Add functional tests #19
Conversation
mcmire
force-pushed
the
add-functional-tests
branch
from
30abbd6 to
120a622
Compare
This commit adds tests which run the script within a separate shell instance and verifies that the script makes the appropriate changes to the repo. I've tried to make these functional tests as readable as possible. In order to do that, I created a small framework which replicates the environment in which the script will run and then allows that environment to be modified per-test to suit the assertions being made.
mcmire
force-pushed
the
add-functional-tests
branch
from
120a622 to
e23c2e2
Compare
| * @param error - The object to check. | ||
| * @returns True or false, depending on the result. | ||
| */ | ||
| export function isErrorWithCode(error: unknown): error is { code: string } { |
There was a problem hiding this comment.
You'll notice that some of these helpers are repeated from misc-utils. I could have reused them, but it feels a little off to me to reuse implementation code in test code, even if it's the exact same code. Ideally, I think, we could move this to @metamask/action-utils or even @metamask/utils and then I think I'd feel better about it.
|
|
||
| describe('create-release-branch (functional)', () => { | ||
| describe('against a monorepo with independent versions', () => { | ||
| it('updates the version of the root package to be the current date along with the versions of the specified packages', async () => { |
There was a problem hiding this comment.
I'm only testing the happy paths right now. This file is going to grow larger as we add different features — for instance, right now there is no logic around changed packages, so once that happens we need to consider what happens when the repo has no tags, etc. — so I'm hesitant to add error/alternate cases. I figure it's easier to add more test cases later than remove them.
Co-authored-by: Mark Stacey <markjstacey@gmail.com>
2 participants
This commit adds tests which run the script within a separate shell
instance and verifies that the script makes the appropriate changes to
the repo.
I've tried to make these functional tests as readable as possible. In
order to do that, I created a small framework which replicates the
environment in which the script will run and then allows that
environment to be modified per-test to suit the assertions being made.