Update app.py

app.py

@@ -183,16 +183,26 @@ def get_video_id(url):
 def fetch_youtube_transcript(video_id):
     try:
         transcript_list = YouTubeTranscriptApi.list_transcripts(video_id)
-        transcript = transcript_list.find_transcript(['en']).fetch()
-        full_text = " ".join([item['text'] for item in transcript])
-        return full_text
-    except TranscriptsDisabled:
-        return None
-    except NoTranscriptFound:
+        # Try English variants first
+        for lang in ['en', 'en-US', 'en-GB']:
+            try:
+                transcript = transcript_list.find_transcript([lang]).fetch()
+                full_text = " ".join([item['text'] for item in transcript])
+                return full_text
+            except NoTranscriptFound:
+                continue
+        
+        # If no English transcript, try any available transcript and translate to English
         for transcript in transcript_list:
             if transcript.is_translatable:
                 translated_transcript = transcript.translate('en').fetch()
                 return " ".join([item['text'] for item in translated_transcript])
+        
+        return None
+    except TranscriptsDisabled:
+        return None
+    except Exception as e:
+        st.error(f"Error fetching transcript with youtube-transcript-api: {str(e)}")
         return None
 
 # Function to extract subtitles using yt-dlp with cookies
@@ -205,10 +215,17 @@ def extract_subtitles_with_ytdlp(video_url):
         'subtitlesformat': 'vtt',
         'outtmpl': 'subtitle.%(ext)s',
         'http_headers': {
-            'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36',
-            'Accept-Language': 'en-US,en;q=0.9',
+            'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/126.0.0.0 Safari/537.36',
+            'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8',
+            'Accept-Language': 'en-US,en;q=0.5',
+            'Referer': 'https://www.youtube.com/',
+            'Origin': 'https://www.youtube.com',
         },
         'cookiefile': 'cookies.txt',
+        'retries': 5,  # Increased retries
+        'retry_sleep': 3,
+        'geo_bypass': True,  # Attempt to bypass geo-restrictions
+        'no_check_certificate': True,  # Bypass certificate issues
     }
     try:
         if not os.path.exists('cookies.txt'):
@@ -216,9 +233,10 @@ def extract_subtitles_with_ytdlp(video_url):
                 "cookies.txt file not found. Please upload a valid cookies.txt file to the root directory of your Space. "
                 "To generate it:\n"
                 "1. Open Chrome and log in to YouTube.\n"
-                "2. Install the 'Export Cookies' extension.\n"
+                "2. Install the 'Export Cookies' extension (or use a tool like 'cookies.txt' for Firefox).\n"
                 "3. Export cookies for 'youtube.com' and save as 'cookies.txt'.\n"
-                "4. Upload the file to your Space via the Files tab."
+                "4. Upload the file to your Space via the Files tab.\n"
+                "Alternative: If this fails, test locally to rule out Spaces IP restrictions."
             )
             return None
 
@@ -227,11 +245,9 @@ def extract_subtitles_with_ytdlp(video_url):
             available_subs = info.get('subtitles', {})
             auto_subs = info.get('automatic_captions', {})
 
-            st.text(f"Available subtitles: {list(available_subs.keys())}")
-            st.text(f"Available auto-captions: {list(auto_subs.keys())}")
-
             subtitle_langs = list(available_subs.keys()) or list(auto_subs.keys())
             if not subtitle_langs:
+                st.warning("No subtitles or auto-captions available in any language.")
                 return None
 
             ydl.params['subtitleslangs'] = subtitle_langs
@@ -245,6 +261,7 @@ def extract_subtitles_with_ytdlp(video_url):
                 break
 
         if not subtitle_file:
+            st.warning("No subtitle files were downloaded.")
             return None
 
         with open(subtitle_file, 'r', encoding='utf-8') as f: