generative_audio_service.py

import wsl_shell
import simpleaudio as sa
import os


class GenerativeAudioService:
    """
    Create and manage playback with text to speech audio files
    """

    def __init__(self) -> None:
        # todo. check if piper is installed
        # piper --update-voices -m en_US-libritts-high
        pass

    def generative(self, text: str):
        import time
        import shlex
        import os.path
        temp_filename = 'generative.wav'
        if os.path.isfile(temp_filename):
            os.remove(temp_filename)

        encoded_text = shlex.quote(text)
        # en_US-lessac-medium
        shell_command = f"echo {encoded_text} | piper --model en_US-libritts-high --output_file {temp_filename}"
        # print(shell_command)
        _, stderr, returncode = wsl_shell.shell(shell_command)
        print(stderr)
        if returncode == 0:
            while not os.path.isfile(temp_filename):
                time.sleep(1)

        GenerativeAudioService.play(temp_filename)

    @staticmethod
    def play(filename: str, blocking: bool = True):
        wave_obj = sa.WaveObject.from_wave_file(filename)
        play_obj = wave_obj.play()
        if blocking:
            play_obj.wait_done()

    @staticmethod
    def ding():
        GenerativeAudioService.play('ding.wav', False)

wsl_shell.py

@staticmethod
def shell(command: str, distribution: str = "Ubuntu-22.04"):
    """
    Command to start an interactive shell in WSL
    """
    import subprocess
    shell = ["wsl", "-d", distribution, "/bin/bash"]
    p = subprocess.run(shell, input=command, capture_output=True, text=True)
    stdout = p.stdout
    stderr = p.stderr
    returncode = p.returncode
    return stdout, stderr, returncode