As my wife said last night "aren't you reinventing the horse"? - talk about a mixed metaphor :-)
I'm putting the king james version bible into a database, with a breakdown to the verse level. It has a main object table which could submit to a breakdown to the word and punctuation mark level. However it could also be a superset and hold paragraphs, sections, etc..
The idea is that I want to be able to hold everything from a Bible - any version - to a Tom Clancy novel, and allow for sensible XML output and efficient searches.
I'll be posting my code here pretty soon, but I thought I'd open this up and see if anyone had any links to some ideas on this or examples.
The main two areas I'm considering are that a book may have many version - esp. the bible, and therefore I have a table bookgroups and books. For example, A Christmas Carol may have the original version and an abridged version. so the object references books, which is always a subset record of bookgroups -even if there's only one version.
The other issue is that a verse belongs to a chapter belongs to a book belongs to a bible. However a verse also belongs to a paragraph which is not technically part of the structure - and there may even be varying interpretations of this structure. I foresee a many-to-many relationship table coming up here soon. If anyone has any thoughts about that.
All that I design here (OK well most) will be public domain.
Sam
