main.py

import asyncio
from fastapi import FastAPI, HTTPException
from fastapi.responses import StreamingResponse
from pydantic import BaseModel
import edge_tts

app = FastAPI()

class TTSRequest(BaseModel):
    text: str
    voice: str = "en-US-AvaNeural"

async def generate_audio_stream(text: str, voice: str):
    try:
        communicate = edge_tts.Communicate(text, voice)
        async for chunk in communicate.stream():
            if chunk["type"] == "audio":
                yield chunk["data"]
    except Exception as e:
        raise HTTPException(status_code=500, detail=str(e))

@app.post("/v1/tts")
async def text_to_speech(request: TTSRequest):
    if not request.text.strip():
        raise HTTPException(status_code=400, detail="Text cannot be empty")
    return StreamingResponse(
        generate_audio_stream(request.text, request.voice),
        media_type="audio/mpeg"
    )

# Root endpoint for checking if the API is awake
@app.get("/")
def home():
    return {"status": "Edge TTS API is running on Hugging Face Spaces!"}

if __name__ == "__main__":
    import uvicorn
    # Hugging Face Spaces strictly requires port 7860
    uvicorn.run(app, host="0.0.0.0", port=7860)