feat: 日报结构化入库（M10 前后端分离数据层）

- 新增 report_db 包: MySQL 连接/建表/幂等写入 (news_report/news_event, myquant 库) - 新增 report_import 包: 历史 178 份日报 HTML 解析入库, 表头驱动列映射 - reporter.py 完全切换: generate_report 结构化入库, 不再生成/上传 HTML - CLI: 新增 report-import 子命令 - 依赖: uv add pymysql; 配置: NEWS_DB_* / REPORT_HISTORY_DIR - 文档: docs/report_db_design.md(实现逻辑), docs/db_schema.md(表结构供 API/前端) - 测试: 24 个单测通过 (parser/builder/models/importer)
2026-08-03 21:32:07 +08:00
parent f2c80c5a9c
commit 366e60e8a9
25 changed files with 1934 additions and 33 deletions
@@ -0,0 +1,87 @@
+"""日报结构化组装单元测试（_build_report_data，纯逻辑）。"""
+
+from __future__ import annotations
+
+import json
+from datetime import date
+
+from scheduler.reporter import _build_report_data
+
+
+def _fake_event(title: str, importance: int, event_type: str = "其他",
+                sentiment: str = "neutral", source_id: str = "cls",
+                url: str = "https://x.com/1") -> dict:
+    return {
+        "title": title,
+        "url": url,
+        "source_id": source_id,
+        "event": {
+            "stock_codes": [],
+            "company_names": [],
+            "industries": [],
+            "sentiment": sentiment,
+            "importance": importance,
+            "event_type": event_type,
+            "summary": f"{title}的摘要",
+        },
+    }
+
+
+class TestBuildReportData:
+    def test_sections_and_ranks(self) -> None:
+        news = {
+            "total": 2, "hi_threshold": 4,
+            "high": [_fake_event("新闻A", 5), _fake_event("新闻B", 4)],
+            "sentiments": {"neutral": 2}, "importances": {5: 1, 4: 1},
+            "event_types": {"其他": 2},
+        }
+        cninfo = {
+            "total": 1, "hi_threshold": 2,
+            "high": [_fake_event("公告C", 3, event_type="公告")],
+            "by_day": {"07月10日": 1}, "announcement": 1, "research": 0, "irm": 0,
+        }
+        pipeline = {"raw_total": 100, "proc": 90}
+        xwlb = {"items": [_fake_event("联播D", 4, event_type="新闻联播", source_id="xwlb")],
+                "date": "07月10日"}
+
+        r = _build_report_data(news, cninfo, pipeline, "AI摘要", "20260710", xwlb=xwlb)
+
+        assert r.report_date == date(2026, 7, 10)
+        assert r.report_type == "finance"
+        assert r.file_name == ""
+        assert r.ai_summary == "AI摘要"
+        assert [(e.section, e.rank) for e in r.events] == [
+            ("news", 1), ("news", 2), ("cninfo", 1), ("xwlb", 1),
+        ]
+        assert r.events[0].source == "cls"
+        assert r.events[3].source == "xwlb"
+
+    def test_stats_snapshot(self) -> None:
+        news = {"total": 1, "hi_threshold": 4, "high": [], "sentiments": {},
+                "importances": {}, "event_types": {}}
+        cninfo = {"total": 0, "hi_threshold": 0, "high": [], "by_day": {},
+                  "announcement": 0, "research": 0, "irm": 0}
+        r = _build_report_data(news, cninfo, {"raw_total": 100}, "s", "20260710")
+        # stats 可 JSON 序列化（入库时 json.dumps）
+        json.dumps(r.stats, ensure_ascii=False)
+        assert r.stats["pipeline"] == {"raw_total": 100}
+        assert r.stats["news"]["total"] == 1
+        assert "xwlb" not in r.stats
+
+    def test_title_truncated(self) -> None:
+        news = {"total": 1, "hi_threshold": 4,
+                "high": [_fake_event("长" * 600, 4)], "sentiments": {},
+                "importances": {}, "event_types": {}}
+        cninfo = {"total": 0, "hi_threshold": 0, "high": [], "by_day": {},
+                  "announcement": 0, "research": 0, "irm": 0}
+        r = _build_report_data(news, cninfo, {}, "s", "20260710")
+        assert len(r.events[0].title) == 512
+
+    def test_empty_events(self) -> None:
+        news = {"total": 0, "hi_threshold": 0, "high": [], "sentiments": {},
+                "importances": {}, "event_types": {}}
+        cninfo = {"total": 0, "hi_threshold": 0, "high": [], "by_day": {},
+                  "announcement": 0, "research": 0, "irm": 0}
+        r = _build_report_data(news, cninfo, {}, None, "20260710")
+        assert r.events == []
+        assert r.ai_summary is None