fix(budget): attribution backfill computes in-process and previews on dry-run (yanxue-20260914-budget-p13p14-activation)

backend/scripts/kpi_alert_attribution_backfill.py

@@ -1,156 +1,177 @@
 #!/usr/bin/env python3
 """预算偏差告警归因回填 —— yanxue-20260914-budget-p13p14-activation（P1-③ 激活）
 
-两类动作:
-  1) attribution 为空的历史 open 告警 → 走现成接口 GET /budget/deviation-alerts/{id}/attribution
-     （端点在响应前现场组装归因并落库，见 app/api/budget.py:1004-1012）
-  2) 已有旧口径归因的告警 → 用修正后的科目口径重算覆盖（旧值见 /root/WB_WS/_backups/p13p14-20260914/）
-另: scenario_id 为空且 alert_type 命中 scenario_suggestions → 补 scenario_id（让告警"能解释、能关"）
+两类动作（都在本进程内组装，不依赖线上服务是否已加载新代码）:
+  1) attribution 为空的历史告警 → 现算归因并落库（等价于现成接口
+     GET /budget/deviation-alerts/{id}/attribution 的组装逻辑，见 app/api/budget.py）
+  2) 已有旧口径归因的告警 → 用修正后的科目口径重算覆盖（旧值备份见
+     /root/WB_WS/_backups/p13p14-20260914/）
+另: alert_type 为空 → 按 KPI 特征推断补上；scenario_id 为空且命中
+    scenario_suggestions → 补 scenario_id（让告警"能解释、能关"）
 
 用法:
-  /usr/bin/python3 scripts/kpi_alert_attribution_backfill.py            # dry-run
-  /usr/bin/python3 scripts/kpi_alert_attribution_backfill.py --apply
+  /usr/bin/python3 scripts/kpi_alert_attribution_backfill.py            # dry-run（只算不写）
+  /usr/bin/python3 scripts/kpi_alert_attribution_backfill.py --apply    # 落库
+
+落库前置: cma_api 需对 budget_deviation_alerts 有 UPDATE 权限
+（SHOW GRANTS 现状: 该表未在授权清单内 → 1142 UPDATE command denied）。
 """
 import json
 import os
 import sys
-import urllib.error
-import urllib.request
+from types import SimpleNamespace
 
 if __name__ == "__main__":
     sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 
 from app.database import get_session_local
-from app.models import BudgetDeviationAlert, KPIDefinition, ScenarioSuggestion
+from app.models import BudgetDeviationAlert, KPIDefinition, KPIValueSource, ScenarioSuggestion
 from app.utils.alert_attribution import build_attribution, match_scenario
 
-API = "http://127.0.0.1:8010"
-SECRET = "/root/.hermes/profiles/wecom-fullstack/.secrets/cma-admin.env"
 OUT = "/tmp/budget-p13p14/attribution-conclusions.json"
 
 
-def load_env(path):
-    vals = {}
-    with open(path) as fh:
-        for line in fh:
-            line = line.strip()
-            if line and not line.startswith("#") and "=" in line:
-                k, v = line.split("=", 1)
-                vals[k.strip()] = v.strip()
-    return vals
+def mapping_status(db, kpi_id):
+    src = db.query(KPIValueSource).filter(
+        KPIValueSource.kpi_id == kpi_id,
+        KPIValueSource.source_table == "voucher_details",
+    ).order_by(KPIValueSource.status.asc()).first()
+    return src.status if src else None
 
