Re: Entities and XPointers

Peter Murray-Rust (peter@ursus.demon.co.uk)
Wed, 03 Dec 1997 00:26:45

At 16:35 02/12/97 -0500, David G. Durand wrote:
>On Dec 2, 5:40pm, Simon St.Laurent wrote:
>> Subject: Entities and XPointers
>> While they don't provide the actuation flexibility or many of the other
>> features of XML-Link, it may be possible to create external entities
that use
>> XPointers in the URL. Of course, this would require that either the
>processing
>> application can cope with XPointers (unlikely in this case), or that the
>> server can interpret the XPointer and return only the chunk requested.
>
>There's no reason that you can't do this at the server side, and a client
that
>was so sophisticated could interpret the URL (but, pracitcally speaking, I
>wouldn't expect many such to exist).

I am probably missing something, but it seems fairly straightforward to
extract something from another document - the question is whether it's
allowed. For example,

<A HREF="foo.xml#DESCENDANT(3,CHAP)..DESCENDANT(4,CHAP)">
or
<A HREF="foo.xml#DESCENDANT(3,CHAP)">

could return a chunk of well-formed XML. (JUMBO is capable of the second
form at present). The question is whether

<!ENTITY chap3 "foo.xml#DESCENDANT(3,CHAP)">
...
&chap3;

is legal in an XML parser. I suspect that this is undefined - however it
must not be 'application-dependent', because otherwise we get different
parser behaviour. (The use of other connectors (| and ?) is presumably
similar - it's the mechanics of how the entity is retrieved.)
>
The only argument I can see against this is that it requires all parser
writers who cope with ENTITYs to resolve XLL - and that is quite a strong
argument :-)

P.

Peter Murray-Rust, Director Virtual School of Molecular Sciences, domestic
net connection
VSMS http://www.nottingham.ac.uk/vsms, Virtual Hyperglossary
http://www.venus.co.uk/vhg