W3C home > Mailing lists > Public > public-rdfa-wg@w3.org > August 2011

Re: ISSUE-101 (xsd:string coercion): Should xsd:string values be coerced to plain literals? [LC Comment - RDFa Core 1.1]

From: Ivan Herman <ivan@w3.org>
Date: Thu, 11 Aug 2011 18:18:20 +0200
Message-Id: <B56579C2-EF67-4B79-A17C-1D43C86F9D0B@w3.org>
To: RDF Web Applications Working Group WG <public-rdfa-wg@w3.org>
First of all: this is relevant for plain literals _without language tags_. Plain literals with language tags are a different ballgame and it is not yet decided in the RDF WG how to handle them

The important reference is: 

http://www.w3.org/2011/rdf-wg/wiki/StringLiterals/AbolishUntaggedPlain

which will find its way to the new version of the RDF Concept document, to be publish soonish. Important points:

[[[
Proposal

1. Abolish plain literals without language tag from the abstract syntax

2. "foo" and corresponding forms in other concrete syntaxes are syntactic sugar for "foo"^^xsd:string. In general, both forms MAY be used and represent identical literals in the abstract syntax.

...

Serializing an RDF graph SHOULD use the plain literal syntax "foo" in preference to the "foo"^^xsd:string form.
]]]

RDFa is defined as a mapping from RDFa+HTML/XML/XHTML to an abstract RDF graph. If we take that as a basis, then all literals should be mapped onto Literal(val, datatype=xsd:string), or whatever the syntax of your particular implementation uses. 

The problem I see is that then if you _serialize_ this, then the serialization should, on the fly, turn all those literals into plain literals. And because RDF environments do not really do that yet, I know I would have to process all the Literals explicitly before letting rdflib's serialization work or I would generate xsd:string literals only, which would be ugly and bad.

Ie, I am not sure. We could:

1. define the mapping following the latest WG resolutions and say that all plain literals without language tags are mapped against xsd:string literals in the default graph
2. we could live it as is, and trust that later rdf environment would take care of this at some point, and this is not what the RDFa processor should care about

Note that we cannot have a normative reference to that new resolution yet, and it would not be finalized before (hopefully) RDFa 1.1 becomes final.

I am tempted to say: no, we should not change anything and the RDFa document should be silent of the issue.

Ivan

On Aug 11, 2011, at 15:10 , RDF Web Applications Working Group Issue Tracker wrote:

> 
> ISSUE-101 (xsd:string coercion): Should xsd:string values be coerced to plain literals? [LC Comment - RDFa Core 1.1]
> 
> http://www.w3.org/2010/02/rdfa/track/issues/101
> 
> Raised by: Manu Sporny
> On product: LC Comment - RDFa Core 1.1
> 
> RDF WG has just issued a decision on the long-standing xsd:string vs.
> plain literal debate:
> 
> http://www.w3.org/blog/SW/2011/06/29/reconciling_various_forms_of_string_lite
> 
> Should we update RDFa 1.1 Core Processing Sequence to ensure that
> anything typed as "xsd:string" generates a plain literal?
> 
> Raised in: 
> http://lists.w3.org/Archives/Public/public-rdfa-wg/2011Jun/0064.html
> 
> 
> 
> 


----
Ivan Herman, W3C Semantic Web Activity Lead
Home: http://www.w3.org/People/Ivan/
mobile: +31-641044153
PGP Key: http://www.ivan-herman.net/pgpkey.html
FOAF: http://www.ivan-herman.net/foaf.rdf
Received on Thursday, 11 August 2011 16:16:27 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:19:52 UTC