Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Always run the test suite when building the wheel
Unfortunately there's no Python-wide way of saying "this is how you run tests", and no Python-wide way of saying "please run the tests when building the wheel". If you ask me, this is obviously wrong because tests are usually written for a reason, and not just for the lulz, and it is a good thing to know that the build did not produce utter crap. Since there's no Python-ish way of achieving this, let's give up and hardcode everything. That way the test suite will be run as a part of each build. In the CI we're still taking the wheel in another VM and executing the test suite again to make sure that whatever wheel magic which had to happen happened correctly. That, however, is a different test from "is the build OK" (e.g., `auditwheel repair`), and there's still point in having an "early warning" should the build produce a non-working result. Also, it (neatly) prevents a possible scenario where a wheel was left unchecked by a second, testing CI job. Bug: pypa/setuptools#1684
- Loading branch information