feat: refactor summarizer and PDF extraction pipeline

- Split summarizer into summary_generator and summary_persister modules
- Refactor pdf_image_extractor to two-phase pipeline with PicoDet layout detection
- Add layout_detector service for PicoDet-S_layout_3cls integration
- Add exceptions module with ConflictError and NotFoundError
- Improve admin dashboard with better statistics and task management
- Add design review document with system optimization suggestions
- Add new tests for crawler, pdf_downloader, pipeline, and summary_utils
- Update dependencies and configuration
- Clean up dead code and improve error handling

tests/test_summarizer.py

@@ -23,12 +23,8 @@ from app.services.pdf_downloader import (
 )
 from app.services.pi_client import PiTimeoutError
 from app.services.schemas import SummarySchema
-from app.services.summarizer import (
-    _save_files,
-    _update_summary_in_db,
-    summarize_batch,
-    summarize_one,
-)
+from app.services.summarizer import summarize_batch, summarize_one
+from app.services.summary_persister import _save_files, _update_summary_in_db
 from app.utils import utc_now
 
 
@@ -39,7 +35,14 @@ from app.utils import utc_now
 def _summarize_tmp_paths(tmp_path):
     """将 data 目录重定向到 tmp_path（供 summarizer 测试使用）。"""
     with (
-        patch("app.services.summarizer.paper_dir", lambda aid: tmp_path / "papers" / aid),
+        patch(
+            "app.services.summary_persister.paper_dir",
+            lambda aid: tmp_path / "papers" / aid,
+        ),
+        patch(
+            "app.services.summary_generator.paper_dir",
+            lambda aid: tmp_path / "papers" / aid,
+        ),
         patch("app.services.pdf_downloader.PAPERS_DIR", tmp_path / "papers"),
         patch("app.services.pdf_downloader.TMP_DIR", tmp_path / "tmp"),
         patch("app.utils.PAPERS_DIR", tmp_path / "papers"),
@@ -134,7 +137,9 @@ class TestFileOperations:
 
     def test_save_files(self, tmp_path, sample_summary_dict):
         schema = SummarySchema.model_validate(sample_summary_dict)
-        with patch("app.services.summarizer.paper_dir", lambda aid: tmp_path / aid):
+        with patch(
+            "app.services.summary_persister.paper_dir", lambda aid: tmp_path / aid
+        ):
             _save_files("2401.12345", schema, "raw output text")
 
             paper_dir = tmp_path / "2401.12345"
@@ -144,7 +149,9 @@ class TestFileOperations:
             assert saved["title_zh"] == "测试论文中文标题"
 
     def test_save_raw_output_only(self, tmp_path):
-        with patch("app.services.summarizer.paper_dir", lambda aid: tmp_path / aid):
+        with patch(
+            "app.services.summary_persister.paper_dir", lambda aid: tmp_path / aid
+        ):
             _save_files("2401.12345", None, "raw output")
             paper_dir = tmp_path / "2401.12345"
             assert (paper_dir / "raw_output.txt").exists()
@@ -180,7 +187,7 @@ class TestSummarizeOneFlow:
         with (
             patch("app.services.summarizer.download_pdf", new_callable=AsyncMock),
             patch(
-                "app.services.summarizer.call_pi",
+                "app.services.summary_generator.call_pi",
                 new_callable=AsyncMock,
                 return_value=(mock_pi_output, "test-session-id"),
             ),
@@ -209,7 +216,9 @@ class TestSummarizeOneFlow:
         assert fts_row[0] == "测试论文中文标题"
 
