app.py

import gradio as gr
import git
import os
import tempfile
import shutil

def is_text_file(file_path):
    """Check if a file is likely to be a text file."""
    try:
        with open(file_path, 'r', encoding='utf-8') as f:
            f.read(1024)  # Try to read the first 1024 bytes
            return True
    except UnicodeDecodeError:
        return False
    except Exception:
        return False

def clone_and_concatenate(repo_url):
    # Create a temporary directory
    temp_dir = tempfile.mkdtemp()
    output_file = os.path.join(temp_dir, "concatenated_files.txt")
    
    try:
        # Clone the repository
        git.Repo.clone_from(repo_url, temp_dir)
        
        # Walk through all files in the repository
        with open(output_file, 'w', encoding='utf-8') as outfile:
            for root, dirs, files in os.walk(temp_dir):
                # Skip .git directory
                if '.git' in dirs:
                    dirs.remove('.git')
                
                for file in files:
                    if file == "concatenated_files.txt":
                        continue
                        
                    file_path = os.path.join(root, file)
                    relative_path = os.path.relpath(file_path, temp_dir)
                    
                    # Check if the file is a text file
                    if is_text_file(file_path):
                        try:
                            with open(file_path, 'r', encoding='utf-8') as infile:
                                outfile.write(f"\n\n{'='*80}\n")
                                outfile.write(f"File: {relative_path}\n")
                                outfile.write(f"{'='*80}\n\n")
                                outfile.write(infile.read())
                        except Exception as e:
                            outfile.write(f"\nError reading file {relative_path}: {str(e)}\n")
    
    except Exception as e:
        # Clean up the temporary directory
        shutil.rmtree(temp_dir)
        return None, f"Error: {str(e)}"
    
    return output_file, "Repository cloned and files concatenated successfully!"

# Create the Gradio interface
iface = gr.Interface(
    fn=clone_and_concatenate,
    inputs=gr.Textbox(label="Git Repository URL"),
    outputs=[
        gr.File(label="Concatenated Files"),
        gr.Textbox(label="Status")
    ],
    title="Git Repository File Concatenator",
    description="Enter a Git repository URL to clone it and concatenate all text files into a single downloadable file.",
    examples=[["https://github.com/username/repository"]]
)

# Launch the app
if __name__ == "__main__":
    iface.launch()