- From: Dave Beckett <dave.beckett@bristol.ac.uk>
- Date: Mon, 11 Mar 2002 14:50:35 +0000
- To: Jeremy Carroll <jjc@hplb.hpl.hp.com>
- cc: w3c-rdfcore-wg <w3c-rdfcore-wg@w3.org>
>>>Jeremy Carroll said:
> >
> > Can't be (a) since the "" content is just a sequence of characters.
>
> Oh, that's interesting. I had been assuming that it was well-balanced xml.
No, that's why I posted the proposed new productions to make it quite
clear:
langString ::= '"' string '"' ('-' language)
xmlString ::= 'xml' langString
and in the existing grammar at http://www.w3.org/TR/rdf-testcases/#string
string ::= character* with escapes as defined in section Strings
> So for you
>
> xml"<"
>
> is a plausible value. (Not representable in RDF/XML)?
Legal - yes, this changed NTriples doesn't care. However: the
NTriples that are generated from RDF/XML will *not* generate such a
literal.
If we went with
xmlString ::= 'xml"' xmlExclusiveC14Nstring '"' ('-' language)
xmlExclusiveC14Nstring ::= as defined by http://www.w3.org/TR/xml-exc-c14n
then there would be further requirements
> That's fine.
> If that's the case then:
>
> > b) legal but different from xml("<b>foo</b>") (and not representable in
> > RDF/XML?)
>
> looks like the attractive answer to my question.
It is an answer from my current defn :)
Dave
Received on Monday, 11 March 2002 09:53:47 UTC