Re: [xsl] XSLT 1.1 comments

[Francis Norton <francis@xxxxxxxxxxx>]

Michael Kay wrote:
> 
> > <saxon:function> looks just right, now I've looked it up.
> >
> > And I'm just hoping and guessing - again without research -
> > that implementing this feature is largely syntax-sugar - and if not,
> > let's make it optional.
> >
> It wasn't difficult to implement but it wasn't pure syntactic sugar either.
> There were three things that needed care: using positional arguments instead
> of named arguments in the call; preventing the function writing to the
> output tree; and ensuring that lazy evaluation of node-sets still worked in
> the case where a node-set was returned from a function. But apart from the
> questions of data typing mentioned earlier (and those are no worse than any
> other implications of data typing on XSLT), I don't think there are any
> reasons to regard the facility as being risky, or bad coding practice.
> 
How would you feel, with your author-of-saxon hat on, about defining a
mapping for xsl:script/@language="saxon:xslt"?

This would have the advantages of letting the portability enthusiasts
write platform-independent code for all engines that provide this
mapping (and I'm betting that most would), and of providing a
demonstration that this can be done for other languages, while not
giving the XSLT 2.0 implementation of XSLT extension functions any
offical baggage to handicap its design freedom.

How's that for an offer you can't refuse!

Francis.

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