VocabularySize()

oT = new stzNaiveBayes()
oT.Train("don't 3.14 word2vec", "technical")
oT.Train("apple banana cherry", "fruit")
# a query of the exact tricky tokens finds its label
? oT.Classify("don't 3.14 word2vec")
# ...and plain words find theirs
? oT.Classify("apple banana")
# six distinct tokens were learned, not more
? oT.VocabularySize()
# stzText sees three words in the tricky line too
? len((new stzText("don't 3.14 word2vec")).Words())

oF = new stzNaiveBayes()
oF.Train("Hello World", "a")
oF.Train("hello world", "a")
# two spellings are one vocabulary
? oF.VocabularySize()
# ...and both documents counted
? oF.NumberOfDocuments()
oU = new stzNaiveBayes()
oU.Train("ÉCOLE", "x")
oU.Train("école", "x")
# an accented pair folds together too
? oU.VocabularySize()