app.py

import gradio as gr
import os
import zipfile
import shutil
from pathlib import Path
import git
import re

def is_macos_metadata(rel_path):
    return (
        rel_path.startswith("__MACOSX") or
        "/__MACOSX" in rel_path or
        os.path.basename(rel_path).startswith("._") or
        os.path.basename(rel_path) == ".DS_Store"
    )

def is_git_url(input_str):
    # Simple regex to check if input is a Git URL (supports HTTP/HTTPS and SSH)
    return bool(re.match(r"^(https?://|git@).*\.git$", input_str))

def clone_git_repo(repo_url, extract_dir):
    try:
        # Clone the repository to the specified directory
        repo = git.Repo.clone_from(repo_url, extract_dir)
        return True, None
    except Exception as e:
        return False, f"Failed to clone repository: {str(e)}"

def process_input(input_source, skip_macos):
    if not input_source:
        return "Please provide a ZIP file or a Git repository URL.", None

    extract_dir = "working_dir/extracted"
    output_file = "working_dir/combined_output.txt"
    shutil.rmtree("working_dir", ignore_errors=True)
    os.makedirs(extract_dir, exist_ok=True)

    # Determine if input is a Git URL or a ZIP file
    if is_git_url(input_source):
        # Handle Git repository
        success, error = clone_git_repo(input_source, extract_dir)
        if not success:
            return error, None
        folder_name = input_source.rstrip("/").rstrip(".git").split("/")[-1] + "/"
    else:
        # Handle ZIP file
        if not os.path.exists(input_source):
            return "Invalid ZIP file path.", None
        zip_path = "working_dir/uploaded.zip"
        shutil.copy(input_source, zip_path)
        try:
            with zipfile.ZipFile(zip_path, 'r') as zip_ref:
                zip_ref.extractall(extract_dir)
        except zipfile.BadZipFile:
            return "Invalid ZIP file.", None
        folder_name = Path(input_source).stem + "/"

    tree_lines = []
    file_list = []

    def build_tree(path, prefix=""):
        entries = sorted(os.listdir(path))
        for idx, entry in enumerate(entries):
            full_path = os.path.join(path, entry)
            rel_path = os.path.relpath(full_path, extract_dir)
            if skip_macos and is_macos_metadata(rel_path):
                continue
            if os.path.basename(rel_path) == ".git":  # Skip .git directory for repos
                continue
            connector = "└── " if idx == len(entries) - 1 else "├── "
            tree_lines.append(f"{prefix}{connector}{entry}")
            if os.path.isdir(full_path):
                extension = "    " if idx == len(entries) - 1 else "│   "
                build_tree(full_path, prefix + extension)
            else:
                file_list.append((rel_path, full_path))

    build_tree(extract_dir)

    combined_text = []
    combined_text.append(f"Folder {folder_name}")
    combined_text.extend(tree_lines)
    combined_text.append("")

    for i, (rel_path, full_path) in enumerate(file_list, 1):
        if skip_macos and is_macos_metadata(rel_path):
            continue
        combined_text.append(f"{i}. {rel_path}")
        combined_text.append(f"-- content of the {rel_path} file --")
        try:
            with open(full_path, "r", encoding="utf-8") as source_file:
                combined_text.append(source_file.read())
        except Exception as e:
            combined_text.append(f"[Could not read file: {e}]")
        combined_text.append("")

    final_text = "\n".join(combined_text)

    with open(output_file, "w", encoding="utf-8") as f:
        f.write(final_text)

    return final_text, output_file

# Gradio UI
with gr.Blocks() as demo:
    gr.Markdown(" ")  # Required to avoid HF title error
    with gr.Row():
        input_source = gr.Textbox(label="Enter Git Repository URL or Upload ZIP File", placeholder="e.g., https://github.com/user/repo.git")
        zip_input = gr.File(file_types=[".zip"], label="Upload ZIP File (optional)")
        skip_toggle = gr.Checkbox(label="Skip macOS metadata files (e.g., .DS_Store, __MACOSX)", value=True)
    with gr.Row():
        process_btn = gr.Button("Generate Output")

    status = gr.Textbox(label="Output Text", lines=30, interactive=False)
    file_output = gr.File(label="Download Output File")

    def handle_inputs(input_text, zip_file, skip_macos):
        # Use ZIP file if provided, otherwise use text input
        source = zip_file if zip_file else input_text
        return process_input(source, skip_macos)

    process_btn.click(fn=handle_inputs, inputs=[input_source, zip_input, skip_toggle], outputs=[status, file_output])

if __name__ == "__main__":
    demo.launch()