Analysis of Algorithms for Listing Equivalence Classes of k-ary Strings

We give eecient algorithms for listing equivalence classes of k-ary strings under reversal and permutation of alphabet symbols. As representative of each equivalence class we choose that string which is lexicographically smallest. These algorithms use space O(n) and time O(p kN), where N is the total number of strings generated and n is the length of each string. For k = 2, we obtain a recursive decomposition of the set of binary strings that allows the strings to be generated without rejecting any strings. For k 3, some strings must be rejected. The algorithm is simple but its exact analysis is rather complicated. In the analysis we determine a quantity of independent interest: the average length of the common preex of two randomly chosen innnite length \restricted-growth" strings. (k) j : falling factorial (k) j = k(k-1) (k-j + 1). n k : Stirling number of the second kind.