[osis-core] Proposed works syntax

[Steve DeRose]

At 03:26 PM -0400 07/03/02, Harry Plantinga wrote:
>  > I think I'd suggest a work name be like a namespace prefix: can
>>  contain the usual XML name stuff (at least; but no colon) -- and is
>>  sep. from the ref by a colon.
>>
>>  Should we make it a validity requirement that all works cited be
>>  declared in this thing up top, or should we be loose and not
>>  *require* it?
>
>I don't think you mean validity in the XML sense -- I don't see how you
>could check this proposed requirement with an XML parser. Are you
>envisioning another, separate parsing stage or program that would make
>additional checks on the document to see if it is "osis-valid" or some such?

I think you could, actually. I think schema (though not DTD, if 
that's what you were getting at perhaps) could check that all later 
values of 'work' attributes lead to an existing header object with 
that value on the right attribute. Certainly possible in Schematron. 
Oh, wherever the work is stuck in as a prefix, you'd have to parse to 
the colon and then do the same thing with the value. I'm not fluent 
enough in schema to be sure, but I think that should be possible.

Once we have a schema for declaring ref systems and their ranges and 
such, then I'm pretty sure we're in territory that couldn't be done 
via off the shelf schema languages. Like if we declare that there are 
150 Psalms (I wonder if it's significant that that's also how many 
Pokemon there are?), getting schema to read the file declaring that, 
and use it to parameterize validating another document seems 
infeasible. So for that, i think we would need some special app.

-- 

Steve DeRose -- http://www.stg.brown.edu/~sjd
Chair, Bible Technologies Group -- http://www.bibletechnologies.net
Email: sderose@speakeasy.net
Backup email: sderose@mac.com, sjd@stg.brown.edu