Overhauled and updated file extension definition & MAG audio file transcription, fixed file metadata editing

application/single_app/config.py

@@ -122,11 +122,49 @@
 CLIENTS = {}
 CLIENTS_LOCK = threading.Lock()
 
-ALLOWED_EXTENSIONS = {
- 'txt', 'pdf', 'doc', 'docm', 'docx', 'xlsx', 'xlsm', 'xls', 'csv', 'pptx', 'html', 'jpg', 'jpeg', 'png', 'bmp', 'tiff', 'tif', 'heif', 'md', 'json', 
- 'mp4', 'mov', 'avi', 'mkv', 'flv', 'mxf', 'gxf', 'ts', 'ps', '3gp', '3gpp', 'mpg', 'wmv', 'asf', 'm4a', 'm4v', 'isma', 'ismv', 
- 'dvr-ms', 'wav', 'xml', 'yaml', 'yml', 'log'
+# Base allowed extensions (always available)
+BASE_ALLOWED_EXTENSIONS = {'txt', 'doc', 'docm', 'html', 'md', 'json', 'xml', 'yaml', 'yml', 'log'}
+DOCUMENT_EXTENSIONS = {'pdf', 'docx', 'pptx', 'ppt'}
+TABULAR_EXTENSIONS = {'csv', 'xlsx', 'xls', 'xlsm'}
+
+# Updates to image, video, or audio extensions should also be made in static/js/chat/chat-enhanced-citations.js if the new file types can be natively rendered in the browser.
+IMAGE_EXTENSIONS = {'jpg', 'jpeg', 'png', 'bmp', 'tiff', 'tif', 'heif', 'heic'}
+
+# Optional extensions by feature
+VIDEO_EXTENSIONS = {
+ 'mp4', 'mov', 'avi', 'mkv', 'flv', 'mxf', 'gxf', 'ts', 'ps', '3gp', '3gpp',
+ 'mpg', 'wmv', 'asf', 'm4v', 'isma', 'ismv', 'dvr-ms', 'webm', 'mpeg'
 }
+
+AUDIO_EXTENSIONS = {'mp3', 'wav', 'ogg', 'aac', 'flac', 'm4a'}
+
+def get_allowed_extensions(enable_video=False, enable_audio=False):
+ """
+ Get allowed file extensions based on feature flags.
+
+ Args:
+ enable_video: Whether video file support is enabled
+ enable_audio: Whether audio file support is enabled
+
+ Returns:
+ set: Allowed file extensions
+ """
+ extensions = BASE_ALLOWED_EXTENSIONS.copy()
+ extensions.update(DOCUMENT_EXTENSIONS)
+ extensions.update(IMAGE_EXTENSIONS)
+ extensions.update(TABULAR_EXTENSIONS)
+
+ if enable_video:
+ extensions.update(VIDEO_EXTENSIONS)
+
+ if enable_audio:
+ extensions.update(AUDIO_EXTENSIONS)
+
+ return extensions
+
+ALLOWED_EXTENSIONS = get_allowed_extensions(enable_video=True, enable_audio=True)
+
+# Admin UI specific extensions (for logo/favicon uploads)
 ALLOWED_EXTENSIONS_IMG = {'png', 'jpg', 'jpeg'}
 MAX_CONTENT_LENGTH = 5000 * 1024 * 1024 # 5000 MB AKA 5 GB
 

application/single_app/functions_documents.py

@@ -4851,7 +4851,7 @@ def process_di_document(document_id, user_id, temp_file_path, original_filename,
  is_pdf = file_ext == '.pdf'
  is_word = file_ext in ('.docx', '.doc')
  is_ppt = file_ext in ('.pptx', '.ppt')
- is_image = file_ext in ('.jpg', '.jpeg', '.png', '.bmp', '.tiff', '.tif', '.heif')
+ is_image = file_ext in tuple('.' + ext for ext in IMAGE_EXTENSIONS)
 
  try:
  if is_pdf:
@@ -5233,23 +5233,138 @@ def process_audio_document(
  print(f"[Debug] Transcribing chunk {idx}: {chunk_path}")
 
