The document which was here has been moved to termgenerator.html in xapian-core.
Still To Do
Bug #22:
- [DEFERRED] Some special handling for '-' and '.' with one and perhaps two character components is still to think about.
Bug #113:
- [DEFERRED] Predicate functions should be user-specifiable via the API so users can configure the characters which fullfil various roles - so if for example you wanted filenames to be single terms you could set that up. We still need good defaults categories though.
Bug #119:
- We now have an adequate number of TermGenerator testcases - more can be added over time.
