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)

tests/test_report_builder.py

@@ -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

tests/test_report_db.py

@@ -0,0 +1,51 @@
+"""日报数据模型单元测试。"""
+
+from __future__ import annotations
+
+from datetime import date, datetime
+
+import pytest
+from pydantic import ValidationError
+
+from report_db.models import EventRow, ReportData
+
+
+class TestEventRow:
+    def test_minimal(self) -> None:
+        ev = EventRow(section="news", rank=1, title="标题")
+        assert ev.importance is None
+        assert ev.sentiment is None
+
+    def test_full(self) -> None:
+        ev = EventRow(
+            section="intl", rank=2, importance=4, event_type="地缘政治",
+            title="t", summary="s", sentiment="negative", source="ForexLive",
+            url="https://x.com/1",
+        )
+        assert ev.sentiment == "negative"
+
+    def test_missing_title_raises(self) -> None:
+        with pytest.raises(ValidationError):
+            EventRow(section="news", rank=1)  # type: ignore[call-arg]
+
+
+class TestReportData:
+    def test_defaults(self) -> None:
+        r = ReportData(
+            report_date=date(2026, 7, 11),
+            report_type="finance",
+            generated_at=datetime(2026, 7, 11, 7, 0),
+        )
+        assert r.file_name == ""
+        assert r.stats == {}
+        assert r.events == []
+
+    def test_with_events(self) -> None:
+        r = ReportData(
+            report_date=date(2026, 7, 11),
+            report_type="finance",
+            generated_at=datetime(2026, 7, 11, 7, 0),
+            ai_summary="摘要",
+            events=[EventRow(section="xwlb", rank=1, title="t")],
+        )
+        assert len(r.events) == 1

tests/test_report_import.py

@@ -0,0 +1,32 @@
+"""历史导入器单元测试（文件匹配/扫描逻辑，不依赖真实 DB）。"""
+
+from __future__ import annotations
+
+from pathlib import Path
+
+from report_import.importer import _match_file
+
+
+class TestMatchFile:
+    def test_finance(self) -> None:
+        p = Path("20260711/finance_news_daily_20260710_0720.html")
+        assert _match_file(p, None, None)
+        assert _match_file(p, "20260710", None)
+        assert _match_file(p, None, "finance")
+        assert not _match_file(p, "20260711", None)  # 文件名日期不含 20260711
+        assert not _match_file(p, None, "intl")
+
+    def test_no_timestamp_suffix(self) -> None:
+        # 早期文件无时间戳后缀，也应匹配
+        p = Path("20260616/finance_news_daily_20260616.html")
+        assert _match_file(p, None, None)
+        assert _match_file(p, "20260616", "finance")
+
+    def test_intl(self) -> None:
+        p = Path("20260711/intl_news_daily_20260711_070304.html")
+        assert _match_file(p, "20260711", "intl")
+        assert not _match_file(p, None, "finance")
+
+    def test_non_report_ignored(self) -> None:
+        assert not _match_file(Path("20260711/002714.SZ_0724.html"), None, None)
+        assert not _match_file(Path("20260711/readme.md"), None, None)

