RE: [xsl] Getting the Base Character of Character with Diacritic

Subject: RE: [xsl] Getting the Base Character of Character with Diacritic
From: "Michael Kay" <mike@xxxxxxxxxxxx>
Date: Tue, 19 Sep 2006 09:02:41 +0100
> Colin Adams wrote:
> > as the  C forms consist of de-composition followed by composition.
> > So it would be strange to implement NFC, and not implement NFD.
> 
> I see. I never looked at it that way. I always thought that 
> de-composition was the hard way, and never realized that you 
> actually needed to de-compose before you can compose. But 
> come to think of it, it sounds rather logical.
> 
> Hmm, that sounds counter-intuitive towards the fact that the 
> W3C committee for XPath decided that only NFC be mandatory. 

The fact that the specification describes the process in this way doesn't
mean the implementation has to work this way.

And remember (as would-be contributors to Saxon often seem to forget) that
testing is more work than coding.

Michael Kay
http://www.saxonica.com/

Current Thread