import streamlit as st
from transformers import pipeline


from scipy.io.wavfile import write
import gradio as gr
import wavio

input_file = 'recorded.wav'
output_file = 'output_filtered_receiver.wav'

low_frequency = 18000
high_frequency = 19000
bit_duration = 0.007
sample_rate = 44100
amplitude_scaling_factor = 10.0

def record(audio):
    """
    This function records audio and writes it to a .wav file.
    Parameters:
    audio (tuple): A tuple containing the sample rate and the audio data.
    Returns:
    str: A success message if the audio is recorded correctly, otherwise an error message.
    """
    try:
        # Check if the audio tuple contains exactly two elements
        if len(audio) != 2:
            return f"Error: Expected a tuple with 2 elements, but got {len(audio)}"

        # Unpack the sample rate and data from the audio tuple
        sr, data = audio

        # Write the audio data to a .wav file
        wavio.write("recorded.wav", data, sr)

        # Call the filtered function to apply the bandpass filter to the audio data
        filtered()

        # Return a success message
        return f"Audio receive correctly"
    except Exception as e:
        # If an error occurs, return an error message
        return f"Error: {str(e)}"

with gr.Blocks() as demo:
    btn_record = gr.Button(value="record")
    btn_record.click(fn=record, inputs=input_audio, outputs=output_text)


demo.launch()



########################     models
# model = pipeline("sentiment-analysis")
# st.title("Hugging Face Model Demo")
# input_text = st.text_input("Enter your text", "")
# if st.button("Analyze"):
#     # Perform inference using the loaded model
#     result = model(input_text)
#     st.write("Prediction:", result[0]['label'], "| Score:", result[0]['score'])