feat: add admin dashboard, pipeline service, lightbox, and update dependencies

app/services/scheduler.py

@@ -3,7 +3,6 @@
 from __future__ import annotations
 
 import logging
-from datetime import datetime, timezone
 
 from apscheduler.schedulers.asyncio import AsyncIOScheduler
 from apscheduler.triggers.cron import CronTrigger
@@ -12,10 +11,8 @@ from zoneinfo import ZoneInfo
 
 from app.config import settings
 from app.database import SessionLocal
-from app.models import CrawlLog, TaskLock
-from app.services.cleaner import cleanup_tmp
-from app.services.crawler import crawl_daily
-from app.services.summarizer import summarize_batch
+from app.services.pipeline import run_pipeline
+from app.utils import today_str
 
 logger = logging.getLogger(__name__)
 
@@ -92,85 +89,15 @@ def stop_scheduler() -> None:
 async def _daily_pipeline() -> None:
     """每日流水线：抓取 → 总结 → 清理。
 
-    使用 task_locks 表防止重入：同一天的 pipeline 任务只有一个能运行。
+    委托给 pipeline.run_pipeline 执行，使用 task_locks 防重入。
     """
-    tz = ZoneInfo(settings.APP_TIMEZONE)
-    today = datetime.now(tz).strftime("%Y-%m-%d")
-    now = datetime.now(timezone.utc)
-    lock_key = f"pipeline-{today}"
+    today = today_str()
 
     db: Session = SessionLocal()
     try:
-        # 尝试获取锁
-        lock = TaskLock(
-            task="scheduler",
-            lock_key=lock_key,
-            status="running",
-            owner="daily_pipeline",
-            acquired_at=now,
-        )
-        try:
-            db.add(lock)
-            db.commit()
-        except Exception:
-            db.rollback()
-            logger.warning("Daily pipeline already running for %s, skipping", today)
-            return
-
-        # 写调度日志
-        log_entry = CrawlLog(
-            task="scheduler",
-            status="running",
-            date=datetime.now(tz).date(),
-            started_at=now,
-        )
-        db.add(log_entry)
-        db.commit()
-
-        error_msg = None
-        try:
-            # Step 1: 抓取
-            logger.info("Scheduler pipeline: crawl %s", today)
-            crawl_result = await crawl_daily(db, today)
-            logger.info(
-                "Scheduler pipeline: crawl done, found=%d new=%d",
-                crawl_result.get("found", 0),
-                crawl_result.get("new", 0),
-            )
-
-            # Step 2: 总结 pending 论文
-            logger.info("Scheduler pipeline: summarize batch")
-            summarize_result = await summarize_batch(db)
-            logger.info(
-                "Scheduler pipeline: summarize done, result=%s", summarize_result
-            )
-
-            # Step 3: 清理临时文件
-            logger.info("Scheduler pipeline: cleanup tmp")
-            cleanup_result = cleanup_tmp()
-            logger.info(
-                "Scheduler pipeline: cleanup done, removed=%d",
-                cleanup_result.get("removed", 0),
-            )
-
-            log_entry.status = "success"
-
-        except Exception as exc:
-            logger.exception("Scheduler pipeline failed for %s", today)
-            log_entry.status = "failed"
-            error_msg = str(exc)[:2000]
-
-        finally:
-            log_entry.completed_at = datetime.now(timezone.utc)
-            if error_msg:
-                log_entry.error = error_msg
-            db.commit()
-
-            # 释放锁
-            lock.status = "finished"
-            lock.released_at = datetime.now(timezone.utc)
-            db.commit()
-
+        await run_pipeline(db, today, owner="daily_pipeline")
+    except RuntimeError:
+        logger.warning("Daily pipeline already running for %s, skipping", today)
     except Exception:
         logger.exception("Unexpected error in daily pipeline")
     finally: