W3C home > Mailing lists > Public > public-i18n-core@w3.org > January to March 2009

Re: [CSS21][css3-namespace][css3-page][css3-selectors][css3-content] Unicode Normalization

From: Zack Weinberg <zweinberg@mozilla.com>
Date: Thu, 29 Jan 2009 21:29:01 -0800
To: "L. David Baron" <dbaron@dbaron.org>
Cc: public-i18n-core@w3.org, www-style@w3.org
Message-ID: <20090129212901.0aec27cd@mrtock>

"L. David Baron" <dbaron@dbaron.org> wrote:
> On Thursday 2009-01-29 16:15 -0800, fantasai wrote:
> >   page names
> >   counter names
> >   namespace prefixes
> 
> These three are all analogous to variable names in a computer
> program.  Do we really need to apply Unicode normalization to them?

I would say that we do.  As Jonathan Kew pointed out upthread, whether
or not a particular identifier is normalized (and which normalization
form it's in) can be completely invisible to the user; it depends on
things like which text editor was last used to modify the style sheet.
It would be very surprising if such invisible transformations on the
style sheet or the HTML changed its meaning.

> Are there other languages that do similar?

C99 and C++98 place restrictions on which Unicode characters can appear
in identifiers, that have the effect of forcing them to be NFC. I dimly
recall hearing that some revision of the Fortran standard did or would
require NFKC as well as case folding.  

zw
Received on Friday, 30 January 2009 05:29:50 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 30 January 2009 05:29:51 GMT