[tei-council] more on constraint

Sebastian Rahtz sebastian.rahtz at oucs.ox.ac.uk
Tue Apr 21 17:44:51 EDT 2009


I think I am mostly agreeing with Syd now, with one big difference - I
think we should forget about the alternative modelling of the
same constraint in different languages. How can we be sure
the two alternatives do the same thing? why not just make all
constraints additive? I also think grouping is a luxury we can
add later

so my idea is now settled on

   * no container list element, no functional grouping
   * zero or more <constraint> elements to follow <content>
   * <content> is syntactic sugar for <constraint scheme="relaxng">
   * model of <constraint> is
       @scheme,
       att.typed,
       att.identifiable,
       macro.glossLike,
       data
  * Content model of <data> is macro.anyXML.
  * normal I18N for <desc> and <gloss> (ie repeatable with multiple 
languages)
  * I18N method for <data> is undefined, scheme-dependent
  * semantics are that
     - application of <content> is mandatory
     - application of all <constraints> is optional,
       but do as many as you can. the more you do,
       the closer you'll get to conformance to the
       abstract model. constraints are there to help you,
       but do not fully test the abstract model.
     - all constraints are additive. if you choose
       to write the same rule in two languages, good for you -
       a processor should do as many as it can.

example:

<constraint schema="isoschematron" ident="c1" type="accessibility">
   <desc> stuff</desc>
   <desc xml:lang="fr">choses</desc>
   <data>
     <s:rule>..... </s:rule>
   </data>
</constraint>


note  the type attribute which you may choose to use
and implement how you like.

The bit I am not happy with is the name "<data>". Any other suggestions?

Would the above make most people more or less happy? could you
live without
    a) grouping
    b) alternatives in different languages
?

Sorry for going round and round in circles.

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

Sólo le pido a Dios
que el futuro no me sea indiferente


More information about the tei-council mailing list