stock-analyzer-skill 1.1.0

package/scripts/refresh_pool.py

@@ -0,0 +1,328 @@
+#!/usr/bin/env python3
+"""
+股票池自动刷新脚本 — 从东财 push2 API 拉取板块成分股。
+
+用法:
+  python3 scripts/refresh_pool.py                     # 刷新全部板块
+  python3 scripts/refresh_pool.py --sector 机器人      # 只刷新指定板块
+  python3 scripts/refresh_pool.py --top 30             # 每板块取 Top 30
+  python3 scripts/refresh_pool.py --sort cap           # 按市值排序（默认 amount）
+  python3 scripts/refresh_pool.py --dry-run            # 只打印不写入
+  python3 scripts/refresh_pool.py --diff               # 对比当前池显示变更
+
+数据源: 东财 push2.eastmoney.com 板块成分股 API
+"""
+
+import argparse
+import json
+import os
+import sys
+import time
+from datetime import datetime
+
+# 复用 common.py 的 HTTP 和编码工具
+sys.path.insert(0, os.path.dirname(__file__))
+from common import http_get_cached, normalize_quote_code, board_type
+
+# ---------- 常量 ----------
+
+SCRIPT_DIR = os.path.dirname(os.path.abspath(__file__))
+DATA_DIR = os.path.join(SCRIPT_DIR, "data")
+MAPPING_FILE = os.path.join(DATA_DIR, "sector_mapping.json")
+POOL_FILE = os.path.join(DATA_DIR, "sector_stocks.json")
+
+API_BASE = "https://push2.eastmoney.com/api/qt/clist/get"
+API_TOKEN = os.environ.get("EASTMONEY_API_TOKEN", "")
+if not API_TOKEN:
+    print("警告: 未设置 EASTMONEY_API_TOKEN 环境变量，refresh_pool 功能不可用", file=sys.stderr)
+FIELDS = "f12,f14,f2,f3,f6,f8,f9,f20"  # code,name,price,chg%,amount,turnover,pe,cap
+
+# 硬过滤阈值
+FILTER = {
+    "min_amount": {"主板": 5000, "创业板": 3500, "科创板": 3500, "北交所": 7500},  # 万元
+    "min_cap":    {"主板": 40,   "创业板": 24,   "科创板": 24,   "北交所": 16},     # 亿元
+}
+
+
+# ---------- API 调用 ----------
+
+def fetch_board_stocks(bk_code: str, max_retries: int = 2) -> list[dict]:
+    """获取板块成分股，返回 [{code, name, price, change_pct, amount, turnover, pe, cap}]"""
+    url = (
+        f"{API_BASE}?pn=1&pz=500&np=1&ut={API_TOKEN}"
+        f"&fltt=2&invt=2&fid=f3&fs=b:{bk_code}&fields={FIELDS}"
+    )
+    for attempt in range(max_retries + 1):
+        try:
+            raw = http_get_cached(url, ttl=3600)  # 缓存 1 小时
+            data = json.loads(raw)
+            if not data or "data" not in data or not data["data"]:
+                return []
+            items = data["data"].get("diff", [])
+            results = []
+            for item in items:
+                code6 = str(item.get("f12", ""))
+                if not code6 or len(code6) != 6:
+                    continue
+                # 推断交易所前缀
+                if code6.startswith(("6",)):
+                    full_code = f"sh{code6}"
+                elif code6.startswith(("0", "3")):
+                    full_code = f"sz{code6}"
+                elif code6.startswith(("4", "8")):
+                    full_code = f"bj{code6}"
+                else:
+                    full_code = f"sz{code6}"
+                results.append({
+                    "code": full_code,
+                    "name": item.get("f14", ""),
+                    "price": item.get("f2"),
+                    "change_pct": item.get("f3"),
+                    "amount": item.get("f6"),      # 成交额（元）
+                    "turnover": item.get("f8"),     # 换手率
+                    "pe": item.get("f9"),           # PE(动)
+                    "cap": item.get("f20"),         # 总市值（元）
+                })
+            return results
+        except Exception as e:
+            if attempt < max_retries:
+                time.sleep(1)
+                continue
+            print(f"  ⚠ API 请求失败 ({bk_code}): {e}", file=sys.stderr)
+            return []
+
+
+def fetch_multiple_boards(bk_codes: list[str]) -> list[dict]:
+    """合并多个板块的成分股，去重"""
+    seen = {}
