"You shall know a word by the company it keeps."

This is a common task in linguistic analysis. There is no mention of regular expressions anywhere in the question.
(Indeed, I would approach this problem by tokenizing the text into array of words.)