[tei-council] pdf version of P5
James Cummings
James.Cummings at oucs.ox.ac.uk
Wed Nov 14 18:22:31 EST 2007
Gabriel Bodard wrote:
> Um. I guess each attribute would be listed, alongside a link to whatever
> class or element (s) contain the definition of it. Some will belong to
> more than one, I guess.
>
> e.g.
>
> n =>class:global
> target =>element:certainty (etc.)
> scribe =>class:att.handFeatures
>
> etc.
>
> Is that not what people were imagining an index of attributes to look like?
I guess when you suggested it I was thinking more of a reverse attribute
reference than a simple index per se. So for each attribute one would list a
set of information concerning it, similar to what appears in the attribute
classes reference. so:
n
classes: att.global
desc: its desc here
status: optional
Datatype: 1 - infinity occurences of: data.word separated by whitespace
note: notes here
target
classes: att.whatever
elements: individual elements with same(?) attribute
desc: ...
status: ...
Datatype: ...
note: ...
My worry with envisioning that, however, is that it might be a) impossible,
especially when we have the same named attributes on different elements and they
have different dataypes. (i.e. we have provided a closed valList for one and not
for another). (No examples spring to mind, but I'm sure compiling such a list
would find them.)
and b) If it is just a join table of this attribute name is used in this class
or on these elements, I worry about how much use that will be. The only use, I
guess for it is when I'm looking at @ref on persName and wondering "what other
elements have a @ref attribute?" ... The question is whether I really should be
wondering that or "what other elements have *this* @ref attribute?" (i.e. are
members of att.naming, and the Appendix B should already answer that.)
So I originally misunderstood what you meant and thought "A great idea, but what
is it feasible", but now am not entirely convinced the simpler one buys the
reader much more (but I'm willing to be convinced).
-James
--
Dr James Cummings, Oxford Text Archive, University of Oxford
James dot Cummings at oucs dot ox dot ac dot uk
More information about the tei-council
mailing list