W3C home > Mailing lists > Public > xsl-editors@w3.org > July to September 2000

Re: Support of XPointer by XPath/XSLT

From: Eric van der Vlist <vdv@dyomedea.com>
Date: Wed, 13 Sep 2000 10:10:51 +0200
Message-ID: <39BF368B.3C0F9B70@dyomedea.com>
To: James Clark <jjc@jclark.com>
CC: xsl-editors@w3.org

James Clark wrote:
> document() already supports XPointer:
> "If the URI reference does contain a fragment identifier, the function
> returns a node-set containing the nodes in the tree identified by the
> fragment identifier of the URI reference. The semantics of the
> fragment identifier are dependent on the media type of the result of
> retrieving the URI."
> The spec for text/xml says fragment identifiers use XPointer.
> XSLT implementations are allowed but not required to support XPointer:
> "An XSLT processor is not
> required to support any particular media types.  The documentation for
> an XSLT processor should specify for which media types the XSLT
> processor supports fragment identifiers."

Ooops, focused on the specific syntax for XPointer, I had forgotten they
were part of a bigger picture and coherent with the text/xml media type

Thanks for the clarification.

If I understand all the implications correctly, this means that a XSLT
processor supporting the text/xml media type should also support

This support would indirectly give to the users of such a processor the
ability to use variable XPath expressions over any text/xml document
including the source document through writing :


where the variable value can have the form "#xpointer(XPath

This is a feature which has been requested many times and many XSLT
users should be happy if it happens.

> > However, the fact that a XPointer expression equal to "#id" is pointing
> > to the node whose ID is "id" in the current document while document("")
> > is pointing to the current stylesheet could, IMHO, make it quite
> > confusing if we said that the current document() function was to be used
> > with XPointers as an argument.
> Huh?

I was just meaning that if document("#id") is identifying the fragment
named "id" in the source document, one could expect document("") itself
to identify the source document instead of the current stylesheet...

Anyway, it's clearly documented and the document itself (or at least its
document element) can be referenced through document("#/1").



> James

Eric van der Vlist       Dyomedea                    http://dyomedea.com
http://xmlfr.org         http://4xt.org              http://ducotede.com
Received on Wednesday, 13 September 2000 04:09:50 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:44:20 UTC