update:合并最新代码

main/xiaozhi-server/core/handle/intentHandler.py

@@ -3,6 +3,8 @@ import json
 import uuid
 from core.handle.sendAudioHandle import send_stt_message
 from core.utils.util import remove_punctuation_and_length
+from core.utils.dialogue import Message
+from loguru import logger
 
 TAG = __name__
 logger = setup_logging()
@@ -12,15 +14,20 @@ async def handle_user_intent(conn, text):
     # 检查是否有明确的退出命令
     if await check_direct_exit(conn, text):
         return True
+    # 4月4日因流式改造暂时关闭唤醒词加速功能
+    # # 检查是否是唤醒词
+    # if await checkWakeupWords(conn, text):
+    #     return True
+
     if conn.use_function_call_mode:
         # 使用支持function calling的聊天方法,不再进行意图分析
         return False
     # 使用LLM进行意图分析
-    intent = await analyze_intent_with_llm(conn, text)
-    if not intent:
+    intent_result = await analyze_intent_with_llm(conn, text)
+    if not intent_result:
         return False
     # 处理各种意图
-    return await process_intent_result(conn, intent, text)
+    return await process_intent_result(conn, intent_result, text)
 
 
 async def check_direct_exit(conn, text):
@@ -30,6 +37,7 @@ async def check_direct_exit(conn, text):
     for cmd in cmd_exit:
         if text == cmd:
             logger.bind(tag=TAG).info(f"识别到明确的退出命令: {text}")
+            await send_stt_message(conn, text)
             await conn.close()
             return True
     return False
@@ -37,7 +45,7 @@ async def check_direct_exit(conn, text):
 
 async def analyze_intent_with_llm(conn, text):
     """使用LLM分析用户意图"""
-    if not hasattr(conn, 'intent') or not conn.intent:
+    if not hasattr(conn, "intent") or not conn.intent:
         logger.bind(tag=TAG).warning("意图识别服务未初始化")
         return None
 
@@ -45,52 +53,64 @@ async def analyze_intent_with_llm(conn, text):
     dialogue = conn.dialogue
     try:
         intent_result = await conn.intent.detect_intent(conn, dialogue.dialogue, text)
-        # 尝试解析JSON结果
-        try:
-            intent_data = json.loads(intent_result)
-            if "intent" in intent_data:
-                return intent_data["intent"]
-        except json.JSONDecodeError:
-            # 如果不是JSON格式，尝试直接获取意图文本
-            return intent_result.strip()
-
+        return intent_result
     except Exception as e:
         logger.bind(tag=TAG).error(f"意图识别失败: {str(e)}")
 
     return None
 
 
-async def process_intent_result(conn, intent, original_text):
+async def process_intent_result(conn, intent_result, original_text):
     """处理意图识别结果"""
-    # 处理退出意图
-    if "结束聊天" in intent:
-        logger.bind(tag=TAG).info(f"识别到退出意图: {intent}")
-        # 如果是明确的离别意图，发送告别语并关闭连接
-        await send_stt_message(conn, original_text)
-        conn.executor.submit(conn.chat_and_close, original_text)
-        return True
+    try:
+        # 尝试将结果解析为JSON
+        intent_data = json.loads(intent_result)
 
-    # 处理播放音乐意图
-    if "播放音乐" in intent:
-        logger.bind(tag=TAG).info(f"识别到音乐播放意图: {intent}")
-        # 调用play_music函数来播放音乐
-        song_name = extract_text_in_brackets(intent)
-        function_id = str(uuid.uuid4().hex)
-        function_name = "play_music"
-        function_arguments = '{ "song_name": "' + song_name + '" }'
+        # 检查是否有function_call
+        if "function_call" in intent_data:
+            # 直接从意图识别获取了function_call
+            logger.bind(tag=TAG).info(
+                f"检测到function_call格式的意图结果: {intent_data['function_call']['name']}"
+            )
+            function_name = intent_data["function_call"]["name"]
+            if function_name == "continue_chat":
+                return False
+            function_args = None
+            if "arguments" in intent_data["function_call"]:
+                function_args = intent_data["function_call"]["arguments"]
+            # 确保参数是字符串格式的JSON
+            if isinstance(function_args, dict):
+                function_args = json.dumps(function_args)
 
-        function_call_data = {
-            "name": function_name,
-            "id": function_id,
-            "arguments": function_arguments
-        }
-        conn.func_handler.handle_llm_function_call(conn, function_call_data)
-        return True
+            function_call_data = {
+                "name": function_name,
+                "id": str(uuid.uuid4().hex),
+                "arguments": function_args,
+            }
 
-    # 其他意图处理可以在这里扩展
+            await send_stt_message(conn, original_text)
 
-    # 默认返回False，表示继续常规聊天流程
-    return False
+            # 使用executor执行函数调用和结果处理
+            def process_function_call():
+                conn.dialogue.put(Message(role="user", content=original_text))
+                result = conn.func_handler.handle_llm_function_call(
+                    conn, function_call_data
+                )
+                if result and function_name != "play_music":
+                    # 获取当前最新的文本索引
+                    text = result.response
+                    if text is None:
+                        text = result.result
+                    if text is not None:
+                        # TODO 将text转换为tts_message_dto
+                        pass
+            # 将函数执行放在线程池中
+            conn.executor.submit(process_function_call)
+            return True
+        return False
+    except json.JSONDecodeError as e:
+        logger.bind(tag=TAG).error(f"处理意图结果时出错: {e}")
+        return False
 
 
 def extract_text_in_brackets(s):
@@ -100,10 +120,14 @@ def extract_text_in_brackets(s):
     :param s: 输入字符串
     :return: 中括号内的文字，如果不存在则返回空字符串
     """
-    left_bracket_index = s.find('[')
-    right_bracket_index = s.find(']')
+    left_bracket_index = s.find("[")
+    right_bracket_index = s.find("]")
 
-    if left_bracket_index != -1 and right_bracket_index != -1 and left_bracket_index < right_bracket_index:
-        return s[left_bracket_index + 1:right_bracket_index]
+    if (
+        left_bracket_index != -1
+        and right_bracket_index != -1
+        and left_bracket_index < right_bracket_index
+    ):
+        return s[left_bracket_index + 1 : right_bracket_index]
     else:
-        return ""
+        return ""