Hass/xiaozhi-esp32-server
1
0
Fork 0
mirror of https://github.com/xinnan-tech/xiaozhi-esp32-server.git synced 2026-07-30 03:43:55 +08:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'main' into tts-response

Browse Source 
# Conflicts:
#	main/xiaozhi-server/core/handle/intentHandler.py
This commit is contained in:
hrz
2025-05-23 09:15:15 +08:00
parent fdc6dcb26e 0a9e5d2cea
commit 8119897818
23 changed files with 224 additions and 108 deletions
Download Patch File Download Diff File Expand all files Collapse all files
main/xiaozhi-server/core/handle/functionHandler.py
+9 -2
View File
@@ -1,6 +1,12 @@
from config.logger import setup_logging
import json
from plugins_func.register import FunctionRegistry, ActionResponse, Action, ToolType
from plugins_func.register import (
FunctionRegistry,
ActionResponse,
Action,
ToolType,
DeviceTypeRegistry,
)
from plugins_func.functions.hass_init import append_devices_to_prompt
TAG = __name__
@@ -10,6 +16,7 @@ class FunctionHandler:
def __init__(self, conn):
self.conn = conn
self.config = conn.config
self.device_type_registry = DeviceTypeRegistry()
self.function_registry = FunctionRegistry()
self.register_nessary_functions()
self.register_config_functions()
@@ -54,7 +61,7 @@ class FunctionHandler:
self.function_registry.register_function("plugin_loader")
self.function_registry.register_function("get_time")
self.function_registry.register_function("get_lunar")
self.function_registry.register_function("handle_device")
self.function_registry.register_function("handle_speaker_volume_or_screen_brightness")
def register_config_functions(self):
"""注册配置中的函数,可以不同客户端使用不同的配置"""
main/xiaozhi-server/core/handle/helloHandle.py
+2 -2
View File
@@ -40,8 +40,8 @@ async def checkWakeupWords(conn, text):
if not enable_wakeup_words_response_cache:
return False
"""检查是否是唤醒词"""
_, text = remove_punctuation_and_length(text)
if text in conn.config.get("wakeup_words"):
_, filtered_text = remove_punctuation_and_length(text)
if filtered_text in conn.config.get("wakeup_words"):
await send_stt_message(conn, text)
conn.tts_first_text_index = 0
conn.tts_last_text_index = 0
main/xiaozhi-server/core/handle/intentHandler.py
+4 -3
View File
@@ -12,11 +12,12 @@ TAG = __name__
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):
filtered_text = remove_punctuation_and_length(text)[1]
if await check_direct_exit(conn, filtered_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):
# 检查是否是唤醒词
# if await checkWakeupWords(conn, filtered_text):
# return True
if conn.intent_type == "function_call":
main/xiaozhi-server/core/handle/iotHandle.py
+19 -14
View File
@@ -1,9 +1,8 @@
import json
import asyncio
from config.logger import setup_logging
from plugins_func.register import (
device_type_registry,
register_function,
FunctionItem,
register_device_function,
ActionResponse,
Action,
ToolType,
@@ -177,7 +176,7 @@ class IotDescriptor:
self.methods.append(method)
def register_device_type(descriptor):
def register_device_type(descriptor, device_type_registry):
"""注册设备类型及其功能"""
device_name = descriptor["name"]
type_id = device_type_registry.generate_device_type_id(descriptor)
@@ -213,10 +212,12 @@ def register_device_type(descriptor):
},
}
query_func = create_iot_query_function(device_name, prop_name, prop_info)
decorated_func = register_function(func_name, func_desc, ToolType.IOT_CTL)(
query_func
decorated_func = register_device_function(
func_name, func_desc, ToolType.IOT_CTL
)(query_func)
functions[func_name] = FunctionItem(
func_name, func_desc, decorated_func, ToolType.IOT_CTL
)
functions[func_name] = decorated_func
# 为每个方法创建控制函数
for method_name, method_info in descriptor["methods"].items():
