[tei-council] resolving the Birnbaum Biznai

Sebastian Rahtz sebastian.rahtz at oucs.ox.ac.uk
Fri May 19 22:22:21 EDT 2006


You will recall that yesterday we debated how to resolve the fact that
David and Matthew wanted a specific set of elements to be allowed
in <msIdentifier>, in order, but optional; but that this was not
what they got if the elements were simply put into a model class.

I propose that the idea of a model class be extended to allow
an extra piece of information which says how it is to be implemented.
Currently, if a class foo has members a, b and c, it is instantiated
as:

  model.foo = a | b | c

and it is always used as:

  (model.foo)*

or

  (model.foo)+

I propose an attribute on <classSpec> which
will allow it to be additionally instantiated as either

  model.foo = a, b, c

or
 
  model.foo = a?, b?, c?

and then in real content models we could opt NOT to
have the surrounding (...)*.

I have implemented and tested this, using <msIdentifier>
and <physDesc> as a testcase.  I have thus been able to
refute the Cummings Claim (that msdescription
depends on namesdates) by making a new (ordered) class
containing settlement, region, and country.

If anyone sees a flaw in my thinking, or thinks this is a
bad idea, please speak up.

-- 
Sebastian Rahtz      

Information Manager, Oxford University Computing Services
13 Banbury Road, Oxford OX2 6NN. Phone +44 1865 283431

OSS Watch: JISC Open Source Advisory Service
http://www.oss-watch.ac.uk




More information about the tei-council mailing list