增加对voice类型消息的支持

src/chat/message_receive/message.py

@@ -9,6 +9,7 @@ from maim_message import Seg, UserInfo, BaseMessageInfo, MessageBase
 
 from src.common.logger import get_logger
 from src.chat.utils.utils_image import get_image_manager
+from src.chat.utils.utils_voice import get_voice_text
 from .chat_stream import ChatStream
 
 install(extra_lines=3)
@@ -106,6 +107,7 @@ class MessageRecv(Message):
         self.has_emoji = False
         self.is_picid = False
         self.has_picid = False
+        self.is_voice = False
         self.is_mentioned = None
 
         self.is_command = False
@@ -156,6 +158,14 @@ class MessageRecv(Message):
                 if isinstance(segment.data, str):
                     return await get_image_manager().get_emoji_description(segment.data)
                 return "[发了一个表情包，网卡了加载不出来]"
+            elif segment.type == "voice":
+                self.has_picid = False
+                self.is_picid = False
+                self.is_emoji = False
+                self.is_voice == True
+                if isinstance(segment.data, str):
+                    return await get_voice_text(segment.data)
+                return "[发了一段语音，网卡了加载不出来]"
             elif segment.type == "mention_bot":
                 self.is_picid = False
                 self.is_emoji = False
@@ -233,6 +243,14 @@ class MessageRecvS4U(MessageRecv):
                 if isinstance(segment.data, str):
                     return await get_image_manager().get_emoji_description(segment.data)
                 return "[发了一个表情包，网卡了加载不出来]"
+            elif segment.type == "voice":
+                self.has_picid = False
+                self.is_picid = False
+                self.is_emoji = False
+                self.is_voice == True
+                if isinstance(segment.data, str):
+                    return await get_voice_text(segment.data)
+                return "[发了一段语音，网卡了加载不出来]"
             elif segment.type == "mention_bot":
                 self.is_picid = False
                 self.is_emoji = False
@@ -343,6 +361,10 @@ class MessageProcessBase(Message):
                 if isinstance(seg.data, str):
                     return await get_image_manager().get_emoji_description(seg.data)
                 return "[表情，网卡了加载不出来]"
+            elif seg.type == "voice":
+                if isinstance(seg.data, str):
+                    return await get_voice_text(seg.data)
+                return "[发了一段语音，网卡了加载不出来]"
             elif seg.type == "at":
                 return f"[@{seg.data}]"
             elif seg.type == "reply":

src/chat/utils/utils_voice.py

@@ -0,0 +1,46 @@
+import base64
+import os
+import time
+import hashlib
+import uuid
+from typing import Optional, Tuple
+from PIL import Image
+import io
+import numpy as np
+import asyncio
+
+
+from src.common.database.database import db
+from src.common.database.database_model import Images, ImageDescriptions
+from src.config.config import global_config
+from src.llm_models.utils_model import LLMRequest
+
+from src.common.logger import get_logger
+from rich.traceback import install
+import traceback
+install(extra_lines=3)
+
+logger = get_logger("chat_voice")
+
+async def get_voice_text(voice_base64: str) -> str:
+    """获取音频文件描述"""
+    try:
+        # 计算图片哈希
+        # 确保base64字符串只包含ASCII字符
+        if isinstance(voice_base64, str):
+            voice_base64 = voice_base64.encode("ascii", errors="ignore").decode("ascii")
+        voice_bytes = base64.b64decode(voice_base64)
+        _llm = LLMRequest(model=global_config.model.voice, request_type="voice")
+        text = await _llm.generate_response_for_voice(voice_bytes)
+        if text is None:
+            logger.warning("未能生成语音文本")
+            return "[语音(文本生成失败)]"
+        
+        logger.debug(f"描述是{text}")
+
+        return f"[语音：{text}]"
+    except Exception as e:
+        traceback.print_exc()
+        logger.error(f"语音转文字失败: {str(e)}")
+        return "[语音]"
+