feat(dupont): period-basis design (cumulative flow x average stock, optional annualize, identity guard)

Task: finance-20261003-dupont-period-basis-design-fullstack (P1; proposer wecom-finance)
Basis ruling: finance 2026-10-03 (Boss 09:49 CST DM approval of the design task)

- /reports/dupont accepts period/start/end/annualize (default YTD), returns
  period_start/period_end/basis/annualized metadata; identity guard
  |ROE - npm x turnover x multiplier| <= 0.05 pp, else identity_broken (no values).
- new app/utils/dupont_basis.py = single source of truth (window resolve, flow
  cumulative via voucher_details, stock average via subject_balances level=1,
  fail-closed basis_fallback, pure calc + identity check).
- scripts/kpi_dupont_deriver.py (idempotent derived writer, dry-run/check/write),
  scripts/apply_dupont_kpi_defs.py + sql/dupont_kpi_defs_20261003.sql (seed:
  F_ASSET_TOTAL/F_EQUITY_TOTAL/F_EQUITY_MULTIPLIER as derived; F_ASSET_TURNOVER
  manual -> derived). Seed already applied under approval 296cc1fd.
- scripts/verify_dupont_period_basis.py: read-only acceptance battery A1-A9
  (independent SQL recompute). BEFORE (base tree) 1/10 PASS -> AFTER 10/13 PASS;
  remaining 3 need the deriver value write (approval-gated).
- frontend DupontAnalysis.vue: period selector (YTD default) + annualize switch
  + caliber metadata bar + identity_broken alert.

No DDL. No ERP direct connection (old register_dupont_kpis.py caliber dropped).

backend/scripts/apply_dupont_kpi_defs.py

@@ -0,0 +1,92 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""杜邦 KPI 定义种子应用器 —— 执行 scripts/sql/dupont_kpi_defs_20261003.sql
+
+任务：finance-20261003-dupont-period-basis-design-fullstack.md
+动作：① 两账套注册 F_ASSET_TOTAL/F_EQUITY_TOTAL/F_EQUITY_MULTIPLIER（derived）
+      ② F_ASSET_TURNOVER data_source_type: manual → derived
+幂等：SQL 自带条件 INSERT（NOT EXISTS）+ 条件 UPDATE，可安全重跑。
+
+用法（**写库**，须先过 approval-gate 放行）:
+  venv/bin/python3 scripts/apply_dupont_kpi_defs.py --check     # 只读预检（不执行任何语句）
+  venv/bin/python3 scripts/apply_dupont_kpi_defs.py --apply     # 执行（需放行）
+"""
+from __future__ import annotations
+
+import argparse
+import os
+import sys
+
+sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+
+from sqlalchemy import text  # noqa: E402
+
+from app.database import get_engine  # noqa: E402
+
+SQL_FILE = os.path.join(os.path.dirname(os.path.abspath(__file__)),
+                        "sql", "dupont_kpi_defs_20261003.sql")
+TARGETS = ("F_ASSET_TOTAL", "F_EQUITY_TOTAL", "F_EQUITY_MULTIPLIER")
+TARGET_ENTITIES = (1, 2)
+
+
+def statements(sql_text: str):
+    """按 ASCII ';' 切分（本文件的字符串字面量内不含 ';'，已人工核对）"""
+    out = []
+    for chunk in sql_text.split(";"):
+        stmt = chunk.strip()
+        if stmt and not all(ln.strip().startswith("--") or not ln.strip() for ln in stmt.splitlines()):
+            out.append(stmt)
+    return out
+
+
+def snapshot(conn) -> dict:
+    rows = conn.execute(text("""SELECT entity_id, kpi_code, data_source_type, status
+                                FROM kpi_definitions
+                                WHERE entity_id IN (1,2) AND kpi_code IN
+                                ('F_ASSET_TOTAL','F_EQUITY_TOTAL','F_EQUITY_MULTIPLIER',
+                                 'F_ASSET_TURNOVER')
+                                ORDER BY entity_id, kpi_code""")).fetchall()
+    return {f"{r[0]}/{r[1]}": (r[2], r[3]) for r in rows}
+
+
+def main() -> int:
+    ap = argparse.ArgumentParser()
+    ap.add_argument("--check", action="store_true", help="只读预检（不执行写语句）")
+    ap.add_argument("--apply", action="store_true", help="执行种子 SQL（写库）")
+    args = ap.parse_args()
+    if not (args.check or args.apply):
+        print("须指定 --check 或 --apply", file=sys.stderr)
+        return 2
+
+    eng = get_engine()
+    with open(SQL_FILE, encoding="utf-8") as f:
+        stmts = statements(f.read())
+    print(f"SQL 文件: {SQL_FILE}")
+    print(f"语句数: {len(stmts)}")
+    for i, s in enumerate(stmts, 1):
+        head = " ".join(s.split())[:110]
+        print(f"  [{i}] {head}...")
+
+    with eng.connect() as conn:
+        before = snapshot(conn)
+        print(f"\n写前像({len(before)} 行): {before}")
+        if args.check:
+            print("\n[--check] 只读预检：未执行任何语句（写路径未触发）")
+            return 0
+        for i, s in enumerate(stmts, 1):
+            r = conn.execute(text(s))
+            print(f"  [{i}] rowcount={r.rowcount}")
+        conn.commit()
+        after = snapshot(conn)
+        print(f"\n写后像({len(after)} 行): {after}")
+        added = {k: v for k, v in after.items() if k not in before}
+        changed = {k: (before[k], v) for k, v in after.items() if k in before and before[k] != v}
+        print(f"新增: {added}")
+        print(f"变更: {changed}")
+        missing = [f"{e}/{c}" for e in TARGET_ENTITIES for c in TARGETS if f"{e}/{c}" not in after]
+        print(f"缺件: {missing or '无'}")
+    return 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())