xiaozhi-sdk 0.2.7__tar.gz → 0.2.8__tar.gz

{xiaozhi_sdk-0.2.7/xiaozhi_sdk.egg-info → xiaozhi_sdk-0.2.8}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: xiaozhi-sdk
-Version: 0.2.7
+Version: 0.2.8
 Summary: 一个用于连接和控制小智智能设备的Python SDK，支持实时音频通信、MCP工具集成和设备管理功能。
 Author-email: dairoot <623815825@qq.com>
 License-Expression: MIT

{xiaozhi_sdk-0.2.7 → xiaozhi_sdk-0.2.8}/pyproject.toml

@@ -66,5 +66,3 @@ omit = [
     "xiaozhi_sdk/cli.py",
     "tests/*",
 ]
-
-

{xiaozhi_sdk-0.2.7 → xiaozhi_sdk-0.2.8}/xiaozhi_sdk/__init__.py

@@ -1,3 +1,3 @@
-__version__ = "0.2.7"
+__version__ = "0.2.8"
 
 from xiaozhi_sdk.core import XiaoZhiWebsocket  # noqa

{xiaozhi_sdk-0.2.7 → xiaozhi_sdk-0.2.8}/xiaozhi_sdk/cli.py

@@ -4,6 +4,7 @@ import logging
 import time
 import uuid
 from collections import deque
+from tkinter import NO
 from typing import Optional
 
 import click
@@ -76,7 +77,8 @@ logger.setLevel(logging.DEBUG)
 
 # 全局状态
 input_audio_buffer: deque[bytes] = deque()
-is_playing_audio = False
+device_stauts = "listen"  # "speak" or "listen"
+
 is_end = False
 human_speak_time = None
 
@@ -109,7 +111,7 @@ def get_image_byte(data):
 
 
 async def handle_message(message):
-    global is_playing_audio
+    global device_stauts
     global human_speak_time
 
     """处理接收到的消息"""
@@ -122,11 +124,11 @@ async def handle_message(message):
         logger.info1("human: %s", message["text"])
 
     elif message["type"] == "tts" and message["state"] == "sentence_start":  # AI语音
-        is_playing_audio = True  # 防止打断
+        device_stauts = "speak"  # 防止打断
         logger.info2("AI: %s", message["text"])
 
     elif message["type"] == "tts" and message["state"] == "stop":
-        is_playing_audio = False
+        device_stauts = "listen"
         # logger.info2("播放结束")
         logger.info("聆听中...")
     elif message["type"] == "llm":  # 表情
@@ -141,27 +143,33 @@ async def handle_message(message):
 
 async def play_assistant_audio(audio_queue: deque[bytes], enable_audio, audio_samplerate):
     """播放音频流"""
-    global is_playing_audio
+    global device_stauts
     global human_speak_time
 
     stream = None
     if enable_audio:
         stream = sd.OutputStream(samplerate=audio_samplerate, channels=INPUT_AUDIO_CHANNELS, dtype=np.int16)
         stream.start()
+
     last_audio_time = None
 
     while True:
         if is_end:
             return
 
+        if device_stauts == "listen":
+            last_audio_time = None
+
         if not audio_queue:
-            if last_audio_time and time.time() - last_audio_time > 2:
-                last_audio_time = time.time()
-                is_playing_audio = False
+            # 空音频 超过 2s ，将device_stauts 设置为listen，代表聆听中
+            if device_stauts == "speak" and last_audio_time and time.time() - last_audio_time > 2:
+                device_stauts = "listen"
 
             await asyncio.sleep(0.01)
             continue
 
+        last_audio_time = time.time()
+
         if human_speak_time:
             logger.debug("首个音频包响应时间：%s 秒", time.time() - human_speak_time)
             human_speak_time = None
@@ -169,7 +177,6 @@ async def play_assistant_audio(audio_queue: deque[bytes], enable_audio, audio_sa
         pcm_data = audio_queue.popleft()
         if stream:
             stream.write(pcm_data)
-        last_audio_time = time.time()
 
 
 class XiaoZhiClient:
@@ -225,7 +232,7 @@ class XiaoZhiClient:
                 continue
 
             pcm_data = input_audio_buffer.popleft()
-            if not is_playing_audio:
+            if device_stauts == "listen":
 
                 await self.xiaozhi.send_audio(pcm_data)
             else:

{xiaozhi_sdk-0.2.7 → xiaozhi_sdk-0.2.8/xiaozhi_sdk.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: xiaozhi-sdk
-Version: 0.2.7
+Version: 0.2.8
 Summary: 一个用于连接和控制小智智能设备的Python SDK，支持实时音频通信、MCP工具集成和设备管理功能。
 Author-email: dairoot <623815825@qq.com>
 License-Expression: MIT