Multikey Quickselect

We present multikey quickselect: an efficient, in-place, easy-to-implement algorithm for the selection problem for strings. We analyze its expected cost under a uniform model, measured as the number of character comparisons and pointer swaps, depending on the alphabet cardinality. From the analysis, we derive a binary variant of the algorithm, which is more efficient when the range of values for the alphabet is known. This variant can also be applied to multikey quicksort.