app.py

import gradio as gr
import os
from acrcloud.recognizer import ACRCloudRecognizer

# Retrieve ACRCloud credentials from environment variables
acr_access_key = os.environ.get('ACR_ACCESS_KEY')
acr_access_secret = os.environ.get('ACR_ACCESS_SECRET')
acr_host = 'identify-ap-southeast-1.acrcloud.com'

# ACRCloud recognizer configuration
config = {
    'host': acr_host,
    'access_key': acr_access_key,
    'access_secret': acr_access_secret,
    'timeout': 10  # seconds
}

# Initialize ACRCloud recognizer
acr = ACRCloudRecognizer(config)

def identify_audio(file):
    # Save the uploaded file temporarily
    file_path = "temp_audio_file" + os.path.splitext(file.name)[-1]  # Preserve the file extension

    # Write the uploaded file to disk
    with open(file_path, 'wb') as f:
        f.write(file.file.read())  # Write the content of the uploaded file

    # Get the duration of the audio file in milliseconds
    duration_ms = int(acr.get_duration_ms_by_file(file_path))

    results = []

    # Process audio in 10-second chunks
    for i in range(0, duration_ms // 1000, 10):
        res = acr.recognize_by_file(file_path, i, 10)
        results.append(f"Time {i}s: {res.strip()}")

    # Full recognition result
    full_result = acr.recognize_by_file(file_path, 0)

    # Recognize using file buffer
    with open(file_path, 'rb') as f:
        buf = f.read()
        buffer_result = acr.recognize_by_filebuffer(buf, 0)

    return {
        "Partial Results": results,
        "Full Result": full_result,
        "Buffer Result": buffer_result
    }

# Create Gradio interface
iface = gr.Interface(
    fn=identify_audio,
    inputs=gr.File(label="Upload Audio File"),
    outputs=gr.JSON(label="Audio Metadata"),
    title="Audio Search by File",
    description="Upload an audio file to identify it using ACRCloud."
)

# Launch the Gradio interface
iface.launch()