models/STTTTS.py

from . import *

# Settings
SAMPLE_RATE = 16000
CHANNELS = 1
RECORD_FILE = './recording.wav'
SPEECH_FILE = './speech.wav'

def transcribe(audio_data):
    # Save to an in-memory WAV file
    # buffer = io.BytesIO()
    # sf.write(buffer, audio_data, SAMPLE_RATE, format="WAV")
    # buffer.seek(0)

    # with open(RECORD_FILE, "wb") as f:
    #     f.write(buffer.getbuffer())

    # print("Transcribing...")

    RECORD_FILE = audio_data
    with open(RECORD_FILE, 'rb') as file:
        response = client.audio.transcriptions.create(
            file=file,
            model="whisper-large-v3",
            prompt="You are transcribing questions about a company called 'Summit' in English or 'سَمِت' in Arabic",  # Optional
            response_format="verbose_json",  # Optional
            temperature=0.0  # Optional
        )

    print("Language:", response.language.strip())
    print("Transcription:", response.text.strip())
    return response.text.strip(), response.language.strip()

def synthesize(text, lang):
    # Choose the right model and voice based on the language
    if lang == 'English':
        model = 'playai-tts'
        voice = 'Fritz-PlayAI'
    elif lang == 'Arabic':
        model = 'playai-tts-arabic'
        voice = 'Ahmad-PlayAI'

    # Synthesize
    response = client.audio.speech.create(
        model=model,
        voice=voice,
        input=text,
        response_format='wav'
    )
    # Save to file
    response.write_to_file(SPEECH_FILE)

    # Read the file and play it
    data, samplerate = sf.read(SPEECH_FILE)
    sd.play(data, samplerate)
    sd.wait() # Busy wait until speech is played