+    for bk in bk_codes:
+        stocks = fetch_board_stocks(bk)
+        for s in stocks:
+            if s["code"] not in seen:
+                seen[s["code"]] = s
+        time.sleep(0.5)  # 限流
+    return list(seen.values())
+
+
+# ---------- 过滤 ----------
+
+def is_st(name: str) -> bool:
+    return "ST" in name.upper() or "*ST" in name
+
+
+def passes_filter(stock: dict) -> tuple[bool, str]:
+    """硬过滤，返回 (通过, 原因)"""
+    name = stock.get("name", "")
+    code = stock.get("code", "")
+    bt = board_type(code)
+
+    if is_st(name):
+        return False, "ST"
+
+    amount = stock.get("amount")
+    if amount is not None:
+        amount_wan = amount / 10000  # 元→万元
+        min_amt = FILTER["min_amount"].get(bt, 5000)
+        if amount_wan < min_amt:
+            return False, f"成交额{amount_wan:.0f}万<{min_amt}万"
+
+    cap = stock.get("cap")
+    if cap is not None:
+        cap_yi = cap / 100000000  # 元→亿元
+        min_cap = FILTER["min_cap"].get(bt, 40)
+        if cap_yi < min_cap:
+            return False, f"市值{cap_yi:.0f}亿<{min_cap}亿"
+
+    return True, ""
+
+
+# ---------- 排序与筛选 ----------
+
+def sort_stocks(stocks: list[dict], key: str = "amount") -> list[dict]:
+    """排序，降序"""
+    keys = {
+        "amount": lambda s: s.get("amount") or 0,
+        "cap": lambda s: s.get("cap") or 0,
+        "pe": lambda s: s.get("pe") or 9999,
+        "turnover": lambda s: s.get("turnover") or 0,
+    }
+    return sorted(stocks, key=keys.get(key, keys["amount"]), reverse=True)
+
+
+def build_sector_pool(stocks: list[dict], top_n: int = 20, sort_by: str = "amount") -> list[str]:
+    """过滤+排序+截取，返回代码列表"""
+    filtered = []
+    reasons = {}
+    for s in stocks:
+        ok, reason = passes_filter(s)
+        if ok:
+            filtered.append(s)
+        else:
+            reasons[s["code"]] = reason
+
+    sorted_stocks = sort_stocks(filtered, sort_by)
+    return [s["code"] for s in sorted_stocks[:top_n]]
+
+
+# ---------- 高股息特殊处理 ----------
+
+def build_dividend_pool(all_pools: dict[str, list[str]], all_stocks: dict[str, dict]) -> list[str]:
+    """从所有板块中筛选 PE<20 且 ROE>8% 的标的"""
+    candidates = []
+    seen = set()
+    for sector, codes in all_pools.items():
+        if sector == "高股息":
+            continue
+        for code in codes:
+            if code in seen:
+                continue
+            seen.add(code)
+            s = all_stocks.get(code, {})
+            pe = s.get("pe")
+            # ROE 需要额外查询，这里用 PE 作为第一轮筛选
+            if pe is not None and 0 < pe < 20:
+                candidates.append(s)
+    # 按成交额排序取 Top 15
+    sorted_c = sort_stocks(candidates, "amount")
+    return [s["code"] for s in sorted_c[:20]]
+
+
+# ---------- 主流程 ----------
+
+def load_mapping() -> dict:
+    with open(MAPPING_FILE, "r", encoding="utf-8") as f:
+        return json.load(f)
+
+
+def load_current_pool() -> dict:
+    if not os.path.exists(POOL_FILE):
+        return {}
+    with open(POOL_FILE, "r", encoding="utf-8") as f:
+        data = json.load(f)
+    return {k: v for k, v in data.items() if not k.startswith("_")}
+
+
+def refresh_pool(sectors: list[str] | None = None, top_n: int = 20,
+                 sort_by: str = "amount", dry_run: bool = False,
+                 show_diff: bool = False) -> dict:
+    """刷新股票池，返回新池"""
+    mapping = load_mapping()
+    current = load_current_pool()
+
+    target_sectors = sectors or [k for k in mapping if not k.startswith("_")]
+    new_pool = dict(current)  # 以当前池为基础，只更新目标板块
+    all_stocks_raw = {}  # code → stock dict，用于高股息筛选
+
+    for sector in target_sectors:
+        if sector not in mapping:
