Re: [xsl] question about identity transform

Subject: Re: [xsl] question about identity transform
From: Abel Braaksma <>
Date: Wed, 01 Nov 2006 00:46:34 +0100
Dimitre Novatchev wrote:
I think this has been discussed quite often in this list and should
probably be well-reflected in the FAQ.

A node-set is a set.

A set, by definition, is unordered.

Indeed. And I know of the theory. But I find myself every now and then back on slippery ground, especially when it comes to explaining something of what I remember to have had difficulties with in the past, now finding myself making the same mistakes again. And on the same every now and then, when I am coding my stylesheet crafting new xpaths, it so often looks like it works one way ( A | B) but not the other way ( B | A ), eventually finding out that there's a glitch in the expression making it so (and I never seem to learn: keep those expressions less than a couple of lines in size!). Just being a human being, it keeps being hard, even after some time of xslt programming, unraveling the mixture of ordered XML with unordered (well, partially) XSLT, resulting, in - hey - an ordered result set.

But, I must admit, it's a joy to mind and brains to do it, and any XSLT stylesheet (esp. 2.0), once finished, sure looks like a piece of art when you do it well ;-)

(Bill, sorry for being off-topic)

-- Abel

Current Thread