AutoJudge is a simple script that automates the process of running test cases for different programming languages. It can be used to evaluate the correctness of solutions for various programming problems.

• You can use this project as a CLI code runner. For this, check the runner-cli branch.

• Check Autojudge online version at autojudge.werlang.site.

• Supports multiple programming languages, including C, JavaScript, PHP, Python.
• Automatically detects the language based on the script's file extension.
• Compiles and runs code with test cases from the input directory.
• Compares the output with expected results from the output directory.
• Provides pass/fail statistics for each test case.
• Web interface allowing to save problems on local storage and run autojudge in the browser.

• Docker

Build the Docker images and start the containers:

docker compose up -d

Pull the images for the languages you will use. With the api service running, you can pull the images with the following command:

docker exec api npm run compilers

Build the front-end css and js files. call the webpack command inside the container:

docker exec web npx webpack

Hit Ctrl+C when the process finishes.

Nginx will serve the front-end at https://autojudge.localhost. You can access the web interface to run the autojudge. From there, you can check the instructions on how to use the tool.