app.py

import gradio as gr
import subprocess
import os

def run_katana(url):
    # Set a custom configuration directory within /app
    katana_config_dir = "/app/katana"
    os.makedirs(katana_config_dir, exist_ok=True)
    
    # Ensure the config file exists, or create a dummy one if needed
    config_file_path = os.path.join(katana_config_dir, "config.yaml")
    if not os.path.exists(config_file_path):
        with open(config_file_path, 'w') as f:
            f.write("# Dummy configuration\n")

    output_file = "/app/urls.txt"  # Use /app for writable directory
    command = f"katana -config {config_file_path} -u {url} > {output_file}"
    try:
        subprocess.run(command, shell=True, check=True)
        return output_file
    except subprocess.CalledProcessError as e:
        return f"Error running katana: {e}"

# Ensure the /app/flagged directory exists
os.makedirs("/app/flagged", exist_ok=True)

# Gradio interface
interface = gr.Interface(
    fn=run_katana,
    inputs="text",
    outputs="file",
    title="Katana URL Crawler",
    description="Enter a URL to crawl with Katana.",
    flagging_dir="/app/flagged"  # Specify a writable directory for flagging
)

interface.launch(server_name="0.0.0.0", server_port=7860)