fix(nas): bs_5m失败股尾部补扫一轮——03:0x同步链持写锁60s busy_timeout耗尽database is locked(每晚1-5只)+偶发网络错(600247),原形态=该片窗口永久小洞;修=_sweep返failed清单,run尾部(07:40+同步链早已收工)chunk/inc各自失败集各补扫一轮自愈,补扫query计入swept配额、rlr并入exit语义;+2测试(双窗口失败→双retry=6调用自愈/零失败无retry回归);venv310 644绿 [nas]

scripts/data_platform/bs_5m_eod.py

@@ -211,10 +211,11 @@ def _sweep(conn, stocks, start, end, swept, label):
     """扫全 A 一个窗口: per-stock 短事务 + 周期 relogin + 间隔。
 
     swept 为跨窗口累计股数(=query 数, 每股恰 1 次调用), 由调用方持有并回填;
-    返 (stats, limit_reached)。
+    返 (stats, limit_reached, failed); failed=[(code,prefix)...] 供 run 尾部补扫。
     """
     stats = {"ok": 0, "empty": 0, "failed": 0, "db_rows": 0}
     limit_reached = False
+    failed = []
     t0 = time.time()
     for i, (code, prefix) in enumerate(stocks):
         if swept[0] >= DAILY_LIMIT:
@@ -232,6 +233,7 @@ def _sweep(conn, stocks, start, end, swept, label):
                 stats["empty"] += 1
         except Exception as e:
             stats["failed"] += 1
+            failed.append((code, prefix))
             if stats["failed"] <= 5 or stats["failed"] % 100 == 0:
                 log.warning("[%s] %s err: %s", label, code, e)
             if not relogin():
@@ -244,7 +246,7 @@ def _sweep(conn, stocks, start, end, swept, label):
                 log.warning("[%s] 周期 relogin 失败, 继续跑", label)
         if i < len(stocks) - 1:
             time.sleep(BS_INTERVAL)
-    return stats, limit_reached
+    return stats, limit_reached, failed
 
 
 def _run(args):
@@ -276,9 +278,10 @@ def _run(args):
     limit_reached = False
     t0 = time.time()
     try:
+        failed_chunk = []
         if chunk:
-            cs, lr = _sweep(conn, stocks, chunk[0], chunk[1], swept,
-                            "chunk %s~%s" % chunk)
+            cs, lr, failed_chunk = _sweep(conn, stocks, chunk[0], chunk[1], swept,
+                                          "chunk %s~%s" % chunk)
             limit_reached = limit_reached or lr
             log.info("[CHUNK] %s~%s ok=%d empty=%d failed=%d rows=%d",
                      chunk[0], chunk[1], cs["ok"], cs["empty"], cs["failed"],
@@ -289,11 +292,25 @@ def _run(args):
                 # 实锤多推一片跳过半年窗口)
                 save_chunk_state({"next_end": chunk[0]})
         # 每日增量照跑(回灌期也跑, 近端 7 天始终新鲜)
-        istats, lr = _sweep(conn, stocks, inc_start, inc_end, swept, "inc")
+        istats, lr, failed_inc = _sweep(conn, stocks, inc_start, inc_end, swept, "inc")
         limit_reached = limit_reached or lr
         log.info("[INC] ok=%d empty=%d failed=%d rows=%d",
                  istats["ok"], istats["empty"], istats["failed"],
                  istats["db_rows"])
+        # 失败股尾部补扫一轮(2026-08-25): 03:0x 同步链持写锁→60s busy_timeout
+        # 耗尽 database is locked(每晚 1-5 只)+偶发网络错, 原形态=该片窗口永久
+        # 小洞; run 尾部(07:40+)同步链早已收工, 补扫一轮自愈两类失败
+        retries = []
+        if chunk and failed_chunk:
+            retries.append(("chunk-retry", failed_chunk, chunk[0], chunk[1]))
+        if failed_inc:
+            retries.append(("inc-retry", failed_inc, inc_start, inc_end))
+        for rlabel, rcodes, rstart, rend in retries:
+            log.info("[%s] 补扫 %d 只失败股", rlabel, len(rcodes))
+            rs, rlr, _rf = _sweep(conn, rcodes, rstart, rend, swept, rlabel)
+            limit_reached = limit_reached or rlr
+            log.info("[%s] ok=%d empty=%d failed=%d rows=%d",
+                     rlabel, rs["ok"], rs["empty"], rs["failed"], rs["db_rows"])
         if next_chunk(today) is None and not limit_reached:
             save_chunk_state({"next_end": None})  # 回灌完结盖章(幂等)
     finally:

tests/data_platform/test_bs_5m_eod.py

@@ -267,3 +267,34 @@ def test_main_second_instance_skips(main_env, monkeypatch):
     with pytest.raises(SystemExit) as e:
         b5.main()
     assert e.value.code == 0
+
+
+def test_main_failed_stocks_retry_pass(main_env, monkeypatch):
+    """失败股尾部补扫: 首扫抛错的股在 retry 轮重拉, 锁错/网络错自愈不留洞。"""
+    monkeypatch.setattr(sys, "argv", ["bs_5m_eod.py"])
+    seen = set()
+
+    def fake_process(conn, code, prefix, start, end, timeout):
+        key = (code, start)
+        first = key not in seen
+        seen.add(key)
+        if code == "600001" and first:
+            raise RuntimeError("database is locked")
+        return 48
+
+    with patch.object(b5, "_process_one", side_effect=fake_process) as m:
+        with pytest.raises(SystemExit) as e:
+            b5.main()
+    assert e.value.code == 0
+    # A 在 chunk/inc 两窗口各自首调抛错 → chunk-retry + inc-retry 各补 1 次:
+    # chunk(A,B) + inc(A,B) + chunk-retry(A) + inc-retry(A) = 6
+    assert m.call_count == 6
+
+
+def test_main_retry_not_triggered_when_no_failures(main_env):
+    """零失败 → 无补扫轮(回归: 原 4 次调用不变)。"""
+    with patch.object(b5, "_process_one", return_value=48) as m:
+        with pytest.raises(SystemExit) as e:
+            b5.main()
+    assert e.value.code == 0
+    assert m.call_count == 4  # 2 股 × (chunk+inc), 无 retry