RE: Whitespaces against efficency

[Kay Michael <Michael dot Kay at icl dot com>]

> Using Intuitive's OptimizeIt, I saw that the next() method of
> com.jclark.xsl.dom.SiblingNodeIterator is very time consuming.
> I guess that's because my DOM trees have a lot of unuseful (except for
> indenting purposes) whitespaces text nodes.

I seem to recall that when James Clark announced DOM support in the latest
version of xt, he said it was very slow compared with using xt's native tree
format. Presumably this is one of the reasons. XPath's tree model is not the
same as the DOM model, so if the XSLT processor uses a DOM internally it is
going to have to do more work at run time. As a worst case (let me
speculate) the code needs to decide whether a text node is a non-preserved
whitespace text node on each occasion that the node is visited.

Mike Kay


 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list