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

Re: xml:id questions from Dave Pawson

From: Norman Walsh <Norman.Walsh@Sun.COM>
Date: Tue, 26 Oct 2004 12:28:29 -0400
To: Paul Grosso <pgrosso@arbortext.com>
Cc: public-xml-id@w3.org, "Pawson, David" <David.Pawson@rnib.org.uk>
Message-id: <87acu9xx3m.fsf@nwalsh.com>
/ "Paul Grosso" <pgrosso@arbortext.com> was heard to say:
| Question. Can I have attributes of type ID (xml 1.1) and 
| an attribute xml:id as defined?
| xs:ID also mentioned.  What happens when a mix of these 'collide'
| e.g. the values are not unique to the document?

DTD validation and XML Schema validation are independent. If you have
attributes of type ID defined in your DTD and other attributes defined
in your schema and you apply both kinds of processing, your xml:id
processor is probably going to look at only one or the other, not
both. Though I suppose they could look at both.

Implementations SHOULD detect the case where xml:id attribute values
are not unique. That is an xml:id error.

Does that help?

| In the above scenario, the mention (1. Introduction) of 'all 
| conformant processors' begs the question, conformant to which rec?

Conformant to xml 1.[01].

| States 'This specification has been designed to be separate layer in
| processing and to be compatible with existing validation technologies.'
| Is the implication that DTD validation, XSD schema validation 
| and xml:id processing to be considered separately?

They can be considered seperately. The xml:id spec says exactly how
xml:id processing works in valide, partially valid, and only
well-formed instances.

| I'm concerned about a DTD based instance with an idref
| to an element with xml:id attribute?

If you're using DTD validation, then you must define xml:id attributes
as being of type "ID" so you'd expect IDREF to work naturally.

| Clarification please:
| 'Many validation technologies impose the constraint that an XML element
| can have at most one attribute of type ID. That constraint is not
| imposed by xml:id processing.'
| <element id="n1" xml:id="m1" another="p1"/>
|  where id and another are of type ID can now be valid to a DTD?

No. But if they can be valid in your validation technology, xml:id
will still work.

| 'There are four classes of documents in which an xml:id attribute may
| occur: well-formed, non-validated documents, wholly valid documents,
| partially valid documents, and invalid documents.'
| Are these four terms defined anywhere please Paul? They are
| used but not defined in this document.

It isn't clear from 5.1, 5.2, 5.3, and 5.4? Sigh.

| Appendix D is helpful, just wondering if there might
| be any common sense advice on transitioning from DTD use
| with id's to DTD's with xml:id, e.g the id idref pairs
| mentioned above.

Change all occurrences of 



  xml:id ID #IMPLIED


                                        Be seeing you,

Norman.Walsh@Sun.COM / XML Standards Architect / Sun Microsystems, Inc.
NOTICE: This email message is for the sole use of the intended
recipient(s) and may contain confidential and privileged information.
Any unauthorized review, use, disclosure or distribution is prohibited.
If you are not the intended recipient, please contact the sender by
reply email and destroy all copies of the original message.

Received on Tuesday, 26 October 2004 16:28:33 UTC

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