A New String Matching Algorithm

The string matching problem that appears in many applications like word processing, information retrieval, bibliographic search, molecular biology, etc., consists in finding the first or all the occurrences of a pattern in a text, where the pattern and the text are strings over the same alphabet. Many algorithms for solving this problem exist in literature [10, 20]. The naive (brute force) algorithm [7] locates all occurrences in time O(nm), where m is the length of the pattern and n is the length of the text. Hashing provides a simple method [15] that runs in linear time in most practical situations. A minimal DFA recognizing the language Sx, where x[0:m7 1] is the pattern over the alphabet S of size s, can be used to locate all the occurrences of x in the text y[0:n7 1] by examining each text character exactly once [7]. But the construction of the DFA requires O(sþm) time and O(sm) space. Simon [14, 18, 19] showed that this DFA can be constructed in O(m) time and space by introducing a delay bounded by O(s) per text character and his algorithm never performs more than 2n7 1 text character comparisons. However, the first linear time string-matching algorithm was discovered by Morris and Pratt [17], and was later improved by Knuth et al. [16]. The search behaves like a recognition