app.py

import gradio as gr
import google.generativeai as genai
from github import Github, GithubException

def fetch_github_file(github_url, ssh_private_key):
    try:
        # Parse the GitHub URL
        parts = github_url.split('/')
        owner = parts[3]
        repo = parts[4]
        branch = parts[6]
        file_path = '/'.join(parts[7:])

        # Create a Github instance using SSH key or Personal Access Token
        g = Github(login_or_token=ssh_private_key)

        # Get the repository
        repository = g.get_repo(f"{owner}/{repo}")

        try:
            # Get the file content
            file_content = repository.get_contents(file_path, ref=branch).decoded_content.decode('utf-8')
        except GithubException as e:
            if e.status == 404:
                return f"Error: File not found. Please check the path: {file_path}"
            else:
                raise

        if not file_content:
            return "Error: File is empty."

        return file_content
    except GithubException as e:
        if e.status == 401:
            return "Error: Authentication failed. Please check your SSH key or token."
        elif e.status == 404:
            return f"Error: Repository not found: {owner}/{repo}"
        else:
            return f"GitHub Error: {str(e)}"
    except Exception as e:
        return 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')
    
    prompt = f"""
    Analyze the following file content for open-source license information:

    {file_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.decode('utf-8')
    elif github_url and ssh_private_key:
        if not github_url.startswith("https://github.com/"):
            return "Error: Invalid GitHub URL. Please use the format: https://github.com/username/repository/blob/branch/path/to/file"
        if not ssh_private_key.strip():
            return "Error: SSH Private Key or Personal Access Token is empty. Please provide a valid key or token."
        file_content = fetch_github_file(github_url, ssh_private_key)
        if file_content.startswith("Error:"):
            return file_content
    else:
        return "Error: Please either upload a file OR provide both GitHub URL and SSH Private Key or Personal Access Token."
    
    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 File URL (optional)"),
        gr.Textbox(label="SSH Private Key or Personal Access Token (required if using GitHub URL)", type="password"),
        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 file URL to extract and analyze open-source license information.",
)

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