get_questions

app.py

@@ -518,7 +518,8 @@ def process_youtube_link(link):
     TRANSCRIPTS = formatted_transcript
 
     # 基于逐字稿生成其他所需的输出
-    questions = get_questions(video_id, formatted_simple_transcript)
+    source = "gcs"
+    questions = get_questions(video_id, formatted_simple_transcript, source)
     formatted_transcript_json = json.dumps(formatted_transcript, ensure_ascii=False, indent=2)
     summary_json = get_video_id_summary(video_id, formatted_simple_transcript)
     summary = summary_json["summary"]
@@ -824,26 +825,47 @@ def generate_questions(df_string):
 
     return questions
 
-def get_questions(video_id, df_string):
-    # 去 g drive 確認是有有 video_id_questions.json
-    print("===get_questions===")
-    service = init_drive_service()
-    parent_folder_id = '1GgI4YVs0KckwStVQkLa1NZ8IpaEMurkL'
-    folder_id = create_folder_if_not_exists(service, video_id, parent_folder_id)
-    file_name = f'{video_id}_questions.json'
-
-    # 检查檔案是否存在
-    exists, file_id = check_file_exists(service, folder_id, file_name)
-    if not exists:
-        questions = generate_questions(df_string)
-        questions_text = json.dumps(questions, ensure_ascii=False, indent=2)
-        upload_content_directly(service, file_name, folder_id, questions_text)
-        print("questions已上傳到Google Drive")
-    else:
-        # 逐字稿已存在，下载逐字稿内容
-        print("questions已存在于Google Drive中")
-        questions_text = download_file_as_string(service, file_id)
-        questions = json.loads(questions_text)
+def get_questions(video_id, df_string, source="gcs"):
+    if source == "gcs":
+        # 去 gcs 確認是有有 video_id_questions.json
+        print("===get_questions on gcs===")
+        gcs_client = init_gcs_client(GCS_KEY)
+        bucket_name = 'video_ai_assistant'
+        file_name = f'{video_id}_questions.json'
+        blob_name = f"{video_id}/{file_name}"
+        # 检查檔案是否存在
+        is_questions_exists = gcs_check_file_exists(gcs_client, bucket_name, blob_name)
+        if not is_questions_exists:
+            questions = generate_questions(df_string)
+            questions_text = json.dumps(questions, ensure_ascii=False, indent=2)
+            upload_file_to_gcs_with_json_string(gcs_client, bucket_name, blob_name, questions_text)
+            print("questions已上傳到GCS")
+        else:
+            # 逐字稿已存在，下载逐字稿内容
+            print("questions已存在于GCS中")
+            questions_text = download_blob_to_string(gcs_client, bucket_name, blob_name)
+            questions = json.loads(questions_text)
+    
+    else if source == "drive":
+        # 去 g drive 確認是有有 video_id_questions.json
+        print("===get_questions===")
+        service = init_drive_service()
+        parent_folder_id = '1GgI4YVs0KckwStVQkLa1NZ8IpaEMurkL'
+        folder_id = create_folder_if_not_exists(service, video_id, parent_folder_id)
+        file_name = f'{video_id}_questions.json'
+
+        # 检查檔案是否存在
+        exists, file_id = check_file_exists(service, folder_id, file_name)
+        if not exists:
+            questions = generate_questions(df_string)
+            questions_text = json.dumps(questions, ensure_ascii=False, indent=2)
+            upload_content_directly(service, file_name, folder_id, questions_text)
+            print("questions已上傳到Google Drive")
+        else:
+            # 逐字稿已存在，下载逐字稿内容
+            print("questions已存在于Google Drive中")
+            questions_text = download_file_as_string(service, file_id)
+            questions = json.loads(questions_text)
 
     q1 = questions[0] if len(questions) > 0 else ""
     q2 = questions[1] if len(questions) > 1 else ""