add：火山双向tts语音流式输入输出

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

@@ -4,94 +4,31 @@ from config.logger import setup_logging
 import json
 import asyncio
 import time
+
+from core.providers.tts.dto.dto import TTSMessageDTO, SentenceType, MsgType
 from core.utils.util import remove_punctuation_and_length, get_string_no_punctuation_or_emoji
 
 TAG = __name__
 logger = setup_logging()
 
 
-async def sendAudioMessageStream(conn, audios_queue, text, text_index=0, llm_finish_task=False):
+async def sendAudioMessage(conn, ttsMessageDTO: TTSMessageDTO):
+    u_id = None
     # 发送句子开始消息
-    if text_index == conn.tts_first_text_index:
-        logger.bind(tag=TAG).info(f"发送第一段语音: {text}")
-    await send_tts_message(conn, "sentence_start", text)
-
-    # 初始化流控参数
-    frame_duration = 60  # 毫秒
-    start_time = time.time()  # 使用高精度计时器
-    # 初始化流控参数
-    frame_duration = 60  # 毫秒
-    start_time_chunk = time.perf_counter()  # 使用高精度计时器
-    play_position = 0  # 已播放的时长（毫秒）
-    while True:
-        try:
-            start_get_queue = time.time()
-            # 尝试获取数据，如果没有数据，则等待一小段时间再试
-            audio_data_chunke = None
-            try:
-                audio_data_chunke = audios_queue.get(timeout=5)  # 设置超时为1秒
-            except Exception as e:
-                # 如果超时，继续等待
-                logger.bind(tag=TAG).error(f"获取队列超时～{e}")
-
-            audio_opus_datas = audio_data_chunke.get('data') if audio_data_chunke else None
-            duration = audio_data_chunke.get('duration') if audio_data_chunke else 0
-
-            if audio_data_chunke:
-                start_time = time.time()
-            # 检查是否超过 5 秒没有数据
-            if time.time() - start_time > 15:
-                logger.bind(tag=TAG).error("超过15秒没有数据，退出。")
-                break
-
-            if audio_data_chunke and audio_data_chunke.get("end", True):
-                break
-
-            if audio_opus_datas:
-                for opus_packet in audio_opus_datas:
-                    if conn.client_abort:
-                        return
-                    logger.bind(tag=TAG).info(f'发送数据长度：{len(opus_packet)}')
-                    await conn.websocket.send(opus_packet)
-                    play_position += frame_duration  # 更新播放位置
-                start_time = time.time()  # 更新获取数据的时间
-        except Exception as e:
-            logger.bind(tag=TAG).error(f"发生错误: {e}")
-            traceback.print_exc()  # 打印错误堆栈
-    await send_tts_message(conn, "sentence_end", text)
-
-    print(f'{text_index}-{conn.tts_last_text_index}')
-    expected_time = start_time_chunk + (play_position / 1000)
-    current_time = time.perf_counter()
-    # 等待直到预期时间
-    delay = expected_time - current_time
-    if delay > 0:
-        await asyncio.sleep(delay)
-    # 发送结束消息（如果是最后一个文本）
-    logger.bind(tag=TAG).info(f"{conn.llm_finish_task},{text_index},{conn.tts_last_text_index}")
-    if conn.llm_finish_task and text_index == conn.tts_last_text_index:
-        await send_tts_message(conn, 'stop', None)
-        if conn.close_after_chat or "拜拜" in text or "再见" in text:
-            await conn.close()
-
-
-
-async def sendAudioMessage(conn, audios, text, text_index=0):
-    # 发送句子开始消息
-    if text_index == conn.tts_first_text_index:
-        logger.bind(tag=TAG).info(f"发送第一段语音: {text}")
-    await send_tts_message(conn, "sentence_start", text)
+    if SentenceType.SENTENCE_START == ttsMessageDTO.sentence_type:
+        logger.bind(tag=TAG).info(f"发送第一段语音: {ttsMessageDTO.tts_finish_text}")
+        await send_tts_message(conn, "sentence_start", ttsMessageDTO.tts_finish_text)
 
     # 流控参数优化
     original_frame_duration = 60  # 原始帧时长（毫秒）
     adjusted_frame_duration = int(original_frame_duration * 0.8)  # 缩短20%
-    total_frames = len(audios)  # 获取总帧数
+    total_frames = len(ttsMessageDTO.content)  # 获取总帧数
     compensation = total_frames * (original_frame_duration - adjusted_frame_duration) / 1000  # 补偿时间（秒）
 
     start_time = time.perf_counter()
     play_position = 0  # 已播放时长（毫秒）
 
-    for opus_packet in audios:
+    for opus_packet in ttsMessageDTO.content:
         if conn.client_abort:
             return
 
@@ -110,15 +47,17 @@ async def sendAudioMessage(conn, audios, text, text_index=0):
     # 补偿因加速损失的时长
     if compensation > 0:
         await asyncio.sleep(compensation)
-
-    await send_tts_message(conn, "sentence_end", text)
+    if SentenceType.SENTENCE_END == ttsMessageDTO.sentence_type:
+        logger.bind(tag=TAG).info(f"发送最后一段语音: {ttsMessageDTO.tts_finish_text}")
+        await send_tts_message(conn, "sentence_end", ttsMessageDTO.tts_finish_text)
 
     # 发送结束消息（如果是最后一个文本）
-    if conn.llm_finish_task and text_index == conn.tts_last_text_index:
+    if conn.llm_finish_task and MsgType.STOP_TTS_RESPONSE == ttsMessageDTO.msg_type:
         await send_tts_message(conn, 'stop', None)
         if conn.close_after_chat:
             await conn.close()
 
+
 async def send_tts_message(conn, state, text=None):
     """发送 TTS 状态消息"""
     message = {