text_speech_utils.py

import openai
import sounddevice as sd
import audiofile as af
from scipy.io.wavfile import write
from gtts import gTTS

import multiprocessing
import pyttsx3
import keyboard

def say(text):
		p = multiprocessing.Process(target=pyttsx3.speak, args=(text,))
		p.start()
		while p.is_alive():
			if keyboard.is_pressed('enter'):
				p.terminate()
			else:
				continue
		p.join()


import pyaudio
import wave

def record_audio(filename, sec, sr=44100):
    p = pyaudio.PyAudio()

    stream = p.open(format=pyaudio.paInt16, channels=2, rate=sr,
                    input=True, frames_per_buffer=1024)

    frames = []

    for i in range(0, int(sr / 1024 * sec)):
        data = stream.read(1024)
        frames.append(data)

    stream.stop_stream()
    stream.close()
    p.terminate()

    with wave.open(filename, 'wb') as wf:
        wf.setnchannels(2)
        wf.setsampwidth(p.get_sample_size(pyaudio.paInt16))
        wf.setframerate(sr)
        wf.writeframes(b''.join(frames))

def record_audio_manual(filename, sr = 44100):
    input("  ** Press enter to start recording **")
    audio = sd.rec(int(10 * sr), samplerate=sr, channels=2)
    input("  ** Press enter to stop recording **")
    sd.stop()
    write(filename, sr, audio)

def play_audio(filename):
    signal, sr = af.read(filename)
    sd.play(signal, sr)

def transcribe_audio(filename):
    audio_file= open(filename, "rb")
    transcript = openai.Audio.transcribe("whisper-1", audio_file)
    audio_file.close()
    return transcript

def translate_audio(filename):
    audio_file= open(filename, "rb")
    translation = openai.Audio.translate("whisper-1", audio_file)
    audio_file.close()
    return translation

def save_text_as_audio(text, audio_filename):
    myobj = gTTS(text=text, lang='en', slow=False)  
    myobj.save(audio_filename)