Merge branch 'main' of http://git.xintiao100.com/Qyir/rank_backend

backend/Timer_worker.py

@@ -103,12 +103,6 @@ class DouyinAutoScheduler:
 
             # 设置环境变量，确保自动模式
             os.environ['AUTO_CONTINUE'] = '1'
-            # 设置定时器模式环境变量，跳过评论抓取等函数
-            os.environ['TIMER_MODE'] = '1'
-            
-            # 只在定时器模式下设置静默模式（非测试、非单次执行、非仅生成榜单）
-            if hasattr(self, '_is_timer_mode') and self._is_timer_mode:
-                os.environ['QUIET_MODE'] = '1'
 
             # 直接创建并运行 DouyinPlayVVScraper 实例
             scraper = DouyinPlayVVScraper(
@@ -116,11 +110,11 @@ class DouyinAutoScheduler:
                 auto_continue=True,
                 duration_s=60
             )
-
-            logging.warning("📁 开始执行抓取任务...")
+            
+            logging.info("📁 开始执行抓取任务...")
             scraper.run()
-
-            logging.warning("✅ 抖音播放量抓取任务执行成功")
+            
+            logging.info("✅ 抖音播放量抓取任务执行成功")
             
             # 数据抓取完成后，自动生成当日榜单
             self.generate_daily_rankings()
@@ -168,43 +162,35 @@ class DouyinAutoScheduler:
                 today_videos_raw = list(douyin_collection.find({"batch_time": latest_batch_time}).sort("play_vv", -1))
                 logging.info(f"📊 最新批次数据数量: {len(today_videos_raw)}")
                 
-                # 调试：检查原始数据
-                if today_videos_raw:
-                    sample_video = today_videos_raw[0]
-                    logging.info(f"🔍 样本数据检查:")
-                    logging.info(f"   mix_name: {sample_video.get('mix_name')}")
-                    logging.info(f"   play_vv: {sample_video.get('play_vv')} (类型: {type(sample_video.get('play_vv'))})")
-                    logging.info(f"   author: {sample_video.get('author')}")
+                # 按短剧名称去重，每个短剧只保留播放量最高的一条
+                unique_videos = {}
+                for video in today_videos_raw:
+                    mix_name = video.get("mix_name", "")
+                    if mix_name and (mix_name not in unique_videos or video.get("play_vv", 0) > unique_videos[mix_name].get("play_vv", 0)):
+                        unique_videos[mix_name] = video
                 
-                # 按短剧名称去重并确保数据类型正确
-                unique_videos = self._deduplicate_videos_by_mix_name(today_videos_raw, include_rank=False)
                 today_videos = list(unique_videos.values())
                 
                 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）")
                 
-                # 获取昨天最后一批次的数据
-                yesterday_start = datetime(yesterday.year, yesterday.month, yesterday.day)
-                yesterday_end = yesterday_start + timedelta(days=1)
-                yesterday_batch = douyin_collection.find_one({
-                    "batch_time": {"$gte": yesterday_start, "$lt": yesterday_end}
-                }, sort=[("batch_time", -1)])
+                # 获取昨天的榜单数据（如果存在），取最新的计算结果
+                yesterday_ranking = rankings_collection.find_one({
+                    "date": yesterday_str,
+                    "type": "comprehensive"
+                }, sort=[("calculation_sequence", -1)])
                 
                 yesterday_data = {}
-                if yesterday_batch:
-                    # 获取昨天最后一批次的所有数据
-                    yesterday_videos = list(douyin_collection.find({
-                        "batch_time": yesterday_batch["batch_time"]
-                    }).sort("play_vv", -1))
-                    
-                    # 按短剧名称去重，保留播放量最高的记录，并确保数据类型正确
-                    yesterday_data = self._deduplicate_videos_by_mix_name(yesterday_videos, include_rank=True)
-                    
-                    # 计算排名
-                    sorted_videos = sorted(yesterday_data.items(), key=lambda x: x[1]["play_vv"], reverse=True)
-                    for rank, (mix_name, data) in enumerate(sorted_videos, 1):
-                        yesterday_data[mix_name]["rank"] = rank
-                    
-                    logging.info(f"📊 找到昨天的原始数据，共 {len(yesterday_data)} 个短剧")
+                if yesterday_ranking and "data" in yesterday_ranking:
+                    # 将昨天的数据转换为字典，以短剧名称为键
+                    for item in yesterday_ranking["data"]:
+                        title = item.get("title", "")
+                        if title:
+                            yesterday_data[title] = {
+                                "rank": item.get("rank", 0),
+                                "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")
                 else:
                     logging.info("📊 未找到昨天的原始数据，将作为首次生成")