[sword-devel] CCEL, ThML - volunteers?
Dudeck, John
John.Dudeck at sim.org
Thu Nov 15 15:04:46 MST 2018
Troy,
The problem is not parsing references. I have that figured out long ago, and converting from Logos to Sword, etc.
The problem is that when you have a commentary with, for example a milestone for Gen 1:1-31 with some introductory material, followed by a sub-heading with a milestone for Gen 1:1-13, then followed by verse commentaries for Gen 1:1-5, then Gen 1:6-8, then Gen 1:9-13. When compiling with osis2mod, it doesn't quite know what to do with the nested ranges, and stuff does not display properly.
John
>
> Hi John,
> Do you have an example? Our parser handles some pretty crazy verse references. You can
> give it a try here:
>
> http://crosswire.org/study/examples/parsevs.jsp
>
>
>
> On 11/15/18 10:35 AM, Dudeck, John wrote:
> The main problem with commentaries is getting the annotateRef attribute right without
> manual intervention. Logos supports nested levels of reference, eg. book-level,
> chapter-level, verse-level, and verse ranges (that can span mutiple chapters). As far as I
> can tell, osis2mod gets terribly confused with nested ranges.
>
> Generally speaking, black box converters are pretty much a lost cause. I have used a .conf
> file approach where variable stuff can be configured, for example how to map class and
> style attributes to the limited rendering capabilities of Sword. And I have separate scripts
> for genbooks, bibles, commentaries, and dictionaries/lexicons.
>
> John
>
> > Any time your list of exceptional conditions contains more text than the
> > conversion itself, you have to start questioning whether 'automation' is
> > even possible.
> >
> > I spent time developing similar scripts (something like wordxml ->
> > USFM) for single use cases. Doing the conversion was usually about
> > 10 steps, but then dealing with the exceptions (both before and after
> > the conversion) took 100+ steps. Without automation, the 10 steps
> > was good enough. but not keeping a human involved, all the human
> > decisions had to be mapped.
> >
> >
> >
> >
> > On Thu, Nov 15, 2018 at 10:39 AM Greg Hellings <greg.hellings at gmail.com> wrote:
> >
> >
> > On Thu, Nov 15, 2018 at 10:35 AM Dudeck, John <John.Dudeck at sim.org> wrote:
> > I am developing a set of Perl scripts that convert Logos XML to OSIS for
> > commentaries. It has proven to be a non-trivial exercise.
> >
> > I undertook a similar exercise over a decade ago as part of an internship. It was a beast
> of
> > a task and one that was never fully completed.
> >
> > --Greg
John Dudeck
Programmer at Editions Cle Lyon, France
john.dudeck at sim.org john at editionscle.com
--
"If we knew what it was we were doing, it would not be called
research, would it?" -- Albert Einstein
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.crosswire.org/pipermail/sword-devel/attachments/20181115/e2c74307/attachment-0001.html>
More information about the sword-devel
mailing list