xiaozhi-sdk 0.0.7__tar.gz → 0.0.9__tar.gz

{xiaozhi_sdk-0.0.7/xiaozhi_sdk.egg-info → xiaozhi_sdk-0.0.9}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: xiaozhi-sdk
-Version: 0.0.7
+Version: 0.0.9
 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-0.0.7 → xiaozhi_sdk-0.0.9}/tests/test_xiaozhi.py

@@ -69,9 +69,6 @@ async def message_handler_callback(message):
 
 MAC_ADDR = "00:22:44:66:88:00"
 
-ota_url = "http://localhost:3080/api/ota"
-URL = "ws://120.79.156.134:8380"
-
 ota_url = None
 URL = None
 
@@ -86,21 +83,21 @@ async def test_main():
     await xiaozhi.init_connection(MAC_ADDR)
 
     # # say hellow
-    # for pcm in read_audio_file("./file/audio/say_hello.wav"):
-    #     await xiaozhi.send_audio(pcm)
-    # await xiaozhi.send_silence_audio()
-    # await assistant_audio_play(xiaozhi.output_audio_queue)
-
-    # # say take photo
-    # for pcm in read_audio_file("./file/audio/take_photo.wav"):
-    #     await xiaozhi.send_audio(pcm)
-    # await xiaozhi.send_silence_audio()
-    # await assistant_audio_play(xiaozhi.output_audio_queue, 5)
+    for pcm in read_audio_file("./file/audio/say_hello.wav"):
+        await xiaozhi.send_audio(pcm)
+    await xiaozhi.send_silence_audio()
+    await assistant_audio_play(xiaozhi.output_audio_queue)
 
-    # play music
-    for pcm in read_audio_file("./file/audio/play_music.wav"):
+    # say take photo
+    for pcm in read_audio_file("./file/audio/take_photo.wav"):
         await xiaozhi.send_audio(pcm)
     await xiaozhi.send_silence_audio()
-    await assistant_audio_play(xiaozhi.output_audio_queue, 500)
+    await assistant_audio_play(xiaozhi.output_audio_queue, 5)
+
+    # play music
+    # for pcm in read_audio_file("./file/audio/play_music.wav"):
+    #     await xiaozhi.send_audio(pcm)
+    # await xiaozhi.send_silence_audio()
+    # await assistant_audio_play(xiaozhi.output_audio_queue, 500)
 
     await xiaozhi.close()

{xiaozhi_sdk-0.0.7 → xiaozhi_sdk-0.0.9}/xiaozhi_sdk/__init__.py

@@ -1,3 +1,3 @@
-__version__ = "0.0.7"
+__version__ = "0.0.9"
 
 from xiaozhi_sdk.core import XiaoZhiWebsocket  # noqa

{xiaozhi_sdk-0.0.7 → xiaozhi_sdk-0.0.9}/xiaozhi_sdk/cli.py

@@ -46,12 +46,14 @@ async def handle_message(message):
         is_end = True
 
 
-async def play_assistant_audio(audio_queue: deque[bytes]):
+async def play_assistant_audio(audio_queue: deque[bytes], enable_audio):
     """播放音频流"""
     global is_playing_audio
 
-    stream = sd.OutputStream(samplerate=INPUT_SERVER_AUDIO_SAMPLE_RATE, channels=1, dtype=np.int16)
-    stream.start()
+    stream = None
+    if enable_audio:
+        stream = sd.OutputStream(samplerate=INPUT_SERVER_AUDIO_SAMPLE_RATE, channels=1, dtype=np.int16)
+        stream.start()
     last_audio_time = None
 
     while True:
@@ -66,7 +68,8 @@ async def play_assistant_audio(audio_queue: deque[bytes]):
 
         is_playing_audio = True
         pcm_data = audio_queue.popleft()
-        stream.write(pcm_data)
+        if stream:
+            stream.write(pcm_data)
         last_audio_time = time.time()
 
 
@@ -83,7 +86,7 @@ class XiaoZhiClient:
         self.ota_url = ota_url
         self.mac_address = ""
 
-    async def start(self, mac_address: str, serial_number: str = "", license_key: str = ""):
+    async def start(self, mac_address: str, serial_number: str, license_key: str, enable_audio):
         """启动客户端连接"""
         self.mac_address = mac_address
         self.xiaozhi = XiaoZhiWebsocket(handle_message, url=self.url, ota_url=self.ota_url, send_wake=True)
@@ -92,7 +95,7 @@ class XiaoZhiClient:
             self.mac_address, aec=False, serial_number=serial_number, license_key=license_key
         )
 
-        asyncio.create_task(play_assistant_audio(self.xiaozhi.output_audio_queue))
+        asyncio.create_task(play_assistant_audio(self.xiaozhi.output_audio_queue, enable_audio))
 
     def audio_callback(self, indata, frames, time, status):
         """音频输入回调函数"""
@@ -115,11 +118,11 @@ class XiaoZhiClient:
                 await self.xiaozhi.send_audio(pcm_data)
 
