In-place length-restricted prefix coding

Huffman codes, combined with word-based models, are considered efficient compression schemes for full-text retrieval systems. The decoding rate for these schemes can be substantially improved if the maximum length of the codewords is not greater then the machine word size L. However, if the vocabulary is large, simple methods for generating optimal length-restricted codes are either too slow or require a significantly large amount of memory. We present an in-place, simple and fast implementation for the BRCI (Build, Remove, Condense and Insert) algorithm, an approximative method for length-restricted coding. It overwrites a sorted input list of n weights with the corresponding codeword lengths in O(n) time. In addition, the worst-case compression loss introduced by BRCI codes with respect to unrestricted Huffman codes is proved to be negligible for all practical values of both L and n.