updated

main.py

@@ -1,7 +1,8 @@
 from fastapi import FastAPI, HTTPException
 from fastapi.middleware.cors import CORSMiddleware
-from youtube_transcript_api import YouTubeTranscriptApi, TranscriptsDisabled, NoTranscriptFound, VideoUnavailable
+from youtube_transcript_api import YouTubeTranscriptApi, TranscriptsDisabled, NoTranscriptFound, VideoUnavailable, TooManyRequests, NoTranscriptAvailable
 import re
+from pydantic import BaseModel
 
 app = FastAPI()
 
@@ -14,8 +15,11 @@ app.add_middleware(
     allow_headers=["*"],
 )
 
+# Pydantic model to ensure we receive proper JSON data
+class TranscriptRequest(BaseModel):
+    url: str
+
 def extract_video_id(url: str) -> str:
-    # The fix: Changed re.match to re.search so it finds the ID even with https:// in front
     pattern = r'(?:youtube\.com\/(?:[^\/]+\/.+\/|(?:v|e(?:mbed)?)\/|.*[?&]v=)|youtu\.be\/)([^"&?\/\s]{11})'
     match = re.search(pattern, url)
     if match:
@@ -23,16 +27,16 @@ def extract_video_id(url: str) -> str:
     return None
 
 @app.post("/api/transcript")
-async def get_transcript(request: dict):
-    url = request.get("url")
-    if not url:
-        raise HTTPException(status_code=400, detail="Missing URL")
-
+async def get_transcript(req: TranscriptRequest):
+    url = req.url
+    
     video_id = extract_video_id(url)
     if not video_id:
         raise HTTPException(status_code=400, detail="Invalid YouTube URL")
 
     try:
+        print(f"Attempting to fetch transcript for video_id: {video_id}")
+        
         # Fetch available transcripts
         transcript_list = YouTubeTranscriptApi.list_transcripts(video_id)
         
@@ -40,14 +44,18 @@ async def get_transcript(request: dict):
         transcript = None
         try:
             transcript = transcript_list.find_manually_created_transcript(['en'])
+            print("Found manual English transcript.")
         except NoTranscriptFound:
             try:
                 transcript = transcript_list.find_generated_transcript(['en'])
+                print("Found auto-generated English transcript.")
             except NoTranscriptFound:
-                # If no English, get the first available language
-                for t in transcript_list:
-                    transcript = t
-                    break
+                # If no English, get the first available language safely
+                try:
+                    transcript = next(iter(transcript_list))
+                    print(f"Found fallback transcript: {transcript.language_code}")
+                except StopIteration:
+                    pass
         
         if not transcript:
             raise HTTPException(status_code=404, detail="No captions found for this video.")
@@ -55,20 +63,29 @@ async def get_transcript(request: dict):
         # Fetch the actual text data
         transcript_data = transcript.fetch()
         
-        # Combine all text lines
-        full_text = " ".join([line['text'] for line in transcript_data])
+        # 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("'", "'")
         
+        print(f"Successfully fetched transcript. Length: {len(full_text)} chars")
         return {"text": full_text}
 
+    except HTTPException:
+        raise # Re-raise our custom HTTP exceptions
+    except TooManyRequests:
+        raise HTTPException(status_code=429, detail="YouTube is rate-limiting requests from this server. Please try again in a minute.")
     except TranscriptsDisabled:
         raise HTTPException(status_code=403, detail="Captions are disabled for this video.")
     except VideoUnavailable:
         raise HTTPException(status_code=404, detail="Video is unavailable.")
+    except NoTranscriptAvailable:
+        raise HTTPException(status_code=404, detail="No transcript is available for this video.")
     except Exception as e:
-        raise HTTPException(status_code=500, detail=f"Failed to fetch transcript: {str(e)}")
+        # Catch-all: Log the exact error to Hugging Face logs so we can see what went wrong!
+        print(f"UNHANDLED ERROR for {video_id}: {type(e).__name__} - {str(e)}")
+        raise HTTPException(status_code=500, detail=f"Failed to fetch transcript: {type(e).__name__} - {str(e)}")
 
 # KEEP-ALIVE ENDPOINT
 @app.get("/api/health")