Update app.py

app.py

@@ -6,6 +6,7 @@ import asyncio
 from huggingface_hub import InferenceClient
 from googleapiclient.discovery import build
 from youtube_transcript_api import YouTubeTranscriptApi
+from youtube_transcript_api.formatters import TextFormatter
 from dotenv import load_dotenv
 
 # 환경 변수 로드
@@ -55,7 +56,7 @@ class MyClient(discord.Client):
         try:
             video_id = extract_video_id(message.content)
             if video_id:
-                transcript = await get_video_transcript(video_id)
+                transcript = await get_best_available_transcript(video_id)
                 comments = await get_video_comments(video_id)
                 if comments and transcript:
                     replies = await generate_replies(comments, transcript)
@@ -89,18 +90,28 @@ def extract_video_id(url):
     logging.debug(f'Extracted video ID: {video_id}')
     return video_id
 
-async def get_video_transcript(video_id):
+async def get_best_available_transcript(video_id):
     """
     YouTube 비디오의 자막을 가져옵니다.
     """
     try:
-        transcript = YouTubeTranscriptApi.get_transcript(video_id)
-        transcript_text = " ".join([entry['text'] for entry in transcript])
-        logging.debug(f'Fetched transcript: {transcript_text}')
-        return transcript_text
+        # 영어 자막 시도
+        transcript = YouTubeTranscriptApi.get_transcript(video_id, languages=['en'])
     except Exception as e:
-        logging.error(f'Error fetching transcript: {e}')
-        return None
+        logging.warning(f'Error fetching English transcript: {e}')
+        try:
+            # 영어 자막이 없으면 다른 언어 자막을 시도
+            transcripts = YouTubeTranscriptApi.list_transcripts(video_id)
+            transcript = transcripts.find_manually_created_transcript(['ko', 'ja', 'zh-Hans', 'zh-Hant']).fetch()
+        except Exception as e:
+            logging.error(f'Error fetching alternative transcript: {e}')
+            return None
+
+    # 자막 포맷팅
+    formatter = TextFormatter()
+    transcript_text = formatter.format_transcript(transcript)
+    logging.debug(f'Fetched transcript: {transcript_text}')
+    return transcript_text
 
 async def get_video_comments(video_id):
     """