W3C home > Mailing lists > Public > public-xml-id@w3.org > April 2004

Re: XPath 2 without WXS (XSD)

From: Tobias Reif <tobiasreif@pinkjuice.com>
Date: Sun, 18 Apr 2004 15:49:29 +0200
To: Daniel Veillard <veillard@redhat.com>
Cc: public-xml-id@w3.org, Norman.Walsh@Sun.COM, jmarsh@microsoft.com, jim.melton@acm.org, ashokma@microsoft.com
Message-ID: <20040418134929.GA3087@linux>

[thread:
http://lists.w3.org/Archives/Public/public-xml-id/2004Apr/thread.html#18
http://lists.w3.org/Archives/Public/public-xml-id/2004Apr/0018.html
]

On Sun 2004-04-18 Daniel Veillard wrote:
[...]
>   [2] states
>   "those elements that are in the same document as the context node,
>   and that have an ID value equal to one or more of the IDREFs in
>   the list"
> 
> How is "element having an ID value defined ?

The XPath 2 guys might answer that.

> If the definition is that if hosts an attribute with an [attribute
> type] property of type ID or a [type definition] property of type
> ID, then there is no change needed.

So in the scenario I describe, id('foo') would return the element with
xml:id="foo"? That would be great (especially since I think the main
advantage of xml:id will be to have IDs without relying on any schema
code (neither external nor internal) or any specific schema language).

> If they define IDness in a different way

I don't really know, in this specific case I'm merely a user
describing a scenario and an expectation.

> then yes that would not work and require coordination.

If coordination is required then last call seems to be a good period
for you xml:id guys to work this out with the XPath/XSLT people.

> Since I expect the XPAth2 data model to be based on the Infoset,

I thought it's different (XPath 1 was diff IIRC), but can't remember;
and don't forget that at least one type of conforming XSLT 2 processor
is not schema-aware.

>   http://www.w3.org/TR/xpath-datamodel/
>  section 6.3.3.  lists the [attribute type] property and the ID
>  value but I could not find any definition of "element ID value" in
>  the datamodel spec, it is unclear what the id() function really
>  references.

If that's the case then additional specification and clarification is
needed as well.

I think it's reasonable to expect id() to return elements with
matching xml:id values (as it does for attributes which are defined
(eg in an internal DTD subset) to be IDs), this should get specified
somewhere.

Tobi

-- 
Vim users, don't forget to
   http://iccf-holland.org/donate.html
Received on Sunday, 18 April 2004 09:47:08 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 19:53:49 UTC