translate_speak.py

from gtts import gTTS
from deep_translator import GoogleTranslator
import soundfile as sf
import tempfile
import numpy as np
import gtts
import datetime

output_path = 'Audio/output.wav'
translate_path = 'Audio/translate.wav'

def get_lang(lang):
    translate = GoogleTranslator().get_supported_languages()

    speak = list(gtts.lang.tts_langs().values())
    speak = list(map(lambda x: x.lower(), speak))
    speak.sort()

    set_lang = list(set(speak) & set(translate))
    set_lang.sort()

    if lang in set_lang:
        for key, value in gtts.lang.tts_langs().items():
            if value.lower() == lang:
                return key
        else:
            return 'en'
    else:
        return 'en'

def audio_streaming(txt=None, lang='en', to=None):
    # If an audio file is provided as input, use it; otherwise, use the direct file path
    speak = gTTS(text=txt, lang=lang, slow=False)
    if to == 1:
        audio = output_path
    else:
        audio = translate_path
    speak.save(audio)

    # Load the audio file
    data, samplerate = sf.read(audio)

    # Ensure data is in float32 format
    data = np.array(data, dtype=np.float32)

    # Save to a temporary file that Gradio can use for audio playback
    with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as tmp_file:
        sf.write(tmp_file.name, data, samplerate)
        temp_audio_path = tmp_file.name

    # Return the file path to Gradio
    return temp_audio_path

def translate_txt(lang, text):
    translator = GoogleTranslator(source="en", target=lang)
    translated_text = translator.translate(text)

    # with open("Request.txt", 'a', encoding="utf-8") as f:
    #     f.write(f"{datetime.datetime.now()}: {text}->{lang}->{translated_text}\n")
    print(f"\n{datetime.datetime.now()}: {text}->{lang}->{translated_text}\n")
    audio_path = audio_streaming(translated_text, lang=get_lang(lang), to=2)

    return translated_text, audio_path

if __name__ == "__main__":
    # print(audio_streaming("hello world"))
    # os.system(f"start {audio_streaming('hello world!')}")
    translate = GoogleTranslator().get_supported_languages()
    # print(f"Can translate: {translate}", len(translate))
    # print()

    speak = list(gtts.lang.tts_langs().values())
    speak = list(map(lambda x: x.lower(), speak))
    speak.sort()
    # print(f"Can speak: {speak}", len(speak))
    # print()

    set_lang = list(set(speak)&set(translate))
    set_lang.sort()
    # print(f"Set of lang: {set_lang}", len(set_lang))

    not_in_speak = list(set(translate) - set(speak))
    print(not_in_speak)