fix: 修复补跑历史日期的静默空转与日报日期错位 (P0-3/P0-4)

- run_step: 补跑保护——历史日期跳过 crawler(首页只含当天内容)并告警,复用已有 raw 数据
- run_step: report 步骤改用传入的 date_str,不再硬编码 date.today()
- 新增 4 个补跑相关测试,修正 3 个既有测试适配新行为
- 全量 259 passed(3 个 crawler 基线失败与本次无关)

tests/test_scheduler.py

@@ -56,9 +56,12 @@ def test_pipeline_result_partial_failure() -> None:
 # --------------------------------------------------------------------------- #
 
 def test_run_step_success() -> None:
+    # crawler 补跑保护:仅当天日期才执行子进程,故用今天的日期
+    from datetime import date
+    today = date.today().strftime("%Y%m%d")
     with patch("subprocess.run", return_value=_mock_proc(returncode=0,
                                                           stderr="INFO | 完成: 成功 20/20")):
-        sr = run_step("crawler", "20260616")
+        sr = run_step("crawler", today)
     assert sr.success is True
     assert sr.exit_code == 0
 
@@ -109,6 +112,8 @@ def test_run_pipeline_all_success() -> None:
 
 def test_run_pipeline_continues_on_failure() -> None:
     """中间步骤失败,后续继续执行(不阻断)。"""
+    from datetime import date
+    today = date.today().strftime("%Y%m%d")  # crawler 补跑保护:需当天日期才执行
     call_count = {"n": 0}
 
     def _side_effect(*args, **kwargs):
@@ -118,7 +123,7 @@ def test_run_pipeline_continues_on_failure() -> None:
         return _mock_proc(returncode=0, stderr="ok")
 
     with patch("subprocess.run", side_effect=_side_effect):
-        result = run_pipeline("20260616", steps=["crawler", "extractor", "dedup", "llm"])
+        result = run_pipeline(today, steps=["crawler", "extractor", "dedup", "llm"])
     assert len(result.steps) == 4
     # extractor 失败,但后续仍执行
     assert result.steps[1].success is False
@@ -126,7 +131,9 @@ def test_run_pipeline_continues_on_failure() -> None:
 
 
 def test_run_pipeline_custom_steps() -> None:
+    from datetime import date
+    today = date.today().strftime("%Y%m%d")
     with patch("subprocess.run", return_value=_mock_proc(returncode=0, stderr="ok")):
-        result = run_pipeline("20260616", steps=["crawler", "extractor"])
+        result = run_pipeline(today, steps=["crawler", "extractor"])
     assert len(result.steps) == 2
     assert result.all_success is True