@@ -267,10 +268,12 @@ def register_device_type(descriptor):
},
}
control_func = create_iot_function(device_name, method_name, method_info)
decorated_func = register_function(func_name, func_desc, ToolType.IOT_CTL)(
control_func
decorated_func = register_device_function(
func_name, func_desc, ToolType.IOT_CTL
)(control_func)
functions[func_name] = FunctionItem(
func_name, func_desc, decorated_func, ToolType.IOT_CTL
)
functions[func_name] = decorated_func
device_type_registry.register_device_type(type_id, functions)
return type_id
@@ -289,7 +292,6 @@ async def handleIotDescriptors(conn, descriptors):
functions_changed = False
for descriptor in descriptors:
# 如果descriptor没有properties和methods，则直接跳过
if "properties" not in descriptor and "methods" not in descriptor:
continue
@@ -319,13 +321,16 @@ async def handleIotDescriptors(conn, descriptors):
if conn.load_function_plugin:
# 注册或获取设备类型
type_id = register_device_type(descriptor)
device_type_registry = conn.func_handler.device_type_registry
type_id = register_device_type(descriptor, device_type_registry)
device_functions = device_type_registry.get_device_functions(type_id)
# 在连接级注册设备函数
if hasattr(conn, "func_handler"):
for func_name in device_functions:
conn.func_handler.function_registry.register_function(func_name)
for func_name, func_item in device_functions.items():
conn.func_handler.function_registry.register_function(
func_name, func_item
)
conn.logger.bind(tag=TAG).info(
f"注册IOT函数到function handler: {func_name}"
)
main/xiaozhi-server/core/handle/receiveAudioHandle.py
+5 -5
View File
@@ -39,14 +39,14 @@ async def handleAudioMessage(conn, audio):
if len(conn.asr_audio) < 15:
conn.asr_server_receive = True
else:
text, _ = await conn.asr.speech_to_text(conn.asr_audio, conn.session_id)
conn.logger.bind(tag=TAG).info(f"识别文本: {text}")
text_len, _ = remove_punctuation_and_length(text)
raw_text, _ = await conn.asr.speech_to_text(conn.asr_audio, conn.session_id) # 确保ASR模块返回原始文本
conn.logger.bind(tag=TAG).info(f"识别文本: {raw_text}")
text_len, _ = remove_punctuation_and_length(raw_text)
if text_len > 0:
# 使用自定义模块进行上报
enqueue_asr_report(conn, text, copy.deepcopy(conn.asr_audio))
enqueue_asr_report(conn, raw_text, copy.deepcopy(conn.asr_audio))
await startToChat(conn, text)
await startToChat(conn, raw_text)
else:
conn.asr_server_receive = True
conn.asr_audio.clear()
main/xiaozhi-server/core/handle/textHandle.py
+6 -6
View File
@@ -45,17 +45,17 @@ async def handleTextMessage(conn, message):
conn.client_have_voice = False
conn.asr_audio.clear()
if "text" in msg_json:
text = msg_json["text"]
_, text = remove_punctuation_and_length(text)
original_text = msg_json["text"] # 保留原始文本
filtered_len, filtered_text = remove_punctuation_and_length(original_text)
# 识别是否是唤醒词
is_wakeup_words = text in conn.config.get("wakeup_words")
is_wakeup_words = filtered_text in conn.config.get("wakeup_words")
# 是否开启唤醒词回复
enable_greeting = conn.config.get("enable_greeting", True)
if is_wakeup_words and not enable_greeting:
# 如果是唤醒词，且关闭了唤醒词回复，就不用回答
await send_stt_message(conn, text)
await send_stt_message(conn, original_text)
await send_tts_message(conn, "stop", None)
elif is_wakeup_words:
# 上报纯文字数据（复用ASR上报功能，但不提供音频数据）
@@ -63,9 +63,9 @@ async def handleTextMessage(conn, message):
await startToChat(conn, "嘿，你好呀")
else:
# 上报纯文字数据（复用ASR上报功能，但不提供音频数据）
enqueue_asr_report(conn, text, [])
enqueue_asr_report(conn, original_text, [])
# 否则需要LLM对文字内容进行答复
await startToChat(conn, text)
await startToChat(conn, original_text)
elif msg_json["type"] == "iot":
conn.logger.bind(tag=TAG).info(f"收到iot消息：{message}")
if "descriptors" in msg_json: