Quoting Mike Taylor <[log in to unmask]>:
> I do, but I am still sure this isn't correct. Otherwise the XHTML document:
>
> <xhtml>
> <p>Hello</p>
> <p>Goodbye</p>
> </xhtml>
>
> Could be legitimately rendered as:
>
> Goodbye
>
> Hello
- Firstly, document representation and rendering are two different things and
should NEVER been confused with one another (that's the damnation of HTML
and its still broken descendants XHTML)
- Secondly, XHTML is not XML.
- In XML its the processing that applies interpretation.
- The XHTML standard DOES appeal to a "document order". Not all XML is XHTML.
- The XML specification (have a close look) does not demand that element order
be anything specific. A specification (or parser) can use their (its) own
model to present siblings and "do the right thing".
- Many XML processing specifications DO appeal to document order--- many
derived from the Infoset.
- They are not always the same as some specifications define their own data
models with their own set of rules for siblings.
- That XML does not specify the order was INTENT. This allows us to write
specifications that can indeed require the order to matter and it to be
alphabetical or geospatial or .. as per our needs and wants.
In my football example (which I choose for YOU)
<item>Manchester United</item>
<item>Arsenal</item>
<item>Tottenham Hotspur</item>
<item>Liverpool</item>
<item>Aston Villa</item>
can indeed well be reported in the order in which they won Premiership titles
or alphabetically or .. If we don't set a specification then it could literally
be (according to the XML) anything. Each processing system can do what they
please or think is right (even if most will indeed return the siblings in
document order). Its specifications for the processing layer (what to do) that
may hammer things down.
Of course beyond all this we need to grasp that SRU/W is not just about XML
but is (or should be) intended for generic search inclusive of systems that
don't even have a (persistent) document model or a hierarchical model (trees,
forests of trees or even loops) that is not consistent with XML (even on a
primitive level, see my talk of our concept of lines and sentences as an
example of overlapping content objects).
--
--
Edward C. Zimmermann, Basis Systeme netzwerk, Munich
Office Leo (R&D):
Leopoldstrasse 53-55, D-80802 Munich,
Federal Republic of Germany
http://www.nonmonotonic.net
|