fix: 优化docker 镜像

crawler/realtime_conflict_service.py

@@ -14,13 +14,13 @@ from datetime import datetime
 from pathlib import Path
 from typing import List, Optional
 
+import asyncio
 import logging
 import requests
 from fastapi import FastAPI
 from fastapi.middleware.cors import CORSMiddleware
-from apscheduler.schedulers.background import BackgroundScheduler
 
-logging.getLogger("apscheduler.scheduler").setLevel(logging.ERROR)
+logging.getLogger("uvicorn").setLevel(logging.INFO)
 app = FastAPI(title="GDELT Conflict Service")
 app.add_middleware(CORSMiddleware, allow_origins=["*"], allow_methods=["*"])
 
@@ -263,8 +263,9 @@ def _extract_and_merge_panel_data(items: list) -> None:
             from extractor_ai import extract_from_news
         from datetime import timezone
         merged_any = False
-        # 只对前几条有足够文本的新闻做提取，避免 Ollama 调用过多
-        for it in items[:5]:
+        # 规则模式可多处理几条（无 Ollama）；AI 模式限制 5 条避免调用过多
+        limit = 10 if os.environ.get("CLEANER_AI_DISABLED", "0") == "1" else 5
+        for it in items[:limit]:
             text = (it.get("title", "") or "") + " " + (it.get("summary", "") or "")
             if len(text.strip()) < 20:
                 continue
@@ -292,12 +293,22 @@ def _extract_and_merge_panel_data(items: list) -> None:
 
 
 # ==========================
-# 定时任务（RSS 更频繁，优先保证事件脉络实时）
+# 定时任务（asyncio 后台任务，避免 APScheduler executor 关闭竞态）
 # ==========================
-scheduler = BackgroundScheduler()
-scheduler.add_job(fetch_news, "interval", seconds=RSS_INTERVAL_SEC, max_instances=2, coalesce=True)
-scheduler.add_job(fetch_gdelt_events, "interval", seconds=FETCH_INTERVAL_SEC, max_instances=2, coalesce=True)
-scheduler.start()
+_bg_task: Optional[asyncio.Task] = None
+
+
+async def _periodic_fetch() -> None:
+    loop = asyncio.get_event_loop()
+    while True:
+        try:
+            await loop.run_in_executor(None, fetch_news)
+            await loop.run_in_executor(None, fetch_gdelt_events)
+        except asyncio.CancelledError:
+            break
+        except Exception as e:
+            print(f"  [warn] 定时抓取: {e}")
+        await asyncio.sleep(min(RSS_INTERVAL_SEC, FETCH_INTERVAL_SEC))
 
 
 # ==========================
@@ -356,18 +367,23 @@ def _get_conflict_stats() -> dict:
 
 
 @app.on_event("startup")
-def startup():
-    # 新闻优先启动，确保事件脉络有数据
-    fetch_news()
-    fetch_gdelt_events()
+async def startup():
+    global _bg_task
+    loop = asyncio.get_event_loop()
+    await loop.run_in_executor(None, fetch_news)
+    await loop.run_in_executor(None, fetch_gdelt_events)
+    _bg_task = asyncio.create_task(_periodic_fetch())
 
 
 @app.on_event("shutdown")
-def shutdown():
-    try:
-        scheduler.shutdown(wait=False)
-    except Exception:
-        pass
+async def shutdown():
+    global _bg_task
+    if _bg_task and not _bg_task.done():
+        _bg_task.cancel()
+        try:
+            await _bg_task
+        except asyncio.CancelledError:
+            pass
 
 
 if __name__ == "__main__":