 
-async def run_client(mac_address: str, url: str, ota_url: str, serial_number: str, license_key: str):
+async def run_client(mac_address: str, url: str, ota_url: str, serial_number: str, license_key: str, enable_audio: bool):
     """运行客户端的异步函数"""
     logger.debug("Recording... Press Ctrl+C to stop.")
     client = XiaoZhiClient(url, ota_url)
-    await client.start(mac_address, serial_number, license_key)
+    await client.start(mac_address, serial_number, license_key, enable_audio)
 
     with sd.InputStream(callback=client.audio_callback, channels=1, samplerate=16000, blocksize=960):
         await client.process_audio_input()
@@ -131,9 +134,10 @@ async def run_client(mac_address: str, url: str, ota_url: str, serial_number: st
 @click.option("--ota_url", help="OTA地址")
 @click.option("--serial_number", default="", help="设备的序列号")
 @click.option("--license_key", default="", help="设备的授权密钥")
-def main(mac_address: str, url: str, ota_url: str, serial_number: str, license_key: str):
+@click.option("--enable_audio", default=True, help="是否开启音频播放")
+def main(mac_address: str, url: str, ota_url: str, serial_number: str, license_key: str, enable_audio: bool):
     """小智SDK客户端
 
     MAC_ADDRESS: 设备的MAC地址 (格式: XX:XX:XX:XX:XX:XX)
     """
-    asyncio.run(run_client(mac_address, url, ota_url, serial_number, license_key))
+    asyncio.run(run_client(mac_address, url, ota_url, serial_number, license_key, enable_audio))

{xiaozhi_sdk-0.0.7 → xiaozhi_sdk-0.0.9}/xiaozhi_sdk/mcp.py

@@ -43,7 +43,7 @@ class McpTool(object):
         try:
             response = requests.post(self.explain_url, files=files, data=payload, headers=headers, timeout=5)
             res_json = response.json()
-        except Exception as e:
+        except Exception:
             return "网络异常", True
         if res_json.get("error"):
             return res_json, True
@@ -72,10 +72,11 @@ class McpTool(object):
                 else:
                     tool_res, is_error = {"message": "正在为你播放: {}".format(arguments["music_name"])}, False
                     data = {
-                        "type": "music", "state": "start",
+                        "type": "music",
+                        "state": "start",
                         "url": music_info["url"],
                         "text": arguments["music_name"],
-                        "source": "sdk.mcp_music_tool"
+                        "source": "sdk.mcp_music_tool",
                     }
                     await self.message_handler_callback(data)
 

{xiaozhi_sdk-0.0.7 → xiaozhi_sdk-0.0.9}/xiaozhi_sdk/utils/mcp_data.py

@@ -1,4 +1,4 @@
-from typing import Any, Dict, List
+from typing import Any, Dict
 
 mcp_initialize_payload: Dict[str, Any] = {
     "jsonrpc": "2.0",
@@ -23,7 +23,10 @@ mcp_tool_conf: Dict[str, Dict[str, Any]] = {
         "description": "Play music using music IDs. IMPORTANT: You must call `search_custom_music` first to get the music IDs before using this tool. Use this tool after getting music IDs from search results. Args:\n  `id_list`: The id list of the music to play (obtained from search_custom_music results). The list must contain more than 2 music IDs, and the system will randomly select one to play.\n  `music_name`: The name of the music (obtained from search_custom_music results)",
         "inputSchema": {
             "type": "object",
-            "properties": {"music_name": {"type": "string"}, "id_list": {"type": "array", "items": {"type": "string"}, "minItems": 3}},
+            "properties": {
+                "music_name": {"type": "string"},
+                "id_list": {"type": "array", "items": {"type": "string"}, "minItems": 3},
+            },
             "required": ["music_name", "id_list"],
         },
     },

{xiaozhi_sdk-0.0.7 → xiaozhi_sdk-0.0.9}/xiaozhi_sdk/utils/mcp_tool.py

@@ -3,7 +3,6 @@ import random
 
 import aiohttp
 import numpy as np
-from pydub import AudioSegment
 
 
 async def async_search_custom_music(data) -> tuple[dict, bool]:
@@ -57,6 +56,11 @@ async def _get_random_music_info(id_list: list) -> dict:
 
 
 async def async_mcp_play_music(data) -> tuple[list, bool]:
+    try:
+        from pydub import AudioSegment
+    except ImportError:
+        return [], True
+
     id_list = data["id_list"]
     res_json = await _get_random_music_info(id_list)
 
@@ -79,7 +83,7 @@ async def async_mcp_play_music(data) -> tuple[list, bool]:
 
     chunk_size = 960 * 2
     for i in range(0, len(pcm_data), chunk_size):
-        chunk = pcm_data[i: i + chunk_size]
+        chunk = pcm_data[i : i + chunk_size]
 
         if chunk:  # 确保不添加空块
             chunk = np.frombuffer(chunk, dtype=np.int16)

{xiaozhi_sdk-0.0.7 → xiaozhi_sdk-0.0.9/xiaozhi_sdk.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: xiaozhi-sdk
-Version: 0.0.7
+Version: 0.0.9
 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