  # Get fresh config (tokens expire after ~1 hour)
- speech_config = _get_speech_config(settings, endpoint, locale)
+ try:
+ speech_config = _get_speech_config(settings, endpoint, locale)
+ except Exception as e:
+ print(f"[Error] Failed to get speech config for chunk {idx}: {e}")
+ raise RuntimeError(f"Speech configuration failed for chunk {idx}: {e}")
 
-audio_config = speechsdk.AudioConfig(filename=chunk_path)
-speech_recognizer = speechsdk.SpeechRecognizer(
- speech_config=speech_config,
-audio_config=audio_config
- )
+try:
+ audio_config = speechsdk.AudioConfig(filename=chunk_path)
+except Exception as e:
+print(f"[Error] Failed to load audio file {chunk_path}: {e}")
+ raise RuntimeError(f"Audio file loading failed: {e}")
 
- result = speech_recognizer.recognize_once()
- if result.reason == speechsdk.ResultReason.RecognizedSpeech:
- print(f"[Debug] Recognized: {result.text}")
- all_phrases.append(result.text)
- elif result.reason == speechsdk.ResultReason.NoMatch:
- print(f"[Warning] No speech in {chunk_path}")
- elif result.reason == speechsdk.ResultReason.Canceled:
- print(f"[Error] {result.cancellation_details.reason}: {result.cancellation_details.error_details}")
- raise RuntimeError(f"Transcription canceled for {chunk_path}: {result.cancellation_details.error_details}")
+ try:
+ speech_recognizer = speechsdk.SpeechRecognizer(
+ speech_config=speech_config,
+ audio_config=audio_config
+ )
+ except Exception as e:
+ print(f"[Error] Failed to create speech recognizer for chunk {idx}: {e}")
+ raise RuntimeError(f"Speech recognizer creation failed: {e}")
+
+ # Use continuous recognition instead of recognize_once
+ all_results = []
+ done = False
+ error_occurred = False
+ error_message = None
+
+ def stop_cb(evt):
+ nonlocal done
+ print(f"[Debug] Session stopped for chunk {idx}")
+ done = True
+
+ def recognized_cb(evt):
+ try:
+ if evt.result.reason == speechsdk.ResultReason.RecognizedSpeech:
+ all_results.append(evt.result.text)
+ print(f"[Debug] Recognized: {evt.result.text}")
+ elif evt.result.reason == speechsdk.ResultReason.NoMatch:
+ print(f"[Debug] No speech recognized in segment")
+ except Exception as e:
+ print(f"[Error] Error in recognized callback: {e}")
+ # Don't fail on individual recognition errors
+
+ def canceled_cb(evt):
+ nonlocal done, error_occurred, error_message
+ print(f"[Debug] Recognition canceled for chunk {idx}: {evt.cancellation_details.reason}")
+
+ if evt.cancellation_details.reason == speechsdk.CancellationReason.Error:
+ error_occurred = True
+ error_message = evt.cancellation_details.error_details
+ print(f"[Error] Recognition error: {error_message}")
+ elif evt.cancellation_details.reason == speechsdk.CancellationReason.EndOfStream:
+ print(f"[Debug] End of audio stream reached")
+
+ done = True
+
+ try:
+ # Connect callbacks
+ speech_recognizer.recognized.connect(recognized_cb)
+ speech_recognizer.session_stopped.connect(stop_cb)
+ speech_recognizer.canceled.connect(canceled_cb)
+
+ # Start continuous recognition
+ print(f"[Debug] Starting continuous recognition for chunk {idx}")
+ speech_recognizer.start_continuous_recognition()
+
+ # Wait for completion with timeout
+ import time
+ timeout_seconds = 600 # 10 minutes max per chunk
+ start_time = time.time()
+
+ while not done:
+ if time.time() - start_time > timeout_seconds:
+ print(f"[Error] Recognition timeout for chunk {idx}")
+ error_occurred = True
+ error_message = f"Recognition timed out after {timeout_seconds} seconds"
+ break
+ time.sleep(0.5)
+
+ # Stop recognition
+ try:
+ speech_recognizer.stop_continuous_recognition()
+ print(f"[Debug] Stopped continuous recognition for chunk {idx}")
+ except Exception as e:
+ print(f"[Warning] Error stopping recognition for chunk {idx}: {e}")
+ # Continue even if stop fails
+
+ # Check for errors after completion
+ if error_occurred:
+ raise RuntimeError(f"Recognition failed for chunk {idx}: {error_message}")
+
+ # Add all recognized phrases to the overall list
+ if all_results:
+ all_phrases.extend(all_results)
+ print(f"[Debug] Total phrases from chunk {idx}: {len(all_results)}")
+ else:
+ print(f"[Warning] No speech recognized in {chunk_path}")
+ # Continue to next chunk - empty result is not necessarily an error
+
+ except RuntimeError as e:
+ # Re-raise runtime errors (these are our custom errors)
+ raise
+ except Exception as e:
+ print(f"[Error] Unexpected error during recognition for chunk {idx}: {e}")
+ raise RuntimeError(f"Recognition failed unexpectedly for chunk {idx}: {e}")
+ finally:
+ # Cleanup: disconnect callbacks and dispose recognizer
+ try:
+ speech_recognizer.recognized.disconnect_all()
+ speech_recognizer.session_stopped.disconnect_all()
+ speech_recognizer.canceled.disconnect_all()
+ except Exception as e:
+ print(f"[Warning] Error disconnecting callbacks for chunk {idx}: {e}")
+
+ # # Get fresh config (tokens expire after ~1 hour)
+ # speech_config = _get_speech_config(settings, endpoint, locale)
+
+ # audio_config = speechsdk.AudioConfig(filename=chunk_path)
+ # speech_recognizer = speechsdk.SpeechRecognizer(
+ # speech_config=speech_config,
+ # audio_config=audio_config
+ # )
+
+ # result = speech_recognizer.recognize_once()
+ # if result.reason == speechsdk.ResultReason.RecognizedSpeech:
+ # print(f"[Debug] Recognized: {result.text}")
+ # all_phrases.append(result.text)
+ # elif result.reason == speechsdk.ResultReason.NoMatch:
+ # print(f"[Warning] No speech in {chunk_path}")
+ # elif result.reason == speechsdk.ResultReason.Canceled:
+ # print(f"[Error] {result.cancellation_details.reason}: {result.cancellation_details.error_details}")
+ # raise RuntimeError(f"Transcription canceled for {chunk_path}: {result.cancellation_details.error_details}")
 
