fix(HEL-459): 影子比较按请求字段投影，盘后整批原子发布

比较侧只对网站本次请求字段计业务差异，忽略数据中枢额外列；盘后 A/B/重发改为先整批暂存与交叉校验，再单事务切换公开版本。 Co-authored-by: Cursor <cursoragent@cursor.com> Co-authored-by: multica-agent <github@multica.ai>
2026-09-05 08:39:31 +08:00
parent bed6450992
commit 16841e9ae3
7 changed files with 720 additions and 78 deletions
@@ -201,6 +201,87 @@ class DatahubBridgeTests(unittest.TestCase):
        skew = compare_rows("daily", [LEGACY_DAILY], [HUB_DAILY], {"stale": False, "staleness_seconds": 12})
        self.assertTrue(skew["time_skew"])

+    def test_shadow_extra_hub_columns_are_not_false_diffs_when_projected(self) -> None:
+        hub_full = {**HUB_DAILY, "adj_factor": 1.1}
+        legacy_close_only = {k: LEGACY_DAILY[k] for k in ("ts_code", "trade_date", "close")}
+        report = compare_rows(
+            "daily", [legacy_close_only], [hub_full],
+            {"stale": False, "staleness_seconds": 0},
+            fields="ts_code,trade_date,close",
+        )
+        self.assertTrue(report["equal"])
+        self.assertEqual(report["value_diff_count"], 0)
+        self.assertEqual(report["fields_compared"], ["close", "trade_date", "ts_code"])
+        # without projection the same pair shows the historic false diff
+        unprojected = compare_rows("daily", [legacy_close_only], [hub_full])
+        self.assertFalse(unprojected["equal"])
+
+        legacy_stocks = {"ts_code": "600000.SH", "name": "浦发银行"}
+        hub_stocks = {
+            "ts_code": "600000.SH", "symbol": "600000", "name": "浦发银行", "area": "上海",
+            "industry": "银行", "market": "主板", "list_status": "L", "list_date": "19991110",
+        }
+        stocks = compare_rows("stocks", [legacy_stocks], [hub_stocks], {}, fields="ts_code,name")
+        self.assertTrue(stocks["equal"])
+
+        legacy_cal = {"cal_date": "20240902", "is_open": 1}
+        hub_cal = {
+            "cal_date": "20240902", "is_open": True,
+            "pretrade_date": "20240830", "prev_open": "20240830",
+        }
+        calendar = compare_rows(
+            "calendar", [legacy_cal], [hub_cal], {}, fields="cal_date,is_open"
+        )
+        self.assertTrue(calendar["equal"])
+
+    def test_shadow_projection_still_alarms_on_requested_field_problems(self) -> None:
+        hub_missing_field = {k: v for k, v in HUB_DAILY.items() if k != "close"}
+        legacy_close_only = {k: LEGACY_DAILY[k] for k in ("ts_code", "trade_date", "close")}
+        lost = compare_rows(
+            "daily", [legacy_close_only], [hub_missing_field], fields="ts_code,trade_date,close"
+        )
+        self.assertFalse(lost["equal"])
+        self.assertEqual(lost["value_diff_count"], 1)
+
+        changed = compare_rows(
+            "daily", [legacy_close_only], [{**HUB_DAILY, "close": 99.0}],
+            fields="ts_code,trade_date,close",
+        )
+        self.assertFalse(changed["equal"])
+        self.assertEqual(changed["value_diff_count"], 1)
+        self.assertEqual(changed["value_diffs"][0]["fields"][0]["field"], "close")
+
+        gone = compare_rows("daily", [LEGACY_DAILY], [], fields="ts_code,trade_date,close")
+        self.assertEqual(gone["missing_hub_count"], 1)
+        self.assertFalse(gone["equal"])
+
+        unit = compare_rows(
+            "daily", [LEGACY_DAILY], [{**HUB_DAILY, "amount": 2000.0, "volume": 1000.0}],
+            fields="ts_code,trade_date,vol,amount",
+        )
+        self.assertGreater(unit["unit_conversion_count"], 0)
+        self.assertFalse(unit["equal"])
+
+    def test_bridge_shadow_report_uses_website_request_fields(self) -> None:
+        hub_full = {**HUB_DAILY, "adj_factor": 1.1}
+        legacy_close_only = {k: LEGACY_DAILY[k] for k in ("ts_code", "trade_date", "close", "vol", "amount")}
+        reports: list[dict[str, Any]] = []
+        client = FakeClient(
+            response=DatahubResponse(
+                data=[hub_full],
+                meta={"tier": "official", "trade_date": "20240902", "stale": False, "staleness_seconds": 0},
+            )
+        )
+        wrapped = DatahubAwareTushareClient(
+            FakeLegacy([legacy_close_only]),
+            DatahubBridge(flags(daily=(False, True)), client, shadow_sink=reports.append),
+        )
+        rows = wrapped.query("daily", {"trade_date": "20240902"}, "ts_code,trade_date,close,vol,amount")
+        self.assertEqual(rows[0]["close"], 10.20)
+        self.assertEqual(rows[0]["vol"], 1000.0)
+        self.assertTrue(reports[0]["equal"])
+        self.assertEqual(reports[0]["matched"], 1)
+
    def test_native_roundtrip_matches_known_scales(self) -> None:
        native = to_native_row("daily", HUB_DAILY)
        self.assertEqual(native["vol"], 1000.0)