* refactor: use configurable running service for Python integration tests
Instead of configuring and starting server instances within Python as
test fixtures, simply testing against a running server. A mix of make
target and docker-compose changes is used to achieve the same level of
test coverage that previously relied on the (re-)configure and
(re-)start of the services in conftest.py.
A new `make` target, 'run_local_e2e_tests', can be used to run the
integration tests locally. However, the stage FxA JWT validation tests
in test_e2e.py are excluded. Those tests rely on the JWK configuration
of the Token Server and the stage FxA API, making them less "local".
Anyone working on that specific integration can certainly invoke those
tests themselves.
This patch also:
- deletes some duplicate docs
- moves the docker-compose yamls into a dir name 'docker'
* Apply suggestion from @pjenvey
Co-authored-by: Philip Jenvey <pjenvey@underboss.org>
* Apply suggestion from @pjenvey
Co-authored-by: Philip Jenvey <pjenvey@underboss.org>
* Clean-up based on feedback.
* Remove integration_tests/conftest.py
* Prune yamls
---------
Co-authored-by: Philip Jenvey <pjenvey@underboss.org>
