[sword-devel] Re: [osis-editors] OSIS 2.0.1 modules updated

[Patrick Durusau]

Michael,

Slaving away on the users manual but wanted to take a moment to reply to 
your points on quotation marks:

Michael Paul Johnson wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> At 00:29 18-03-04, Patrick Durusau wrote:
> ...
> 
<snip>

> 
>>Are we disagreeing about where the rules for rendering should be 
>>placed? 
> 
> 
> We are disagreeing as to the nature of quotation mark insertion. You 
> seem to be of the opinion that this is a rendering issue, much as 
> selection of font parameters for the various kinds of encoded text 
> elements would be, or how verses are marked (or not marked) in the 
> rendered text. I strongly disagree with that perspective. Punctuation 
> marks, including quotation marks, are part of the Bible text. If you 
> want to include metadata in the XML markup about the quotations, 
> perhaps to enable analysis of the text or even to regenerate a 
> different version of the text punctuated differently, that is a 
> separate issue.
> 
<snip>
> 
> The wording of the current documentation for OSIS pretty much demands 
> that I treat quotation marks as a rendering issue instead of part of 
> the text. I am unwilling to do that. I would rather introduce a 
> competing standard than do that. Treating quotation marks as rendering 
> issues may make sense when dealing with one or even a small number of 
> languages, but the very idea of doing so is repulsive to me when 
> dealing with any significant fraction of the world's languages.
> 
> 
Your point about nature of punctuation was what I was missing. You kept 
saying it and I kept hearing it as a rendering issue.

Appreciate your patience!

The odd thing is that I argue the same point for medieval and ANE texts, 
where Unicode wants to treat all glyphs that represent the same 
"character" as the same, when in fact that is, as you say, a lossy 
representation. Glyphs, even if they represent the same "character" vary 
by genre, time period, etc.

Well, then the question is how to preserve punctuation, as seen in the 
text, and at the same time delimit what are quotations (or other textual 
features) in the text? In other words, I want to have quotes delimited 
by elements for searching, something you don't get with punctuation, at 
least not in an XML environment, and at the same time, you want to avoid 
double rendering of punctuation in XML based on the same elements.

The only way I can think of offhand is to somehow mark the original 
punctuation inside the quote or other elements so that I could have:

(WARNING: untested)

<xsl:template match="q/punc[@style="dq"] >

It would be required to be the first child of the q element so we could 
reliably find it. But optional whether it was present or not.

Can you suggest other places where this would be necessary?

This might (emphasis on might) solve Troy's problem with multiple 
embedded quotes as well, where he does not (and we have not provided) a 
reliable rendering solution.

Though I tend towards using a traditional milestone for some cases.

I really was not trying to be difficult but the problem this poses is 
far from trivial. At least it doesn't look so after spending more than 
the last decade with a paradigm that sees all punctuation as rendering. ;-)

<snip>
>>>
>>I recall some recent discussion of footnote start anchors but don't 
>>have 
>>it at my finger tips. Can you say a few words about that?
> 
> 
> I thought that I already did that. I look at footnotes as a note that 
> pertains to either a range of text or a point in the text. This note 
> may be rendered at the bottom of the page, in a pop-up window, or 
> whatever, but it contains information about the main text that may be 
> helpful but that is not part of the main text. Since the note may 
> pertain to a range of text, I have found it useful to mark the 
> beginning of the text with a "begin reference" marker (<RB> in GBF), 
> then mark the end of the text with an element containing the note 
> itself. This way, it is easy to render the text to which the footnote 
> pertains as a hyperlink. Also, if you wanted to treat footnotes like 
> the JPS Tanach did in print (with superscripted markers at both 
> places), you can. OSIS has no equivalent marker, so I put in a generic 
> milestone. In rendering footnotes, if there is no beginning footnoted 
> text marker, I just render it as a point, for example as a hyperlinked 
> asterisk pointing to the note. In print rendering, I usually ignore 
> the first marker, but could do something with it.
> 
> This isn't a major problem, just a feature that I miss that would be 
> easy to supply. I can live with my current work-around in the current 
> OSIS version just fine.
> 
> 
Hmmm, might I suggest (again off the top of my head) the use of a 
<reference> element that has its osisRef attribute pointing to the 
<note> with its content? Seems to me if you put a type on <reference> 
that said, type="note" that would do the trick. (We would have to add 
note to the list of osisRefernce types.)

For the case you suggest, I think this would capture the behavior you want.

<snip>

> May God bless you with wisdom and insight.

Appreciate that! I don't use the ones I have as well as I should so I 
don't ask for more, just His guidance in using what I have already.

Hope you are having a great day!

Patrick

-- 
Patrick Durusau
Director of Research and Development
Society of Biblical Literature
Patrick.Durusau@sbl-site.org
Chair, V1 - Text Processing: Office and Publishing Systems Interface
Co-Editor, ISO 13250, Topic Maps -- Reference Model

Topic Maps: Human, not artificial, intelligence at work!