import gradio as gr

from google.cloud import speech
from microphone import MicrophoneStream
from utils import listen_print_loop


# Audio recording parameters
RATE = 16000
CHUNK = int(RATE / 10)  # 100ms
LANGUAGE = "id-ID"

transcribe_client = speech.SpeechClient()
config = speech.RecognitionConfig(
    encoding=speech.RecognitionConfig.AudioEncoding.LINEAR16,
    sample_rate_hertz=RATE,
    language_code=LANGUAGE,
)

streaming_config = speech.StreamingRecognitionConfig(
    config=config, interim_results=True
)
    
async def main(audio) -> None:
    
    print("Streaming started ...")

    with MicrophoneStream(RATE, CHUNK) as stream:
        audio_generator = stream.generator()
        requests = (
            speech.StreamingRecognizeRequest(audio_content=content)
            for content in audio_generator
        )

        responses = transcribe_client.streaming_recognize(streaming_config, requests)

        return await listen_print_loop(responses)
    
demo = gr.Interface(
    fn=main, 
    inputs=[
        gr.Audio(sources="microphone", streaming=True, label="Input Speech")
    ],
    outputs=[
        gr.Textbox(label="Transcription"),
        gr.Audio(label="Audio")
    ],
    live=True)

if __name__ == "__main__":
    demo.launch()