  else:
  # Use the fast-transcription API if not in sovereign or custom cloud
@@ -5357,8 +5472,12 @@ def process_document_upload_background(document_id, user_id, temp_file_path, ori
  enable_extract_meta_data = settings.get('enable_extract_meta_data', False) # Used by DI flow
  max_file_size_bytes = settings.get('max_file_size_mb', 16) * 1024 * 1024
 
- video_extensions = ('.mp4', '.mov', '.avi', '.mkv', '.flv')
- audio_extensions = ('.mp3', '.wav', '.ogg', '.aac', '.flac', '.m4a')
+ # Get allowed extensions from config.py to determine which processing function to call
+ tabular_extensions = tuple('.' + ext for ext in TABULAR_EXTENSIONS)
+ image_extensions = tuple('.' + ext for ext in IMAGE_EXTENSIONS)
+ di_supported_extensions = tuple('.' + ext for ext in DOCUMENT_EXTENSIONS | IMAGE_EXTENSIONS)
+ video_extensions = tuple('.' + ext for ext in VIDEO_EXTENSIONS)
+ audio_extensions = tuple('.' + ext for ext in AUDIO_EXTENSIONS)
 
  # --- Define update_document callback wrapper ---
  # This makes it easier to pass the update function to helpers without repeating args
@@ -5402,8 +5521,6 @@ def update_doc_callback(**kwargs):
 
