ISSUE-54: Which URIs should replace null relative URIs provided in LDPR representations?

NullURIReplacement

Which URIs should replace null relative URIs provided in LDPR representations?

State:
CLOSED
Product:
Linked Data Platform Spec
Raised by:
David Wood
Opened on:
2013-03-15
Description:
This issue was raised by James Leigh on the public-ldp mailing list:

http://www.w3.org/TR/2013/WD-ldp-20130307/#http-post-1 says:
5.4.8 In RDF representations, LDPC servers must interpret the
null relative URI for the subject of triples in the LDPR
representation in the request entity body as referring to the
entity in the request body. Commonly, that entity is the model
for the “to be created” LDPR, so triples whose subject is the
null relative URI will usually result in triples in the created
resource whose subject is the created resource.

According to the above the term <> in turtle should be replaced with the
to-be-created URI. However, the term <#adr> would still be resolved
against the base URI of the document (either in the @base directive,
Content-Location, or the request-uri). This will be hard to implement as
most Turtle parsers do not expose the relative lexical term used in the
document, but often only the absolute URI.

In Callimachus we experimented with overriding the base URI while
parsing, but that proved problematic as many turtle writers don't allow
explicit term representations and it prevented the use of general
purpose entity handling (on either client or server). In the end we
realized that by overriding the base URI we were essentially /forking/
Turtle and only parsers/writers that were aware of this could be trusted
preserve the null relative URI.

Callimachus now requires the client to create a URI and use it in the
RDF document. However, the server may end up substituting the primary
URI with a canonical variant. I suggest the LDP spec adopt a similar
approach.

Henry Story proposed to resolve this issue thusly:

Proposal:
A LDPC receiving a POST of an RDF document D MUST resolve all relative URIs in D using the URI of the created resource.

This is implied by the current spec, but is not spelled out clearly enough. The current spec text makes it look like arbitrary decisions on relative URIs are being made.
Related Actions Items:
Related emails:
  1. LDP Rec (from eric@w3.org on 2015-02-20)
  2. Re: Specification edits for Issue-54 resolution (from sspeiche@gmail.com on 2013-05-23)
  3. Specification edits for Issue-54 resolution (from nmihindu@fi.upm.es on 2013-05-23)
  4. Re: Section 5.4.8 null relative URI (from henry.story@bblfish.net on 2013-03-15)
  5. Re: Section 5.4.8 null relative URI (from david@3roundstones.com on 2013-03-15)
  6. ldp-ISSUE-54 (NullURIReplacement): Which URIs should replace null relative URIs provided in LDPR representations? [Linked Data Platform core] (from sysbot+tracker@w3.org on 2013-03-15)

Related notes:

Resolution: close ISSUE-54 changing the first sentence of 5.4.8 to "In RDF representations, the base URI for RFC3987 relative URI resolution is the IRI of the created resource."
See http://www.w3.org/2012/ldp/meeting/2013-04-01#resolution_3

Arnaud Le Hors, 1 Apr 2013, 17:25:06

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: 54.html,v 1.1 2015/08/17 04:43:10 denis Exp $