|
Subject: Re: [xsl] Multiple elements condition From: Florent Georges <lists@xxxxxxxxxxxx> Date: Fri, 31 Aug 2007 11:11:42 +0200 (CEST) |
David Carlisle wrote:
Hi
> > exists(para[2]) and not(exists(para[3]))
> > or
> > exists(para[2]) and empty(para[3])
> they both always look for the 2nd and 3rd element but you
> don't need t0 look for the 3rd if the 2nd isn't there....
I am not sure what you mean by /always/. I expect my
processor to use short-cut boolean connectors. And if I
understand correctly net.sf.saxon.expr.BooleanExpression,
Saxon does so.
On the other hand, it seems (if I understand correctly the
code) that the optimizer will rewrite:
xxx and my:user-function(yyy)
into:
if (xxx) then my:user-function(yyy) else false()
I don't understand why this rewrite is intended to, except
for optimization, but if we have short-cut evaluation yet I
don't see the point here...
I should have missed something.
Regards,
--drkm
_____________________________________________________________________________
Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers Yahoo! Mail
| Current Thread |
|---|
|
| <- Previous | Index | Next -> |
|---|---|---|
| Re: [xsl] Multiple elements conditi, David Carlisle | Thread | RE: [xsl] Multiple elements conditi, Michael Kay |
| Re: [xsl] Multiple elements conditi, Abel Braaksma | Date | Re: [xsl] Multiple elements conditi, Andrew Welch |
| Month |