WIP: Use husky to execute linters

[MartijnCuppens]

Closes #27973

The linting taks are now ran before anything is pushed which will help us with builds that fail because of trivial lint errors.

I preferred the pre-push option above the pre-commit option because all linting tasks take about 9 seconds to complete which could be annoying.

This PR also contains a commit which parallelizes the linting tasks. This decreases the time from around 9 seconds to 4-5 seconds.

Not sure if we should add the --fix command which automatically fixes fixable linting errors (like fixing the property order). Maybe we could add an additional command for it?

TODO:

• Add documentation once we've decided how we're going to implement it.

[XhmikosR]

I still find it too slow.

[XhmikosR]

You can split the parallel patch to a separate PR so that we try it and merge it regardless of this.

[Johann-S]

It's a bit slow 😟 do you know if there is a way to skip that on push ?

[MartijnCuppens]

#27989 will increase speed to around 1.8s (for me) which is acceptable for me as pre-push, don't know what you guys think? Imo it's worth it because figuring out what's wrong with some one else's PR takes a lot more time.

--no-verify can be used to skip the tests.

[XhmikosR]

Or we could just leave things as is, and you can add the hook manually.

[MartijnCuppens]

It's not only for me I opt for this. A lot of Travis tests are failing because of linting issues and these hooks can spare contributors some time because it will give linting feedback before anything is pushed. It would also save us some time figuring out what's wrong with the Travis tests and replying they should check their linting or fixing the code ourselfs.

On the other hand it slows down our current workflow and is yet another dependency where thing could go wrong with.

[XhmikosR]

The thing is, many people don't even check out the repo locally and even if they do they don't run the test scripts.

Secondly, we do have CI for this anyway.

I'm not against it as long as it doesn't disrupt my current workflow, and as it seems it does since it slows us down. The dep I don't mind, it's a small one anyway.

Also, IMO this should be in precommit if we ever go with it. You could try calling the individual css and js lint scripts if that's faster instead of the wrapper lint script.

[XhmikosR]

That's not a rebase.

[MartijnCuppens]

Fixed it

[XhmikosR]

You missed the update package-lock.json.

[XhmikosR]

OK, I tried this again, and it's just too slow for me...

[MartijnCuppens]

Ok fine, I'll close it than