ISSUE-20: XMLLiteral content isn't processed for RDFa attributes in RDFa 1.0 - should this change in RDFa 1.1?

Deep Processing of XMLLiterals

XMLLiteral content isn't processed for RDFa attributes in RDFa 1.0 - should this change in RDFa 1.1?

State:
CLOSED
Product:
RDFa 1.1 Core
Raised by:
Manu Sporny
Opened on:
2010-03-15
Description:
The issue was re-raised in the RDFa WG here:

http://lists.w3.org/Archives/Public/public-rdfa-wg/2010Mar/0092.html

When an XMLLiteral is generated, in RDFa 1.0 descendant elements are
skipped for parsing. In RDFa 1.1, should we require descendant elements
to be parsed? Should we keep the RDFa 1.0 behaviour, or should we
provide a mechanism for page authors to decide?

Related Actions Items:
No related actions
Related emails:
  1. RDFa WG telecon minutes for 2010-10-07 (from msporny@digitalbazaar.com on 2010-10-07)
  2. Re: Telecon Agenda - October 7th 2010, 1400 UTC (from mark.birbeck@webbackplane.com on 2010-10-07)
  3. Telecon Agenda - October 7th 2010, 1400 UTC (from msporny@digitalbazaar.com on 2010-10-04)
  4. Re: Comment on RDFa 1.1 Core: Nested RDFa markup (from msporny@digitalbazaar.com on 2010-08-05)
  5. RDFa WG telecon minutes for 2010-07-15 (from msporny@digitalbazaar.com on 2010-07-15)
  6. Re: Telecon Agenda - July 15th 2010, 1400 UTC (from benjamin.adrian@dfki.de on 2010-07-15)
  7. Telecon Agenda - July 15th 2010, 1400 UTC (from msporny@digitalbazaar.com on 2010-07-13)
  8. Telecon Agenda - July 8th 2010, 1400 UTC (from msporny@digitalbazaar.com on 2010-07-05)
  9. Re: Telecon Agenda - July 1st 2010, 1400 UTC (from mark.birbeck@webbackplane.com on 2010-07-01)
  10. Re: Telecon Agenda - July 1st 2010, 1400 UTC (from benjamin.adrian@dfki.de on 2010-07-01)
  11. Telecon Agenda - July 1st 2010, 1400 UTC (from msporny@digitalbazaar.com on 2010-06-28)
  12. Re: Telecon Agenda - June 24th 2010, 1400 UTC (from mark.birbeck@webbackplane.com on 2010-06-24)
  13. Telecon Agenda - June 24th 2010, 1400 UTC (from msporny@digitalbazaar.com on 2010-06-24)
  14. Re: Telecon Agenda - June 17th 2010, 1400 UTC (from knud.moeller@deri.org on 2010-06-16)
  15. Re: Telecon Agenda - June 17th 2010, 1400 UTC (from ivan@w3.org on 2010-06-14)
  16. Telecon Agenda - June 17th 2010, 1400 UTC (from msporny@digitalbazaar.com on 2010-06-13)
  17. Re: Telecon Agenda - June 10th 2010, 1400 UTC (from mark.birbeck@webbackplane.com on 2010-06-10)
  18. Re: Telecon Agenda - June 10th 2010, 1400 UTC (from shane@aptest.com on 2010-06-09)
  19. Telecon Agenda - June 10th 2010, 1400 UTC (from msporny@digitalbazaar.com on 2010-06-09)
  20. Re: Does RDFa processing recurse into content:encoded? (from scorlosquet@gmail.com on 2010-03-15)
  21. ISSUE-20 (Deep Processing of XMLLiterals): XMLLiteral content isn't processed for RDFa attributes in RDFa 1.0 - should this change in RDFa 1.1? [RDFa 1.1 Core] (from sysbot+tracker@w3.org on 2010-03-15)
  22. Re: Does RDFa processing recurse into content:encoded? (from msporny@digitalbazaar.com on 2010-03-14)

Related notes:

From Richard Cyganiak:

This is a comment on Profiles in the latest RDFa WD [1]. It concerns the processing of RDFa markup that is nested inside an element that carries @property.

My understanding is that the latest draft prescribes the following behaviour: Nested RDFa markup inside the element is ignored and will not produce any triples if the element carries @datatype="rdf:XMLLiteral". Otherwise, any nested RDFa markup is processed as usual.

This behaviour causes a problem in the following situation. On web pages that contain blog posts or other “prose content”, it is often tempting to mark up the content with @property="ex:content" or something similar, so that the actual text content is accessible to processors. (Cf. SIOC, AtomOWL and hAtom.)

For the content to be useful, it has to be presented including HTML markup, so we can assume that @datatype="rdf:XMLLiteral" will and should be used in these situations.

This means however that any RDFa markup that may be present *inside* the content of the blog post or article will not be processed.

The effect is that implementers of content management systems, blogging software and similar tools who want to deploy RDFa have to make a choice between the following two options:

a) Make the entire content available to RDFa processors as an rdf:XMLLiteral, but authors using the software cannot add their own RDFa markup inside their posts.
b) Authors can use RDFa markup inside their posts, but the system does not make the entire content available to RDFa processors (or only as an unformatted string, which isn't very useful).

This is not a satisfactory choice. I believe that RDFa should provide a mechanism for capturing the content of an element as an rdf:XMLLiteral, while still processing markup inside that content so that it generates triples.

(Offhand I can't think of a reason why RDFa processing is suppressed at all inside XMLLiterals.)

Manu Sporny, 5 Aug 2010, 13:32:51

Recurse flag has been removed from the RDFa Processing rules - RDFa now supports deep processing of XMLLiterals:

http://www.w3.org/2010/02/rdfa/meetings/2010-10-07#resolution_1

Manu Sporny, 11 Oct 2010, 01:58:21

Display change log ATOM feed


Chair, Staff Contact
Tracker: documentation, (configuration for this group), originally developed by Dean Jackson, is developed and maintained by the Systems Team <w3t-sys@w3.org>.
$Id: 20.html,v 1.1 2015/03/27 14:12:22 vivien Exp $