Re: XML literals

----- Original Message ----- 
From: "ext Martin Duerst" <duerst@w3.org>
To: <Patrick.Stickler@nokia.com>
Cc: <w3c-rdfcore-wg@w3.org>
Sent: 07 August, 2003 23:51
Subject: RE: XML literals


> At 14:34 03/08/05 +0300, Patrick.Stickler@nokia.com wrote:
> 
> >What exactly are you proposing we say the value space of XML literals 
> >contains?
> 
> Here is my list of preferences (best first):
> 
> 1) XML Literals are sequences of characters and markup (so that
>     XML Literals not containing markup are the same as corresponding
>     plain literals.

How about my recent proposal:

1. The lexical space of rdfs:XMLLiteral consists of canonicalized
   XML fragments encoded as octet sequences (as presently defined
   by the RDF specs).

2. The value space of rdfs:XMLLiteral consists of XML Infosets
   which can be serialized as canonicalized XML fragments encoded
   as octet sequences.

3. There is a 1:1 correspondence between the lexical and value spaces.
   A given lexical form always denotes the same Infoset and a given
   Infoset has a single canonicalized lexical form (this needs to
   be checked, but I think it's doable).

4. The equality test for members of the value space corresponds to
   the literal octet sequence comparison of their lexical forms.

Thus, the graph contains canonicalized octet sequences as
the lexical forms of rdfs:XMLLiteral typed literals.

The values denoted by those rdfs:XMLLiteral typed literals
are Infosets, where the semantics of the XML markup is
richly, and consistently, provided.

-- 

I believe this would satisfy your #1 preference above
while still allowing for a little bit of wriggle room when
clarifying the relationships between rdfs:XMLLiteral's and
XSD types (simple or complex), plain literals, etc.

Eh?

Patrick

Received on Friday, 8 August 2003 09:20:55 UTC