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
This commit is contained in:
@@ -71,6 +71,20 @@
|
|||||||
|
|
||||||
> 历史文件与新生成日报的 stats 结构存在差异(历史为 HTML 解析快照,新生成为结构化组装),前端建议按 key 防御性读取。
|
> 历史文件与新生成日报的 stats 结构存在差异(历史为 HTML 解析快照,新生成为结构化组装),前端建议按 key 防御性读取。
|
||||||
|
|
||||||
|
### 3.1 口径说明(重要,避免误解)
|
||||||
|
|
||||||
|
`stats` 内各数字口径不同,请勿直接互相比较:
|
||||||
|
|
||||||
|
| 字段 | 口径 |
|
||||||
|
| --- | --- |
|
||||||
|
| `pipeline.raw_total` | **日报日期当天**抓取的文章数(`data/raw/{src}/{date}/index.jsonl` 中 `stage=article 且 success` 的条目)。`raw_by_source` 是各源明细,**其和 = raw_total**;当天未抓取/无文章的源显示 0 |
|
||||||
|
| `pipeline.raw_total_24h` | 最近 24 小时内**抓取**(按 `fetched_at`)的文章数;`raw_by_source_24h` 为各源明细,和 = raw_total_24h。当天 07:00 抓取的数据其值 ≈ raw_total(并非"24h 内发布的新闻",raw 层无发布时间的可靠字段) |
|
||||||
|
| `pipeline.proc / deduped / dups / emb_count / qdrant_count` | 抽取 / 去重后 / 重复 / 向量化 / Qdrant 总量(`qdrant_count` 为全量累计,非当天) |
|
||||||
|
| `news.total` | **过去 30 小时窗口内**经 LLM 抽取的新闻事件数。**≠ raw_total**:raw 是抓取的文章数,news 是抽取后的事件数(会有过滤/合并),两者不可互相验证 |
|
||||||
|
| `news.importances` | `{重要度等级(1-5): 事件数}`,**各等级之和 = news.total** |
|
||||||
|
| `news.sentiments` | `{情绪: 事件数}`(positive/negative/neutral),和 = news.total |
|
||||||
|
| `news.event_types` | `{事件类型: 事件数}`(TOP 10) |
|
||||||
|
|
||||||
---
|
---
|
||||||
|
|
||||||
## 4. 常用查询示例(API 实现参考)
|
## 4. 常用查询示例(API 实现参考)
|
||||||
|
|||||||
+11
-24
@@ -414,27 +414,8 @@ def _load_article_urls_from_index(index_path: Path) -> list[dict]:
|
|||||||
return articles
|
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]:
|
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()
|
now = datetime.now()
|
||||||
cutoff_24h = now - timedelta(hours=24)
|
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_by_source[name] = n
|
||||||
raw_total += 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
|
n_24h = 0
|
||||||
for art in articles:
|
for art in articles:
|
||||||
dt = _extract_date_from_url(art.get("url", ""))
|
fa = art.get("fetched_at")
|
||||||
if dt and dt >= cutoff_24h:
|
if fa:
|
||||||
n_24h += 1
|
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_by_source_24h[name] = n_24h
|
||||||
raw_total_24h += n_24h
|
raw_total_24h += n_24h
|
||||||
|
|
||||||
|
|||||||
Reference in New Issue
Block a user