Re: Why is xml:lang not allowed on typed literals?

On Thu, 15 Oct 2009, Bernard Vatant wrote:

> Hi all
> This list does not seem very active, but hopefully someone is still
> monitoring it and will be able to answer
> In
> I
> read
> Plain literals have a lexical
> form<>and
> optionally a language
> tag as defined by
> [RFC-3066<>],
> normalized to lowercase.
> Typed literals have a lexical
> form<>and
> a datatype
> URI being an RDF URI
> reference<>
> .
> Between the lines I read that the language tag xml:lang is not allowed on
> typed literals. Actually I just tried to do this. The rationale is to define
> a datatype "One Sentence" which must contain a single sentence, starting
> with a upper-case, ending with a dot etc ... and using this datatype for a
> "tagLine" property - which of course has also a language.
> So I tried the syntax below and proposed it to various tools
> - W3C validator validates it, seems to ignore the xml:lang tag
> - Prot?g? does the same, imports the file and ignores the xml:lang tag when
> saving
> - SWOOP does the other way round, ignores the rdf:datatype but keeps the
> language tag.
> My question is, just out of curiosity, what is the rationale behind not
> allowing xml:lang on typed literals?
> Thanks for any clue

I believe the rationale was along the lines that if the value of a typed 
literal was represented by an XML construction, the xml:lang belonged _in_ 
the representation, not _on_ it. That is, that if a literal's values are 
represented in infoset terms, the xlm:lang belongs in the representation.

jan grant, ISYS, University of Bristol.
Tel +44 (0)117 3317661
Usenet: The separation of content AND presentation - simultaneously.

Received on Thursday, 15 October 2009 11:13:03 UTC