i need to add a feature to limit the markov chain generation to portions of the text file. it's just so big now.