A construct that identifies token boundaries (or breaks) in text that is based on regular expressions and defined in terms of Unicode characters and character properties.