W3C home > Mailing lists > Public > www-rdf-comments@w3.org > July to September 2003

Re: Test case regarding XML Literals and octets

From: Graham Klyne <GK-lists@ninebynine.org>
Date: Thu, 31 Jul 2003 21:42:09 +0100
Message-Id: <>
To: Martin Duerst <duerst@w3.org>, pat hayes <phayes@ihmc.us>
Cc: "Peter F. Patel-Schneider" <pfps@research.bell-labs.com>, www-rdf-comments@w3.org, w3c-i18n-ig@w3.org, msm@w3.org, w3c-rdf-core-wg@w3.org


as far as I can tell, you're contradicting the XML canonicalization spec.

Is canonical XML a sequence of octets or something else?

The XML canonicalization spec, I understand, says it's a sequence of octets.

Maybe, you want to say it's a sequence of octets that is to be interpreted 
in specific way, in which case it's not *just* a sequence of octets, but 
must also carry some distinguishing datum that indicates that this special 
processing is required.

Specifically, if I have the values denoted by:

    <eg:bar rdf:parseType="Literal"><br/></eg:bar>


    <eg:bar rdf:datatype="http://www.w3.org/2001/XMLSchema#hexBinary"

what is it that tells me the first is to be treated as markup, but not the 


At 14:58 31/07/03 -0400, Martin Duerst wrote:
>At 09:50 03/07/29 +0100, Graham Klyne wrote:
>>At 00:46 29/07/03 -0500, pat hayes wrote:
>>>>Are 'binary octets' different from 'octets'?
>>>I have absolutely no idea. :-)
>>Anyway, returning to the original question (Are 'binary octets' different 
>>from 'octets'?), I think the answer is:  not for any meaningful purpose 
>>as far as RDF is concerned.
>Here is a test case that I am proposing to get clarity on this.
>I guess this test would be classified as a datatype-aware entailment
>Do the following two RDF fragments entail each other?
><rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
>          xmlns:eg="http://example.org/">
>  <rdf:Description rdf:about="http://example.org/foo">
>    <eg:bar rdf:parseType="Literal">XML</eg:bar>
>  </rdf:Description>
><rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
>          xmlns:eg="http://example.org/">
>  <rdf:Description rdf:about="http://example.org/foo">
>    <eg:bar rdf:datatype="http://www.w3.org/2001/XMLSchema#hexBinary"
>        >584D4C</eg:bar>
>  </rdf:Description>
>Comment: "584D4C" is the hexBinary representation of "XML" encoded
>in UTF-8 (which for this case is the same as ASCII). The current
>(post-lastcall) RDF spec says that XML fragments denote their
>exclusive canonicalization, a sequence of octets after encoding
>with UTF-8. The value space of hexBinary is sequences of (binary)
>octets (http://www.w3.org/TR/xmlschema-2/#hexBinary).
>This seems to indicate that the current spec says that this test
>is true (positive entailment test). However, I think equating these
>conceptually very different things (XML complex types and a specific
>simple type) is highly problematic. I propose that this test be
>added to the negative entailment tests (with a corresponding one
>making the same statement with regards to parseType="Literal"
>and base64Binary, which needs a bit more work for the base64
>calculation), and that the spec be changed if necessary to make
>this clear.
>Regards,    Martin.

Graham Klyne  <GK@NineByNine.net>
Nine by Nine
Received on Thursday, 31 July 2003 16:45:35 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:15:21 UTC