优化了实时保存的功能

backend/Timer_worker.py

@@ -179,24 +179,44 @@ class DouyinAutoScheduler:
                 
                 logging.info(f"📊 今日数据去重后：{len(today_videos)} 个独特短剧（原始数据：{len(today_videos_raw)} 条）")
                 
-                # 获取昨天的榜单数据（如果存在），取最新的计算结果
+                # 从Ranking_storage_list中获取昨天最后一次抓取的数据
-                yesterday_ranking = rankings_collection.find_one({
+                yesterday_start = datetime.combine(yesterday, datetime.min.time())
-                    "date": yesterday_str,
+                yesterday_end = datetime.combine(yesterday, datetime.max.time())
-                    "type": "comprehensive"
+                
-                }, sort=[("calculation_sequence", -1)])
+                # 获取昨天的最后一次抓取数据（按batch_time排序取最新的）
+                yesterday_latest_batch = douyin_collection.find_one({
+                    "batch_time": {
+                        "$gte": yesterday_start,
+                        "$lte": yesterday_end
+                    }
+                }, sort=[("batch_time", -1)])
                 
                 yesterday_data = {}
-                if yesterday_ranking and "data" in yesterday_ranking:
+                if yesterday_latest_batch:
+                    yesterday_batch_time = yesterday_latest_batch.get("batch_time")
+                    logging.info(f"📊 找到昨天最后一次抓取时间: {yesterday_batch_time}")
+                    
+                    # 获取昨天最后一次抓取的所有数据
+                    yesterday_videos_raw = list(douyin_collection.find({
+                        "batch_time": yesterday_batch_time
+                    }).sort("play_vv", -1))
+                    
+                    # 按短剧名称去重，每个短剧只保留播放量最高的一条
+                    unique_yesterday_videos = {}
+                    for video in yesterday_videos_raw:
+                        mix_name = video.get("mix_name", "")
+                        if mix_name and (mix_name not in unique_yesterday_videos or video.get("play_vv", 0) > unique_yesterday_videos[mix_name].get("play_vv", 0)):
+                            unique_yesterday_videos[mix_name] = video
+                    
                     # 将昨天的数据转换为字典，以短剧名称为键
-                    for item in yesterday_ranking["data"]:
+                    for mix_name, video in unique_yesterday_videos.items():
-                        title = item.get("title", "")
+                        yesterday_data[mix_name] = {
-                        if title:
+                            "rank": 0,  # 原始数据没有排名，设为0
-                            yesterday_data[title] = {
+                            "play_vv": video.get("play_vv", 0),
-                                "rank": item.get("rank", 0),
+                            "video_id": str(video.get("_id", ""))
-                                "play_vv": item.get("play_vv", 0),
-                                "video_id": item.get("video_id", "")
                         }
-                    logging.info(f"📊 找到昨天的榜单数据，共 {len(yesterday_data)} 个短剧")
+                    
+                    logging.info(f"📊 找到昨天的原始数据，共 {len(yesterday_data)} 个短剧（原始数据：{len(yesterday_videos_raw)} 条）")
                 else:
                     logging.info("📊 未找到昨天的原始数据，将作为首次生成")
                 

backend/config.py

@@ -3,9 +3,9 @@ import importlib
 
 # 数据库配置
 MONGO_URI = "mongodb://localhost:27017"
-MONGO_DB_NAME = "kemeng_media"
+# MONGO_DB_NAME = "Rankings"
 # MONGO_URI = "mongodb://mongouser:Jdei2243afN@172.16.0.6:27017,172.16.0.4:27017/test?replicaSet=cmgo-r6qkaern_0&authSource=admin"
-# MONGO_DB_NAME = "kemeng_media"
+MONGO_DB_NAME = "kemeng_media"
 
 # 应用配置
 APP_ENV = os.getenv('APP_ENV', 'development')