Can you explain how the `split` method operates in this string tokenizer code