 
-def api(method, path, body=None, token=None):
-    data = json.dumps(body).encode() if body is not None else None
-    req = urllib.request.Request(API + path, data=data, method=method)
-    req.add_header("Content-Type", "application/json")
-    if token:
-        req.add_header("Authorization", f"Bearer {token}")
-    try:
-        with urllib.request.urlopen(req, timeout=90) as r:
-            return r.status, json.loads(r.read().decode())
-    except urllib.error.HTTPError as e:
-        return e.code, {"_error": e.read().decode()[:200]}
-
-
-def conclusion(alert, kpi, attribution, scenario):
+def conclusion(alert, kpi, attribution, scenario, map_status):
     dims = attribution.get("dimensions") or []
     subs = attribution.get("subjects") or []
     trend = attribution.get("trend") or {}
     vtype = attribution.get("variance_type")
     hot = [s for s in subs if abs(s.get("amount_diff") or 0) > 0.01][:3]
-    sub_txt = "; ".join(
-        f"{s['subject_code']} {s.get('subject_name','')} {s['amount_diff']:+.2f}元(权重{s.get('weight')})"
-        for s in hot) or "（科目明细为空：该KPI无可用取数映射，见报告§未接通清单）"
+    if hot:
+        sub_txt = "; ".join(
+            f"{s['subject_code']} {s.get('subject_name', '')} "
+            f"{s['amount_diff']:+.2f}元(x{s.get('weight')})" for s in hot)
+        depth = f"科目级({map_status or 'kpi_subject_map原口径'})"
+    elif map_status:
+        sub_txt = (f"科目明细为空/金额为0（映射={map_status}，当期凭证明细无该科目发生额）")
+        depth = f"KPI级(映射={map_status})"
+    else:
+        sub_txt = "无取数映射（kpi_subject_map 科目未机械桥接到 voucher_details 科目）→ 仅 KPI 级结论"
+        depth = "KPI级(无映射)"
     dim_txt = "; ".join(
         f"{d.get('kpi_name')} 偏差{d.get('deviation_value')}" for d in dims
         if d.get("deviation_value") is not None and abs(d["deviation_value"]) > 0.01) or "无下级KPI层级"
     trend_txt = (f"{trend.get('type')}({trend.get('message')})" if trend.get("anomaly")
                  else "无连续3期异常")
-    return (
-        f"[{alert.id}] {kpi.kpi_code if kpi else alert.kpi_id} {alert.period}: "
+    view = (
+        f"[{alert.id}] {kpi.kpi_code if kpi else alert.kpi_id} {kpi.kpi_name if kpi else ''} {alert.period}: "
         f"实际{alert.actual_value} vs 预算{alert.budget_value}（偏差{alert.deviation_rate}%, {alert.alert_level}）| "
         f"场景={alert.alert_type or '未分类'}{'(建议:' + scenario['title'] + ')' if scenario else ''} | "
-        f"量价={vtype} | 主科目={sub_txt} | 子KPI={dim_txt} | 趋势={trend_txt}"
+        f"深度={depth} | 量价={vtype} | 主科目={sub_txt} | 子KPI={dim_txt} | 趋势={trend_txt}"
     )
+    return view, depth
 
 
 def main():
     apply = "--apply" in sys.argv
-    sec = load_env(SECRET)
-    st, login = api("POST", "/api/cma/auth/login",
-                    {"username": sec["CMA_ADMIN_USER"], "password": sec["CMA_ADMIN_PASS"], "entity_id": 1})
-    token = login.get("token")
-    print(f"login: HTTP {st} entity={login.get('entity_short_name')}")
-
     db = get_session_local()()
     try:
         alerts = db.query(BudgetDeviationAlert).filter(
             BudgetDeviationAlert.entity_id == 1).order_by(BudgetDeviationAlert.id).all()
-        print(f"== 告警: {len(alerts)} 条（open {sum(1 for a in alerts if a.status == 'open')}）")
+        print(f"== 告警: {len(alerts)} 条（open {sum(1 for a in alerts if a.status == 'open')}）"
+              f" | mode={'APPLY' if apply else 'DRY-RUN'}")
 
-        api_filled, recomputed, scenario_filled = 0, 0, 0
+        filled, recomputed, type_filled, scenario_filled = 0, 0, 0, 0
         lines = []
         for a in alerts:
             kpi = db.query(KPIDefinition).filter(KPIDefinition.id == a.kpi_id).first()
+            mstat = mapping_status(db, a.kpi_id)
+            att, inferred = build_attribution(db, a.kpi_id, a.period, a.alert_type)
             had = a.attribution is not None
 
-            if not had and apply:
-                # 1) 现成接口路径：现场组装并落库
-                st_a, resp = api("GET", f"/api/cma/budget/deviation-alerts/{a.id}/attribution", token=token)
-                if st_a == 200:
-                    api_filled += 1
-                db.expire(a)  # 端点已 commit，重读
-                a = db.query(BudgetDeviationAlert).filter(BudgetDeviationAlert.id == a.id).first()
-            elif had and apply:
-                # 2) 旧口径 → 按修正科目口径重算
-                att, inferred = build_attribution(db, a.kpi_id, a.period, a.alert_type)
-                a.attribution = att
-                if a.alert_type is None:
+            if apply:
+                if not had:
+                    a.attribution = att
+                    filled += 1
+                else:
+                    if (a.attribution or {}) != att:
+                        a.attribution = att
+                        recomputed += 1
+                if a.alert_type is None and inferred:
                     a.alert_type = inferred
-                recomputed += 1
+                    type_filled += 1
 
-            scenario = match_scenario(db, a.alert_type)
-            if scenario and a.scenario_id is None and apply:
+            eff_type = a.alert_type or inferred
+            scenario = match_scenario(db, eff_type)
+            if apply and scenario and a.scenario_id is None:
                 a.scenario_id = scenario["scenario_id"]
                 scenario_filled += 1
 
