W3C home > Mailing lists > Public > public-css-testsuite@w3.org > October 2010

Re: first-letter-non-punctuation-001 invalid

From: Geoffrey Sneddon <gsneddon@opera.com>
Date: Sat, 16 Oct 2010 20:17:55 +0100
Message-ID: <4CB9FA63.9060700@opera.com>
To: "L. David Baron" <dbaron@dbaron.org>
CC: public-css-testsuite@w3.org
On 15/10/10 20:21, L. David Baron wrote:
> http://test.csswg.org/suites/css2.1/20101001/html4/first-letter-non-punctuation-001.htm
> http://test.csswg.org/suites/css2.1/20101001/xhtml1/first-letter-non-punctuation-001.xht
>
> This test is invalid, I believe, since it is testing that
> :first-letter applies to a character that:
>   * is not a letter or a digit
>   * is not in one of the punctuation classes through which
>     :first-letter is extended
>
> The relevant section of the spec,
> http://www.w3.org/TR/CSS21/selector.html#first-letter  , says that
> :first-letter applies to the first letter or digit, and is also
> extended through punctuation.
>
> However, Gecko does not apply it to other types of characters, such
> as symbols (character classes S*) and the types of punctuation
> through which it is not extended (Pd, in this case).  This appears
> to be the behavior the spec calls for, which leads me to conclude
> that this test is invalid, and should actually be testing that the
> :first-letter does not apply to any characters at all (if that's the
> correct interpretation of the spec, which, I admit, is a little
> uncclear).

I agree that under Unicode 5.1 (and above) that test is invalid (and 
:first-letter should apply to nothing); however, the testsuite doesn't 
define what version of Unicode it expects to be supported (which has 
implications for first-letter-* tests as well as others), which it 
probably should. (Otherwise, I'd expect it to test Unicode 4.0, as the 
spec requires, excluding anything that has changed in any future version 
of Unicode.)

-- 
Geoffrey Sneddon  Opera Software
<http://gsnedders.com>
<http://opera.com>
Received on Saturday, 16 October 2010 19:18:36 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 16 October 2010 19:18:42 GMT