- From: Christoph Päper <christoph.paeper@crissov.de>
- Date: Thu, 15 Dec 2011 17:15:12 +0100
- To: www-style list <www-style@w3.org>
Håkon Wium Lie: > > Yes, I think we can find a simple generic solution that can handle the > documented use cases [1]. Some of them require digraph support, e.g. romanized Greek: ‘θ’ → ‘th’. You cannot do that (in any sane way) without a separator, whether it’s a space or something else. It might be possible to introduce one later, however, because in @text-transform foo { convert: "a b c" to "d e f"} you would have specified a mapping from ‘ ’ to ‘ ’ in the basic level, twice though. Dupicate conversions are not covered yet, but I assume the last would win, i.e. ‘f’ in the following example @text-transform foo { convert: "aaa" to "def"} > @text-transform german-lowercase { convert: "ẞ" to "ß" } I’m still a fan of extensible keywords, i.e. @text-transform lowercase { convert: "ẞ" to "ß" } By the way, does any CSS module have a normative note to the effect that UAs should copy and paste its styles only into rich-text environments, not into plain-text ones? That means <foo style="text-transform: lowercase"> B<baz style="display: none"> </baz>A/R </foo> should be pasted as “B A/R”, not “ba/r” into form fields and the like. I think most browsers are broken in this regard, especially concerning the two properties I used in the example. > @text-transform foo { convert: "abc" to "def", "ghi" to "jkl" } I’ve suggested, for a later level, to optionally specify the scope, like so: @text-transform foo { convert: all "abc" to "def", initial "abc" to "DEF"; } That could of course be done with ‘convert’ becoming a “shorthand descriptor” in a later level: @text-transform foo { convert: "abc" to "def"; /* CSS3 only handles ‘all’ */ convert-initial: "abc" to "DEF"; /* CSS4, specific override */ } > @text-transform foo { convert: "a-c" to "d-f" } /* ranges, … */ I agree that this, like <urange>, is better than ‘convert-range’. Nevertheless, ranges won’t be all that helpful in many scenarios, because they’re mostly arbitrary after all.
Received on Thursday, 15 December 2011 16:17:44 UTC