feat: openvoice重构，添加语音克隆功能

app/controllers/user_controller.py

@@ -1,14 +1,20 @@
 from ..schemas.user_schema import *
 from ..dependencies.logger import get_logger
+from ..dependencies.tts import get_tts
 from ..models import User, Hardware, Audio
 from datetime import datetime
 from sqlalchemy.orm import Session
 from fastapi import HTTPException, status
+from pydub import AudioSegment
+import numpy as np
+import io
 
 
 #依赖注入获取logger
 logger = get_logger()
 
+#依赖注入获取tts
+tts = get_tts()
 
 #创建用户
 async def create_user_handler(user:UserCrateRequest, db: Session):
@@ -156,13 +162,17 @@ async def get_hardware_handler(hardware_id, db):
 #用户上传音频
 async def upload_audio_handler(user_id, audio, db):
     try:
-        audio_data = audio.file.read()
-        new_audio = Audio(user_id=user_id, audio_data=audio_data)
+        audio_data = await audio.read()
+        raw_data = AudioSegment.from_file(io.BytesIO(audio_data), format="mp3").raw_data
+        numpy_data = np.frombuffer(raw_data, dtype=np.int32)
+        emb_data = tts.audio2emb(numpy_data,rate=44100,vad=True).tobytes()
+        new_audio = Audio(user_id=user_id, audio_data=audio_data,emb_data=emb_data)
         db.add(new_audio)
         db.commit()
         db.refresh(new_audio)
     except Exception as e:
         db.rollback()
+        print(str(e))
         raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail=str(e))
     audio_upload_data = AudioUploadData(audio_id=new_audio.id, uploadedAt=datetime.now().isoformat())
     return AudioUploadResponse(status="success", message="用户上传音频成功", data=audio_upload_data)
@@ -174,8 +184,11 @@ async def update_audio_handler(audio_id, audio_file, db):
         existing_audio = db.query(Audio).filter(Audio.id == audio_id).first()
         if existing_audio is None:
             raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="音频不存在")
-        audio_data = audio_file.file.read()
+        audio_data = await audio_file.read()
+        raw_data = AudioSegment.from_file(io.BytesIO(audio_data), format="mp3").raw_data
+        emb_data = tts.audio2emb(np.frombuffer(raw_data, dtype=np.int32),rate=44100,vad=True).tobytes()
         existing_audio.audio_data = audio_data
+        existing_audio.emb_data = emb_data
         db.commit()
     except Exception as e:
         db.rollback()