Re: [xsl] RE: Designs for XSLT functions

[David dot Rosenborg at pantor dot com]

Hi Joe,

I agree fully with you here. An apply() function is inherently easier
to implement than an eval () in general. In my XSLT implementation the
stylesheet gets compiled into a program where all the information about
names of variables are gone. However, top level objects like templates
and extension functions are accessible by name at runtime so an apply
function would be a walk in the park. I'm not saying eval () isn't
doable, it defentiely is, it just isn't as trivial as apply () is.

And the idea to mix apply () and eval () behaviour seems like an overkill.
If you need to dynamically specify pairs of paramter names and values on a
lexical level, I think the way to go would be a plain exsl:eval (string)
function.

Cheers,

</David>

David Rosenborg
Pantor Engineering AB


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