app.py

from flask import Flask, send_file, request, render_template
import io
import zipfile
import os

app = Flask(__name__)

import subprocess
import re
from urllib.parse import urlparse, urlunparse
import shutil

def sanitize_git_url(url):
  """Sanitizes a Git repository URL."""

  # Validate URL structure using urllib.parse
  try:
    parsed_url = urlparse(url)
    if not all([parsed_url.scheme, parsed_url.netloc]):
        print(f"Error: Invalid URL structure: {url}")
        return None
    if parsed_url.scheme not in ("https", "http", "git", "ssh"):
        print(f"Error: Invalid URL scheme: {parsed_url.scheme}")
        return None
  except Exception:
      print(f"Error: Could not parse url: {url}")
      return None


  # check input for suspicious characters
  if re.search(r'[;&|]', url):
      print(f"Error: Suspicious character in url {url}")
      return None


  # Normalize the url. Remove trailing slashes and .git suffix
  normalized_url = url.rstrip("/").rstrip(".git")

  return normalized_url


def git_clone(repo_url):
  """Clones a Git repository, using the repo name as the destination folder."""

  sanitized_url = sanitize_git_url(repo_url)
  if not sanitized_url:
    return # If sanitize_git_url returned None, don't continue.

  try:
    # Extract the repository name from the URL
    repo_name = extract_repo_name(sanitized_url)
    if not repo_name:
        print(f"Error: Unable to extract repository name from URL: {repo_url}")
        return

    destination_dir = repo_name  # Use the extracted name for the destination folder
    if os.path.isdir(destination_dir):
        return repo_name

    subprocess.run(
        ["git", "clone", sanitized_url, destination_dir],
        check=True,
        capture_output=True,
        text=True
    )
    print(f"Successfully cloned {sanitized_url} to {destination_dir}")
    return repo_name

  except subprocess.CalledProcessError as e:
      print(f"Error cloning {sanitized_url}:")
      print(f"  Return code: {e.returncode}")
      print(f"  Stdout:\n{e.stdout}")
      print(f"  Stderr:\n{e.stderr}")

def extract_repo_name(repo_url):
  """Extracts the repository name from a Git URL."""
  # Example formats:
  #  https://github.com/user/repo.git
  #  [email protected]:user/repo.git
  #  https://gitlab.com/user/repo
  match = re.search(r"([^/:]+)\/?$", repo_url.rstrip(".git/"))
  if match:
     return match.group(1)
  else:
     return None

def zip_folder_to_buffer(folder_path):
    """
    Zips a folder and returns the zipped content in a BytesIO buffer.

    Args:
        folder_path: The path to the folder you want to zip.

    Returns:
        A io.BytesIO buffer containing the zipped folder data, or None if
        there is an error.
    """
    try:
        buffer = io.BytesIO()
        with zipfile.ZipFile(buffer, 'w', zipfile.ZIP_DEFLATED) as zf:
            for root, _, files in os.walk(folder_path):
                for file in files:
                    file_path = os.path.join(root, file)
                    arcname = os.path.relpath(file_path, folder_path)
                    zf.write(file_path, arcname=arcname)
        buffer.seek(0)  # Reset buffer position to the beginning for reading
        return buffer
    except Exception as e:
        print(f"Error creating zip file: {e}")
        return None


@app.route('/', methods=['GET', 'POST'])
def index():
    if request.method == 'POST':
        # folder_to_zip = "./my_folder"  # Replace with the actual folder path
        # deltemp = remove_all_files_and_directories("temp_folder") #clean temp folder

        # # Create an example folder and files
        # os.makedirs(folder_to_zip, exist_ok=True)
        # with open(os.path.join(folder_to_zip, "file1.txt"), "w") as f:
        #     f.write("This is file1")
        # with open(os.path.join(folder_to_zip, "file2.txt"), "w") as f:
        #     f.write("This is file2")

        repo_url = request.form['giturl'] #"https://huggingface.co/spaces/stepenZEN/gitgud"  # Example repo, can be very large
        
        folder_to_zip = git_clone(repo_url)
        # if not folder_to_zip:
        #     return

        zipped_buffer = zip_folder_to_buffer(folder_to_zip)
        if zipped_buffer:
            # Send the zipped buffer as a file download
            response = send_file(
                zipped_buffer,
                mimetype="application/zip",  # Important: Set the correct MIME type
                as_attachment=True,
                download_name=f"{folder_to_zip}.zip"  # Optional: Sets the suggested file name
            )
            # Clean up the example folder
            # os.remove(os.path.join(folder_to_zip, "file1.txt"))
            # os.remove(os.path.join(folder_to_zip, "file2.txt"))
            # os.rmdir(folder_to_zip)
            # shutil.rmtree(folder_to_zip)
            return response
        else:
            return "Failed to create zip file", 500

    return render_template('form.html') # Render form template

if __name__ == '__main__':
    # app = create_app()
    
    app.run(debug=True, port=5500)