feat: 提交热榜评论分析工具 MVP 基线

2026-07-03 11:12:12 +08:00
parent 0231be6338
commit 9bad7b6e63
60 changed files with 4058 additions and 0 deletions
@@ -0,0 +1,76 @@
+from datetime import UTC, datetime
+
+from sqlalchemy import create_engine
+from sqlalchemy.orm import Session
+
+from app.db import Base
+from app.models import Comment, ContentItem, Hotspot, Task
+from app.services.report_service import build_comment_metrics, generate_item_report
+
+
+def test_build_comment_metrics_counts_sentiments_and_top_labels():
+    comments = [
+        Comment(content="好", sentiment="positive", labels='["质量好", "价格好"]', like_count=5),
+        Comment(content="差", sentiment="negative", labels='["质量好"]', like_count=2),
+        Comment(content="一般", sentiment="neutral", labels="[]", like_count=1),
+        Comment(content="未知", sentiment="unknown", labels="[]", like_count=0),
+    ]
+
+    metrics = build_comment_metrics(comments)
+
+    assert metrics["sample_count"] == 4
+    assert metrics["sentiment"]["positive"]["count"] == 1
+    assert metrics["sentiment"]["negative"]["count"] == 1
+    assert metrics["sentiment"]["neutral"]["count"] == 1
+    assert metrics["sentiment"]["unknown"]["count"] == 1
+    assert metrics["top_labels"][0] == {"name": "质量好", "count": 2}
+
+
+def test_generate_item_report_persists_markdown_and_default_summary():
+    engine = create_engine("sqlite:///:memory:", connect_args={"check_same_thread": False})
+    Base.metadata.create_all(engine)
+    try:
+        with Session(engine) as session:
+            task = Task(platform="douyin", status="success")
+            session.add(task)
+            session.flush()
+            hotspot = Hotspot(task_id=task.id, platform="douyin", title="热点", raw_data="{}")
+            session.add(hotspot)
+            session.flush()
+            item = ContentItem(
+                task_id=task.id,
+                hotspot_id=hotspot.id,
+                platform="douyin",
+                source_item_id="v1",
+                item_type="video",
+                title="视频",
+                status="success",
+                raw_data="{}",
+            )
+            session.add(item)
+            session.flush()
+            session.add(
+                Comment(
+                    task_id=task.id,
+                    hotspot_id=hotspot.id,
+                    content_item_id=item.id,
+                    platform="douyin",
+                    source_comment_id="c1",
+                    content="好评",
+                    sentiment="positive",
+                    labels='["认可"]',
+                    like_count=10,
+                    comment_time=datetime(2026, 1, 1, tzinfo=UTC),
+                    raw_data="{}",
+                )
+            )
+            session.commit()
+
+            report = generate_item_report(session, item.id, summary_provider=lambda *_args, **_kwargs: (_ for _ in ()).throw(RuntimeError("ai failed")))
+
+            assert report.report_type == "item"
+            assert report.content_item_id == item.id
+            assert "总结生成失败，请查看上方统计数据。" in report.markdown_content
+            assert "样本评论数量" in report.markdown_content
+    finally:
+        engine.dispose()