Major update. Support for 15 LLMs, World Flora Online taxonomy validation, geolocation, 2 OCR methods, significant UI changes, stability improvements, consistent JSON parsing

.gitignore

@@ -24,7 +24,6 @@ runs/
 KP_Test/
 
 # VV Specific
-.streamlit*/
 demo/demo_output/*
 demo/validation_configs/*
 /bin/*

.streamlit/config.toml

@@ -0,0 +1,3 @@
+[theme]
+    base = "dark"
+    primaryColor = "#00ff00"

.streamlit/secrets.toml

@@ -0,0 +1,25 @@
+# Streamlit secrets
+
+GOOGLE_APPLICATION_CREDENTIALS_JSON = '''
+{
+  "type": "service_account",
+  "project_id": "directed-curve-401601",
+  "private_key_id": "53457e69bc362dd27bacb83228839a93c7361640",
+  "private_key": "-----BEGIN PRIVATE KEY-----\nMIIEvQIBADANBgkqhkiG9w0BAQEFAASCBKcwggSjAgEAAoIBAQCnnRop1c/YAwQZ\nWuXJn9brvNUVJDuf+zmSuhZt9J3J4xnbeGDVrqJ78YOfB/W9HfdywtSXP/m0q2LC\nIeatg+Kc4aFrfjQueexNr6pYV+TklU+vhFMEJP8AJcUCLC5YymYntqUWHTMy4swg\n/JsfDQPB5KgxIC7yo907QbF7KAD/VHiPqRje8J5OV+5yAnvEmdcc3E/lXvMEKVo3\nfpDf3IO6knpnvlxtcMlKCLFb23m070jJhuUqjbBS8rI8qKZI9mUutMPXGOeVucIV\nkFfRoTCCIlRlSnrq0DewPx+RQkXEJUVSQbTi9MLsXIqmu2RkwpusmBX3GFuUXBwk\n33/aJwqfAgMBAAECggEASE+JFYpWECGA97VU1C6+iFYfcW7/wC2gcXIRsh8poYr1\nwU7GK/maxLD1cTr8Y7nVYq+2CoiBoFWeFhWNEMeIkNjwjFDWj0RvUzaZXWKR76GQ\nzq5eKDgTwQsKWr0Xw5uSBckQoKE5UNcybHq+K1+wgso0W3lAd9KE44ccW/7uqzgg\n/ls/rSlFff0QU5yLawBHWHlNwDhAhElPeZHQNxJ4Cc46iy9z2cSBHMpOWxUBWh1Z\nOpRBsH8mnUzzYG7kiDQd6+wS1IJ5zPSM5b28GGChAbziWV0nVMKwvmWHShLk11N4\nxsSLLNFlA2nbRL1FxFDyMLfgBDbN517eTEyjlbhoNQKBgQDj0Cf+L+JxFQvcNPAR\nE/I7b2K0qDKiTBWsi7LLQxqUpGoVCAL0kixtFr8Lr+Jm+0ih0VBJUVYMePDqj3C1\nTeidAmeP0UPxDwDNh2VystrRKpRh9Q9tx0vvtvWhDfsZmTawjoC+/HFonppsQW+k\n3LMCzaXJAewIF5MhOa3Bw/MnuwKBgQC8WipEf3dBm4OwjazuGdQ6ywlNbfxFO7GD\nOZ4A+V+dldTpP20KFkV9D4hUXTgJw8fzhVJtEQs1gqm3E5AKhbY4VrsJFyw/0DWf\nUafcadMtKeTgr1p+28KiY+/DZLHwPTNwYun0dkhyPabSkrneLzVBIKmvxbKsqayl\nuWZuZBtgbQKBgGmo1aOzGddQ55tOgjmpgqPGWYVorXNQzoykjlxv1xgx2XG/sX81\nfJ1r+1e0T9FrophKuG4I51wJi9N+i8vLQqPvxZCSgmx/hOPOdGLqSUrR8NguDOXh\nlXqs5Oe50EhjI1ZoAZaTBxZ62Sx0fQ99vRkDwGymE479QCQc2FYSR+iZAoGADRgH\nkDN8SudQxFlVCEhrZJrwzp0yyG1Z7dF3loG/j+j2oDB0mBrsQcFurM6+Mo+Vuuwt\nyr9AlDu9KKkHfnPFMy10WzFtNhJ3Z9f5X21L7PvYmXYXSHJmmXtxE5hQdm8KwGlf\n8kq2xppampDwsu19OkXXBg0sWqyBABRfdpUnAhECgYEAme0nq52gw9TbDqrfqQ/r\nT4NzjX7AaI0/oXekeW2Dm+NIx+umgggsx1/Hf3q5PWiy/x8ViTyoWYFdgaM72yCW\n9gVCTXkE/IHsTq9zX4khsv/UHTkHFf3VADCEjW+ht2nYeCBN1HwtLjdRidwvrSrC\nfL666ill9fXvSFnC4KsEfxE=\n-----END PRIVATE KEY-----\n",
+  "client_email": "vouchervision@directed-curve-401601.iam.gserviceaccount.com",
+  "client_id": "100706991621579323653",
+  "auth_uri": "https://accounts.google.com/o/oauth2/auth",
+  "token_uri": "https://oauth2.googleapis.com/token",
+  "auth_provider_x509_cert_url": "https://www.googleapis.com/oauth2/v1/certs",
+  "client_x509_cert_url": "https://www.googleapis.com/robot/v1/metadata/x509/vouchervision%40directed-curve-401601.iam.gserviceaccount.com",
+  "universe_domain": "googleapis.com"
+}
+'''
+
+OPENAI_API_KEY = "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"
+AZURE_API_VERSION = "2023-05-15"
+AZURE_API_KEY = "2a841cc7fde64109975d84ea0f9685d4"
+AZURE_API_BASE = "https://api.umgpt.umich.edu/azure-openai-api"
+AZURE_ORGANIZATION = "123456"
+AZURE_API_TYPE = "azure"
+GOOGLE_PALM_API = "AIzaSyCiTxmzxi6vHih0jbAf8wtSX9_dBBXttKA"

