Like a tokenizer's purpose is to convert text into an array of words, right?