新增自定义时间段批量打分功能：支持设置日期区间，对区间内每天自动拉取数据并打分

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

tushare/src/api.py

@@ -1,6 +1,6 @@
 from typing import Optional
 
-from datetime import date
+from datetime import date, datetime, timedelta
 
 from fastapi import FastAPI, HTTPException, Query
 from pydantic import BaseModel
@@ -16,6 +16,12 @@ class RunRequest(BaseModel):
     trade_date: Optional[str] = None
 
 
+class RunRangeRequest(BaseModel):
+    symbol: str = "FG"
+    start_date: str
+    end_date: str
+
+
 class RunResponse(BaseModel):
     ts_code: str
     trade_date: str
@@ -88,6 +94,42 @@ def run_batch():
     return {"results": results, "errors": errors}
 
 
+@app.post("/api/v1/run/range")
+def run_range(req: RunRangeRequest):
+    """对指定日期区间内的每一天分别打分。"""
+    ts_code = contracts.active_contract(req.symbol)
+
+    # 为确保区间开始日有足够前置数据,拉取时 start_date 前推 60 天
+    start_dt = datetime.strptime(req.start_date, "%Y%m%d")
+    fetch_start = (start_dt - timedelta(days=60)).strftime("%Y%m%d")
+
+    df = fetcher.fetch_contract(ts_code, start_date=fetch_start, end_date=req.end_date)
+    storage.save_candles(df)
+
+    results, warnings = scorer.score_range(df, req.start_date, req.end_date)
+
+    for r in results:
+        storage.save_score(r)
+
+    return {
+        "ts_code": ts_code,
+        "start_date": req.start_date,
+        "end_date": req.end_date,
+        "scored": len(results),
+        "skipped": len(warnings),
+        "warnings": warnings,
+        "results": [
+            {
+                "trade_date": r.trade_date,
+                "close": r.close,
+                "composite": r.composite,
+                "signal": r.signal,
+            }
+            for r in results
+        ],
+    }
+
+
 @app.get("/api/v1/scores")
 def list_scores(
     ts_code: Optional[str] = Query(None),

tushare/src/fetcher.py

@@ -11,10 +11,19 @@ def _init_api():
     return ts.pro_api()
 
 
-def fetch_contract(ts_code: str, limit: int = 100) -> pd.DataFrame:
+def fetch_contract(
+    ts_code: str,
+    start_date: Optional[str] = None,
+    end_date: Optional[str] = None,
+) -> pd.DataFrame:
     """拉取指定期货合约的日线数据,返回按 trade_date 升序排列的 DataFrame。"""
     pro = _init_api()
-    df = pro.fut_daily(ts_code=ts_code)
+    kwargs: dict = {"ts_code": ts_code}
+    if start_date:
+        kwargs["start_date"] = start_date
+    if end_date:
+        kwargs["end_date"] = end_date
+    df = pro.fut_daily(**kwargs)
 
     if df.empty:
         raise RuntimeError(f"未返回 {ts_code} 的任何数据,可能合约代码错误或 token 积分不足")

tushare/src/main.py

@@ -81,6 +81,52 @@ def run(ts_code: str, trade_date: Optional[str] = None) -> int:
     return 0
 
 
+def run_range(ts_code: str, start_date: str, end_date: str) -> int:
+    from datetime import datetime, timedelta
+
+    storage.init_db()
+
+    print(f"[1/4] 拉取 {ts_code} 数据 ({start_date} ~ {end_date})...")
+    start_dt = datetime.strptime(start_date, "%Y%m%d")
+    fetch_start = (start_dt - timedelta(days=60)).strftime("%Y%m%d")
+    df = fetcher.fetch_contract(ts_code, start_date=fetch_start, end_date=end_date)
+    print(f"      返回 {len(df)} 行")
+
+    print(f"[2/4] 写入/更新 PostgreSQL...")
+    storage.save_candles(df)
+
+    print(f"[3/4] 批量计算打分...")
+    results, warnings = scorer.score_range(df, start_date, end_date)
+
+    print(f"[4/4] 保存打分结果...")
+    for r in results:
+        storage.save_score(r)
+
+    print("\n" + "=" * 65)
+    print(f"合约: {ts_code}")
+    print(f"区间: {start_date} ~ {end_date}")
+    print(f"成功打分: {len(results)} 条")
+    if warnings:
+        print(f"跳过: {len(warnings)} 条")
+        for w in warnings[:5]:
+            print(f"  - {w}")
+        if len(warnings) > 5:
+            print(f"  ... 还有 {len(warnings) - 5} 条")
+    print("=" * 65)
+
+    quadrant_names = {
+        "accumulation": "增仓上涨", "distribution": "增仓下跌",
+        "covering": "减仓上涨", "liquidation": "减仓下跌", "flat": "持仓持平",
+    }
+    print(f"\n{'日期':<12} {'收盘':>10} {'综合':>8} {'信号':<20}")
+    print("-" * 55)
+    for r in results:
+        print(f"{r.trade_date:<12} {r.close:>10.2f} {r.composite:>8.1f} {r.signal:<20}")
+
+    print(f"\n[OK] {len(results)} 条打分已持久化到 PostgreSQL")
+    return 0
+
+
 def main() -> int:
     parser = argparse.ArgumentParser(description="期货合约三层打分模型")
     parser.add_argument(
@@ -97,11 +143,26 @@ def main() -> int:
         "--date",
         help="指定打分日期,格式 YYYYMMDD,不传则对最新日期打分",
     )
+    parser.add_argument(
+        "--start-date",
+        dest="start_date",
+        help="区间打分开始日期,格式 YYYYMMDD(与 --end-date 同时使用时生效)",
+    )
+    parser.add_argument(
+        "--end-date",
+        dest="end_date",
+        help="区间打分结束日期,格式 YYYYMMDD(与 --start-date 同时使用时生效)",
+    )
     args = parser.parse_args()
 
     ts_code = args.ts_code or contracts.active_contract(args.symbol)
     if not args.ts_code:
         print(f"[AUTO] {args.symbol} 当月主力 -> {ts_code}")
+
+    if args.start_date and args.end_date:
+        print(f"[RANGE] 区间打分: {args.start_date} ~ {args.end_date}")
+        return run_range(ts_code, args.start_date, args.end_date)
+
     if args.date:
         print(f"[DATE] 指定打分日期: {args.date}")
     return run(ts_code, args.date)

tushare/src/scorer.py

@@ -229,3 +229,28 @@ def score_daily(df: pd.DataFrame, trade_date: Optional[str] = None) -> ScoreResu
             },
         ),
     )
+
+
+def score_range(
+    df: pd.DataFrame, start_date: str, end_date: str
+) -> tuple[list[ScoreResult], list[str]]:
+    """对日期区间内的每一天分别打分,返回 (结果列表, 警告列表)。"""
+    if len(df) < 31:
+        raise ValueError(f"数据量不足(仅 {len(df)} 行),需要至少 31 行")
+
+    results: list[ScoreResult] = []
+    warnings: list[str] = []
+
+    target_dates = df[
+        (df["trade_date"].astype(str) >= str(start_date))
+        & (df["trade_date"].astype(str) <= str(end_date))
+    ]["trade_date"].astype(str).tolist()
+
+    for td in target_dates:
+        try:
+            result = score_daily(df, td)
+            results.append(result)
+        except ValueError as e:
+            warnings.append(str(e))
+
+    return results, warnings