with open(chunk_path, "rb") as chunk_file:

app.py

@@ -373,6 +373,9 @@ def get_transcript(video_id):
     for language in languages:
         try:
             transcript = YouTubeTranscriptApi.get_transcript(video_id, languages=[language])
+            print("===transcript===")
+            print(transcript)
+            print("===transcript===")
             return transcript  # 成功獲取字幕，直接返回結果
         except NoTranscriptFound:
             continue  # 當前語言的字幕沒有找到，繼續嘗試下一個語言
@@ -413,73 +416,33 @@ def generate_transcription(video_id):
         chunk_path = f"{OUTPUT_PATH}/{video_id}_part_{i}.{codec_name}"
         chunk.export(chunk_path, format=codec_name)
 
-        with open(chunk_path, "rb") as chunk_file:
-            response = OPEN_AI_CLIENT.audio.transcriptions.create(
-                model="whisper-1",
-                file=chunk_file,
-                response_format="verbose_json",
-                timestamp_granularities=["segment"],
-                prompt="Transcribe the following audio file. if chinese, please using 'language: zh-TW' ",
-            )
-            
-            # Adjusting the timestamps for the chunk based on its position in the full audio
-            adjusted_segments = [{
-                'text': segment['text'],
-                'start': math.ceil(segment['start'] + start_time / 1000.0),  # Converting milliseconds to seconds
-                'end': math.ceil(segment['end'] + start_time / 1000.0),
-                'duration': math.ceil(segment['end'] - segment['start'])
-            } for segment in response.segments]
-
-            transcription.extend(adjusted_segments)
+        try:
+            with open(chunk_path, "rb") as chunk_file:
+                response = OPEN_AI_CLIENT.audio.transcriptions.create(
+                    model="whisper-1",
+                    file=chunk_file,
+                    response_format="verbose_json",
+                    timestamp_granularities=["segment"],
+                    prompt="Transcribe the following audio file. if chinese, please using 'language: zh-TW' ",
+                )
+                
+                # Adjusting the timestamps for the chunk based on its position in the full audio
+                adjusted_segments = [{
+                    'text': segment['text'],
+                    'start': math.ceil(segment['start'] + start_time / 1000.0),  # Converting milliseconds to seconds
+                    'end': math.ceil(segment['end'] + start_time / 1000.0),
+                    'duration': math.ceil(segment['end'] - segment['start'])
+                } for segment in response.segments]
+
+                transcription.extend(adjusted_segments)
+        except Exception as e:
+            print(f"Error processing chunk {i}: {str(e)}")
 
         # Remove temporary chunk files after processing
         os.remove(chunk_path)
 
     return transcription
  
-def process_transcript_and_screenshots(video_id):
-    print("====process_transcript_and_screenshots====")
-
-    # Drive
-    service = init_drive_service()
-    parent_folder_id = '1GgI4YVs0KckwStVQkLa1NZ8IpaEMurkL'
-    folder_id = create_folder_if_not_exists(service, video_id, parent_folder_id)
-    
-    # 逐字稿文件名
-    file_name = f'{video_id}_transcript.json'
-    # 检查逐字稿是否存在
-    exists, file_id = check_file_exists(service, folder_id, file_name)
-    if not exists:
-        # 从YouTube获取逐字稿并上传
-        transcript = get_transcript(video_id)
-        if transcript:
-            print("成功獲取字幕")
-        else:
-            print("沒有找到字幕")
-        transcript_text = json.dumps(transcript, ensure_ascii=False, indent=2)
-        file_id = upload_content_directly(service, file_name, folder_id, transcript_text)
-        print("逐字稿已上传到Google Drive")
-    else:
-        # 逐字稿已存在，下载逐字稿内容
-        print("逐字稿已存在于Google Drive中")
-        transcript_text = download_file_as_string(service, file_id)
-        transcript = json.loads(transcript_text)
-
-    # 处理逐字稿中的每个条目，检查并上传截图
-    for entry in transcript:
-        if 'img_file_id' not in entry:
-            screenshot_path = screenshot_youtube_video(video_id, entry['start'])
-            img_file_id = upload_img_directly(service, f"{video_id}_{entry['start']}.jpg", folder_id, screenshot_path)
-            set_public_permission(service, img_file_id)
-            entry['img_file_id'] = img_file_id
-            print(f"截图已上传到Google Drive: {img_file_id}")
-    
-    # 更新逐字稿文件
-    updated_transcript_text = json.dumps(transcript, ensure_ascii=False, indent=2)
-    update_file_on_drive(service, file_id, updated_transcript_text)
-    print("逐字稿已更新，包括截图链接")
-    return transcript
-
 def process_transcript_and_screenshots_on_gcs(video_id):
     print("====process_transcript_and_screenshots_on_gcs====")
     # GCS