fix：aliyun,cozecn,custom,doubao,gpt_sovits_v2,gpt_sovits_v3,minimax,openai,siliconflow,ttson,这些tts兼容

main/xiaozhi-server/core/providers/tts/aliyun.py

@@ -6,6 +6,9 @@ import hashlib
 import base64
 import requests
 from datetime import datetime
+
+from pydub import AudioSegment
+
 from core.providers.tts.base import TTSProviderBase
 
 import http.client
@@ -13,6 +16,10 @@ import urllib.parse
 import time
 import uuid
 from urllib import parse
+
+from core.providers.tts.dto.dto import TTSMessageDTO, MsgType, SentenceType
+
+
 class AccessToken:
     @staticmethod
     def _encode_text(text):
@@ -105,7 +112,7 @@ class TTSProvider(TTSProviderBase):
     def generate_filename(self, extension=".wav"):
         return os.path.join(self.output_file, f"tts-{__name__}{datetime.now().date()}@{uuid.uuid4().hex}{extension}")
 
-    async def text_to_speak(self, text, output_file):
+    async def text_to_speak(self, u_id, text, is_last_text=False, is_first_text=False):
         request_json = {
             "appkey": self.appkey,
             "token": self.token,
@@ -119,14 +126,26 @@ class TTSProvider(TTSProviderBase):
         }
 
         print(self.api_url, json.dumps(request_json, ensure_ascii=False))
+        tmp_file = self.generate_filename()
         try:
             resp = requests.post(self.api_url, json.dumps(request_json), headers=self.header)
             # 检查返回请求数据的mime类型是否是audio/***，是则保存到指定路径下；返回的是binary格式的
             if resp.headers['Content-Type'].startswith('audio/'):
-                with open(output_file, 'wb') as f:
+                with open(tmp_file, 'wb') as f:
                     f.write(resp.content)
-                return output_file
             else:
                 raise Exception(f"{__name__} status_code: {resp.status_code} response: {resp.content}")
+            # 使用 pydub 读取临时文件
+            audio = AudioSegment.from_file(tmp_file, format="wav")
+            audio = audio.set_channels(1).set_frame_rate(16000)
+            opus_datas = self.wav_to_opus_data_audio_raw(audio.raw_data)
+            yield TTSMessageDTO(u_id=u_id, msg_type=MsgType.TTS_TEXT_RESPONSE, content=opus_datas,
+                                tts_finish_text=text, sentence_type=SentenceType.SENTENCE_START)
+            # 用完后删除临时文件
+            try:
+                os.remove(tmp_file)
+            except FileNotFoundError:
+                # 若文件不存在，忽略该异常
+                pass
         except Exception as e:
             raise Exception(f"{__name__} error: {e}")