refactor: split monolithic phase tests into per-module test files

- rename test_admin_phase4.py -> test_admin.py, test_search.py -> test_searcher.py
- split test_phase5.py into test_cleaner, test_embedder, test_image_extractor, test_pages
- move schema tests from test_summarizer.py into dedicated test_schemas.py
- add sample_papers_range and sample_papers_with_summary fixtures in conftest
- update .gitignore to exclude all of data/

tests/conftest.py

@@ -209,3 +209,122 @@ def admin_token():
 def admin_headers(admin_token):
     """带 Bearer token 的请求头。"""
     return {"Authorization": f"Bearer {admin_token}"}
+
+
+@pytest.fixture
+def wrong_admin_headers():
+    """错误的 Authorization 请求头。"""
+    return {"Authorization": "Bearer wrong-token"}
+
+
+# ── 多样例数据 ────────────────────────────────────────────────────────────
+
+
+@pytest.fixture
+def sample_papers_range(db_session):
+    """插入 5 篇不同日期的论文（用于 admin / cleaner 测试）。"""
+    now = datetime.now(timezone.utc)
+    papers = []
+    for i, (arxiv_id, paper_date_str) in enumerate([
+        ("2401.10001", "2024-01-10"),
+        ("2401.10002", "2024-01-11"),
+        ("2401.10003", "2024-01-12"),
+        ("2401.10004", "2024-01-13"),
+        ("2401.10005", "2024-01-14"),
+    ]):
+        paper_date = date.fromisoformat(paper_date_str)
+        p = Paper(
+            arxiv_id=arxiv_id,
+            title_en=f"Test Paper {i+1}",
+            abstract=f"Abstract for paper {i+1}.",
+            paper_date=paper_date,
+            crawled_at=now,
+            upvotes=i * 10,
+        )
+        db_session.add(p)
+        db_session.flush()
+        db_session.add(PaperAuthor(paper_id=p.id, name=f"Author {i+1}", position=0))
+        db_session.add(PaperTag(paper_id=p.id, tag=f"Tag{i+1}", source="hf"))
+        db_session.add(SummaryStatus(paper_id=p.id, status="pending"))
+        # FTS5
+        db_session.execute(
+            __import__("sqlalchemy").text(
+                "INSERT INTO papers_fts(rowid, title_en, abstract, authors, tags) "
+                "VALUES (:id, :title, :abstract, :authors, :tags)"
+            ),
+            {"id": p.id, "title": p.title_en, "abstract": p.abstract,
+             "authors": f"Author {i+1}", "tags": f"Tag{i+1}"},
+        )
+        papers.append(p)
+    db_session.commit()
+    return papers
+
+
+@pytest.fixture
+def sample_papers_with_summary(db_session):
+    """插入 5 篇带总结的论文（用于 search / pages / trends 测试）。"""
+    now = datetime.now(timezone.utc)
+    papers = []
+    for i, (arxiv_id, paper_date_str) in enumerate([
+        ("2401.20001", "2024-01-10"),
+        ("2401.20002", "2024-01-11"),
+        ("2401.20003", "2024-01-12"),
+        ("2401.20004", "2024-01-13"),
+        ("2401.20005", "2024-01-14"),
+    ]):
+        paper_date = date.fromisoformat(paper_date_str)
+        p = Paper(
+            arxiv_id=arxiv_id,
+            title_en=f"Test Paper {i+1}",
+            title_zh=f"测试论文 {i+1}",
+            abstract=f"Abstract for paper {i+1}.",
+            paper_date=paper_date,
+            crawled_at=now,
+            upvotes=i * 10 + 5,
+        )
+        db_session.add(p)
+        db_session.flush()
+
+        db_session.add(PaperAuthor(paper_id=p.id, name=f"Author {i+1}", position=0))
+        db_session.add(PaperTag(paper_id=p.id, tag="NLP", source="hf"))
+        db_session.add(PaperTag(paper_id=p.id, tag=f"Tag{i+1}", source="hf"))
+
+        db_session.add(SummaryStatus(
+            paper_id=p.id,
+            status="done" if i < 4 else "pending",
+            quality="normal",
+        ))
+
+        # 添加总结（前 4 篇）
+        if i < 4:
+            summary = PaperSummary(
+                paper_id=p.id,
+                one_line=f"这是论文{i+1}的一句话摘要",
+                difficulty="中级",
+                motivation_problem=f"论文{i+1}的研究问题",
+                motivation_goal=f"论文{i+1}的研究目标",
+                method_key_idea=f"论文{i+1}的关键思路",
+                method_overview=f"论文{i+1}的方法概述",
+                updated_at=now,
+                full_json=json.dumps({"title_zh": f"测试论文 {i+1}"}),
+            )
+            db_session.add(summary)
+
+        # FTS5
+        db_session.execute(
+            __import__("sqlalchemy").text(
+                "INSERT INTO papers_fts(rowid, title_en, title_zh, abstract, authors, tags) "
+                "VALUES (:id, :title_en, :title_zh, :abstract, :authors, :tags)"
+            ),
+            {
+                "id": p.id,
+                "title_en": p.title_en,
+                "title_zh": p.title_zh or "",
+                "abstract": p.abstract or "",
+                "authors": f"Author {i+1}",
+                "tags": f"NLP, Tag{i+1}",
+            },
+        )
+        papers.append(p)
+    db_session.commit()
+    return papers