Word frequency statistics

Tokenize English by word boundaries, treat consecutive Han as one token, then rank counts with an adjustable list size — useful for keyword spotting and wording checks.