app.py

import gradio as gr
import subprocess
import os

def run_katana(url):
    katana_config_dir = "/tmp/katana"
    os.makedirs(katana_config_dir, exist_ok=True)
       
    config_file_path = os.path.join(katana_config_dir, "config.yaml")
    with open(config_file_path, 'w') as f:
        f.write("headless: true\n")  # Add a valid configuration option
   
    output_file = "/tmp/urls.txt"
    command = f"katana -config {config_file_path} -u {url} -o {output_file}"
    try:
        subprocess.run(command, shell=True, check=True, stderr=subprocess.PIPE, text=True)
        if os.path.exists(output_file):
            return output_file
        else:
            return "Katana completed but no output file was created."
    except subprocess.CalledProcessError as e:
        return f"Error running katana: {e.stderr}"

# 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)