- From: Simon Sapin <simon.sapin@exyr.org>
- Date: Tue, 27 Aug 2013 10:37:35 +0100
- To: www-style@w3.org
Le 27/08/2013 08:52, Cameron McCormack a écrit :
> And a related question: should we preserve the fact that a variable
> value might have two adjacent white space tokens? For example:
>
> p {
> var-a: ;
> var-b: var(a);
> }
>
> Do we need to ensure that the computed value of var-b is reported as "
> /**/ "?
This would be necessary to round-trip the exact token stream, but in
this case I think it is not necessary (as all CSS grammars do and should
treat 2+ whitespace tokens as identical to one) and not desirable (it
would add noise.)
I added a note in Syntax to clarify:
http://dev.w3.org/csswg/css-syntax/#serialization
--
Simon Sapin
Received on Tuesday, 27 August 2013 09:37:58 UTC