Update app.py

app.py

@@ -133,54 +133,51 @@ async def get_transcript(youtube_url: str):
             'skip_download': True,
             'writesubtitles': True,
             'writeautomaticsub': True,
-            'subtitleslangs': ['en'],  # Adjust regex if needed, e.g., 'en.*'
-            'subtitlesformat': 'srt',   # Use 'srt' for easier parsing
-            'outtmpl': '%(id)s',       # Output template without ext to avoid conflicts
+            'subtitleslangs': ['en'],  # Try 'en.*' for all English variants
+            'subtitlesformat': 'best',  # Let yt-dlp choose best available format
+            'outtmpl': '%(id)s',        # Output template without extension
             'noplaylist': True,
             'cookiefile': "firefox-cookies.txt",
         }
 
-        # Load cookies (ensure this function correctly sets up the cookie file)
         env_to_cookies_from_env("firefox-cookies.txt")
 
         with yt_dlp.YoutubeDL(ydl_opts) as ydl:
             info = ydl.extract_info(youtube_url, download=False)
             video_id = info['id']
 
-            # Check for manually created subtitles
-            subtitles = info.get('subtitles', {})
-            # Check for auto-generated captions
-            auto_subtitles = info.get('automatic_captions', {})
-
-            transcript_lang = None
-            # Prioritize manual subs
+            # Find available subtitle format
+            sub_ext = None
             for lang in ydl_opts['subtitleslangs']:
-                if lang in subtitles:
-                    transcript_lang = lang
-                    break
-            # Fallback to auto subs
-            if not transcript_lang:
-                for lang in ydl_opts['subtitleslangs']:
-                    if lang in auto_subtitles:
-                        transcript_lang = lang
+                for sub_type in ['subtitles', 'automatic_captions']:
+                    subs = info.get(sub_type, {}).get(lang, [])
+                    if subs:
+                        sub_ext = subs[0].get('ext', 'vtt')
                         break
+                if sub_ext:
+                    break
 
-            if transcript_lang:
-                # Look for the downloaded subtitle file
-                subtitle_file = f"{video_id}.{transcript_lang}.srt"
+            if sub_ext:
+                subtitle_file = f"{video_id}.{lang}.{sub_ext}"
                 if os.path.exists(subtitle_file):
                     with open(subtitle_file, 'r', encoding='utf-8') as f:
-                        srt_content = f.read()
-                    # Simple parsing to extract text (consider using a library like pysrt)
-                    text = ' '.join(line.strip() for line in srt_content.split('\n') 
-                                   if not line.strip().isdigit() 
-                                   and '-->' not in line 
-                                   and line.strip())
-                    return {'transcript': text}
+                        content = f.read()
+                    
+                    # Parse both VTT and SRT formats
+                    lines = content.split('\n')
+                    text_lines = []
+                    for line in lines:
+                        # Skip timestamps and metadata
+                        if '-->' in line or line.strip().isdigit() or line.startswith('WEBVTT'):
+                            continue
+                        if line.strip():
+                            text_lines.append(line.strip())
+                    
+                    return {"transcript": ' '.join(text_lines)}
                 else:
-                    return {'transcript': 'Subtitle file not found'}
+                    return {"transcript": f"Found subtitles but file {subtitle_file} missing"}
             else:
-                return {'transcript': 'No transcript available'}
+                return {"transcript": "No subtitles available in requested languages"}
 
     except Exception as e:
         raise HTTPException(status_code=500, detail=str(e))