[osis-core] osisCore_Candidate.1.1_006 - 2 - extra attributes on elements by way of globalAttributes

[Todd Tillinghast]

Patrick,

Several elements have attributes that they don't need by way of
globalAttributes?

We need to remove globalAttributes from several elements and/or create
two attribute groups out of globalAttributes and apply the only the
needed attribute groups to elements.

Todd


> -----Original Message-----
> From: owner-osis-core@bibletechnologieswg.org [mailto:owner-osis-
> core@bibletechnologieswg.org] On Behalf Of Patrick Durusau
> Sent: Friday, August 30, 2002 5:36 AM
> To: osis-core
> Subject: [osis-core] osisCore_Candidate.1.1_006!
> 
> Greetings,
> 
> Hopefully much improved and clearer schema!
> 
> Changes:
> 
> 1. Dropped osisWork from globalAttributes (error noted by
> Harry/HotMetal, skipped by XMLSpy)
> 
> 2. Dropped refSystem and refSystems
> 
> 3. Added refSystem as optional element in <work>
> 
> 4. Now have elements back in header (header is optional on osisCorpus,
> required in osisText)
> 
> 5. Default osisWork set to Bible.
> 
> 6. Only <work> appears in the header, occurs a minimum of once but is
> unbounded. Has the requested <refSystem> element.
> 
> Question: Do I need to rework the regex to allow the wildcard
character
> "*" in the osisRef? (suspect I do since it only allows letters/numbers
> and underscore
> 
> Question: Do I need to drop the "[ ]" in the namespace prefix for
> osisWork in osisRef? (This was a reference system specifier. That is
how
> I interpret Harry's post but want to be sure.)
> 
> Note on the question of the "bohu" attribute (I had to search for the
> explanation):
> 
> > The set of Dublin Core elements allows you to bind the meaning of
> > "bible.nrsva" or whatever, as closely or as loosely as you desire to
> > the content of those elements. Note that these elements have a
"bohu"
> > attribute which defaults to "true." This is so we can distinguish
> > between simple omission and "I don't care" about language for
example.
> > The bohu attribute forces the default to "I don't care" if not
> > otherwise specified.
> 
> 7. I have dropped the bohu attribute (temporarily), at least for
> everyone to consider along with Harry's reference syntax. It looks
like
> overkill at this point.
> 
> 8. Note that the osisWork attribute now defaults to Bible but is
> optional (this is required by XML Schema since the default value has
> been fixed, the user has to have the option to not include it in their
> document. Still means it occurs in any case, but you have the default,
> Bible or you can insert your own.)
> 
> Note that the IBM SQC 2.0 reports nesting of divs, title, q, and title
> as "ambiguous content models" and says they violate the contraints
part
> of Part 1 Chapter 3. I have read that section but got bogged down in
the
> syntax of the restrictions.
> 
> Steve, the content models look only possibly ambiguous to me (if you
did
> not close a nested div for example). Comments?
> 
> Well, all for now.
> 
> I can try to post 007 early this evening if I get some feedback on the
> two questions I posed above. That will give us all day tomorrow to
work
> on documentation and bang on it some more.
> 
> Thanks!
> 
> Patrick
> 
> 
> --
> Patrick Durusau
> Director of Research and Development
> Society of Biblical Literature
> pdurusau@emory.edu