README

This is an attempt to create an open-source, MIT-license-friendly, content runner for HSPC-style and similar coding contests.

This project was born as a project at RubyforGood 2017.

Submission Runners

A SubmissionRunner is used to compile and run the language of choice in order to score a Submission. SubmissionRunners::Base handles the logic for building docker containers used in compiling the language and running the submission. Each language's runner should inherit from Base and provide methods for the image, how to compile the language, and how to run it.

`docker_run`

We are using TTY::Command to interface with the docker command line. SubmissionRunners::Base offers a docker_run method, which a language runner's build and run methods should utilize. docker_run accepts a splatted command parameter. This gets passed through to the docker container as its run command.

For example, a program Foobar is compiled via

$ javac Foobar.java

Acceptable inputs to docker_run might look like:

docker_run('javac', 'Foobar.java')

After compilation, running a class Foobar is accomplished with

$ java Foobar

Acceptable inputs might look like (where submission_dir and input are provided by Base):

docker_run('java', 'FooBar', chdir: submission_dir, in: input)

Running tests

$ bundle exec rake                # all specs
$ bundle exec rake ci             # all specs + rubocop
$ bundle exec rake minus_docker   # non-docker specs
$ bundle exec rake dev_specs      # non-docker specs + rubocop

Contributors

David Bock (dbock@loudouncodes.org, @bokmann), Team Lead
Daniel P. Clark (6ftdan@gmail.com, @danielpclark)
Brandon Rice (brandon@blrice.net, @brandonlrice)

(contributors, please add your information here.)

Getting Started

Install Ruby v2.4.1
Install (and start) Redis
mkdir -p /var/lib/milton - You must ensure that the user running the app has permissions in this directory.
Install Docker curl -fsSL https://get.docker.com/ | sh; sudo usermod -aG docker $(whoami)
Pull docker images rake docker
git clone https://github.com/rubyforgood/loudoun_codes.git
cd loudoun_codes
bundle install
bundle exec rake db:setup
bundle exec foreman start

If you are running the application for an actual competition, you probably want to use RAILS_ENV=production bundle exec foreman start in step 7.

After starting the application, you will see this line (or something similar) in your output:

14:25:13 web.1 | * Listening on tcp://0.0.0.0:5100

Browse to the provided address and you will see the web application. Browse to /sidekiq to see statistics about currently running jobs.

Future Deployment

The 'getting started' instructions are above are while this project is in active development. A near-future goal of this project is to make deployment as brain-dead simple as possible. We may, for instance, wrap this code in a mini custom linux distribution so it could be put on a thumb drive and take over a machine for the purpose of the running content. We had previously considered a docker container mimicking the setup of the jenkins docker container, but considering our use of docker containers for the submission judging, we believe the inception scenario there would hurt our heads too much.