fix: 合规审查卡死修复 + 小红书复制格式 + today-only过滤

platform/backend/app/core/scheduler.py

@@ -31,10 +31,38 @@ MODULES = {
     "scheduled_task_monitor": {"name": "⏰ 任务监控", "cron": "*"},
 }
 
+LOG_FILE_MAP = {
+    "scheduled_refresh_search_cache": "opencode_search",
+    "scheduled_fetch_trends": "trends",
+    "scheduled_collect": "collector",
+    "scheduled_generate": "creator",
+    "scheduled_optimize": "optimizer",
+    "scheduled_optimize_sources": "optimizer_sources",
+    "scheduled_metrics_sync": "metrics_sync",
+    "scheduled_reset_search_usage": "reset_search_usage",
+    "scheduled_task_monitor": "task_monitor",
+}
+
+def _log_to_file(module_id: str, status: str, message: str = None, error_trace: str = None):
+    log_name = LOG_FILE_MAP.get(module_id, module_id)
+    log_dir = PROJECT_ROOT / "automation" / "logs"
+    log_dir.mkdir(parents=True, exist_ok=True)
+    log_file = log_dir / f"{log_name}_{datetime.now().strftime('%Y-%m-%d')}.log"
+    ts = datetime.now().strftime('%Y-%m-%d %H:%M:%S')
+    line = f"[{ts}] [{status.upper()}] {message or ''}"
+    if error_trace:
+        line += f"\n{error_trace[:500]}"
+    try:
+        with open(log_file, 'a', encoding='utf-8') as f:
+            f.write(line + "\n")
+    except Exception:
+        pass
+
 def _log_task(module_id: str, status: str, message: str = None,
               error_trace: str = None, result_data: dict = None,
               started_at: datetime = None, finished_at: datetime = None,
               triggered_by: str = "scheduler", next_run_time: datetime = None):
+    _log_to_file(module_id, status, message, error_trace)
     try:
         from ..database import SessionLocal
         from ..models import TaskLog
@@ -159,7 +187,7 @@ class TaskScheduler:
             logger.info("[Scheduled] Fetching hot trends...")
             import subprocess
             result = subprocess.run(
-                [sys.executable, str(Path(__file__).parent.parent.parent.parent / "scripts" / "trends.py")],
+                [sys.executable, str(PROJECT_ROOT / "scripts" / "trends.py")],
                 capture_output=True, text=True, timeout=120
             )
             if result.returncode == 0:
@@ -190,7 +218,7 @@ class TaskScheduler:
             logger.info("[Scheduled] Refreshing search cache via opencode...")
             import subprocess
             result = subprocess.run(
-                [sys.executable, str(Path(__file__).parent.parent.parent.parent / "scripts" / "opencode_search.py"), "--refresh-cache"],
+                [sys.executable, str(PROJECT_ROOT / "scripts" / "opencode_search.py"), "--refresh-cache"],
                 capture_output=True, text=True, timeout=600
             )
             for line in result.stdout.strip().split("\n"):
@@ -285,10 +313,10 @@ class TaskScheduler:
                       started_at=started, finished_at=datetime.now(timezone.utc))
             logger.exception("[Scheduled] Collection failed: %s", e)
 
-    def _run_optimize_sources(self):
+    def _run_optimize_sources(self, triggered_by="scheduler"):
         """AI自动优化采集类别与信息源：对比市场热点和当前配置，给出调整建议"""
         started = datetime.now(timezone.utc)
-        _log_task("scheduled_optimize_sources", "running", started_at=started)
+        _log_task("scheduled_optimize_sources", "running", started_at=started, triggered_by=triggered_by)
         try:
             logger.info("[Scheduled] Starting source optimization with AI...")
             from .nvidia_client import call_llm
@@ -337,19 +365,19 @@ class TaskScheduler:
                                   "sources_assessed": len(result.get("source_assessment", [])),
                                   "suggested_cats": len(result.get("suggested_new_categories", [])),
                                   "suggested_srcs": len(result.get("suggested_new_sources", []))},
-                      started_at=started, finished_at=datetime.now(timezone.utc))
+                      started_at=started, finished_at=datetime.now(timezone.utc), triggered_by=triggered_by)
             db.close()
         except Exception as e:
             _log_task("scheduled_optimize_sources", "failed",
                       message=str(e),
                       error_trace=traceback.format_exc(),
-                      started_at=started, finished_at=datetime.now(timezone.utc))
+                      started_at=started, finished_at=datetime.now(timezone.utc), triggered_by=triggered_by)
             logger.exception("[Scheduled] Source AI optimization failed: %s", e)
 
-    def _run_metrics_sync(self):
+    def _run_metrics_sync(self, triggered_by="scheduler"):
         """定时从各平台公开API获取发布文章的效果数据（当前仅支持知乎）"""
         started = datetime.now(timezone.utc)
-        _log_task("scheduled_metrics_sync", "running", started_at=started)
+        _log_task("scheduled_metrics_sync", "running", started_at=started, triggered_by=triggered_by)
         try:
             logger.info("[Scheduled] Starting metrics sync (zhihu auto-fetch)...")
             from ..database import SessionLocal
@@ -411,7 +439,7 @@ class TaskScheduler:
                 _log_task("scheduled_metrics_sync", "success",
                           message=f"同步完成，{count} 篇知乎文章",
                           result_data={"articles_synced": count},
-                          started_at=started, finished_at=datetime.now(timezone.utc))
+                          started_at=started, finished_at=datetime.now(timezone.utc), triggered_by=triggered_by)
                 # 生成指标反馈：按 field 聚合表现，写入 metrics_feedback.json 供 collector 读取
                 try:
                     import json as json_mod
@@ -436,7 +464,7 @@ class TaskScheduler:
                             "top_domains": [(f, s) for f, s, _ in scored[:5]],
                             "detail": [{"field": f, "score": s, "articles": c} for f, s, c in scored],
                         }
-                        feedback_file = Path(__file__).parent.parent.parent.parent / "automation" / "data" / "metrics_feedback.json"
+                        feedback_file = PROJECT_ROOT / "automation" / "data" / "metrics_feedback.json"
                         feedback_file.parent.mkdir(parents=True, exist_ok=True)
                         feedback_file.write_text(json_mod.dumps(feedback_data, ensure_ascii=False, indent=2), encoding='utf-8')
                         logger.info("[Scheduled] Metrics feedback written: top domain %s (score %.1f)", scored[0][0], scored[0][1])
@@ -445,14 +473,14 @@ class TaskScheduler:
             else:
                 _log_task("scheduled_metrics_sync", "success",
                           message="无已发布的知乎文章",
-                          started_at=started, finished_at=datetime.now(timezone.utc))
+                          started_at=started, finished_at=datetime.now(timezone.utc), triggered_by=triggered_by)
                 logger.info("[Scheduled] Metrics sync: no zhihu articles to sync")
             db.close()
         except Exception as e:
             _log_task("scheduled_metrics_sync", "failed",
                       message=str(e),
                       error_trace=traceback.format_exc(),
-                      started_at=started, finished_at=datetime.now(timezone.utc))
+                      started_at=started, finished_at=datetime.now(timezone.utc), triggered_by=triggered_by)
             logger.exception("[Scheduled] Metrics sync failed: %s", e)
 
     def _run_reset_search_usage(self):