/AutoFactor

running Codex in a Docker Compose file to auto refactor a local repo, allowing higher token count, custom temperature control and mostl seeing the code wholestically rather than copy pasted back and forth in chat with 150 tokens just summarizing a copy of a copy each time.

Primary LanguagePythonApache License 2.0Apache-2.0

AutoFactor

GPT-3.5 Code Refactor

This repository contains a Python application that uses GPT-3.5, an AI language model by OpenAI, to automatically refactor code in a given repository. The application processes code files in various languages, refactors them using the GPT-3.5 API, and saves the refactored code along with a summary of the changes.

Prerequisites

  • Docker
  • Docker Compose
  • An OpenAI API key

Setup

  1. Clone the repository:
git clone https://github.com/TortoiseWolfe/AutoFactor
cd AutoFactor
  1. Create a .env file in the root directory of the project and add your OpenAI API key:
OPENAI_API_KEY=your_actual_api_key_here

Replace your_actual_api_key_here with your actual API key.

Usage

  1. Copy the code you want to refactor into the your_repo directory.

  2. Build the Docker container:

docker-compose build
  1. Run the container:
docker-compose up -d

  1. The refactored code will replace the original code in the your_repo directory. The original code will be backed up in the original_code directory.

  2. Check the logs for the summary of changes:

cat your_repo/logs/output.log

Customization

To add support for more languages, update the extension_to_language function in the main.py file with the desired file extension and language name.

def extension_to_language(extension):
    languages_map = {
        "cs": "C#",
        "js": "JavaScript",
        "py": "Python",
        # Add more file extensions and languages as needed
    }
    return languages_map.get(extension.lower())

License

This project is open source under the MIT License. See the LICENSE file for details.