CONFLUENTLY PERSISTENT DEQUES VIA DATA-STRUCTURAL BOOTSTRAPPING

We introduce data-structural bootstrapping, a technique to design data structures recursively, and use it to design confluently persistent d eques. Our data structure requires O(log() k) worst-case time and spa ce per deletion, where k is the total number of deque operations, and constant worst-case time and space for other operations. Further, the data structure allows a purely functional implementation, with no side effects. This improves a previous result of Driscoll, Sleator, and Ta jan (in ''Proceedings 2nd Annual ACM-SIAM Symposium on Discrete Algori thms, 1991,'' pp. 89-99). (C) 1995 Academic Press, Inc.