W3C home > Mailing lists > Public > www-style@w3.org > December 2011

Re: [css3-text] Splitting CSS Text into Level 3 and Level 4

From: Brad Kemper <brad.kemper@gmail.com>
Date: Fri, 16 Dec 2011 08:41:38 -0800
Cc: MURATA Makoto <eb2m-mrt@asahi-net.or.jp>, www-style@w3.org
Message-Id: <9485B136-C377-4C4C-BD44-07624F368254@gmail.com>
To: Håkon Wium Lie <howcome@opera.com>

On Dec 14, 2011, at 6:03 AM, Håkon Wium Lie wrote:

> Also sprach MURATA Makoto:
> 
>> The smallest generic solution is one-to-one mapping of UCS code values.
>> I would be a small subset of your "convert".  I think that it would be
>> very appropriate as Level 1 of text transformation.
> 
> Yes, I think we can find a simple generic solution that can handle the
> documented use cases [1]. Personally, I would suggest that we do:
> 
>  @text-transform german-lowercase { convert: "ẞ" to "ß" }
>  @text-transform foo { convert: "abc" to "def" }
>  @text-transform foo { convert: "abc" to "def", "ghi" to "jkl" } /* multiple tranformations */
>  @text-transform foo { convert: "a-c" to "d-f" } /* ranges, use \- to represent the '-' character */

That's a good start... actually, nothing on the wiki looked too complicated to me feature-wise. But I like your simpler syntax.

I think for your german-lowercase you need a way to combine it with predefined lowercase. Rather than having a separate 'convert-predefined' though, I would rather see it worked into the "multiple transformations", like this:

@text-transform german-lowercase { convert: lowercase, "ẞ" to "ß" }

Then you could also use that syntax for a"convert previously-defined" effect too:

@text-transform latin-only-uppercase { convert-range: "a","z" to "A","Z"; }
@text-transform latin-german-lowercase { convert: latin-only-uppercase, "ẞ" to "ß" }

And lastly, space separating the characters into character groups seems like a good idea, even if it requires a little more effort in defining the spec (not that bad, I feel, Christopher already did good there, with his "greedy" matching in author order). It makes it a little more readable, allows conversions of "one to some"and "some to one" (the following example is from Christoph Päper's earlier e-mail), and would seem to make it easier to extend in future versions.

convert: "Th th Dh dh" to 
         "Þ  þ  Ð  ð";

All of that still seems very simple to spec and accomplish. I would also love to be able to describe even a super-basic version of title case, but maybe boundary-based text-transforms could wait for a future level, if it would slow down this level to include it.
Received on Friday, 16 December 2011 16:42:15 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:47 GMT