RE: Re: . in for

> The picture that Mike paints here is of an XSLT that is basically a
> language for defining functions, defining sort keys, defining decimal
> formats, defining variables and so on. XPath becomes the main driving
> force of the transformation, with XSLT simply backing it up.

I'd characterize it differently. XSLT declarations are used for defining
things as you describe; XPath expressions are used for computing information
based on the contents of the source tree(s); XSLT instructions are used for
creating nodes in the result tree, with the contents of those nodes being
computed using XPath expressions. If we get the design right, I don't think
it should ever be necessary to use an XSLT instruction unless you are trying
to write something to the result tree.

I take your point about the limitations of named sort keys. To solve that I
think we would have to allow named sort keys to be declared locally within a
template, so that the sort expression could refer to variables.

Mike Kay


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