fix: 提示尾音播放和单模块第一次唤醒时等待tts初始化

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

@@ -56,7 +56,16 @@ async def checkWakeupWords(conn, text):
         "enable_wakeup_words_response_cache"
     ]
 
-    if not enable_wakeup_words_response_cache or not conn.tts:
+    # 等待tts初始化，最多等待3秒
+    start_time = time.time()
+    while time.time() - start_time < 3:
+        if conn.tts:
+            break
+        await asyncio.sleep(0.1)
+    else:
+        return False
+
+    if not enable_wakeup_words_response_cache:
         return False
 
     _, filtered_text = remove_punctuation_and_length(text)

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

@@ -2,6 +2,7 @@ import json
 import time
 import asyncio
 from core.utils import textUtils
+from core.utils.util import audio_to_data_stream
 from core.providers.tts.dto.dto import SentenceType
 
 TAG = __name__
@@ -30,7 +31,7 @@ async def sendAudioMessage(conn, sentenceType, audios, text):
 
 
 # 播放音频
-async def sendAudio(conn, audios, pre_buffer=False, frame_duration=60):
+async def sendAudio(conn, audios, frame_duration=60):
     """
     发送单个opus包，支持流控
     Args:
@@ -46,11 +47,6 @@ async def sendAudio(conn, audios, pre_buffer=False, frame_duration=60):
         if conn.client_abort:
             return
 
-        # 短音频直接发送（例如：提示音）
-        if pre_buffer:
-            await conn.websocket.send(audios)
-            return
-
         conn.last_activity_time = time.time() * 1000
 
         # 获取或初始化流控状态
@@ -78,7 +74,7 @@ async def sendAudio(conn, audios, pre_buffer=False, frame_duration=60):
         flow_control["packet_count"] += 1
         flow_control["last_send_time"] = time.perf_counter()
     else:
-        # 流控参数优化
+        # 提示音和唤醒音频走普通播放
         start_time = time.perf_counter()
         play_position = 0
 
@@ -124,12 +120,13 @@ async def send_tts_message(conn, state, text=None):
             stop_tts_notify_voice = conn.config.get(
                 "stop_tts_notify_voice", "config/assets/tts_notify.mp3"
             )
-            conn.tts.audio_to_opus_data_stream(
-                stop_tts_notify_voice,
-                callback=lambda audio_data: asyncio.run_coroutine_threadsafe(
-                    sendAudio(conn, audio_data, True), conn.loop
-                ),
-            )
+            # 获取音频数据
+            audios = []
+            def handle_audio_frame(frame_data):
+                audios.append(frame_data)
+
+            audio_to_data_stream(stop_tts_notify_voice, is_opus=True, callback=handle_audio_frame)
+            await sendAudio(conn, audios)
         # 清除服务端讲话状态
         conn.clearSpeakStatus()