Subject: Re: [xsl] Testing if all child nodes match (based on the value of certain elements) From: David Carlisle <davidc@xxxxxxxxx> Date: Mon, 23 Apr 2007 22:26:30 +0100 |
> Did you never use something like: > > somenode[position() != 1] Yes, but (some of the time) I know what I'm doing. The semantics of != when both operands are atomic values, as in your example, are useful, in fact they are the same as not(position() = 1) so the only advantage to != is that it saves 4 characters. If one operand is atomic and the other is a node set, != is less useful, and if they are both node sets then != is almost never useful (this thread being the exception). 99 times out of a hundred when a beginner uses != the stylesheet does the wrong thing and not(...=...) was intended. Rather than try to remember the circumstances in which it's safe to use != I think it's simpler just to always use not (..= ..) (although, as you suggest above, I don't always follow my own advice). David
Current Thread |
---|
|
<- Previous | Index | Next -> |
---|---|---|
Re: [xsl] Testing if all child node, Abel Braaksma | Thread | Re: [xsl] Testing if all child node, Abel Braaksma |
RE: [xsl] Testing if all child node, Michael Kay | Date | Re: [xsl] XSLT Text Processing: Fun, Dimitre Novatchev |
Month |