Japanese text tokenization with deinflection support, reimplementing the algorithm from 10ten Reader
Japanese text tokenizer with deinflection support