"Krzysztof Bialas" <[EMAIL PROTECTED]> wrote on Sun, 20 Mar 2005 15:01:02 +0100:
I've invested quite some time into developing an example osis file showing off the problems with osis2mod/Sword for windows 1.5.8pre3 that are (in my opinion) 1.5.8 blockers. These are stated in the forum message http://www.crosswire.org/forums/mvnforum/viewthread?thread=274.
I have converted another Bible to OSIS 2.0.1 format, and xml2gbs (1.5.7 Linux) crashes half-way through. If I take only parts of the Bible, it completes, but all verses are empty.
After running it in the debugger and looking at the source, it seems that the problem is due to unhandled elements or unanticipated ways of nesting elements.
At this point I stopped investigating because I don't think there's any hope of ever parsing OSIS XML correctly using an ad-hoc parser; OSIS is way too complex and very flexible. A clean solution would require that xml2gbs be rewritten to employ an external XML library (e.g., libxml2 from http://xmlsoft.org/).
It is not up to me to judge whether this blocks 1.5.8.
Justus
-- Justus H. Piater, Ph.D. http://www.montefiore.ulg.ac.be/~piater/ Institut Montefiore, B28 Phone: +32-4-366-2279 Universit� de Li�ge, Belgium Fax: +32-4-366-2620
I was also a bit suprised to see hand-made OSIS parser. I understand that it is there becouse of 'historical reasons' and surely it will be replaced by something more decent (xml2, xerces, ...) in future when someone will find time to do it. Looking at BibleTime makes me optimistic as this application already follows the clean solution - and (in my personal opinion) should also read plain OSIS Bible xml modules, not only those transformed by osis2mod. But that was just a side note, as I'm not a Sword developer, but a Sword module developer. So from my perspective, it does not matter whether it is done the clean way or the rough way as soon it does the job correctly and in my opinion other tasks as alternative versification systems and apocrypha support is more needed then the 'niceness' of parser. My job is to prepare Bible module as close to the original as possible - including all the formatting peculiarities the printed text posseses. I believe it is also the developers aim to have quality modules available, and I'm ready to help them to test the Sword so it would support broader range of OSIS texts.
Greetings
Christopher
_______________________________________________ sword-devel mailing list: [email protected] http://www.crosswire.org/mailman/listinfo/sword-devel Instructions to unsubscribe/change your settings at above page
