W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > August 2001

Re: rdf:parseType="Literal" and XML Fragment interchange

From: Dan Connolly <connolly@w3.org>
Date: Thu, 23 Aug 2001 11:56:27 -0500
Message-ID: <3B8535BB.D3A5B962@w3.org>
To: Paul Grosso <pgrosso@arbortext.com>
CC: www-xml-fragment-comments@w3.org, w3c-rdfcore-wg@w3.org
Paul Grosso wrote:
> 
> At 11:15 2001 08 23 -0500, Dan Connolly wrote:
> >Paul Grosso wrote:
> >[...]
> >>   <p:body>
> >>   <dc:Title>...</dc:Title>
> >>   </p:body>
> >> </p:package>
> >
> >[...]
> >
> >> In the case that you have no fragment context information
> >
> >If that can capture xml:lang and/or namespace bindings, we are.
> 
> I see, yes, it can.
> 
> If you had any element context of interest, you could include
> those elements in the fcs and put the lang and ns attrs on
> those elements.  However, given that you have no elements of
> interest to put into the fcs, you can put such attributes on
> the f:fcs tag itself or even on the p:body tag.
> 
> >> to interchange
> >> and your fragment body is already single rooted
> >
> >no, I'm asking to put the *content* of the element, not the
> >element itself, in a fragment thingy:
> 
> Okay, then that is what the p:body element gives you.  You
> can make just the content of the dc:Title element (without
> the dc:Title tags) the content of the p:body element.

Like this? (let's pretend there was an xml:lang="en"
on the original rdf:Description element...)

<p:package xmlns:p="http://www.w3.org/2001/02/xml-package"
           xmlns:f="http://www.w3.org/2001/02/xml-fragment"
           xmlns:dc="http://purl.org/metadata/dublin_core#"
           xmlns="http://www.w3.org/TR/REC-mathml">
  <f:fcs>
    <f:fragbody/>
  </f:fcs>
  <p:body xml:lang="en">Ramifications of
       <apply>
      <power/>
      <apply>
        <plus/>
        <ci>a</ci>
        <ci>b</ci>
      </apply>
      <cn>2</cn>
    </apply>
    to World Peace...</p:body>
</p:package>


Hmm... it seems there are a number of degrees of freedom
in how to construct a fragment interchange package
to express what we're interested in. I think we'd
need to choose a canonical form: exactly where
to put the namespace declarations and xml:lang, etc.
We'd probably have to be careful about whitespace too.

Ugh... we'd have to be careful about how we choose
the f: and p: prefixes too, since they might clash.
We could move them down to the p:body element, but
that still leaves a possibility of clashing there.

It might be simpler to just use an rdf:value element
as the wrapper in stead, since, as you say, we're
not really using the interesting part of the
fragment interchange spec, which is the f:fcs
bit.

-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
Received on Thursday, 23 August 2001 12:56:32 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:38:48 EDT