Re: Designs for XSLT functions (Was: Re: [xsl] RE: syntax sugar for call-template)

[Jeni Tennison <mail at jenitennison dot com>]

Hi Uche,

> Speaking of gnarly, I'm assuming xsl:call-template can be used
> within exsl:function. If so, is it permissible for the exsl:result
> to be in the called template, or so on down the chain?

We could say that xsl:call-template is only allowed within
xsl:variable or xsl:params in exsl:function (i.e. isn't allowed as a
direct child of exsl:function, xsl:if, xsl:when, xsl:otherwise or
xsl:for-each).

We could then ban exsl:result from being instantiated within either
xsl:variable or xsl:param, which I think is a good move anyway. If we
do that, we've effectively banned exsl:result from called templates.

> I'd say "scary thought: no", but how to best restrict this?
> Structurally? ("Every exsl:function must have an exsl:result within
> the nodes in its descendant axis").

There's a good point there about what to do if no exsl:result element
is instantiated.  Returning an empty node set would at least stop
function calls from breaking if the result of the function was used as
a node set.  Or should it be an unrecoverable error if no exsl:result
is instantiated?

Cheers,

Jeni

---
Jeni Tennison
http://www.jenitennison.com/



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