fix: correctly extract baseline support for CSS functions

[xbinaryx]

Prerequisites checklist

• I have read the contributing guidelines.

AI acknowledgment

• I did not use AI to generate this PR.
• (If the above is not checked) I have reviewed the AI-generated content before submitting.

What did you do?

a {
	background-image: image("foo.png", red);
}

What did you expect to happen?

use-baseline should report image() as a non-baseline CSS function.

What actually happened?

No lint error was reported for image().

What is the purpose of this pull request?

This PR fixes CSS function baseline extraction in the generator.

The previous logic would treat CSS data types as CSS functions when reading css.types.* compat keys. In practice, that meant image() could fail to report correctly because css.types.image (<image>) was being conflated with the image() function.

What changes did you make? (Give an overview)

Refactored function extraction so CSS functions are handled explicitly, regenerated baseline-data.js, and added regression coverage.

Related Issues

Is there anything you'd like reviewers to focus on?

https://developer.mozilla.org/en-US/docs/Web/CSS/Reference/Values/image/image

[DMartens]

Thank you for spotting this.
In the future please split up the PR in the fix and refactor as combining them makes it harder to review.
How did you determine that only image and color are affected?
For example I tried sibling-index which is also non-baseline but use-baseline does not report it.

[xbinaryx]

In the future please split up the PR in the fix and refactor as combining them makes it harder to review.

Yeah sorry about that. I had to refactor first to avoid making the code uglier with this fix.

How did you determine that only image and color are affected?

They’re the direct bare css.types.<name> collisions I identified.

For example I tried sibling-index which is also non-baseline but use-baseline does not report it.

sibling-index() is part of a broader gap here, same as sibling-count(), shape(), progress(), if(), etc. Those exist in web-features, but not in mdn-data.css.functions, so use-baseline doesn’t currently recognize/report them.

This should be addressed by the mdn-data -> @webref/css migration in #376, which I’m working on.

[nzakas]

LGTM. Would like @DMartens to verify before merging.

[DMartens]

Changes LGTM, thanks.