'commit'

dsLightRag/WxGzh/T2_GetArticleList.py

@@ -27,6 +27,8 @@ async def get_wechat_sources():
             return [dict(row) for row in rows]
     finally:
         await pool.close()
+
+
 """
 # 查看selenium版本
 pip show selenium
@@ -161,6 +163,8 @@ if __name__ == '__main__':
             publish_time = datetime.datetime.fromtimestamp(int(item.get("update_time"))).strftime('%Y-%m-%d %H:%M:%S')
 
             # 直接获取并显示文章内容
+            if '试卷' in article_title: # 过滤掉试卷
+                continue
             print(f"正在处理文章: {article_title} ({publish_time})")
             content = get_article_content(article_url)
             print(f"文章内容预览: {content[:200]}...")