Merge branch '2025-06-10-category-experiment' of luojunhui/LongArticlesJob into master

cold_start/publish/publish_single_video_pool_videos.py

@@ -23,6 +23,11 @@ video_pool_config = json.loads(
 video_category_list = json.loads(apollo_api.get_config_value(key="category_list"))
 platform_list = list(video_pool_config.keys())
 
+experimental_account = json.loads(
+    apollo_api.get_config_value(key="experimental_account")
+)
+experimental_account_set = set(experimental_account)
+
 
 class PublishSingleVideoPoolVideos:
     def __init__(self):
@@ -32,7 +37,7 @@ class PublishSingleVideoPoolVideos:
     def get_task_list(self, platform: str) -> list[dict]:
         daily_limit = video_pool_config[platform]["process_num_each_day"]
         fetch_query = f"""
-            select t1.id, t1.content_trace_id, t1.pq_vid, t1.score, t2.category
+            select t1.id, t1.content_trace_id, t1.pq_vid, t1.score, t2.category, t2.out_account_id
             from single_video_transform_queue t1
             join publish_single_video_source t2 on t1.content_trace_id = t2.content_trace_id
             where 
@@ -59,11 +64,20 @@ class PublishSingleVideoPoolVideos:
         return affected_rows
 
     def create_crawler_plan(
-        self, vid_list: list, platform: str, task_id_tuple: tuple, category: str
+        self,
+        vid_list: list,
+        platform: str,
+        task_id_tuple: tuple,
+        category: str,
+        experiment_tag: str = None,
     ) -> None:
         try:
             # create video crawler plan
-            plan_name = f"{video_pool_config[platform]['nick_name']}-{category}-{datetime.datetime.today().strftime('%Y-%m-%d')}-视频数量： {len(vid_list)}"
+            date_info = datetime.datetime.today().strftime("%Y-%m-%d")
+            if experiment_tag:
+                plan_name = f"{video_pool_config[platform]['nick_name']}-{category}-{date_info}-视频数量： {len(vid_list): {experiment_tag}}"
+            else:
+                plan_name = f"{video_pool_config[platform]['nick_name']}-{category}-{date_info}-视频数量： {len(vid_list)}"
             crawler_plan_response = aiditApi.auto_create_single_video_crawler_task(
                 plan_name=plan_name,
                 plan_tag="单视频供给冷启动",
@@ -145,17 +159,48 @@ class PublishSingleVideoPoolVideos:
                     task_id_tuple = tuple(
                         [task["id"] for task in task_list_with_category]
                     )
-                    vid_list = [task["pq_vid"] for task in task_list_with_category]
-                    if vid_list:
+                    # 区分账号ID 是否属于实验账号
+                    experimental_vid_list = [
+                        task["pq_vid"]
+                        for task in task_list_with_category
+                        if task["out_account_id"] in experimental_account_set
+                    ]
+
+                    normal_vid_list = [
+                        task["pq_vid"]
+                        for task in task_list_with_category
+                        if task["out_account_id"] not in experimental_account_set
+                    ]
+
+                    if normal_vid_list:
+                        self.create_crawler_plan(
+                            normal_vid_list, platform, task_id_tuple, category
+                        )
+                        if publish_detail_table.get(platform):
+                            publish_detail_table[platform][category] = len(
+                                normal_vid_list
+                            )
+                        else:
+                            publish_detail_table[platform] = {
+                                category: len(normal_vid_list)
+                            }
+
+                    if experimental_vid_list:
                         self.create_crawler_plan(
-                            vid_list, platform, task_id_tuple, category
+                            experimental_vid_list,
+                            platform,
+                            task_id_tuple,
+                            category,
+                            "20250610-品类账号实验",
                         )
                         if publish_detail_table.get(platform):
-                            publish_detail_table[platform][category] = len(vid_list)
+                            publish_detail_table[platform][category] += len(
+                                experimental_vid_list
+                            )
                         else:
-                            publish_detail_table[platform] = {category: len(vid_list)}
-                    else:
-                        continue
+                            publish_detail_table[platform] = {
+                                category: len(experimental_vid_list)
+                            }
             else:
                 feishu_bot_api.bot(
                     title="视频内容池发布任务",