-            att_now = a.attribution or {}
+            alert_copy = a
+            if not apply:
+                # 只读预览：不改 ORM 对象（避免 autoflush 落库）
+                alert_copy = SimpleNamespace(
+                    id=a.id, kpi_id=a.kpi_id, period=a.period, status=a.status,
+                    alert_level=a.alert_level, budget_value=a.budget_value,
+                    actual_value=a.actual_value, deviation_rate=a.deviation_rate,
+                    deviation_value=a.deviation_value, alert_type=eff_type,
+                    scenario_id=a.scenario_id)
+
+            view, depth = conclusion(alert_copy, kpi, att, scenario, mstat)
             lines.append({
                 "id": a.id, "kpi_id": a.kpi_id, "kpi_code": kpi.kpi_code if kpi else None,
+                "kpi_name": kpi.kpi_name if kpi else None,
                 "period": a.period, "status": a.status, "alert_level": a.alert_level,
                 "budget_value": a.budget_value, "actual_value": a.actual_value,
-                "deviation_rate": a.deviation_rate, "alert_type": a.alert_type,
-                "scenario_id": a.scenario_id, "has_attribution": bool(a.attribution),
-                "attribution_keys": sorted(att_now.keys()),
-                "view": conclusion(a, kpi, att_now, scenario),
+                "deviation_rate": a.deviation_rate, "alert_type": eff_type,
+                "scenario_id": a.scenario_id,
+                "mapping_status": mstat, "depth": depth,
+                "n_subjects": len(att.get("subjects") or []),
+                "n_subject_moves": sum(1 for s in (att.get("subjects") or [])
+                                       if abs(s.get("amount_diff") or 0) > 0.01),
+                "variance_type": att.get("variance_type"),
+                "had_attribution": had,
+                "view": view,
             })
 
         if apply:
-            db.commit()
+            try:
+                db.commit()
+            except Exception as e:
+                db.rollback()
+                print(f"!! 落库失败（未写任何行）: {type(e).__name__}: {str(e)[:200]}")
+                return 2
 
         for row in lines:
             print(row["view"])
 
+        os.makedirs(os.path.dirname(OUT), exist_ok=True)
         with open(OUT, "w") as fh:
             json.dump(lines, fh, ensure_ascii=False, indent=1)
-        print(f"\n== 输出结论: {OUT} ({len(lines)} 条)")
+        print(f"\n== 结论落盘: {OUT} ({len(lines)} 条)")
 
-        n_attr = db.query(BudgetDeviationAlert).filter(
-            BudgetDeviationAlert.entity_id == 1,
-            BudgetDeviationAlert.attribution.isnot(None)).count()
-        total = db.query(BudgetDeviationAlert).filter(BudgetDeviationAlert.entity_id == 1).count()
-        n_open = db.query(BudgetDeviationAlert).filter(
-            BudgetDeviationAlert.entity_id == 1, BudgetDeviationAlert.status == "open").count()
-        n_scen = db.query(BudgetDeviationAlert).filter(
-            BudgetDeviationAlert.entity_id == 1,
-            BudgetDeviationAlert.scenario_id.isnot(None)).count()
-        print(f"== 归因覆盖: {n_attr}/{total}（open {n_open}）| scenario_id 已补: {n_scen} | "
-              f"接口落库 {api_filled} / 重算 {recomputed} / 场景补全 {scenario_filled}"
-              f"{'' if apply else ' [dry-run 未写库]'}")
+        depths = {}
+        for row in lines:
+            depths[row["depth"]] = depths.get(row["depth"], 0) + 1
+        print(f"== 归因深度分布: {json.dumps(depths, ensure_ascii=False)}")
+        opens = [r for r in lines if r["status"] == "open"]
+        print(f"== open 告警有结论: {len(opens)}/{len(opens)}"
+              f" | 其中科目级可解释: {sum(1 for r in opens if r['depth'].startswith('科目级'))}")
+        if apply:
+            n_attr = db.query(BudgetDeviationAlert).filter(
+                BudgetDeviationAlert.entity_id == 1,
+                BudgetDeviationAlert.attribution.isnot(None)).count()
+            total = db.query(BudgetDeviationAlert).filter(BudgetDeviationAlert.entity_id == 1).count()
+            n_scen = db.query(BudgetDeviationAlert).filter(
+                BudgetDeviationAlert.entity_id == 1,
+                BudgetDeviationAlert.scenario_id.isnot(None)).count()
+            print(f"== 落库结果: attribution {n_attr}/{total} | scenario_id {n_scen} | "
+                  f"新填 {filled} / 重算 {recomputed} / 场景类型补全 {type_filled} / "
+                  f"场景建议补全 {scenario_filled}")
+        else:
+            print("== DRY-RUN: 未写库（落库需 --apply，且 cma_api 需 UPDATE 授权）")
+        return 0
     finally:
         db.close()
 
 
 if __name__ == "__main__":
-    main()
+    sys.exit(main())