fix: 日报数据总览统计修复与口径说明

- _collect_pipeline_stats 24h 统计改用 fetched_at(抓取时间), 原实现从 URL 猜日期对多数源失效(实测 804 篇文章仅 3 条计入)
- 删除废弃的 _extract_date_from_url
- docs/db_schema.md 新增 3.1 口径说明: raw_total=当天抓取文章(sum=raw_by_source), raw_total_24h=24h 内抓取, news.total=30h 窗口 LLM 事件数(≠raw_total), importances/sentiments 之和=news.total

scheduler/reporter.py

@@ -414,27 +414,8 @@ def _load_article_urls_from_index(index_path: Path) -> list[dict]:
     return articles
 
 
-def _extract_date_from_url(url: str) -> datetime | None:
-    """从 URL 中提取发布日期（用于估算 24h 新鲜度）。"""
-    import re as _re2
-    patterns = [
-        _re2.compile(r'/(\d{4})[-/](\d{2})[-/](\d{2})/'),
-        _re2.compile(r'/(\d{4})(\d{2})(\d{2})/'),
-        _re2.compile(r'(\d{4})(\d{2})(\d{2})\.(?:s?html|pdf)'),
-        _re2.compile(r'/t(\d{4})(\d{2})(\d{2})_'),
-    ]
-    for pat in patterns:
-        m = pat.search(url)
-        if m:
-            try:
-                return datetime(int(m.group(1)), int(m.group(2)), int(m.group(3)))
-            except ValueError:
-                pass
-    return None
-
-
 def _collect_pipeline_stats(day_str: str) -> dict[str, Any]:
-    """收集管道统计数据（仅计文章级条目 + 24h 新鲜度）。"""
+    """收集管道统计数据（仅计文章级条目 + 24h 抓取新鲜度, 用 fetched_at）。"""
     now = datetime.now()
     cutoff_24h = now - timedelta(hours=24)
 
@@ -451,12 +432,18 @@ def _collect_pipeline_stats(day_str: str) -> dict[str, Any]:
         raw_by_source[name] = n
         raw_total += n
 
-        # 统计 24h 内文章
+        # 统计 24h 内抓取的文章（用 fetched_at；原实现从 URL 猜日期, 对多数源失效导致全 0）
         n_24h = 0
         for art in articles:
-            dt = _extract_date_from_url(art.get("url", ""))
-            if dt and dt >= cutoff_24h:
-                n_24h += 1
+            fa = art.get("fetched_at")
+            if fa:
+                try:
+                    if datetime.fromisoformat(fa) >= cutoff_24h:
+                        n_24h += 1
+                except ValueError:
+                    n_24h += 1  # 时间格式异常时保守计入
+            else:
+                n_24h += 1  # 时间缺失时保守计入
         raw_by_source_24h[name] = n_24h
         raw_total_24h += n_24h