sysnthetyze.py

# coding: utf-8

def synthesize_text(text, path):
    import os
    from google.cloud import texttospeech
    from pydub import AudioSegment
    import pathlib

    client = texttospeech.TextToSpeechClient()

    input_text = texttospeech.types.SynthesisInput(text=text)

    voice = texttospeech.types.VoiceSelectionParams(
            language_code='ja-jp', name='ja-jp-Wavenet-B')

    audio_config = texttospeech.types.AudioConfig(
        audio_encoding=texttospeech.enums.AudioEncoding.LINEAR16,
        speaking_rate=1.5)

    response = client.synthesize_speech(input_text, voice, audio_config)

    with open(path + '.wav', 'wb') as out:
        out.write(response.audio_content)
    audio1 = AudioSegment.from_wav(path + '.wav')
    audio2 = AudioSegment.from_wav(path + '_all.wav')
    audio = audio2 + audio1
    audio.export(path + '.mp3', format='mp3')
    # os.remove(path + '.wav')

if __name__ == '__main__':
    with open('scripts.txt', mode='rt', encoding='utf-8') as fi:
        print(fi)
        path = "/home/sayaka/sample01"
        synthesize_text(fi, path)

import os
os.remove(path + '.wav')