charsets {
    lowerletter = 'a'..'z';
    upperletter = 'A'..'Z';
}

tokens {
    FirstUpper = upperletter (lowerletter)*;
    FirstLower = (lowerletter)+;
    Number     = (digit)+ ['.' (digit)+];
    Word       = FirstUpper | FirstLower | Number;

    Colon       = ':';

    MidSentence = ',' | ';';
    EndSentence = '.' | '!' | '?'.
}

productions {
    Text     = (Sentence)+;
    Sentence = (FirstUpper (Word | MidSentence (Word)+)* Colon)*
               FirstUpper (Word | MidSentence (Word)+)* EndSentence;
}