feat: refactor PDF extraction to caption-based screenshots, add upvote refresh, clean up UI

- PDF extractor: rewrite from embedded bitmap extraction to caption-based
  page region screenshots. Finds Figure/Table captions via regex,截取上方/下方
  page region, handles compound figures and vector graphics.
- Upvote refresh: new crawler.refresh_upvotes() re-fetches upvotes for recent
  N days without inserting new papers. Scheduler runs daily 30min after pipeline.
- Admin: add /admin/refresh-upvotes endpoint and dashboard button.
- UI: remove date quick nav, show upvote update time on detail/card pages,
  clean up CSS date-chip styles.
- Utils: add recent_date_strs() helper.

app/routes/admin.py

@@ -26,7 +26,7 @@ from app.models import (
 )
 from app.services.admin import get_admin_stats
 from app.services.cleaner import cleanup_tmp, delete_papers_by_date_range
-from app.services.crawler import crawl_daily
+from app.services.crawler import crawl_daily, refresh_upvotes
 from app.services.pipeline import run_pipeline
 from app.services.scheduler import get_scheduler
 from app.services.summarizer import summarize_batch, summarize_single
@@ -129,16 +129,20 @@ async def admin_scheduler_status(_admin: None = Depends(verify_admin)):
     """调度器运行状态（JSON）。"""
     scheduler = get_scheduler()
     next_run = None
+    upvote_next_run = None
     if scheduler:
         for job in scheduler.get_jobs():
             if job.id == "daily_pipeline":
                 next_run = job.next_run_time
-                break
+            elif job.id == "upvote_refresh":
+                upvote_next_run = job.next_run_time
     return {
         "enabled": scheduler is not None,
         "schedule_time": f"{settings.SCHEDULE_HOUR:02d}:{settings.SCHEDULE_MINUTE:02d}",
         "timezone": settings.APP_TIMEZONE,
         "next_run": next_run.isoformat() if next_run else None,
+        "upvote_next_run": upvote_next_run.isoformat() if upvote_next_run else None,
+        "upvote_refresh_days": settings.UPVOTE_REFRESH_DAYS,
     }
 
 
@@ -159,6 +163,19 @@ async def admin_trigger_pipeline(
     return {"status": "success", "message": "流水线执行完成"}
 
 
+@router.post("/refresh-upvotes")
+async def admin_refresh_upvotes(
+    _admin: None = Depends(verify_admin),
+    db: Session = Depends(get_db),
+    days: int | None = Query(None, description="刷新最近 N 天，默认使用配置值"),
+):
+    """手动刷新最近 N 天论文的 upvotes。"""
+    result = await refresh_upvotes(db, days=days)
+    if result["status"] == "failed":
+        raise HTTPException(status_code=500, detail=result.get("error"))
+    return result
+
+
 # ── 请求模型 ──────────────────────────────────────────────────────────
 
 

app/routes/pages.py

@@ -315,11 +315,16 @@ def _link_figures_with_images(
     fig_type_unmatched = [f for f in unmatched if _is_figure_type(f.get("id", ""))]
     table_type_unmatched = [f for f in unmatched if not _is_figure_type(f.get("id", ""))]
 
-    # 提取的图片也按类型分流，按文件名排序
+    # 提取的图片按类型分流，按文件名中的编号排序
     def _sort_key(name: str) -> tuple[int, int]:
-        m = re.search(r'page(\d+)_(?:img|table)(\d+)', name)
+        # 新格式：figure_1.png, table_1.png
+        m = re.search(r'(?:figure|table)_(\d+)', name)
         if m:
-            return (int(m.group(1)), int(m.group(2)))
+            return (0, int(m.group(1)))
+        # 旧格式：page2_img1.png, page5_table1.png
+        m2 = re.search(r'page(\d+)_(?:img|table)(\d+)', name)
+        if m2:
+            return (int(m2.group(1)), int(m2.group(2)))
         return (0, 0)
 
     fig_images = sorted(