Three files should exist in this directory for the examples to work.

All samples are assumed to be lowercase.

markov.txt:
    * One line per name
    * Empty lines are ignored
    * Preferred: 100+ samples

cfgrammar.txt
    * One line per name
    * Group headers name token rules
    * Blank line marks new group
    * Asterisk means anonymous token not reused in other rules.
    * example
        SV * END
        e le na
        i ly na

        SCV * END
        ma na va
        re na la