wd

main.py

@@ -35,23 +35,25 @@ async def get_transcript(req: TranscriptRequest):
         raise HTTPException(status_code=400, detail="Invalid YouTube URL")
 
     try:
-        print(f"Attempting to fetch transcript for video_id: {video_id} using youtube-transcript-api v0.6+")
+        print(f"Attempting to fetch transcript for video_id: {video_id} using youtube-transcript-api 0.5.0")
         
-        # Initialize the API client (New syntax for 0.6+)
-        ytt_api = YouTubeTranscriptApi()
-        
-        # Fetch the transcript (prefers English)
+        # Try fetching English transcript first
         try:
-            transcript = ytt_api.fetch(video_id, languages=['en'])
+            print("Trying English transcript...")
+            transcript_data = YouTubeTranscriptApi.get_transcript(video_id, languages=['en'])
             print("Found English transcript.")
         except NoTranscriptFound:
             print("No English found, trying any available language...")
             # Fallback to any language if English is not found
-            transcript = ytt_api.fetch(video_id)
-            print(f"Found fallback transcript.")
+            try:
+                transcript_data = YouTubeTranscriptApi.get_transcript(video_id)
+                print("Found fallback language transcript.")
+            except Exception as e:
+                print(f"Fallback failed: {e}")
+                raise HTTPException(status_code=404, detail="No captions found for this video.")
             
-        # Join the text from the snippets (New syntax for 0.6+)
-        full_text = " ".join([snippet.text for snippet in transcript])
+        # Combine all text lines safely
+        full_text = " ".join([line.get('text', '') for line in transcript_data])
         
         # Clean up HTML entities
         full_text = full_text.replace("'", "'").replace(""", '"').replace("&", "&").replace("'", "'")
@@ -63,8 +65,6 @@ async def get_transcript(req: TranscriptRequest):
         raise HTTPException(status_code=403, detail="Captions are disabled for this video.")
     except VideoUnavailable:
         raise HTTPException(status_code=404, detail="Video is unavailable.")
-    except NoTranscriptFound:
-        raise HTTPException(status_code=404, detail="No captions found for this video.")
     except Exception as e:
         error_name = type(e).__name__
         print(f"UNHANDLED ERROR for {video_id}: {error_name} - {str(e)}")