fix: update yt-dlp and improve logging

app/services/downloader.py

@@ -436,7 +436,11 @@ def stream_extract_frames(url: str, job_id: str, max_frames: int = 5, duration:
         )
         
         if result.returncode != 0:
-            print(f"FFmpeg stderr: {result.stderr.decode()[:500]}")
+            stderr = result.stderr.decode()
+            print(f"[{job_id}] FFmpeg Error (Return Code {result.returncode}):")
+            print(f"[{job_id}] FFmpeg stderr (first 500 chars): {stderr[:500]}")
+        else:
+            print(f"[{job_id}] FFmpeg extraction successful")
         
         # Find extracted frames
         frames = sorted(glob.glob(os.path.join(TEMP_DIR, f"{job_id}_frame_*.jpg")))
@@ -484,7 +488,11 @@ def stream_extract_audio(url: str, job_id: str, duration: int = 30) -> str:
         result = subprocess.run(cmd, capture_output=True, timeout=60)
         
         if result.returncode != 0:
-            print(f"Audio extraction stderr: {result.stderr.decode()[:300]}")
+            stderr = result.stderr.decode()
+            print(f"[{job_id}] Audio extraction Error (Return Code {result.returncode}):")
+            print(f"[{job_id}] Audio stderr (first 300 chars): {stderr[:300]}")
+        else:
+            print(f"[{job_id}] Audio extraction successful")
             
         return audio_path if os.path.exists(audio_path) else None
         

app/services/pipeline.py

@@ -84,20 +84,28 @@ async def run_analysis_pipeline(job_id: str, url: str, uploaded_file_path: str,
         
         # PATH A: URL
         if url and not uploaded_file_path:
+            print(f"[{job_id}] Path A: URL analysis")
             frame_paths = stream_extract_frames(url, job_id, max_frames=8, duration=30)
             
             if not frame_paths:
+                print(f"[{job_id}] Streaming failed, falling back to full download")
                 video_path = download_video(url, job_id)
                 if video_path and os.path.exists(video_path):
+                    print(f"[{job_id}] Full download successful, extracting frames")
                     frame_paths = extract_frames(video_path, job_id, fps=0.5, max_frames=8)
                     audio_path = extract_audio(video_path, job_id)
                 elif is_youtube_url(url):
+                    print(f"[{job_id}] Full download failed for YouTube, trying thumbnail fallback")
                     frame_paths = download_youtube_thumbnail(url, job_id)
                     thumbnail_only = True
-                else:
-                    jobs_db[job_id] = {"status": "failed", "error": "Could not download video or extract frames"}
+                
+                if not frame_paths:
+                    error_msg = "Could not download video or extract frames (All layers failed)"
+                    print(f"[{job_id}] ERROR: {error_msg}")
+                    jobs_db[job_id] = {"status": "failed", "error": error_msg}
                     return
             else:
+                print(f"[{job_id}] Streaming successful, extracting audio")
                 audio_path = stream_extract_audio(url, job_id, duration=30)
         
         # PATH B: Upload

requirements.txt

@@ -3,6 +3,6 @@ uvicorn[standard]>=0.32.0
 python-dotenv>=1.0.0
 pydantic-settings>=2.6.0
 requests>=2.32.0
-yt-dlp>=2024.12.0
+yt-dlp>=2025.1.15
 slowapi>=0.1.9
 python-multipart>=0.0.9