print("逐字稿已上传到GCS")

app.py

@@ -423,17 +423,18 @@ def process_transcript_and_screenshots_on_gcs(video_id):
     gcs_client = init_gcs_client(GCS_KEY)
     bucket_name = 'video_ai_assistant'
     # 检查 folder 是否存在
-    is_gcs_exists = gcs_check_folder_exists(gcs_client, bucket_name, video_id)
-    if not is_gcs_exists:
-        gcs_create_bucket_folder_if_not_exists(gcs_client, bucket_name, video_id)
-        print("GCS folder:{video_id} 已创建")
-    else:
-        print("GCS folder:{video_id} 已存在")
+    # is_gcs_exists = gcs_check_folder_exists(gcs_client, bucket_name, video_id)
+    # if not is_gcs_exists:
+    #     gcs_create_bucket_folder_if_not_exists(gcs_client, bucket_name, video_id)
+    #     print("GCS folder:{video_id} 已创建")
+    # else:
+    #     print("GCS folder:{video_id} 已存在")
 
     # 逐字稿文件名
-    file_name = f'{video_id}_transcript.json'
+    transcript_file_name = f'{video_id}_transcript.json'
+    transcript_blob_name = f"{video_id}/{transcript_file_name}"
     # 检查逐字稿是否存在
-    exists = gcs_check_file_exists(gcs_client, bucket_name, file_name)
+    exists = gcs_check_file_exists(gcs_client, bucket_name, transcript_blob_name)
     if not exists:
         # 从YouTube获取逐字稿并上传
         transcript = get_transcript(video_id)
@@ -442,22 +443,21 @@ def process_transcript_and_screenshots_on_gcs(video_id):
         else:
             print("沒有找到字幕")
         transcript_text = json.dumps(transcript, ensure_ascii=False, indent=2)
-        destination_blob_name = f"{video_id}/{file_name}"
-        upload_file_to_gcs_with_json_string(gcs_client, bucket_name, destination_blob_name, transcript_text)
-    
-    raise gr.Error("test")
-    # # 处理逐字稿中的每个条目，检查并上传截图 到 GCS，然後設定 GCS 權限
-    # 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_and_get_public_url(gcs_client, bucket_name, f"{video_id}_{entry['start']}.jpg", screenshot_path)
-    #         entry['img_file_id'] = img_file_id
-    #         print(f"截图已上传到GCS: {img_file_id}")
+        upload_file_to_gcs_with_json_string(gcs_client, bucket_name, transcript_blob_name, transcript_text)
+        print("逐字稿已上传到GCS")
+
+    for entry in transcript:
+        if 'img_file_id' not in entry:
+            screenshot_path = screenshot_youtube_video(video_id, entry['start'])
+            transcript_blob_name = f"{video_id}/{video_id}_{entry['start']}.jpg"
+            img_file_id = upload_img_and_get_public_url(gcs_client, bucket_name, transcript_blob_name, screenshot_path)
+            entry['img_file_id'] = img_file_id
+            print(f"截图已上传到GCS: {img_file_id}")
         
-    # # 更新逐字稿文件
-    # updated_transcript_text = json.dumps(transcript, ensure_ascii=False, indent=2)
-    # upload_file_to_gcs(gcs_client, bucket_name, file_name, updated_transcript_text)
-    # print("逐字稿已更新，包括截图链接")
+    # 更新逐字稿文件
+    updated_transcript_text = json.dumps(transcript, ensure_ascii=False, indent=2)
+    upload_file_to_gcs_with_json_string(gcs_client, bucket_name, transcript_blob_name, updated_transcript_text)
+    print("逐字稿已更新，包括截图链接")
 
     return transcript