app.py

import streamlit as st
from gtts import gTTS
from io import BytesIO

LANG = "en"


#
# https://gtts.readthedocs.io/en/latest/
#
def tts_gtts(text):
    mp3_fp = BytesIO()
    tts = gTTS(text, lang=LANG)
    tts.write_to_fp(mp3_fp)
    return mp3_fp


def pronounce(text, gender=None):
    if len(text) > 0:
        data1 = tts_gtts(text)
        st.text('gTTS (gender not supported):')
        st.audio(data1, format="audio/wav", start_time=0)


def main():
    st.title('TTS Demo')
    # uploaded_file = st.file_uploader("Upload a recording of you saying the text in .wav format")
    # if uploaded_file is not None:
    #     pass
    text_input = st.text_input("", value="Input the text you are saying in your recording.")
    gender_input = st.radio("Gender", ["M", "F"], captions=["", ""], horizontal=True)
    if st.button("Pronounce"):
        pronounce(text_input, gender_input)


if __name__ == "__main__":
    main()