Subject: Re: Semantics of deep-equal function
From: David Carlisle <davidc@xxxxxxxxx>
Date: Wed, 20 Jul 2005 10:12:08 +0100
|
The results are as expected. It seems deep-equal considers white
spaces significant. Is there any way I can do deep equals comparison
ignoring the white spaces?
The WG have had many comments on the official commentlist to teh efect
taht deep-equal is essentially useless (and has been inconsistently, and
differently, defined in every draft but the current one, as far as I can
rememeber). the problem is that in any real case people don't want a
fixed deep-equals they need to tune whetherwhite space or comments or
attribute or whatever are considered significant.
That said, you have a few choices.
a)
add xsl:strip-space elements="*" so that much of the white space in your
input is ignored
or
b)
before using deep-equal process the nodes (into a variable) with a mode
that normalizes space (ie do a modified identity trasform with a special
template matching text().
or
c)
use a collation that considers white space insignificant. (I'm not sure
what collations saxon currently offers)
or
d)
ignore deep-equal and write your own recursive function (this is likely
to be more efficient than (b) but less efficient than eitherr (a) or (c).
David
________________________________________________________________________
This e-mail has been scanned for all viruses by Star. The
service is powered by MessageLabs. For more information on a proactive
anti-virus service working around the clock, around the globe, visit:
http://www.star.net.uk
________________________________________________________________________
| Current Thread |
|
Michael Kay - 20 Jul 2005 16:49:33 -0000
Message not available
|
|