Re: string-handling functions

This is a response to the comments in
http://lists.w3.org/Archives/Public/public-qt-comments/2004Jan/0037.html
which followed on from
http://lists.w3.org/Archives/Public/public-qt-comments/2004Jan/0020.html

This was considered by the joint working groups yesterday, along with
comment

 http://lists.w3.org/Archives/Public/public-qt-comments/2004Feb/0979.html

from your colleague Henry Zongaro.

We decided that we should clarify the specification of upper-case() and
lower-case() to make the reference to Unicode algorithms absolutely clear,
and to clarify that the case mapping was a locale-independent one. There are
use cases for providing a locale-dependent mapping as well, but we felt we
could leave these to be handled by extension functions if the market needs
them.

Suggestion 2 in your comment
http://lists.w3.org/Archives/Public/public-qt-comments/2004Jan/0020.html was
not actually considered in yesterday's meeting, and I believe this one needs
to stay open. However, I think the correct answer is that characters such as
" " or "-" which are ignored in a collation do not produce any collation
units, and they are therefore ignored by the substring-before function. I
think we have already agreed to rewrite substring-before in terms of the
mapping from characters to collation units, and this should remove the
ambiguity you identify.

Michael Kay

Received on Wednesday, 17 March 2004 09:39:42 UTC