[sword-devel] tei proposal

Chris Little chrislit at crosswire.org
Thu Nov 27 11:03:50 MST 2008



Matthew Talbert wrote:
> I have a proposal to make that could solve two issues at once. In the
> simplest form, the proposal is to organize a group effort to convert
> an existing public domain dictionary to tei. Specifically I would
> suggest this:
> 
> 1. host the source file via svn
> 2. start with the straight output of mod2imp
> 3. slowly, step-by-step, build up a complete module that eventually
> validates and is a correct module.
> 4. still allow minor updates and changes

This sounds like a fine idea, in principle.

> Although I wasn't around for it, I imagine that making the KJV module
> taught everyone involved a good deal of OSIS and exposed lots of bugs
> along the way. This project would have similar expectations.

The OSIS encoding of the KJV module was basically done by Troy and DM. 
The collaborative part did not involve knowing how to encode OSIS.

> I would suggest that this be hosted by crosswire, although I am
> willing to do this myself if people are interested. I would also
> suggest using Easton as a starting point, primarily because I'm
> familiar with it, but also because it has lots of different issues to
> address that would allow trying out a range of tei.

I have no opinion on where to host it. I'm sure Troy would be fine with 
putting it in sword-tools/modules, which is where we've done this 
historically. A greater issue might be having to dole out tons of SVN 
accounts.

Easton's is probably not a good choice if this is anything more than a 
training exercise. I'm planning to do some quick updates to our four 
main Bible dictionaries, as early as this weekend, including clean-up, 
some additional tagging, and conversion to TEI. ISBE, Smith, Nave, and 
Easton are on that list.

What would actually be quite helpful would be to do some tagging of 
ISBE, which has a lot of missing tagging. That includes untagged 
scripture references, cross-references that would be slightly difficult 
to tag programmatically, references to Josephus that aren't tagged, etc. 
The downside, of course, is that it's 10x as long as Easton.

It would allow playing with a full range of cross-reference elements, 
<pron>, rend if you feel like it, the <name> elements, etc.

--Chris




More information about the sword-devel mailing list