feat: dedupe and schedule emoji media downloads

db/message_storage.py

@@ -88,6 +88,25 @@ class MessageStorageDB(BaseDBOperator):
         """
         return self.execute_query(sql, (f'%md5="{md5}"%',), fetch_one=True)
 
+    def get_media_message_by_md5(self, md5: str, current_message_id: int | str | None = None) -> Optional[Dict]:
+        """根据 md5 查找已落盘的图片/表情消息，用于去重复用本地文件"""
+        sql = """
+        SELECT id, group_id, timestamp, sender, content, message_type,
+               attachment_url, message_id, message_xml, message_thumb, image_path
+        FROM messages
+        WHERE attachment_url IS NOT NULL
+          AND attachment_url <> ''
+          AND attachment_url LIKE %s
+          AND image_path IS NOT NULL
+          AND image_path <> ''
+        """
+        params: List = [f'%md5="{md5}"%']
+        if current_message_id is not None:
+            sql += " AND message_id <> %s"
+            params.append(current_message_id)
+        sql += " ORDER BY id DESC LIMIT 1"
+        return self.execute_query(sql, tuple(params), fetch_one=True)
+
     def get_member_recent_messages(self, group_id: str, wxid: str, days: int = 30,
                                    limit: int = 200, include_today: bool = True) -> List[Dict]:
         """获取指定群成员近期消息"""
@@ -513,8 +532,8 @@ class MessageStorageDB(BaseDBOperator):
         
         return self.execute_query(sql, tuple(params)) or []
 
-    def get_pending_image_messages(self, minutes_ago: int = 10, limit: int = 50) -> List[Dict]:
-        """获取最近N分钟内未处理图片的消息（image_path IS NULL）
+    def get_pending_media_messages(self, minutes_ago: int = 10, limit: int = 50) -> List[Dict]:
+        """获取最近N分钟内未处理图片/表情消息（image_path IS NULL）
 
         Args:
             minutes_ago: 查询最近多少分钟的消息，默认10分钟
@@ -524,19 +543,23 @@ class MessageStorageDB(BaseDBOperator):
             包含消息ID、群ID、消息XML等信息的列表
         """
         sql = """
-        SELECT message_id, group_id, message_xml, timestamp,attachment_url
+        SELECT message_id, group_id, sender, message_type, message_xml, timestamp, attachment_url
         FROM messages
-        WHERE message_type = '3'
+        WHERE message_type IN ('3', '47', '1048625', '1090519089')
           AND image_path IS NULL
           AND timestamp >= DATE_SUB(NOW(), INTERVAL %s MINUTE)
-          AND message_xml IS NOT NULL
-          AND message_xml != ''
+          AND attachment_url IS NOT NULL
+          AND attachment_url != ''
         ORDER BY timestamp ASC
         LIMIT %s
         """
         params = (minutes_ago, limit)
         return self.execute_query(sql, params) or []
 
+    def get_pending_image_messages(self, minutes_ago: int = 10, limit: int = 50) -> List[Dict]:
+        """兼容旧方法名，内部复用统一媒体待处理查询"""
+        return self.get_pending_media_messages(minutes_ago, limit)
+
     def get_messages_by_date_range(self, group_id: str, start_date: str, end_date: str = None,
                                    min_content_length: int = 6, max_results: int = 5000) -> List[Dict]:
         """按日期范围获取消息（支持按天总结）