Add ignoreAtRules option to no-invalid-position-at-import

[sylwia-lask]

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

#5492

Is there anything in the PR that needs further explanation?

No, it's self-explanatory.

[jeddy3]

@sylwia-lask Thanks for another pull request!

I've requested some changes (as suggestions, so you can commit them via the GitHub interface).

They align the docs and tests with our conventions, mainly using standard CSS and consistent names. There's a lot of rules in Stylelint, so we do our best to keep things consistent to help our users move through all the documentation.

[jeddy3]

I'm going to create a follow-up issue to update all our custom at-rule, function etc documentation to use dashed idents to be more aligned with the specs (see example 7), where it advocates for @--custom rather than @custom.

[sylwia-lask]

@sylwia-lask Thanks for another pull request!

I've requested some changes (as suggestions, so you can commit them via the GitHub interface).

They align the docs and tests with our conventions, mainly using standard CSS and consistent names. There's a lot of rules in Stylelint, so we do our best to keep things consistent to help our users move through all the documentation.

@jeddy3 thanks a million for the code review :) I fixed everything, hope now everything is ok.

[jeddy3]

thanks a million for the code review

No worries. Thanks for making the changes and for the pull request!

I've requested just two more changes, then I think we're good.

[jeddy3]

Shall we move this into the conditional below to keep everything together, i.e.:

if (
	(node.type === 'atrule' &&
		nodeName !== 'charset' &&
		nodeName !== 'import' &&
		!optionsMatches(options, 'ignoreAtRules', node.name) &&
		isStandardSyntaxAtRule(node)) ||
	(node.type === 'rule' && isStandardSyntaxRule(node))
) {
	invalidPosition = true;

	return;
}

You're right to use node.name rather than nodeName as the ignore* options are case sensitive, and people can use a regex with the i flag if they want case insensitivity.

---

Incidentally, I wonder whether this rule can be refactored to avoid walking every node. Instead, walking just import at rules, and whether that'll affect performance or not. One for another time, though.

[sylwia-lask]

True, maybe it's worth to refactor that rule. For now I've changed it as you suggest - put every contition together

[jeddy3]

LGTM, thanks!

[ybiquitous]

LGTM, thank you! 👍🏼

[jeddy3]

• Added: ignoreAtRules: [] to no-invalid-position-at-import (#5520).