Count the number of documents each term appears in.
Array of tokenized documents (each document is a string[] of tokens)
Mapping of term → document frequency
Count the number of documents each term appears in.