1058408 - support final OM interfaces to @font-face and @font-feature-values rules

Status: NEW

(Core :: CSS Parsing and Computation, defect)

Description

[John Daggett (:jtd)]

The final version of the OM interfaces for @font-face and @font-feature-values rules differ from the present Gecko interfaces:

http://www.w3.org/TR/css3-fonts/#object-model

@font-face: 

interface CSSFontFaceRule : CSSRule {
  attribute DOMString family;
  attribute DOMString src;
  attribute DOMString style;
  attribute DOMString weight;
  attribute DOMString stretch;
  attribute DOMString unicodeRange;
  attribute DOMString variant;
  attribute DOMString featureSettings;
}

@font-feature-values:

interface CSSFontFeatureValuesRule : CSSRule {
  attribute DOMString fontFamily;
  readonly attribute CSSFontFeatureValuesMap annotation;
  readonly attribute CSSFontFeatureValuesMap ornaments;
  readonly attribute CSSFontFeatureValuesMap stylistic;
  readonly attribute CSSFontFeatureValuesMap swash;
  readonly attribute CSSFontFeatureValuesMap characterVariant;
  readonly attribute CSSFontFeatureValuesMap styleset;
}

[MapClass(DOMString, sequence<unsigned long>)]
interface CSSFontFeatureValuesMap {
  void set(DOMString featureValueName,
           (unsigned long or sequence<unsigned long>) values);
}

The MapClass is a tabbism that got negative feedback from js folks so we probably need to convert that into what we actually want to support.

Comment 1

[John Daggett (:jtd)]

Tab notes that Chrome/Webkit still use the old style rule OM model from CSS2:

https://lists.w3.org/Archives/Public/www-style/2015Sep/0055.html

Comment 2

[John Daggett (:jtd)]

Test for simple @font-face rule OM modifications:
http://people.mozilla.org/~jdaggett/tests/fontfaceruleomtest.html

Comment 3

[John Daggett (:jtd)]

David, I'm going to go ahead and work on the implementation of this. This specifically means moving from the existing problematic "style rule within an @-rule" OM model for @font-face rules to something that roughly corresponds with the FontFace interface:

https://drafts.csswg.org/css-fonts/#cssfontfacerule
https://drafts.csswg.org/css-font-loading/#fontface-interface

This will eliminate the problem of having to add new @font-face descriptors to the CSS2Properties interface.

Just wanted to confirm that you are comfortable with this change. If you have any concern/issues, please note them.

Comment 4

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Definitely seems good to untie it from CSS2Properties.

I'm a little worried about the compatibility of getting rid of the .style indirection.  Might it be necessary to continue supporting .style (maybe even with .style being a pointer back to the same object) for compatibility?

Comment 5

[John Daggett (:jtd)]

Yeah, that's a good question. Modifying attributes seems to only be functional in Safari, both Chrome and Firefox don't update the font matching correctly. So I can't see authors really using this interface in its current state. I think this might be something to simply implement and see if anyone reports bugs on. If need be, we can add something to deal with the .style case.

Comment 6

[Cameron McCormack (:heycam)]

The CSSFontFaceRule object model changes seem to have been dropped from the spec, but there is still an interesting  CSSFontFeatureValuesRule interface in css-fonts-4.

Comment 7

[BugBot [:suhaib / :marco/ :calixte]]

The bug assignee didn't login in Bugzilla in the last 7 months, so the assignee is being reset.

Comment 8

[Dominik Röttsches (drott)]

Jonathan, are you planning to add this? I am working on font-variant-alternates in Chromium and it'd be useful to implement parsing tests based on this.

Comment 9

[Jonathan Kew [:jfkthame]]

(In reply to Dominik Röttsches (drott) from comment #8)

Jonathan, are you planning to add this? I am working on font-variant-alternates in Chromium and it'd be useful to implement parsing tests based on this.

In principle, I'd like us to do it; but I don't know of any current plans/timescale for when it might happen, sorry.

(I believe we do have some tests in place, but they test by comparing with a reference rendering, not via the OM.)

Comment 10

[Dominik Röttsches (drott)]

Related: I added a CSSOM test to WPT: https://wpt.fyi/results/css/cssom/CSSFontFeatureValuesRule.html

Comment 12

[Daniel Holbert [:dholbert]]

Here's a simple testcase from dupe bug 1906625, which shows that we just report undefined for various attributes that should in fact be present:
https://bug1906625.bmoattachments.org/attachment.cgi?id=9413136

(Chrome/Safari show the actual values of these attributes, taken from the @font-face {...} rule defined in CSS.)