A Family of Fast Constant-Space Substring Search Algorithms

 Harri Hakonen1  and Timo Raita2

  Acta Cybernetica 14 (1999) 239-250.

Abstract:

 This paper describes a new strategy for searching a substring in a given text. The method is based on the well-known Boyer-Moore algorithm complementing it with a technique called q -slicing, a form of probabilistic q -gram matching. As a result, we get a family of highly parametric algorithms apt for adaptation to the special properties inherent to the source which generates the input strings. The search procedure is independent of the alphabet size and appropriate for efficient and practical on-line implementations. Simulation results show that they are comparable to the fastest currently known Boyer-Moore variants.  

Footnotes

  1  Turku Centre for Computer Science (TUCS), e-mail: hat@cs.utu.fi 2  Department of Computer Science, University of Turku, Lemminkäisenkatu 14 A, SF-20520 Turku, Finland, e-mail: raita@cs.utu.fi. The author acknowledges support by the Academy of Finland under grant No. 865431