Indexing Arbitrary Data with SWISH-E

Fast lookups are crucial to many computer applications and operations. The general problem of indexing and searching on arbitrary data is not a simple one, with many semantic, linguistic, and technical issues to iron out. In this paper we present swish-e, a descendent of Kevin Hughes' SWISH project from 1994. Swish-e provides a fullfeatured and useful toolkit to index and query 8-bit ASCII data. This paper discusses the structure, features, and usage of swish-e, with mentions of possible directions for further development and interesting related work. We also compare swish-e to MySQL's full-text search feature in terms of features and speed, and discuss two real-world swish-e applications, Sman and Swished.