Add min/max function tests

[nt1m]

Test List

https://wpt.fyi/results/css/css-values?label=master&label=experimental&product=chrome&product=firefox&product=safari&aligned&view=interop&q=minmax

Rationale

Some differences in serialization across browsers

[nt1m]

@emilio @tiaanl Can you help review for Gecko?

@xiaochengh / @nairnandu Can you help review for Chromium?

[emilio]

+1 from me

[xiaochengh]

I'm wondering about css/css-values/minmax-percentage-serialize.html, which currently asserts that percentage comparisons are not simplified at parse time.

The spec however seems a bit vague/inconsistent about this.

At one point, it says:

Internal representations of math functions are eagerly simplified to the extent possible

At another point, as a non-normative note, it says:

If a percentage is left at this point, it will usually block simplification of the node, since it needs to be resolved against another value using information not currently available... This includes operations such as "min", since percentages might resolve against a negative basis

However, I find this note not very convincing, because % resolved against a negative value is a very rare case that's currently possible only in the background-position property. So percentage comparisons actually almost never blocks the simplification.

[xiaochengh]

Other tests LGTM

[nt1m]

I was actually wondering the same thing looking at css/css-values/minmax-percentage-serialize.html the other day

[nairnandu]

@nt1m closing this one as the change is merged.

[nt1m]

I'm reopening to discuss css/css-values/minmax-percentage-serialize.html

[Loirooriol]

Would be good to have some test for w3c/csswg-drafts#9631