     @pytest.mark.asyncio
-    async def test_pdf_download_failure(self, db_session, sample_paper, _summarize_tmp_paths):
+    async def test_pdf_download_failure(
+        self, db_session, sample_paper, _summarize_tmp_paths
+    ):
         """PDF 下载失败 → error_type=pdf_download_failed，tmp 被清理。"""
         with (
             patch(
@@ -233,7 +242,7 @@ class TestSummarizeOneFlow:
         with (
             patch("app.services.summarizer.download_pdf", new_callable=AsyncMock),
             patch(
-                "app.services.summarizer.call_pi",
+                "app.services.summary_generator.call_pi",
                 new_callable=AsyncMock,
                 side_effect=PiTimeoutError("timeout after 300s"),
             ),
@@ -250,7 +259,7 @@ class TestSummarizeOneFlow:
         with (
             patch("app.services.summarizer.download_pdf", new_callable=AsyncMock),
             patch(
-                "app.services.summarizer.call_pi",
+                "app.services.summary_generator.call_pi",
                 new_callable=AsyncMock,
                 return_value=("No JSON in this output at all.", "test-session-id"),
             ),
@@ -281,7 +290,7 @@ class TestSummarizeOneFlow:
         with (
             patch("app.services.summarizer.download_pdf", new_callable=AsyncMock),
             patch(
-                "app.services.summarizer.call_pi",
+                "app.services.summary_generator.call_pi",
                 new_callable=AsyncMock,
                 return_value=(bad_output, "test-session-id"),
             ),
@@ -300,7 +309,7 @@ class TestSummarizeOneFlow:
         with (
             patch("app.services.summarizer.download_pdf", new_callable=AsyncMock),
             patch(
-                "app.services.summarizer.call_pi",
+                "app.services.summary_generator.call_pi",
                 new_callable=AsyncMock,
                 return_value=("Some output without JSON", "test-session-id"),
             ),
@@ -319,7 +328,7 @@ class TestSummarizeOneFlow:
         with (
             patch("app.services.summarizer.download_pdf", new_callable=AsyncMock),
             patch(
-                "app.services.summarizer.call_pi",
+                "app.services.summary_generator.call_pi",
                 new_callable=AsyncMock,
                 return_value=(mock_pi_output, "test-session-id"),
             ),
@@ -347,7 +356,9 @@ class TestSummarizeOneFlow:
         assert not tmp_paper.exists()
 
     @pytest.mark.asyncio
-    async def test_skips_done_paper(self, db_session, sample_paper, _summarize_tmp_paths):
+    async def test_skips_done_paper(
+        self, db_session, sample_paper, _summarize_tmp_paths
+    ):
         """已完成的论文跳过。"""
         sample_paper.summary_status.status = "done"
         db_session.commit()
@@ -393,7 +404,7 @@ class TestBatchSummarize:
         with (
             patch("app.services.summarizer.download_pdf", new_callable=AsyncMock),
             patch(
-                "app.services.summarizer.call_pi",
+                "app.services.summary_generator.call_pi",
                 new_callable=AsyncMock,
                 return_value=(mock_pi_output, "test-session-id"),
             ),
@@ -446,7 +457,7 @@ class TestBatchSummarize:
 
         with (
             patch("app.services.summarizer.download_pdf", new_callable=AsyncMock),
-            patch("app.services.summarizer.call_pi", side_effect=_mock_call_pi),
+            patch("app.services.summary_generator.call_pi", side_effect=_mock_call_pi),
         ):
             result = await summarize_batch(db_session, _session_factory=_TestSession)
 
@@ -456,6 +467,8 @@ class TestBatchSummarize:
     @pytest.mark.asyncio
     async def test_task_lock_conflict(self, db_session, _summarize_tmp_paths):
         """TaskLock 防止并发 batch。"""
+        from app.exceptions import ConflictError
+
         # 先插入一个 running 锁
         db_session.add(
             TaskLock(
@@ -467,8 +480,8 @@ class TestBatchSummarize:
         )
         db_session.commit()
 
-        result = await summarize_batch(db_session)
-        assert result["status"] == "conflict"
+        with pytest.raises(ConflictError):
+            await summarize_batch(db_session)
 
     @pytest.mark.asyncio
     async def test_task_lock_released(
@@ -482,7 +495,7 @@ class TestBatchSummarize:
         with (
             patch("app.services.summarizer.download_pdf", new_callable=AsyncMock),
             patch(
-                "app.services.summarizer.call_pi",
+                "app.services.summary_generator.call_pi",
                 new_callable=AsyncMock,
                 return_value=(mock_pi_output, "test-session-id"),
             ),