refactor: 优化相关打断处理

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

@@ -219,8 +219,8 @@ async def process_intent_result(
 
 
 def speak_txt(conn: "ConnectionHandler", text):
-    # 记录文本
-    conn.tts_MessageText = text
+    # 记录文本到 sentence_id 映射
+    conn.tts.store_tts_text(conn.sentence_id, text)
 
     conn.tts.tts_text_queue.put(
         TTSMessageDTO(

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

@@ -26,10 +26,6 @@ async def handleAudioMessage(conn: "ConnectionHandler", audio):
         if not hasattr(conn, "vad_resume_task") or conn.vad_resume_task.done():
             conn.vad_resume_task = asyncio.create_task(resume_vad_detection(conn))
         return
-    # manual 模式下不打断正在播放的内容
-    if have_voice:
-        if conn.client_is_speaking and conn.client_listen_mode != "manual":
-            await handleAbortMessage(conn)
     # 设备长时间空闲检测，用于say goodbye
     await no_voice_close_connect(conn, have_voice)
     # 接收音频
@@ -94,6 +90,10 @@ async def startToChat(conn: "ConnectionHandler", text):
 
     # 意图未被处理，继续常规聊天流程，使用实际文本内容
     await send_stt_message(conn, actual_text)
+
+    # 准备开始新会话
+    conn.client_abort = False
+
     conn.executor.submit(conn.chat, actual_text)
 
 

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

@@ -45,7 +45,6 @@ async def sendAudioMessage(conn: "ConnectionHandler", sentenceType, audios, text
     # 发送结束消息（如果是最后一个文本）
     if sentenceType == SentenceType.LAST:
         await send_tts_message(conn, "stop", None)
-        conn.client_is_speaking = False
         if conn.close_after_chat:
             await conn.close()
 
@@ -270,6 +269,8 @@ async def send_tts_message(conn: "ConnectionHandler", state, text=None):
 
     # TTS播放结束
     if state == "stop":
+        # 保存当前的 sentence_id，用于后续判断是否是当前轮次
+        current_sentence_id = conn.sentence_id
         # 播放提示音
         tts_notify = conn.config.get("enable_stop_tts_notify", False)
         if tts_notify:
@@ -283,8 +284,8 @@ async def send_tts_message(conn: "ConnectionHandler", state, text=None):
         # 停止音频发送循环（仅在流控器已初始化时调用）
         if hasattr(conn, "audio_rate_controller") and conn.audio_rate_controller:
             conn.audio_rate_controller.stop_sending()
-        # 清除服务端讲话状态
-        conn.clearSpeakStatus()
+        # 清除服务端讲话状态，传入 sentence_id 用于判断是否是当前轮次
+        conn.clearSpeakStatus(current_sentence_id)
 
     # 发送消息到客户端
     await conn.websocket.send(json.dumps(message))