Ignore length-zero-no-unit violations found within var() functions if custom properties are ignored

[lanthaler]

Which issue, if any, is this issue related to?

Related to #4945, but limited to the var() function for now

Is there anything in the PR that needs further explanation?

Should be self-explanatory. More than happy to answer questions if there are any

[jeddy3]

@lanthaler Thanks for the pull request.

Two related pull requests were recently merged (#5203 & #5245) which I think will address the problem most users are facing with this rule, i.e. using var within calc or some other math function.

However, users may wish to ignore the contents of functions outright, whether that's var or something else. The convention we use is an ignore* secondary option, e.g. ignoreFunctions.

Rather than overloading the custom-properties keyword option, this option will allow users to make the rule more permissive on a granular scale.

Is it possible to rework this pull request to #4945 (comment)?

[jeddy3]

I hadn't realised the rule was a mesh of style-search and the value parser. I've created an issue to refactor the rule to only use the value parser, which will make adding an ignoreFunction option easier.

[lanthaler]

Thanks for the review @jeddy3. I just discovered that there's already a lingering PR to add ignoreFunction #5082 and that you started refactoring this code altogether #5256. If this change isn't of interest, let's just close this PR. The reason I "overloaded" the custom-properties option is because the var function is exclusively about custom properties.

[jeddy3]

If this change isn't of interest, let's just close this PR.

Yes, let's do that. With the recently merged #5203 & #5245 (which are in the refactoring of #5256), I think the most common problem of using var within calc is solved.

Thanks again for the pull request.