xml-dev - Re: Relational Tables and structured documents

["Michael Kay" <M.H.Kay@eng.icl.co.uk>]

Peter Murray-Rust:

>I have been spending the last two weeks working on a molecular application
>which essentially consists of relational tables.  The application is
>largely hierarchical (a protein molecule) so that it benefits from being
>recast into structured document form. I have therefore found it useful to
>create routines which generate nodes in a tree as a result of joining
tables.


I think there are several things that one could usefully do.

(1) define a recommended way of representing a relational table in XML.
(There
are a number of ways of doing this, the biggest decision is whether to use a
standard DTD for all tables or a DTD that reflects the specific table
definition.)

(2) extend this to a richer data model, e.g. the nested relational model,
that
allows you to represent hierarchic structures, or the ODMG model which
allows
to to represent arbitrary graphs. (Note that the problem then becomes
analogous to the one of using XML as a serialisation format for CORBA
objects)

(3) provide a toolkit that allows you to perform transformations on these
XML
documents that reflect the operators of the relevant data model, e.g.
relational
selection, projection, and join; sorting and grouping; flattening of nested
relations.
Of course one way to implement such operations would be to load the data
into
a database, but many of the operations can be implemented directly without
too
much effort.

I think the key thing is to keep the set of structures and operations very
clean:
base it on a data model with an established formal basis rather than
inventing
something new.

Mike Kay, ICL


xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev@ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/
To (un)subscribe, mailto:majordomo@ic.ac.uk the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo@ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa@ic.ac.uk)