app.py

@@ -71,7 +71,7 @@ def content_input_images_hf():
             st.session_state.config['leafmachine']['project']['dir_images_local'] = st.session_state['dir_uploaded_images']
 
             n_images = len([f for f in os.listdir(st.session_state.config['leafmachine']['project']['dir_images_local']) if os.path.isfile(os.path.join(st.session_state.config['leafmachine']['project']['dir_images_local'], f))])
-            st.session_state['processing_add_on'] = f" {n_images} Images"
+            st.session_state['processing_add_on'] = f" {n_images} images"
             uploaded_files = None
             st.session_state['uploader_idk'] += 1
             st.info(f"Processing **{n_images}** images from {st.session_state.config['leafmachine']['project']['dir_images_local']}")
@@ -96,13 +96,16 @@ def content_input_images_hf():
 
 def create_download_button(zip_filepath, col):
     with col:
+        labale_n_images = f"Download Results for{st.session_state['processing_add_on']} Images"
         with open(zip_filepath, 'rb') as f:
             bytes_io = BytesIO(f.read())
         st.download_button(
-            label=f"Download Results for{st.session_state['processing_add_on']}",type='primary',
+            label=labale_n_images,
+            type='primary',
             data=bytes_io,
             file_name=os.path.basename(zip_filepath),
-            mime='application/zip'
+            mime='application/zip',
+            use_container_width=True
         )
 
 
@@ -1375,14 +1378,27 @@ def show_header_welcome():
 
 def determine_n_images():
     try:
-        # Check if 'dir_uploaded_images' key exists and it is not empty
-        if 'dir_uploaded_images' in st and st['dir_uploaded_images']:
-            dir_path = st['dir_uploaded_images']  # This would be the path to the directory
-            return len([f for f in os.listdir(dir_path) if os.path.isfile(os.path.join(dir_path, f))])
+        # Check if 'dir_uploaded_images' key exists in session state and it is not empty
+        if 'dir_uploaded_images' in st.session_state and st.session_state['dir_uploaded_images']:
+            dir_path = st.session_state['dir_uploaded_images']  # This would be the path to the directory
+            # Count only files (not directories) in the specified directory
+            count = len([f for f in os.listdir(dir_path) if os.path.isfile(os.path.join(dir_path, f))])
+            return count
         else:
-            return None
-    except:
+            return None  # Return 0 if the directory path doesn't exist or is empty
+    except Exception as e:
+        print(e)
         return None
+# def determine_n_images():
+#     try:
+#         # Check if 'dir_uploaded_images' key exists and it is not empty
+#         if 'dir_uploaded_images' in st and st['dir_uploaded_images']:
+#             dir_path = st['dir_uploaded_images']  # This would be the path to the directory
+#             return len([f for f in os.listdir(dir_path) if os.path.isfile(os.path.join(dir_path, f))])
+#         else:
+#             return None
+#     except:
+#         return None
 
 def save_api_status(present_keys, missing_keys, date_of_check):
     with open(os.path.join(st.session_state.dir_home,'api_status.yaml'), 'w') as file:
@@ -1566,8 +1582,9 @@ def content_header():
         # st.subheader('Run VoucherVision')
         N_STEPS = 6
 
-        if determine_n_images():
-            st.session_state['processing_add_on'] = f" {determine_n_images()} Images"
+        count_n_imgs = determine_n_images()
+        if count_n_imgs > 0:
+            st.session_state['processing_add_on'] = f" {count_n_imgs} Images"
         else:
             st.session_state['processing_add_on'] = ''
 
@@ -2360,7 +2377,14 @@ if 'json_report' not in st.session_state:
 if 'hold_output' not in st.session_state:
     st.session_state['hold_output'] = False
 
+if 'dir_uploaded_images' not in st.session_state:
+    st.session_state['dir_uploaded_images'] = os.path.join(st.session_state.dir_home,'uploads')
+    validate_dir(os.path.join(st.session_state.dir_home,'uploads'))
 
+if 'dir_uploaded_images_small' not in st.session_state:
+    st.session_state['dir_uploaded_images_small'] = os.path.join(st.session_state.dir_home,'uploads_small')
+    validate_dir(os.path.join(st.session_state.dir_home,'uploads_small'))
+    
 if 'cost_openai' not in st.session_state:
     st.session_state['cost_openai'] = None
 if 'cost_azure' not in st.session_state: