[osis-core] Schema sample attached

[Steve DeRose]

At 01:42 PM -0500 06/05/02, Todd Tillinghast wrote:
>Kees,
>
>I have not encoded anything other than the test Matt.13.1-Matt.13.23
>[TEV] (and it is not ready for public release). I am moving from TX to
>CO next week so it is unlikely that I will do anything that week.
>
>I can probably finish out a pristine Matt.13 the next week pending
>resolution of the current issues.
>
>It looks to me like there may be a considerable amount of effort needed
>to encode the whole Bible (or even Matt and Ps like we had for the
>pre-Rome schema).  The reason being that the overlapping hierarchies
>that had been handled by milestones will drive the addition of elements
>to replace the milestone pairs currently in encoded in older versions
>(XSEM and pre-Rome OSIS formats).  When the new elements are added there

Isn't this a fairly easy (well, as I wrote it I found I'd upgrade 
that to "slightly annoying but not terrible") algorithm if you're 
just cranking along in Perl or whatever? something like

if start-mileStone:
    close any open verse
    open the stated new verse as newVerseID
    pendingVerse = newVerseID
else if end-milestone:
     if pendingVerse = nil then scream
     else close the open verse and set
else if any other end-tag AND pendingVerse is set
    close the verse (leave pendingVerse set)
    close the thing you just hit the end of
else if any other start-tag AND pendingVerse is set
    open the thing you just hit
    (re-)open verse pendingVerse
    pendingVerse := nil
else if text AND pendingVerse is set
    (re-)open verse pendingVerse
    pendingVerse := nil
end

I think I missed a detail somewhere, but isn't that pretty close? The 
basic idea is to keep track of *both* whether the verse is 
syntactically open (that is, you've issued a start-tag but not an 
end-tag for it), and whether it is logically open (whether you're 
really done with it completely or not). Oh, there is the next/prev 
logic to add, but that's straightforward -- just keep a letter around 
and increment it each time you issue a re-open for a verse, and reset 
it to 'a' when you end a logical verse. Granted the conversion is 
several times harder than with milestones (which I hadn't 
anticipated), but several times a couple lines isn't so bad, perhaps?

We may want to publish the algorithm (once it's right) as a 
convenience for people). Would even make a nice TEI utility for 
rotating between milestone and segmented markup -- hmmm, how would it 
work in XSLT?

>will be more than element for each milestone pair in many cases.  There
>will be several different patterns that arise.  The obvious ones are
>verse and paragraph overlapping as well as quotes and blockQuotes
>overlapping just about everything.  But there will also be overlapping
>related to interactions between <div>, <p>, <q>, <blockQuote>, <verse>,
><list>/<item>, and <lineGroup>/<line>.  For each one the style sheet
>will have to be tweaked to make it produce the appropriate results. 
>
>However, I believe that we should not declare the schema robust enough
>to call final and release until we have encoded the entire Bible in two
>modern versions that INCLUDE the appropriate use of <p>, <q>,
><blockQuote>, <list>/<item>, and <lineGroup>/<line>.

That seems like an excellent idea.

>
>However, the KJV (versions with no paragraph marks) should be pretty
>easy to get going.


Can I have a nickel for every glitch?  :)

>
>Todd
>
>
>>  -----Original Message-----
>>  From: owner-osis-core@bibletechnologieswg.org [mailto:owner-osis-
>>  core@bibletechnologieswg.org] On Behalf Of Kees F. de Blois
>>  Sent: Wednesday, June 05, 2002 5:46 AM
>>  To: osis-core@bibletechnologieswg.org
>>  Subject: RE: [osis-core] Schema sample attached
>>
>>  Dear Patrick and other friends,
>>
>>  I am sending this from Nairobi, having just returned from meetings in
>>  Tanzania where e-mail access was a bit of a problem.
>>
>>  Thanks for all the good work you guys have been doing.
>>
>>  I need some advice. We want to start an internal e-mail discussion
>list
>>  in UBS on USFM and OSIS. The first thing I have promised is to send my
>>  colleagues the latest test-release of the schema and sample
>OSIS-encoded
>>  Scripture text, which will help people not too familiar with XML and
>>  OSIS-schema see how OSIS-encoding looks like. I noticed that you have
>>  done some trial encoding of Mat. 13, I suppose, to highlight some
>>  problems. Is there other material available that conforms to the
>latest
>>  schema, that I could use for my PR work? I could wait, of course,
>until
>>  we have at least a book or a complete discourse unit.
>>
>>  Please advise.
>>
>>  Greetings,
>>
>>  Kees
>>
>>  > -----Original Message-----
>>  > From: owner-osis-core@bibletechnologieswg.org
>>  > [mailto:owner-osis-core@bibletechnologieswg.org] On Behalf Of
>>  > Patrick Durusau
>>  > Sent: 04 June 2002 15:04
>>  > To: osis-core
>>  > Subject: [osis-core] Schema sample attached
>>  >
>>  >
>>  > Guys,
>>  >
>>  > Sorry, forgot to attach.
>>  >
>>  > Patrick
>>  >
>>  > --
>>  > Patrick Durusau
>>  > Director of Research and Development
>>  > Society of Biblical Literature
>>  > pdurusau@emory.edu
>>  >
>>  >


-- 

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