[Boost-users] Question about proper container/algorithm to use for fast dictionary lookup

Hi, This question is whether Boost already has, or perhaps needs to have a container type that is specifically designed for fast word lookup in a UNICODE dictionary. It has occurred to me that an associative container that is specifically designed for word lookup may be much faster than the standard containers such as map, hash_map, associative_vector as they are general purpose containers. So the more detailed questions are: 1) What type of algorithm/container should be used here? 2) Would its performance be much better than a general purpose associative container? Thanks, Elisha Berns e.berns@computer.org tel. (310) 556 - 8332 fax (310) 556 - 2839