持久化表情中文语义资产\n\n- 新增 t_emoji_assets 表及迁移脚本，持久化保存表情发送参数、中文语义与预览图路径\n- 在消息归档与媒体补偿流程中自动回填表情资产，实现收到表情即落语义、补图后回填预览\n- 后台表情库与自动回复优先读取持久化表情资产，仅在空表场景下小范围回补历史数据

utils/wechat/message_to_db.py

@@ -15,12 +15,14 @@ from typing import Dict, Optional
 
 from db.connection import DBConnectionManager
 from db.contacts_db import ContactsDBOperator
+from db.emoji_asset_db import EmojiAssetDB
 from db.levels_db import LevelsDBOperator
 from db.message_storage import MessageStorageDB
 # 导入积分系统
 from db.points_db import PointsDBOperator, PointSource
 from utils.robot_cmd.robot_command import Feature, GroupBotManager, PermissionStatus
 from utils.wechat.contact_manager import ContactManager
+from utils.wechat.emoji_semantic_parser import safe_text
 from wechat_ipad import WechatAPIClient
 from wechat_ipad.models.message import WxMessage, MessageType
 
@@ -36,6 +38,7 @@ class MessageStorage:
         self.db_manager = DBConnectionManager.get_instance()
         self.message_db = MessageStorageDB(self.db_manager)
         self.contacts_db = ContactsDBOperator(self.db_manager)
+        self.emoji_asset_db = EmojiAssetDB(self.db_manager)
 
         self.points_db = PointsDBOperator(self.db_manager)
         # 初始化本地缓存字典，使用 group_id 作为键
@@ -180,6 +183,8 @@ class MessageStorage:
 
             # 使用 MessageStorageDB 类存档消息
             result = self.message_db.archive_message(msg)
+            if result and msg.msg_type in {MessageType.EMOTICON, MessageType.EMOJI}:
+                self._sync_emoji_asset_from_wx_message(msg)
             return {
                 'success': result,
                 'roomid': msg.roomid,
@@ -198,6 +203,25 @@ class MessageStorage:
                 'error': str(e)
             }
 
+    def _sync_emoji_asset_from_wx_message(self, msg: WxMessage):
+        """将表情消息同步到持久化表情资产表。
+
+        说明：
+        1. 群里一旦收到表情，先把 md5/len/语义落表，不必等预览图下载完成；
+        2. 这样自动回复和后台检索都能尽早看到结构化资产；
+        3. 后续定时媒体任务补到 image_path 后，会再次 upsert 同一 md5，把 preview_url 回填完整。
+        """
+        attachment_url = safe_text(getattr(msg.content, "xml_content", "") or getattr(msg.content, "clean_content", "")).strip()
+        if not attachment_url.startswith("<"):
+            return
+        self.emoji_asset_db.upsert_asset_from_message_record({
+            "attachment_url": attachment_url,
+            "image_path": "",
+            "message_id": safe_text(msg.msg_id),
+            "group_id": safe_text(msg.roomid),
+            "sender": safe_text(msg.sender),
+        })
+
     def process_image(self, msg: WxMessage):
         """图片消息已通过 archive_message 存入数据库，不再实时处理
         改为定时任务批量处理，减少对主流程的影响和数据库锁竞争
@@ -243,6 +267,14 @@ class MessageStorage:
                 if existing and existing.get("image_path"):
                     linked_path = existing.get("image_path")
                     success = self.message_db.update_message_image_file_path(message_id, linked_path)
+                    if success and message_type in {str(MessageType.EMOTICON.value), str(MessageType.EMOJI.value)}:
+                        self.emoji_asset_db.upsert_asset_from_message_record({
+                            "attachment_url": xml_content,
+                            "image_path": linked_path,
+                            "message_id": safe_text(message_id),
+                            "group_id": safe_text(group_id),
+                            "sender": safe_text(sender),
+                        })
                     return {
                         "success": bool(success),
                         "message_id": message_id,
@@ -283,6 +315,14 @@ class MessageStorage:
 
             success = self.message_db.update_message_image_file_path(message_id, web_path)
             if success:
+                if message_type in {str(MessageType.EMOTICON.value), str(MessageType.EMOJI.value)}:
+                    self.emoji_asset_db.upsert_asset_from_message_record({
+                        "attachment_url": xml_content,
+                        "image_path": web_path,
+                        "message_id": safe_text(message_id),
+                        "group_id": safe_text(group_id),
+                        "sender": safe_text(sender),
+                    })
                 logger.debug(f"媒体处理成功: message_id={message_id}, path={web_path}")
             else:
                 logger.warning(f"媒体路径更新失败: message_id={message_id}")