INGInious is an intelligent grader that allows secured and automated testing of code made by students.
It is written in Python and uses Docker to run student's code inside a secured environment.
INGInious provides a backend which manages interaction with Docker and grade code, and a frontend which allows students to submit their code in a simple and beautiful interface. The frontend also includes a simple administration interface that allows teachers to check the progression of their students and to modify exercices in a simple way.
INGInious can be used as an external grader for Moodle or edX. The course Constraint Programming uses INGInious to correct students' code.
Simply run:
$ docker compose up --build
And access https://localhost in your browser.
The default login and password are superadmin.
You may alternatively add 127.0.0.1 inginious.local to your /etc/hosts file and access https://inginious.local in your browser.
The superadmin account is configurable at the database instantiation through 6 environment variables to specify in the docker-compose.yml file:
- INGINIOUS_SUPERADMIN_PASSWORD: Sepcifies the superadmin password. MUST BE ENCODED WITH argon2id.
- INGINIOUS_SUPERADMIN_USERNAME: Defines the superadmin username. Defaults to "superadmin" if not provided.
- INGINIOUS_SUPERADMIN_REALNAME: Defines the superadmin real name. Defaults to "INGInious superadmin" if not provided.
- INGINIOUS_SUPERADMIN_EMAIL: Defines the superadmin email address. Defaults to "superadmin@inginious.org".
- INGINIOUS_SUPERADMIN_LANG: Defines the superadmin language. Defaults to "en".
- INIGNIOUS_SUPERADMIN_INDENT: Defines the text indentation (in number of spaces) for the superadmin account. Defaults to "4".
Currently supported INGINIOUS_SUPERADMIN_LANG values are:
| Code | Language |
|---|---|
| en | English |
| el | Greek |
| es | Spanish |
| fr | French |
| he | Hebrew |
| nl | Dutch |
| nb_NO | Norwegian |
| pt | Portuguese |
| vi | Vietnamese |
Currently supported INGINIOUS_SUPERADMIN_INDENT values are: "2", "3", "4" and "tabs".
The INGInious frontend is exposed via the Caddy reverse proxy. A sample configuration file is located at deploy/Caddyfile. Refer to the Caddyfile documentation for more advanced proxy configuration.
The --build argument is optional, use it if you want to rebuild locally the core containers.
If you want to simply pull them from the project's registry, this argument is not required.
Note that you can override the registry and containers version by setting the `REGISTRY` and `VERSION` environment variables.
Docker-compose will create a tasks folder if it does not exist already.
You can then add new courses to your fresh INGInious instance by installing them in the tasks folder.
For example, the INGInious tutorial course is installed with the following commands:
$ git clone https://github.com/INGInious/demo-tasks.git $ mv demo-tasks/tutorial tasks/
If you encounter permission errors, you should run the following command:
$ sudo chown -R <your_user>:<your_user_group> tasks
This can happen when the tasks directory is created by docker-compose.
Note that the configuration.deploy.yaml file provided is a sample configuration, the secret key must be changed by administrators in production deployments.
The provided docker-compose.yaml file builds from sources the basic grading environments env-base and env-default.
It also pulls from the project's registry a basic python3 grading environment.
These three environments are sufficient for a simple INGInious deployment.
Administrators can add grading environments to their instance either by adding them statically to the docker-compose.yaml file or by pulling them manually from the container registry.
Manual installation is also possible with pip.
The documentation is available on Read the Docs:
- For the latest published version on PyPI : https://docs.inginious.org/
- For the master (dev) branch (not always up to date) : http://docs.inginious.org/en/latest/index.html
On Linux, run make html in the directory /doc to create a html version of the documentation.
INGInious is continuously improved. The various Work In Progress tasks are described in the Roadmap of the project.
Docker containers can be used securely with SELinux enabled. Please do not run untrusted code without activating SELinux.