app.py

import gradio as gr
import subprocess
import os

def run_katana(url):
    # Set a custom configuration directory
    katana_config_dir = "/tmp/katana"
    os.makedirs(katana_config_dir, exist_ok=True)
    
    # Set the environment variable for Katana's config directory
    os.environ["KATANA_CONFIG"] = katana_config_dir

    output_file = "/tmp/urls.txt"  # Use /tmp for writable directory
    command = f"katana -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 /tmp/flagged directory exists
os.makedirs("/tmp/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="/tmp/flagged"  # Specify a writable directory for flagging
)

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