  # --- 1. Dispatch to appropriate handler based on file type ---
  # Note: .doc and .docm are handled separately by process_doc() using docx2txt
- di_supported_extensions = ('.pdf', '.docx', '.pptx', '.ppt', '.jpg', '.jpeg', '.png', '.bmp', '.tiff', '.tif', '.heif')
- tabular_extensions = ('.csv', '.xlsx', '.xls', '.xlsm')
 
  is_group = group_id is not None
 
@@ -5512,7 +5629,7 @@ def update_doc_callback(**kwargs):
  final_status = "Processing complete"
  if total_chunks_saved == 0:
  # Provide more specific status if no chunks were saved
- if file_ext in ('.jpg', '.jpeg', '.png', '.bmp', '.tiff', '.tif', '.heif'):
+ if file_ext in image_extensions:
  final_status = "Processing complete - no text found in image"
  elif file_ext in tabular_extensions:
  final_status = "Processing complete - no data rows found or file empty"

application/single_app/route_backend_documents.py

@@ -535,9 +535,23 @@ def api_patch_user_document(document_id):
  try:
  # Log the metadata update transaction if any fields were updated
  if updated_fields:
- # Get document details for logging
- doc = get_document(user_id, document_id)
- if doc:
+ # Get document details for logging - handle tuple return
+ doc_response = get_document(user_id, document_id)
+ doc = None
+
+ # Handle tuple return (response, status_code)
+ if isinstance(doc_response, tuple):
+ resp, status_code = doc_response
+ if hasattr(resp, "get_json"):
+ doc = resp.get_json()
+ else:
+ doc = resp
+ elif hasattr(doc_response, "get_json"):
+ doc = doc_response.get_json()
+ else:
+ doc = doc_response
+
+ if doc and isinstance(doc, dict):
  log_document_metadata_update_transaction(
  user_id=user_id,
  document_id=document_id,

application/single_app/route_backend_group_documents.py

@@ -416,24 +416,43 @@ def api_patch_group_document(document_id):
  )
  updated_fields['authors'] = authors_list
 
- # Log the metadata update transaction if any fields were updated
- if updated_fields:
+ # Save updates back to Cosmos
+ try:
+ # Log the metadata update transaction if any fields were updated
+ if updated_fields:
+ # Get document details for logging - handle tuple return
  # Get document details for logging
- from functions_documents import get_document
- doc = get_document(user_id, document_id, group_id=active_group_id)
- if doc:
- from functions_activity_logging import log_document_metadata_update_transaction
- log_document_metadata_update_transaction(
- user_id=user_id,
- document_id=document_id,
- workspace_type='group',
- file_name=doc.get('file_name', 'Unknown'),
- updated_fields=updated_fields,
- file_type=doc.get('file_type'),
- group_id=active_group_id
- )
-
- return jsonify({'message': 'Group document metadata updated successfully'}), 200
+ from functions_documents import get_document
+ doc_response = get_document(user_id, document_id, group_id=active_group_id)
+ doc = None
+
+ # Handle tuple return (response, status_code)
+ if isinstance(doc_response, tuple):
+ resp, status_code = doc_response
+ if hasattr(resp, "get_json"):
+ doc = resp.get_json()
+ else:
+ doc = resp
+ elif hasattr(doc_response, "get_json"):
+ doc = doc_response.get_json()
+ else:
+ doc = doc_response
+
+ if doc and isinstance(doc, dict):
+ from functions_activity_logging import log_document_metadata_update_transaction
+ log_document_metadata_update_transaction(
+ user_id=user_id,
+ document_id=document_id,
+ workspace_type='group',
+ file_name=doc.get('file_name', 'Unknown'),
+ updated_fields=updated_fields,
+ file_type=doc.get('file_type'),
+ group_id=active_group_id
+ )
+
+ return jsonify({'message': 'Group document metadata updated successfully'}), 200
+ except Exception as e:
+ return jsonify({'Error updating Group document metadata': str(e)}), 500
  except Exception as e:
  return jsonify({'error': str(e)}), 500