Re: xt + lotusxsl (fwd)

Subject: Re: xt + lotusxsl (fwd)
From: James Clark <jjc@xxxxxxxxxx>
Date: Mon, 1 Feb 1999 11:52:45 -0500 (EST)
Conflicts between xsl:id declarations are resolved based on the source
document. xsl:id is less restrictive than with XML ID attributes: in
particular it's not an error if the use of xsl:id leads to an element
having more than one ID. So with your example, the "c" element would
match both id('id15') and id('id16'). It is an error if the use of
xsl:id leads to two different elements having the same ID, and this is
resolved by either signaling an error or treating only the first of the
elements as having that ID. Thus in your example, the "a" element would
*not* match id('id16').

I expect xsl:id will change in a future draft.

> > (3) I am also confused by test no. 15.  The source document contains
> >
> > <c name="id15" id="id16"> *id15*
> > <a id="id16"> *id16*</a> ...
> >
> > The stylesheet declares the following:
> >
> > <xsl:id attribute="id"/>
> > <xsl:id attribute="name" element="c"/>
> >
> You forgot to include the xsl:id declarations from the stylesheet:
>   <xsl:id attribute="id"/>
>   <xsl:id attribute="name" element="c"/>
> > Shouldn't element c has both id15 and id16? Element a also has id16 then the
> > processor should report error or use element c. I got different result from
> > lotusxsl.
> Maybe, maybe not.  I guess I was thinking that <xsl:id attribute="name"
> element="c"/> was more important than <xsl:id attribute="id"/>, and therefore
> would override the ID specification for all "c" elements.  This is probably a
> question for James to see if I've got it right or not.

 XSL-List info and archive:

Current Thread