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

app/services/cleaner.py

@@ -207,11 +207,14 @@ async def delete_papers_by_date_range(
         completed_at=utc_now(),
         papers_found=total,
         papers_new=deleted,
-        details_json=json.dumps({
-            "total_before": total,
-            "deleted": deleted,
-            "failed": len(failed_items),
-        }, ensure_ascii=False),
+        details_json=json.dumps(
+            {
+                "total_before": total,
+                "deleted": deleted,
+                "failed": len(failed_items),
+            },
+            ensure_ascii=False,
+        ),
         error=job_error,
     )
     db.add(log_entry)