过滤非腾讯地址

app/service/recruit_announcement_service.py

@@ -20,6 +20,9 @@ from app.models.recruit_announcement_year import RecruitAnnouncementYear
 from app.service.company_service import find_or_create_company
 from app.tool.page_extract import extract_page
 
+# 微信公众号文章域名，页面提取逻辑只适配了这一种页面结构
+_WECHAT_DOMAIN = "mp.weixin.qq.com"
+
 
 def _parse_datetime(value: str | None) -> datetime | None:
     """将 yyyy-MM-dd HH:mm:ss 字符串解析为 datetime，失败返回 None。"""
@@ -135,7 +138,12 @@ def _save_announcement(announcement_id: int, company_id: int, url: str, data: di
 
 def process_announcement(url: str) -> None:
     """处理单条公告 URL 的完整流程。"""
-    # 1. URL 去重
+    # 1. 只处理微信公众号文章，页面提取逻辑依赖公众号页面结构
+    if _WECHAT_DOMAIN not in url:
+        log.info("非微信公众号文章，跳过: {}", url)
+        return
+
+    # 2. URL 去重
     with MysqlSession() as session:
         row = session.execute(
             text("SELECT id FROM bg_recruit_announcement WHERE announcement_url = :url LIMIT 1"),
@@ -145,23 +153,23 @@ def process_announcement(url: str) -> None:
             log.info("公告已存在，跳过: {}", url)
             return
 
-    # 2. 页面内容提取
+    # 3. 页面内容提取
     result = extract_page(url)
     if not result.content or len(result.content) < 50:
         log.info("公告页内容过短（{}字），跳过: {}", len(result.content) if result.content else 0, url)
         return
 
-    # 3. AI 信息提取
+    # 4. AI 信息提取
     data = extract_announcement(result.content)
     if data is None:
         log.warning("AI 信息提取失败，跳过: {}", url)
         return
 
-    # 4. 公司处理
+    # 5. 公司处理
     company_name = data.get("company_name") or ""
     company_id = find_or_create_company(company_name, result.logo_url)
 
-    # 5. 保存公告
+    # 6. 保存公告
     announcement_id = next_id()
     try:
         _save_announcement(announcement_id, company_id, url, data)