Re: Grouping problem

This seems reasonable, except a stronger condition is possible, as a
for-each with sort children only is in the same position.

However, in Saxon's case, the warning might not be true, as Saxon
allows extension functions to have side effects, so if one were
present in the select attribute, it might indeed have an effect.

Gestalt doesn't allow extension functions to have side effects though,
so I will put this on my todo list.

Presumably a similar warning for xsl:template and xsl:for-each-group
would be equally desirable. Any others?

On 04/12/2007, Andrew Welch <andrew.j.welch@xxxxxxxxx> wrote:
> On 04/12/2007, Colin Adams <colinpauladams@xxxxxxxxxxxxxx> wrote:
> > Andrew,
> >
> > what warning would you hope to see?
>
> Something similar to what Saxon says:
>
> "A variable with no following sibling instructions has no effect"
>
> when you have:
>
>   <xsl:variable name="foo">abc</xsl:variable>
> </xsl:template>
>
> I can't think of a situation off the top of my head when you'd want an
> empty for-each:
>
> <xsl:for-each select="something"/>
>
> ...so it would always be a typo?
>
> Maybe:
>
> "A for-each with no child instructions has no effect"
>
>
> --
> Andrew Welch
> http://andrewjwelch.com
> Kernow: http://kernowforsaxon.sf.net/