feat(data): 15min数据补全—路径参数化+cron入口+回填+设计文档

- download_minute.py: STOCK_ROOT环境变量参数化(Mac/NAS兼容); 修tencent备源amount聚合 last→sum - refresh_15min_daily.py: 新建cron入口(交易日判断baostock/直连pop代理/日志) - 回填470缺口(170 SH/SZ成功, 300 BSE源不支持) - docs/data/15min-data-design.md: 15min数据层设计文档(源/覆盖/脚本/刷新/约束/路径映射)
2026-07-12 08:26:30 +08:00
parent 10508db11d
commit cf4b3a2cb0
3 changed files with 332 additions and 4 deletions
@@ -23,6 +23,7 @@

 import argparse
 import json
+import os
 import re
 import sys
 import time
@@ -42,14 +43,16 @@ logging.basicConfig(
 logger = logging.getLogger(__name__)

 # --- 配置 ---
-OUTPUT_DIR = Path("/Volumes/stock/minute_kline/15min")
+# STOCK_ROOT 可通过环境变量覆盖；默认 Mac 挂载点，NAS 用 STOCK_ROOT=/volume1/stock
+STOCK_ROOT = Path(os.environ.get("STOCK_ROOT", "/Volumes/stock"))
+OUTPUT_DIR = STOCK_ROOT / "minute_kline" / "15min"
 PROGRESS_FILE = OUTPUT_DIR / "download_progress.json"
 REQUEST_INTERVAL = 0.3
 MAX_RETRIES = 3
 CONSECUTIVE_FAIL_PAUSE = 60
 MAX_CONSECUTIVE_FAILS = 5
-HS300_FILE = Path("/Volumes/stock/A股数据/stock_info/hs300_constituents_latest.csv")
-ALL_STOCKS_FILE = Path("/Volumes/stock/A股数据/stock_info/stock_basic_info_raw_20260326_113530.csv")
+HS300_FILE = STOCK_ROOT / "A股数据" / "stock_info" / "hs300_constituents_latest.csv"
+ALL_STOCKS_FILE = STOCK_ROOT / "A股数据" / "stock_info" / "stock_basic_info_raw_20260326_113530.csv"

 HEADERS = {"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7)"}

@@ -141,7 +144,7 @@ def _aggregate_1m_to_15m(df: pd.DataFrame) -> pd.DataFrame:
        low=("price", "min"),
        close=("price", "last"),
        volume=("vol", "sum"),
-        amount=("amount", "last"),
+        amount=("amount", "sum"),
    ).reset_index()

