Closed connectdotz closed 6 months ago
I'm curious why you chose to remove husky
at the same time.
Wow! :clap: :tada:
I'm realizing I won't be able to review a diff this big in my spare time, especially with renaming the files and changes not being picked up https://github.com/jest-community/jest-editor-support/pull/116/commits/3a9d429fc144f492fbe293776a425dcb1c6c3e33#diff-4fa18de8d59e8fbce3df7e88da529ad458bbf37e526d58d30b56707ba4d0d8b6
Sorry. That's as far as I'll get. :face_in_clouds:
@seanpoulter, Thanks for taking a look. I know this is a tough PR to review, sorry. I gave up the piecemeal approach as it made the conversion much harder 😰... You already helped, thank you.🙏 It's nice to know that I am not alone on this!
I figure we will bite the bullet and do this in one PR, but we won't release it to prod (thus the 32.0.0-beta.0 version), so we can have more time to test and fix when needed. Does that sound alright?
I'm curious why you chose to remove husky at the same time.
Mainly to simplify the env and reduce maintenance overhead. Given we already have lint/CI to detect errors and prettier-write to correct them, I thought husky/lint-staged didn't really add any extra value... Besides, we didn't use them in vscode-jest
and it seems to be fine.
Changes
tsc
.babel-jest
withts-test
to run Jest with better type error detection.index.d.ts
from the ts files.Backward compatibility
This PR does not change functionality, and I try to keep the types as backward-compatible as possible. However, there are still some breaking changes (mainly types). Therefore, mark this as "32.0.0-beta.0" to start preparing for the next major release.
Integration Testing
I have done some preliminary tests with
vscode-jest
and it seems to work fine. Would encourage other dependent systems to try out and let me know if there are any other unexpected issues.