[css-cascade][css-transition] Should property aliasing/mapping applies to transition[-property]?

[cdoublev]

I would like to know in which cases a property name should be replaced in the parsing/serialization of transition-property. It does not seem clearly defined in CSS Cascade.

style.transitionProperty = `
  word-wrap,                /* Unprefixed legacy name alias */
  grid-row-gap,             /* Unprefixed legacy shorthand */
  page-break-before,        /* Unprefixed legacy shorthand */
  -webkit-background-size,  /* Legacy name alias */
  -webkit-box-ordinal-group /* Legacy mapping property */
`
style.cssText;

The output in Chrome for style.cssText is transition-property: word-wrap, grid-row-gap, page-break-before, -webkit-background-size, -webkit-box-ordinal-group;.

The output in Firefox is transition-property: overflow-wrap, row-gap, page-break-before, background-size, -moz-box-ordinal-group;.

[Loirooriol]

transition-property just accepts a list of custom idents that may not be properties at all, so applying the mapping seems weird to me.
Just like specifying a shorthand will make the longhands be transitioned without expanding the shorthand in the value, specifying an alias could make the aliased property be transitioned without changing the value.

[cdoublev]

Simple aliases are discussed in #866.

For the alias part, it isn't clear to me that the spec text defines what's supposed to happen in APIs that return the name of the property, such as getting the value of the transition-property property.

Hm, indeed. We should make that clear.

It is clear. The spec says

When the old property’s syntax is identical to or a subset of the value space of the new property’s syntax, the two names are aliased with an operation on par with case-mapping.

[Loirooriol]

Well, I don't see https://drafts.csswg.org/css-transitions-1/#transition-property-property saying that idents should be normalized to lowercase (this should probably be filed as another issue). In fact it doesn't seem much consistent:

document.body.style.transitionProperty = "lOreMipsUm, cOloR";
document.body.style.transitionProperty; // Implementations say "lOreMipsUm, color"

Note that https://drafts.csswg.org/css-values-4/#identifier-value says

Such identifiers are fully case-sensitive

And even if transition-property interprets custom keywords case-insensitively, I still don't think think transition-property: word-wrap becoming transition-property: overflow-wrap follows from the text in CSS Cascade.

Otherwise, imagine we add a property foo with grammar bar | word-wrap. Should foo: word-wrap become foo: overflow-wrap just because foo: BaR becomes foo: bar? I don't think so.

[cdoublev]

I agree. I also do not think a legacy property name should be replaced in a feature query.

[Loirooriol]

Same question for will-change:

document.body.style.willChange = "word-wrap";
document.body.style.willChange;

Firefox and Blink: "word-wrap"
WebKit: "overflow-wrap"