Longest common extension (Java)

From LiteratePrograms
Jump to: navigation, search

Contents

[edit] Overview

Simple computation of the longest common extension (lce) of two given strings, beginning at two given indices. The lce computation is part of different algorithms in inexact string matching (Gusfield 1999:196). This simple implementation has a runtime complexity of O(n). Computation in O(1) is possible through constant-time retrieval of lowest common ancestors in suffix trees (after linear-time preprocessing). See Gusfield 1999:181 for further reference on constant time lca retrival and Gusfield 1999:196 for constant time lce computation using the lca.

[edit] Implementation

The method takes four parameters: t1 and t2 (the two strings) and two indices i1 and i2 (the index to start at in t1 and t2). It returns the length of the longest common extension starting at i1 in t1 and at i2 in t2. The method simply iterates over the given strings, starting at the given indices. It compares the characters, counts equal characters and returns that number as soon as the characters differ.

<<lce>>=

public static int longestCommonExtension(String t1, int i1, String t2, int i2) {
    int res = 0;
    for (int i = i1; i < t1.length() && i2 < t2.length(); i++, i2++) {
        if (t1.charAt(i) == t2.charAt(i2))
            res++;
        else
            return res;
    }
    return res;
}

[edit] Usage

Here's JUnit 4 unit test to demonstrate the functionality. Note that you must have JUnit 4 in your classpath in order to successfully compile.

<<test>>=

@Test
public void testGetLongestCommonExtension() {
    int res = longestCommonExtension("zsdabcdefghj", 3, "abcdezas", 0);
    assertEquals(5, res);
}

The complete program:

<<SimpleLongestCommonExtension.java>>=

import static org.junit.Assert.assertEquals;
import org.junit.Test;

public class SimpleLongestCommonExtension {
	lce
	test
}

[edit] References

  • Gusfield, Dan (1999), Algorithms on Strings, Sequences and Trees. Cambridge: University Press.
Download code
hijacker
hijacker
hijacker
hijacker