str(e)}")
|
591 |
-
finally:
|
592 |
-
os.unlink(temp_file.name)
|
593 |
|
594 |
# =============================================================================
|
595 |
# βββββββββββββ MAIN FUNCTION βββββββββββββ
|
@@ -611,6 +175,14 @@ def main():
|
|
611 |
# Sidebar: Hierarchical Navigation
|
612 |
st.sidebar.title("π Navigator")
|
613 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
614 |
# Databases Section
|
615 |
st.sidebar.subheader("ποΈ Databases")
|
616 |
if "client" not in st.session_state:
|
@@ -649,10 +221,6 @@ def main():
|
|
649 |
|
650 |
# Actions Section
|
651 |
st.sidebar.subheader("βοΈ Actions")
|
652 |
-
if st.session_state.current_container:
|
653 |
-
if st.sidebar.button("π¦ Export Container"):
|
654 |
-
download_link = archive_current_container(st.session_state.selected_database, st.session_state.selected_container, st.session_state.client)
|
655 |
-
st.sidebar.markdown(download_link, unsafe_allow_html=True) if download_link.startswith('<a') else st.sidebar.error(download_link)
|
656 |
|
657 |
# Items Section
|
658 |
st.sidebar.subheader("π Items")
|
|
|
27 |
import requests
|
28 |
import numpy as np
|
29 |
from urllib.parse import quote
|
30 |
+
import logging # Added for logging
|
31 |
|
32 |
+
# Configure logging
|
33 |
+
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
|
34 |
+
logger = logging.getLogger(__name__)
|
35 |
+
|
36 |
+
# ... (Previous imports and external links remain unchanged)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
37 |
|
38 |
# =============================================================================
|
39 |
# βββββββββββββ APP CONFIGURATION βββββββββββββ
|
40 |
# =============================================================================
|
41 |
+
# ... (App configuration remains unchanged)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
42 |
|
43 |
# =============================================================================
|
44 |
# βββββββββββββ HELPER FUNCTIONS βββββββββββββ
|
45 |
# =============================================================================
|
46 |
+
# ... (Previous helper functions remain unchanged)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
47 |
|
48 |
# =============================================================================
|
49 |
# βββββββββββββ COSMOS DB FUNCTIONS βββββββββββββ
|
50 |
# =============================================================================
|
51 |
+
def vector_search(container, query, search_type="basic", limit=10):
|
52 |
+
"""Perform vector search on Cosmos DB container with different search types."""
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
53 |
try:
|
54 |
+
if search_type == "basic":
|
55 |
+
sql_query = f"SELECT TOP {limit} * FROM c WHERE CONTAINS(c.content, '{query}', true) ORDER BY c._ts DESC"
|
56 |
+
elif search_type == "exact":
|
57 |
+
sql_query = f"SELECT TOP {limit} * FROM c WHERE c.content = '{query}' ORDER BY c._ts DESC"
|
58 |
+
elif search_type == "prefix":
|
59 |
+
sql_query = f"SELECT TOP {limit} * FROM c WHERE STARTSWITH(c.content, '{query}', true) ORDER BY c._ts DESC"
|
60 |
+
else:
|
61 |
+
raise ValueError("Unsupported search type")
|
62 |
+
|
63 |
+
logger.info(f"Executing vector search with query: {sql_query}")
|
64 |
+
items = list(container.query_items(query=sql_query, enable_cross_partition_query=True))
|
65 |
+
return items, sql_query
|
66 |
except Exception as e:
|
67 |
+
logger.error(f"Vector search error: {str(e)}")
|
68 |
+
return [], f"Error: {str(e)}"
|
69 |
|
70 |
def delete_record(container, record):
|
71 |
+
"""Delete a record from Cosmos DB with detailed logging."""
|
72 |
try:
|
73 |
doc_id = record["id"]
|
74 |
partition_key_value = record.get("pk", doc_id)
|
75 |
+
logger.info(f"Attempting to delete record {doc_id} with partition key {partition_key_value}")
|
76 |
container.delete_item(item=doc_id, partition_key=partition_key_value)
|
77 |
+
logger.info(f"Successfully deleted record {doc_id}")
|
78 |
return True, f"Record {doc_id} deleted. ποΈ"
|
79 |
+
except exceptions.CosmosResourceNotFoundError as e:
|
80 |
+
logger.warning(f"Record {doc_id} not found (already deleted): {str(e)}")
|
81 |
return True, f"Record {doc_id} not found (already deleted). ποΈ"
|
82 |
except exceptions.CosmosHttpResponseError as e:
|
83 |
+
logger.error(f"HTTP error deleting {doc_id}: {str(e)}")
|
84 |
return False, f"HTTP error deleting {doc_id}: {str(e)} π¨"
|
85 |
except Exception as e:
|
86 |
+
logger.error(f"Unexpected error deleting {doc_id}: {str(e)}")
|
87 |
return False, f"Unexpected error deleting {doc_id}: {str(e)} π±"
|
88 |
|
89 |
+
def import_database(client, file_path):
|
90 |
+
"""Import a Cosmos DB database from a zip file."""
