app.py

import streamlit as st
import edge_tts
import asyncio
import tempfile
import os
from typing import Dict

async def text_to_speech(text: str, voice: str) -> str:
    output_file = tempfile.NamedTemporaryFile(delete=False, suffix=".mp3")
    communicate = edge_tts.Communicate(text, voice)
    await communicate.save(output_file.name)
    return output_file.name

async def list_voices() -> Dict[str, Dict]:
    voices = await edge_tts.list_voices()
    return {v['ShortName']: {'name': v['ShortName'], 'language': v['Locale']} for v in voices}

async def main():
    st.title("OpenSource Text-to-Speech App")
    st.write("Convert text to speech using various voices")

    # Get voices
    voices = await list_voices()

    # Text-to-Speech
    st.header("Text-to-Speech")
    text_input = st.text_area("Enter text to convert to speech:")
    selected_voice = st.selectbox("Select a voice:", list(voices.keys()))

    if st.button("Generate Speech"):
        if not text_input:
            st.error("Please enter some text.")
        else:
            with st.spinner("Generating speech..."):
                output_file = await text_to_speech(text_input, selected_voice)
                st.audio(output_file, format='audio/mp3')
                os.unlink(output_file)  # Delete the temporary file

    # List Available Voices
    st.header("Available Voices")
    st.dataframe(voices)

if __name__ == '__main__':
    asyncio.run(main())