    return pd.DataFrame({
@@ -0,0 +1,130 @@
+#!/usr/bin/env python3
+"""
+15min 日增量刷新 - NAS Synology 任务计划入口
+
+功能：
+  1. 清除代理环境变量（直连，避免被识别为异常流量）
+  2. 判断今天是否 A 股交易日（baostock query_trade_dates；不可用降级为"周六日跳过"）
+  3. 交易日 → 调 download_minute.py --scope all --resume
+  4. 记日志到 $STOCK_ROOT/logs/daily_update/refresh_15min_YYYYMMDD.log
+
+环境变量：
+  STOCK_ROOT  数据根目录（默认 /volume1/stock，Mac 测试用 /Volumes/stock）
+
+Synology 任务计划命令（容器内 Python，依赖已装好）：
+  docker exec sanguo_vnpy_v2 bash -c "cd /app && python3 scripts/data_platform/refresh_15min_daily.py"
+
+或 NAS host Python（需自行 pip install baostock pandas pyarrow）：
+  STOCK_ROOT=/volume1/stock python3 /volume1/homes/admin/.sanguo_projects/sanguo_vnpy_v2/scripts/data_platform/refresh_15min_daily.py
+"""
+
+import logging
+import os
+import subprocess
+import sys
+from datetime import datetime
+from pathlib import Path
+from typing import Tuple
+
+# --- 清除代理（脚本一进来就 pop，保证后续 urllib/baostock 直连）---
+for _k in ("http_proxy", "https_proxy", "HTTP_PROXY", "HTTPS_PROXY",
+           "all_proxy", "ALL_PROXY", "no_proxy", "NO_PROXY"):
+    os.environ.pop(_k, None)
+
+# --- 路径配置 ---
+# NAS: /volume1/stock；Mac: /Volumes/stock。也可通过环境变量覆盖。
+STOCK_ROOT = Path(os.environ.get("STOCK_ROOT", "/volume1/stock"))
+LOG_DIR = STOCK_ROOT / "logs" / "daily_update"
+SCRIPT_DIR = Path(__file__).resolve().parent
+DOWNLOAD_SCRIPT = SCRIPT_DIR / "download_minute.py"
+
+# --- 日志（先建目录，LOG_DIR 在 STOCK_ROOT 下） ---
+LOG_DIR.mkdir(parents=True, exist_ok=True)
+_today_str = datetime.now().strftime("%Y%m%d")
+LOG_FILE = LOG_DIR / f"refresh_15min_{_today_str}.log"
+
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s %(levelname)s %(message)s",
+    handlers=[
+        logging.FileHandler(LOG_FILE, encoding="utf-8"),
+        logging.StreamHandler(),
+    ],
+)
+logger = logging.getLogger(__name__)
+
+
+def is_trading_day() -> Tuple[bool, str]:
+    """
+    判断今天是否 A 股交易日。
+    优先 baostock query_trade_dates；不可用降级为"周六日跳过"（工作日默认交易）。
+    返回 (is_trading, reason)
+    """
+    today_iso = datetime.now().strftime("%Y-%m-%d")
+    today_date = datetime.now().date()
+
+    # 周末一定非交易日（即使 baostock 失败也能正确跳过）
+    if today_date.weekday() >= 5:
+        return False, f"周末({today_date.strftime('%A')})"
+
+    # 工作日 → 用 baostock 查节假日
+    try:
+        import baostock as bs
+        lg = bs.login()
+        if lg.error_code != "0":
+            logger.warning("baostock 登录失败: %s，降级为工作日默认交易日", lg.error_msg)
+            return True, f"baostock登录失败({lg.error_msg})，工作日默认交易日"
+        try:
+            rs = bs.query_trade_dates(start_date=today_iso, end_date=today_iso)
+            if rs.error_code != "0":
+                logger.warning("baostock 查询交易日失败: %s，降级", rs.error_msg)
+                return True, f"baostock查询失败({rs.error_msg})，工作日默认交易日"
+            while rs.next():
+                row = rs.get_row_data()  # [date, is_trading_day]
+                if row[0] == today_iso:
+                    is_trading = (row[1] == "1")
+                    return is_trading, f"baostock is_trading_day={row[1]}"
+            # 没查到今天 → 工作日默认交易
+            return True, "baostock 无今日记录，工作日默认交易日"
+        finally:
+            bs.logout()
+    except Exception as e:
+        logger.warning("baostock 不可用: %s，降级为工作日默认交易日", e)
+        return True, f"baostock异常({e})，工作日默认交易日"
+
+
+def run_download() -> int:
+    """调用 download_minute.py --scope all --resume，继承当前环境（含 STOCK_ROOT）"""
+    cmd = [sys.executable, str(DOWNLOAD_SCRIPT), "--scope", "all", "--resume"]
+    env = dict(os.environ)
+    env["STOCK_ROOT"] = str(STOCK_ROOT)
+    logger.info("执行命令: %s", " ".join(cmd))
+    logger.info("STOCK_ROOT=%s", STOCK_ROOT)
+    result = subprocess.run(cmd, env=env)
+    return result.returncode
+
+
+def main() -> int:
+    logger.info("=" * 60)
+    logger.info("15min 日刷新开始 @ %s", datetime.now().isoformat())
+    logger.info("STOCK_ROOT=%s", STOCK_ROOT)
+    logger.info("LOG_FILE=%s", LOG_FILE)
+
+    trading, reason = is_trading_day()
+    logger.info("交易日判断: %s (%s)", trading, reason)
+    if not trading:
+        logger.info("非交易日, 跳过下载")
+        return 0
+
+    rc = run_download()
+    logger.info("download_minute.py 退出码: %d", rc)
+    if rc != 0:
+        logger.error("❌ 下载脚本失败")
+        return rc
+
+    logger.info("✅ 15min 日刷新完成")
+    return 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())