Unit and integration tests are a core part of TLJH, as important as the code & documentation. They help validate that the code works as we think it does, and continues to do so when changes occur. They also help communicate in precise terms what we expect our code to do.
TLJH is a distribution where the primary value is the many opinionated choices we have made on components to use and how they fit together. Integration tests are perfect for testing that the various components fit together and work as they should. So we write a lot of integration tests, and put in more effort towards them than unit tests.
All integration tests are run in GitHub Actions for each PR and merge, making sure we don’t have broken tests for too long.
The integration tests are in the
in the git repository.
py.test is used to write the integration
tests. Each file should contain tests that can be run in any order
against the same installation of TLJH.
Running integration tests locally#
docker installed and callable by the user running
the integration tests without needing sudo.
You can then run the tests with:
.github/integration-test.py run-test <name-of-run> <test-file-names>
<name-of-run>is an identifier for the tests - you can choose anything you want
<test-file-names>>is list of test files (under
integration-tests) that should be run in one go.
For example, to run all the basic tests, you would write:
.github/integration-test.py run-test basic-tests \ test_hub.py \ test_proxy.py \ test_install.py \ test_extensions.py
This will run the tests in the three files against the same installation of TLJH and report errors.
If you would like to run the tests with a custom pip spec for the bootstrap script, you can use the
.github/integration-test.py run-test <name-of-run> <test-file-names> \ --bootstrap-pip-spec="git+https://github.com/your-username/the-littlest-jupyterhub.git@branch-name"