tests/test_report_parser.py

@@ -0,0 +1,98 @@
+"""历史日报解析器单元测试（基于真实样例 HTML）。"""
+
+from __future__ import annotations
+
+from datetime import date, datetime
+from pathlib import Path
+
+import pytest
+
+from report_import.parser import (
+    ReportParseError,
+    parse_finance_report,
+    parse_intl_report,
+    parse_report,
+)
+
+FIXTURES = Path(__file__).parent / "fixtures"
+FINANCE_HTML = (FIXTURES / "finance_news_daily_20260710_0720.html").read_text(encoding="utf-8")
+INTL_HTML = (FIXTURES / "intl_news_daily_20260711_070304.html").read_text(encoding="utf-8")
+
+
+class TestFinanceParse:
+    def test_metadata(self) -> None:
+        r = parse_finance_report(FINANCE_HTML, "finance_news_daily_20260710_0720.html")
+        assert r.report_date == date(2026, 7, 10)
+        assert r.report_type == "finance"
+        assert r.file_name == "finance_news_daily_20260710_0720.html"
+        assert r.generated_at == datetime(2026, 7, 11, 7, 20, 27)  # header"生成于"优先
+
+    def test_ai_summary_lines(self) -> None:
+        r = parse_finance_report(FINANCE_HTML, "finance_news_daily_20260710_0720.html")
+        assert r.ai_summary is not None
+        assert len(r.ai_summary.splitlines()) >= 5
+        assert "碳达峰" in r.ai_summary
+
+    def test_sections_and_ranks(self) -> None:
+        r = parse_finance_report(FINANCE_HTML, "finance_news_daily_20260710_0720.html")
+        sections = {e.section for e in r.events}
+        assert sections == {"xwlb", "news", "cninfo"}
+        assert sum(1 for e in r.events if e.section == "xwlb") == 16
+        assert sum(1 for e in r.events if e.section == "news") == 20
+        assert sum(1 for e in r.events if e.section == "cninfo") == 20
+
+    def test_event_fields(self) -> None:
+        r = parse_finance_report(FINANCE_HTML, "finance_news_daily_20260710_0720.html")
+        xwlb = next(e for e in r.events if e.section == "xwlb")
+        assert xwlb.importance == 4
+        assert xwlb.event_type == "新闻联播"
+        assert xwlb.sentiment == "neutral"
+        assert "张国清" in xwlb.title
+        news = next(e for e in r.events if e.section == "news" and e.source)
+        assert news.source  # 新闻板块带来源
+
+    def test_stats_keys(self) -> None:
+        r = parse_finance_report(FINANCE_HTML, "finance_news_daily_20260710_0720.html")
+        for key in ("pipeline", "sources", "sentiment", "importance", "event_types"):
+            assert key in r.stats, f"缺少 stats.{key}"
+        assert r.stats["pipeline"]["M1 原始文章"] == 758
+
+
+class TestIntlParse:
+    def test_metadata(self) -> None:
+        r = parse_intl_report(INTL_HTML, "intl_news_daily_20260711_070304.html")
+        assert r.report_date == date(2026, 7, 11)
+        assert r.report_type == "intl"
+        assert r.generated_at == datetime(2026, 7, 11, 7, 3, 30)
+
+    def test_events_and_source_from_small(self) -> None:
+        r = parse_intl_report(INTL_HTML, "intl_news_daily_20260711_070304.html")
+        assert len(r.events) == 19
+        first = r.events[0]
+        assert first.section == "intl"
+        assert first.source == "investinglive.com"  # 从摘要 <small>[来源]</small> 提取
+        assert first.url.startswith("https://investinglive.com/")
+        assert first.importance == 4
+        assert first.event_type == "地缘政治"
+        # 摘要中不应残留 [来源] 标记
+        assert first.summary is not None and "[investinglive.com]" not in first.summary
+
+    def test_stats_keys(self) -> None:
+        r = parse_intl_report(INTL_HTML, "intl_news_daily_20260711_070304.html")
+        for key in ("pipeline", "sentiment", "importance", "event_types", "source_dist"):
+            assert key in r.stats
+        assert r.stats["source_dist"][0]["来源"] == "ForexLive"
+
+
+class TestParseReportDispatch:
+    def test_dispatch_finance(self) -> None:
+        r = parse_report(FINANCE_HTML, "finance_news_daily_20260710_0720.html")
+        assert r.report_type == "finance"
+
+    def test_dispatch_intl(self) -> None:
+        r = parse_report(INTL_HTML, "intl_news_daily_20260711_070304.html")
+        assert r.report_type == "intl"
+
+    def test_bad_filename_raises(self) -> None:
+        with pytest.raises(ReportParseError):
+            parse_report("<html></html>", "not_a_report.html")