this is a file that we can use to test our digram
analyzer it is fairly long and contains 
only lowercase letters hopefully it is long enough
to have a good distribution of english letters
according to the paper the algorithm does not perform
well on texts that are shorter than four hundred characters
so i'm going to keep writing here until i get around
that many letters another twenty five letters should do it
that worked pretty well but there were still four substitutions
that were wrong so we try again with five hundred letters there
were only two mistaken substitutions let's see if these disappear
with a longer text with six hundred letters there was still one
mistaken substitution so we try again with seven hundred the
performance wasn't quite as good so try again