|
91 |
try:
|
92 |
+
with zipfile.ZipFile(file_path, 'r') as zip_ref:
|
93 |
+
zip_ref.extractall("./temp_import")
|
94 |
+
for db_file in glob.glob("./temp_import/*.json"):
|
95 |
+
with open(db_file, 'r') as f:
|
96 |
+
db_data = json.load(f)
|
97 |
+
db_id = db_data.get("id")
|
98 |
+
if db_id:
|
99 |
+
database = client.create_database_if_not_exists(id=db_id)
|
100 |
+
for container_data in db_data.get("containers", []):
|
101 |
+
container_id = container_data.get("id")
|
102 |
+
partition_key = container_data.get("partitionKey", "/pk")
|
103 |
+
container = database.create_container_if_not_exists(
|
104 |
+
id=container_id, partition_key=PartitionKey(path=partition_key)
|
105 |
+
)
|
106 |
+
for item in container_data.get("items", []):
|
107 |
+
container.upsert_item(item)
|
108 |
+
shutil.rmtree("./temp_import")
|
109 |
+
return True, "Database imported successfully! π₯"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
110 |
except Exception as e:
|
111 |
+
logger.error(f"Import error: {str(e)}")
|
112 |
+
return False, f"Import error: {str(e)} π±"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
113 |
|
114 |
+
# ... (Other Cosmos DB functions remain unchanged)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
115 |
|
116 |
# =============================================================================
|
117 |
# βββββββββββββ UI FUNCTIONS βββββββββββββ
|
118 |
# =============================================================================
|
119 |
+
def vector_search_ui(container):
|
120 |
+
st.sidebar.markdown("### π Vector Search")
|
121 |
+
st.sidebar.markdown("*Search documents using vector-based techniques in Cosmos DB.*")
|
122 |
+
with st.sidebar.form("vector_search_form"):
|
123 |
+
query = st.text_input("Enter search query", key="vector_query")
|
124 |
+
search_type = st.selectbox("Search Type", ["basic", "exact", "prefix"], help="Basic: Contains keyword, Exact: Exact match, Prefix: Starts with")
|
125 |
+
search_submitted = st.form_submit_button("π Run Search")
|
|
|
|
|
|
|
|
|
126 |
|
127 |
+
if search_submitted and query:
|
128 |
+
results, executed_query = vector_search(container, query, search_type)
|
129 |
+
st.markdown("#### Vector Search Results")
|
130 |
+
st.text(f"Executed Query: {executed_query}")
|
131 |
+
if isinstance(results, list) and results:
|
132 |
+
for doc in results:
|
133 |
+
with st.expander(f"{doc.get('name', 'Unnamed')} - {doc.get('timestamp', 'No timestamp')}"):
|
134 |
+
st.json(doc)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
135 |
else:
|
136 |
+
st.info("No results found or an error occurred.")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
137 |
|
138 |
+
def import_export_ui(client, database_name, container_name):
|
139 |
+
st.sidebar.markdown("### π¦ Import/Export DB")
|
140 |
+
st.sidebar.markdown("*Easily import or export your Cosmos DB databases.*")
|
141 |
+
col1, col2 = st.sidebar.columns(2)
|
142 |
with col1:
|
143 |
+
if st.button("π€ Export"):
|
144 |
+
download_link = archive_current_container(database_name, container_name, client)
|
145 |
+
st.sidebar.markdown(download_link, unsafe_allow_html=True) if download_link.startswith('<a') else st.sidebar.error(download_link)
|
146 |
with col2:
|
147 |
+
uploaded_file = st.file_uploader("π₯ Import", type=["zip"], key="import_file")
|
148 |
+
if uploaded_file and st.button("Import Now"):
|
149 |
+
success, message = import_database(client, uploaded_file)
|
150 |
+
if success:
|
151 |
+
st.sidebar.success(message)
|
152 |
+
st.rerun()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
153 |
else:
|
154 |
+
st.sidebar.error(message)
|
155 |
+
|
156 |
+
# ... (Other UI functions remain unchanged)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
157 |
|
158 |
# =============================================================================
|
159 |
# βββββββββββββ MAIN FUNCTION βββββββββββββ
|
|
|
175 |
# Sidebar: Hierarchical Navigation
|
176 |
st.sidebar.title("π Navigator")
|
177 |
|
178 |
+
# Vector Search (Moved to top)
|
179 |
+
if st.session_state.current_container:
|
180 |
+
vector_search_ui(st.session_state.current_container)
|
181 |
+
|
182 |
+
# Import/Export (Second feature)
|
183 |
+
if st.session_state.current_container:
|
184 |
+
import_export_ui(st.session_state.client, st.session_state.selected_database, st.session_state.selected_container)
|
185 |
+
|
186 |
# Databases Section
|
187 |
st.sidebar.subheader("ποΈ Databases")
|
188 |
if "client" not in st.session_state:
|
|
|
221 |
|
222 |
# Actions Section
|
223 |
st.sidebar.subheader("βοΈ Actions")
|
|
|
|
|
|
|
|
|
224 |
|
225 |
# Items Section
|
226 |
st.sidebar.subheader("π Items")
|