[sword-devel] virtual modules

[DM Smith]

Chris Little wrote:
>
>
> DM Smith wrote:
>> Didn't think it was going to be easy. But I was going to start with 
>> Bibles, since their structure is more well known and the problem 
>> space would be greatly reduced. Besides, OSIS is not quite there yet 
>> for describing a commentary (though one could use Bible markup, if it 
>> is a verse by verse commentary) or a dictionary.
>
> Commentary structure in OSIS is actually well defined (and stable, I 
> would say). You basically write what you want to write (using divs and 
> p's, primarily, for structure), identify the annotation type (e.g. 
> commentary) via the annotateType attribute, and then identify what it 
> is you're annotating via the annotateRef attribute.
Thanks so much for this tidbit! I have read the manual several times, 
but never grasped this. And though I now see it, I still don't fully 
grasp it.

Now some questions regarding the "proper" construction of a commentary 
that is Testament, Book, Chapter, Verse (TBCV) oriented. (From a Sword 
perspective)
Is annotateType an inherited attribute like xml:lang? (It is listed 
along with xml:lang as a global attribute?) It would be such a pain to 
put it on most elements. If so, how do you turn it off for quotations of 
verses? (which I presume would be marked with the verse element, or is 
that different, too?)

According to the manual an annotateRef has the same syntax as an 
osisRef. And it clearly states that it requires a prefix to refer 
externally. What is the preferred form for such a prefix? Should it be 
Bible: ?

Bible:Gen.1.1-Bible:Gen.22.3
Is the second one in the range needed?

What are the structural elements that should be used to indicate, TBC&V? 
With a bible we have div for T&B, chapter for C and verse for V. Should 
we use these same elements? Or should we use just div?

You have mentioned that it should be normative that milestones be used 
to mark up verses and probably chapters as document structure could 
easily overlap those elements. Should it be normative for commentaries also?

Does osis2mod work for commentaries with this markup? Will it look at 
the annotateRef attribute? So far I have found that encoding a 
commentary as a bible works, but I didn't know if anything else did.
>
> Dictionaries aren't well defined (in terms of structure internal to an 
> entry), but the level with which Sword would likely interact with a 
> dictionary is just the entry level itself (i.e. <div type="entry"/>).
In earlier discussions with the BT folks, they indicated that they were 
going to base the dictionary work on TEI, but they were not there yet. I 
have been looking at it extensively to figure out the best way to 
represent a dictionary (i.e. Naves) and I am closer to actually doing 
something. When I do, I'll probably float it by you and the folks here 
and at BT.

With regard to the level of interaction by Sword, I think that there are 
two or three distinct parts of a dictionary entry that Sword should 
become concerned with. In the general case, it should be the term and 
its definition. These should be marked up distinctly. The one will be 
used as the key. (I understand that at the current time that Sword is 
not concerned with anything more.)

In the special case of Strongs, I see at least one other part. The 
Strong's Number is not actually the term, but it is the key. Or to think 
of it another way, it is an alternate term/key. And, additionally, the 
term is really the actual Greek or Hebrew, but currently it is present 
in a transliteration. Having these separately marked up will provide us 
with the ability to do really neat stuff. For example, one could allow 
the user to see the Greek or Hebrew, or the transliteration, either as a 
replacement, a choice or in addition to the number.