app.py

import gradio as gr
import google.generativeai as genai
import paramiko
import io
import re

def parse_github_url(github_url):
    if github_url.startswith("[email protected]:"):
        # SSH format
        parts = github_url.split(':')[1].split('/')
        owner = parts[0]
        repo = parts[1].split('.git')[0]
        return owner, repo, None, None
    elif github_url.startswith("https://github.com/"):
        # HTTPS format
        parts = github_url.split('/')
        owner = parts[3]
        repo = parts[4]
        branch = parts[6] if len(parts) > 6 else None
        file_path = '/'.join(parts[7:]) if len(parts) > 7 else None
        return owner, repo, branch, file_path
    else:
        raise ValueError("Invalid GitHub URL format")

def fetch_github_file(github_url, ssh_private_key):
    try:
        owner, repo, branch, file_path = parse_github_url(github_url)

        # Set up SSH client
        ssh = paramiko.SSHClient()
        ssh.set_missing_host_key_policy(paramiko.AutoAddPolicy())

        # Load the private key
        try:
            private_key = paramiko.RSAKey.from_private_key(io.StringIO(ssh_private_key.strip()))
        except paramiko.ssh_exception.SSHException as e:
            return f"Error: Invalid SSH key - {str(e)}"

        # Connect to GitHub using SSH
        ssh.connect('github.com', username='git', pkey=private_key)

        # Construct the git command to fetch the file content
        if branch and file_path:
            git_command = f"git archive [email protected]:{owner}/{repo}.git {branch} {file_path} | tar -xO"
        else:
            git_command = f"git ls-tree -r --name-only HEAD | grep -E '(requirements\.txt|package\.json|Gemfile)'"

        # Execute the command
        stdin, stdout, stderr = ssh.exec_command(git_command)
        
        # Read the file content
        file_content = stdout.read().decode('utf-8')

        # Close the SSH connection
        ssh.close()

        if not file_content:
            return {"error": "Error: File not found or empty."}
        
        return {file_path if file_path else "repository_files": file_content}
    except paramiko.AuthenticationException:
        return {"error": "Error: Authentication failed. Please check your SSH key."}
    except paramiko.SSHException as ssh_err:
        return {"error": f"SSH Error: {str(ssh_err)}"}
    except Exception as e:
        return {"error": f"Error accessing GitHub: {str(e)}"}

def process_with_gemini(file_content, gemini_api_key):
    genai.configure(api_key=gemini_api_key)
    model = genai.GenerativeModel('gemini-2.5-pro-preview-03-25')
    
    combined_content = "\n\n".join([f"File: {name}\n{content}" for name, content in file_content.items()])
    
    prompt = f"""
    Analyze the following file content for open-source license information:

    {combined_content}

    Please provide:
    1. A list of dependencies and their versions
    2. The licenses associated with each dependency (if available)
    3. A summary of the project based on these dependencies
    4. Any potential license conflicts or considerations
    5. Suggestions for best practices in open-source license management for this project
    """

    response = model.generate_content(prompt)
    return response.text

def process_input(file, github_url, ssh_private_key, gemini_api_key):
    if file is not None and github_url:
        return "Error: Please either upload a file OR provide a GitHub URL, not both."
    
    if file is not None:
        file_content = {file.name: file.read().decode('utf-8')}
    elif github_url and ssh_private_key:
        if not ssh_private_key.strip():
            return "Error: SSH Private Key is empty. Please provide a valid key."
        if not ssh_private_key.strip().startswith("-----BEGIN"):
            return "Error: Invalid SSH key format. Please ensure you've copied the entire key, including the BEGIN and END lines."
        try:
            file_content = fetch_github_file(github_url, ssh_private_key)
            if "error" in file_content:
                return file_content["error"]
        except ValueError as e:
            return str(e)
    else:
        return "Error: Please either upload a file OR provide both GitHub URL and SSH Private Key."
    
    try:
        # Process the file content with Gemini
        analysis = process_with_gemini(file_content, gemini_api_key)
        return analysis
    except Exception as e:
        return f"Error processing the file: {str(e)}"

iface = gr.Interface(
    fn=process_input,
    inputs=[
        gr.File(label="Upload dependency file (e.g., requirements.txt, package.json, Gemfile)"),
        gr.Textbox(label="GitHub Repository URL (optional, SSH or HTTPS format)"),
        gr.TextArea(label="SSH Private Key (required if using GitHub URL)", lines=10),
        gr.Textbox(label="Gemini API Key", type="password"),
    ],
    outputs=gr.Textbox(label="License Information and Analysis"),
    title="Open Source License Extractor",
    description="Upload a dependency file or provide a GitHub repository URL to analyze open-source licenses.",
)

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