How to efficiently store and access properties of Unicode characters. The technique extends to other problems where sparse or repeating data must be looked up in O(1) time.