- From: Simon Pieters <simonp@opera.com>
- Date: Thu, 25 Apr 2013 14:01:05 +0200
- To: "John Daggett" <jdaggett@mozilla.com>, "Tab Atkins Jr." <jackalmage@gmail.com>
- Cc: "www-style list" <www-style@w3.org>
On Wed, 27 Feb 2013 23:15:29 +0100, Tab Atkins Jr. <jackalmage@gmail.com>
wrote:
> On Wed, Feb 27, 2013 at 5:58 AM, John Daggett <jdaggett@mozilla.com>
> wrote:
>> As was decided at the last F2F, cascading variable properties are
>> referenced via a case-sensitive ident. However, the definition of
>> object model methods referred to in the CSS Variables spec such as
>> setProperty, lookup property values with a case *insensitive* lookup.
>> That needs to be tweaked for the variables case.
>>
>> In other words, the two lines below return different values:
>>
>> a = styleObj.getPropertyValue("var-foo");
>> b = styleObj.getPropertyValue("var-FOO");
>>
>> I think the object model spec just needs to include steps for special
>> handling properties that begin with "var-".
>
> Ah, good catch. Yes, this is something for CSSOM to define, probably.
> I can patch it *in* Variables, but it would be better done in CSSOM,
> I think.
>
> Alternately, the get/set/etc algorithms can define a flag that
> invoking specs can set for CI/CS behavior, and then I can just have
> Variables set the flag to CS. That way future specs that need CS
> matching for whatever reason can specify this without having to get
> CSSOM to update. I'd prefer this.
I made CSSOM dispatch on whether a property is defined to be
case-sensitive or case-insensitive.
https://dvcs.w3.org/hg/csswg/rev/121de33edc34
https://dvcs.w3.org/hg/csswg/rev/8b19669fd05f
--
Simon Pieters
Opera Software
Received on Thursday, 25 April 2013 12:01:43 UTC