This is a simple text parsing library for losely-formalized texts.

The API and the implementation are based on the concept of Hidden Markov
Models (HMMs) (Note: Though HMMs are widely used in machine learning, this
library does not involve the "learning" part of the technique. Instead,
the parameters of the model are specified by the user)

You might not need to understand what HHMs are to use the library,
though: the true power of the library comes from the numerous ways to
combine simpler descriptions of text "chunks" into more complex ones.