interesting. so the tokenized keywords were directly input by the user instead of parsed and transformed into a token form. certainly economical in how the code was composed and transformed. a lot less parsing code I would guess.