refactor: restructure services and add image/pdf extraction utilities

- Add image_extractor, pdf_downloader, pi_client, trends services
- Add shared utils module
- Refactor summarizer, embedder, routes for cleaner separation
- Update tests to match new service structure

app/routes/search.py

@@ -2,14 +2,11 @@
 
 from __future__ import annotations
 
-import math
-from datetime import date, datetime, timedelta, timezone
-from zoneinfo import ZoneInfo
+from datetime import date, timedelta
 from xml.sax.saxutils import escape
 
 from fastapi import APIRouter, Depends, Query, Request
 from fastapi.responses import Response
-from fastapi.templating import Jinja2Templates
 from sqlalchemy import text
 from sqlalchemy.orm import Session, joinedload
 
@@ -17,9 +14,10 @@ from app.config import settings
 from app.database import get_db
 from app.models import Paper, PaperTag, UserReadingStatus
 from app.services.searcher import get_all_tags, search_papers
+from app.services.user_data import query_reading_list
+from app.utils import templates, today_str
 
 router = APIRouter()
-templates = Jinja2Templates(directory="app/templates")
 
 
 # ── 搜索页 ────────────────────────────────────────────────────────────
@@ -56,7 +54,7 @@ def search_page(
             "total_pages": result["total_pages"],
             "all_tags": all_tags,
             "page_title": f"搜索: {q}" if q else "搜索",
-            "today": _today_str(),
+            "today": today_str(),
         },
     )
 
@@ -114,7 +112,7 @@ def reading_list_page(
     db: Session = Depends(get_db),
 ):
     """阅读列表页面。"""
-    papers = _query_reading_list(db, filter, tag or None)
+    papers = query_reading_list(db, filter, tag or None)
     all_tags = get_all_tags(db)
 
     return templates.TemplateResponse(
@@ -126,54 +124,11 @@ def reading_list_page(
             "current_tag": tag,
             "all_tags": all_tags,
             "page_title": "阅读列表",
-            "today": _today_str(),
+            "today": today_str(),
         },
     )
 
 
-def _query_reading_list(
-    db: Session,
-    filter_type: str,
-    tag: str | None,
-) -> list[Paper]:
-    """根据筛选条件查询阅读列表。"""
-    from sqlalchemy import or_
-
-    # 基础：有任意用户数据的论文
-    base = db.query(Paper).filter(
-        or_(
-            Paper.bookmark.has(),
-            Paper.reading_status.has(),
-            Paper.note.has(),
-        )
-    )
-
-    # 应用筛选
-    if filter_type == "has_note":
-        base = base.filter(Paper.note.has())
-    elif filter_type in ("unread", "skimmed", "read_summary", "read_full"):
-        base = base.filter(
-            Paper.reading_status.has(UserReadingStatus.status == filter_type)
-        )
-
-    # 应用标签
-    if tag:
-        base = base.filter(Paper.tags.any(PaperTag.tag == tag))
-
-    return (
-        base.options(
-            joinedload(Paper.authors),
-            joinedload(Paper.tags),
-            joinedload(Paper.summary_status),
-            joinedload(Paper.bookmark),
-            joinedload(Paper.reading_status),
-            joinedload(Paper.note),
-        )
-        .order_by(Paper.paper_date.desc(), Paper.upvotes.desc())
-        .all()
-    )
-
-
 # ── RSS Feed ──────────────────────────────────────────────────────────
 
 
@@ -216,7 +171,7 @@ def _generate_rss_xml(papers: list[Paper], base_url: str, tag: str | None) -> st
     lines.append(f"    <title>{escape(channel_title)}</title>")
     lines.append(f"    <link>{escape(base_url)}</link>")
     lines.append("    <description>HuggingFace Daily Papers — 中文论文导览站</description>")
-    lines.append(f"    <language>zh-CN</language>")
+    lines.append("    <language>zh-CN</language>")
 
     for paper in papers:
         title_text = paper.title_zh or paper.title_en
@@ -245,12 +200,3 @@ def _generate_rss_xml(papers: list[Paper], base_url: str, tag: str | None) -> st
     lines.append("  </channel>")
     lines.append("</rss>")
     return "\n".join(lines)
-
-
-# ── 工具函数 ──────────────────────────────────────────────────────────
-
-
-def _today_str() -> str:
-    """当前日期字符串（按 APP_TIMEZONE）。"""
-    tz = ZoneInfo(settings.APP_TIMEZONE)
-    return datetime.now(tz).strftime("%Y-%m-%d")