The tournament can be configured to handle multiple types of assignments written in multiple languages and so there
are a number of commands, such as running a test suite, that need to be performed differently for different assignment.
The commands are all stored in an assignment class.
To create a configuration for a new assignment create a new class that extends from abstract_assignment.py and fill
out all of the abstract methods.
Make sure that in prep_submission that the original source code for the assignment is not overwritten by the
submitters copy. For example is the ant_assignment the programs/ contains mutant-1, mutant-2, mutant-3
... original. Only mutant-1 through mutant-n are replaced, and original is not overwritten. This is to ensure
that when validate_tests is run the submitters tests are actually checked against the original source code, and
not a submitter modified version.
New assignments should aim for a generic extensible skeleton. For example ant_assignment has a folder called
programs in which all mutants are stored, and a folder tests in which all tests are stored. The tournament
code can then ls the folders to determine the number and name of the PUTs and test suites in the assignment.
All assignments need to have a .gitlab-ci.yml file at the root level so that the Gitlab Runner knows to run a
CI pipeline on the repos. This CI pipeline is how the code is validated and then submitted to the tournament.
An example file can be found here