提交

backend/app/data/etf_sync.py

@@ -20,7 +20,7 @@ from __future__ import annotations
 
 import asyncio
 import time
-from datetime import datetime, timedelta, timezone
+from datetime import datetime, timedelta
 
 from sqlalchemy import delete, func, select, text
 from sqlalchemy.dialects.postgresql import insert as pg_insert
@@ -29,6 +29,7 @@ from sqlalchemy.ext.asyncio import AsyncSession
 from .. import cache
 from ..config import settings
 from . import etf_provider
+from .sync_utils import call_retry, get_pro_lazy, utcnow
 
 # 进程内单例任务状态（uvicorn 单进程场景够用）
 _state: dict = {
@@ -46,34 +47,6 @@ _lock = asyncio.Lock()
 _BATCH = 3000       # upsert 分批行数（asyncpg 单语句参数上限 32766，10 列/行）
 # fund_daily 返回全市场基金 ~2100 行，一天一批远小于上限
 
-# 频率超限特征（等待 62s 重试一次；与 screener.market_sync._call_retry 同款语义）
-_RATE_MARKS = ("频率超限", "每分钟")
-
-
-def _call_retry(fn, *args, **kwargs):
-    """同步调用 tushare 接口；「每分钟」级频率超限等 62s 重试一次。"""
-    try:
-        return fn(*args, **kwargs)
-    except Exception as e:  # noqa: BLE001
-        msg = str(e)
-        if any(m in msg for m in _RATE_MARKS) and "小时" not in msg:
-            time.sleep(62)
-            return fn(*args, **kwargs)
-        raise
-
-
-def _get_pro():
-    """token 检查 + 返回 pro api 客户端（同步对象，调用需 to_thread 包裹）。"""
-    if not settings.tushare_token:
-        raise RuntimeError("未配置 TUSHARE_TOKEN，无法同步 ETF 日线（backend/.env）")
-    from .tushare_provider import get_pro
-
-    return get_pro()
-
-
-def _utcnow() -> datetime:
-    return datetime.now(timezone.utc).replace(tzinfo=None)
-
 
 def _parse_d(s: str) -> datetime:
     return datetime.strptime(str(s), "%Y%m%d")
@@ -85,7 +58,7 @@ async def _sync_spot(session: AsyncSession) -> int:
 
     async with etf_provider.new_client() as client:
         rows = await etf_provider.fetch_etf_spot(client)
-    now = _utcnow()
+    now = utcnow()
     stmt = pg_insert(EtfBasic).values([{**r, "updated_at": now} for r in rows])
     stmt = stmt.on_conflict_do_update(
         index_elements=["ts_code"],
@@ -108,7 +81,7 @@ async def _sync_spot(session: AsyncSession) -> int:
 def _fetch_day_sync(pro, d: str) -> list[dict]:
     """拉某交易日全市场场内基金日线（fund_daily；未生成的日期返回空）。"""
     time.sleep(settings.screener_sync_interval)
-    df = _call_retry(pro.fund_daily, trade_date=d)
+    df = call_retry(pro.fund_daily, trade_date=d)
     if df is None or df.empty:
         return []
     rows = []
@@ -128,7 +101,7 @@ def _fetch_day_sync(pro, d: str) -> list[dict]:
 def _fetch_symbol_sync(pro, ts_code: str, start: str | None, end: str | None) -> list[dict]:
     """按 ts_code 增量/全量拉单只 ETF 日线（start=None 即上市以来全量）。"""
     time.sleep(settings.screener_sync_interval)
-    df = _call_retry(pro.fund_daily, ts_code=ts_code, start_date=start, end_date=end)
+    df = call_retry(pro.fund_daily, ts_code=ts_code, start_date=start, end_date=end)
     if df is None or df.empty:
         return []
     df = df.sort_values("trade_date")
@@ -209,7 +182,7 @@ async def _run_sync(full: bool) -> None:
     from ..models import Candle, EtfBasic, TradeCalendar
 
     try:
-        pro = await asyncio.to_thread(_get_pro)
+        pro = await asyncio.to_thread(get_pro_lazy)
 
         # 1) 快照 -> etf_basic
         _state["step"] = "正在拉取 ETF 列表"