+            print(f"⚠ 板块 '{sector}' 不在映射配置中，跳过", file=sys.stderr)
+            continue
+
+        cfg = mapping[sector]
+        bk_codes = cfg.get("bk_codes", [])
+
+        if cfg.get("filter") == "dividend":
+            # 高股息板块延迟处理
+            continue
+
+        if not bk_codes:
+            print(f"⚠ 板块 '{sector}' 无 BK 代码，跳过", file=sys.stderr)
+            continue
+
+        print(f"📡 获取 {sector} ({', '.join(bk_codes)})...", end=" ", flush=True)
+        stocks = fetch_multiple_boards(bk_codes)
+        print(f"{len(stocks)} 只原始")
+
+        if stocks:
+            pool = build_sector_pool(stocks, top_n, sort_by)
+            new_pool[sector] = pool
+        elif sector in current:
+            new_pool[sector] = current[sector]
+            print(f"  ⚠ API 失败，保留现有 {len(current[sector])} 只")
+
+        # 保存原始数据供高股息筛选
+        for s in stocks:
+            all_stocks_raw[s["code"]] = s
+
+        time.sleep(0.3)
+
+    # 高股息板块
+    if "高股息" in target_sectors and "高股息" in mapping:
+        if all_stocks_raw:
+            dividend_pool = build_dividend_pool(new_pool, all_stocks_raw)
+            new_pool["高股息"] = dividend_pool
+            print(f"📡 高股息: 从 {len(all_stocks_raw)} 只中筛选 {len(dividend_pool)} 只")
+
+    # 对比
+    if show_diff:
+        print_diff(current, new_pool)
+
+    # 写入
+    if not dry_run and new_pool:
+        output = {
+            "_meta": {
+                "updated": datetime.now().strftime("%Y-%m-%dT%H:%M:%S"),
+                "source": "eastmoney_push2",
+                "total_sectors": len(new_pool),
+                "total_stocks": sum(len(v) for v in new_pool.values()),
+            }
+        }
+        output.update(new_pool)
+        with open(POOL_FILE, "w", encoding="utf-8") as f:
+            json.dump(output, f, ensure_ascii=False, indent=2)
+        print(f"\n✅ 已写入 {POOL_FILE} ({output['_meta']['total_stocks']} 只)")
+    elif dry_run:
+        print(f"\n📋 dry-run 模式，未写入")
+
+    return new_pool
+
+
+def print_diff(current: dict, new_pool: dict):
+    """打印新旧池对比"""
+    print("\n" + "=" * 60)
+    print("📊 股票池变更对比")
+    print("=" * 60)
+
+    all_sectors = sorted(set(list(current.keys()) + list(new_pool.keys())))
+
+    for sector in all_sectors:
+        old = set(current.get(sector, []))
+        new = set(new_pool.get(sector, []))
+        added = new - old
+        removed = old - new
+        if not added and not removed:
+            continue
+        print(f"\n【{sector}】")
+        if added:
+            print(f"  + 新增 {len(added)}: {', '.join(sorted(added))}")
+        if removed:
+            print(f"  - 移除 {len(removed)}: {', '.join(sorted(removed))}")
+
+    total_old = sum(len(v) for v in current.values())
+    total_new = sum(len(v) for v in new_pool.values())
+    print(f"\n总计: {total_old} → {total_new} ({total_new - total_old:+d})")
+
+
+# ---------- CLI ----------
+
+def main():
+    parser = argparse.ArgumentParser(description="股票池自动刷新")
+    parser.add_argument("--sector", "-s", nargs="+", help="只刷新指定板块")
+    parser.add_argument("--top", "-n", type=int, default=20, help="每板块取 Top N（默认 20）")
+    parser.add_argument("--sort", choices=["amount", "cap", "pe", "turnover"],
+                        default="amount", help="排序方式（默认 amount 成交额）")
+    parser.add_argument("--dry-run", action="store_true", help="只打印不写入")
+    parser.add_argument("--diff", action="store_true", help="对比当前池显示变更")
+    args = parser.parse_args()
+
+    refresh_pool(
+        sectors=args.sector,
+        top_n=args.top,
+        sort_by=args.sort,
+        dry_run=args.dry_run,
+        show_diff=args.diff,
+    )
+
+
+if __name__ == "__main__":
+    main()