All Downloads are FREE. Search and download functionalities are using the official Maven repository.

org.apache.commons.codec.language.Metaphone Maven / Gradle / Ivy

Go to download

The Apache Commons Codec component contains encoders and decoders for various formats such as Base16, Base32, Base64, digest, and Hexadecimal. In addition to these widely used encoders and decoders, the codec package also maintains a collection of phonetic encoding utilities. This is a port for GWT, which enables program, to use Apache Commons Codec also in the frontend compiled by the gwt compiler to java-script.

There is a newer version: 1.17.1-1
Show newest version
/*
 * Licensed to the Apache Software Foundation (ASF) under one or more
 * contributor license agreements.  See the NOTICE file distributed with
 * this work for additional information regarding copyright ownership.
 * The ASF licenses this file to You under the Apache License, Version 2.0
 * (the "License"); you may not use this file except in compliance with
 * the License.  You may obtain a copy of the License at
 *
 *      http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

package org.apache.commons.codec.language;

import org.apache.commons.codec.EncoderException;
import org.apache.commons.codec.StringEncoder;

/**
 * Encodes a string into a Metaphone value.
 * 

* Initial Java implementation by William B. Brogden. December, 1997. * Permission given by wbrogden for code to be used anywhere. *

* Hanging on the Metaphone by Lawrence Philips in Computer Language of Dec. 1990, * p 39. *

* Note, that this does not match the algorithm that ships with PHP, or the algorithm found in the Perl implementations: *

* *

* They have had undocumented changes from the originally published algorithm. * For more information, see CODEC-57. *

* This class is conditionally thread-safe. * The instance field for maximum code length is mutable {@link #setMaxCodeLen(int)} * but is not volatile, and accesses are not synchronized. * If an instance of the class is shared between threads, the caller needs to ensure that suitable synchronization * is used to ensure safe publication of the value between threads, and must not invoke {@link #setMaxCodeLen(int)} * after initial setup. * */ public class Metaphone implements StringEncoder { /** * Five values in the English language */ private static final String VOWELS = "AEIOU"; /** * Variable used in Metaphone algorithm */ private static final String FRONTV = "EIY"; /** * Variable used in Metaphone algorithm */ private static final String VARSON = "CSPTG"; /** * The max code length for metaphone is 4 */ private int maxCodeLen = 4; /** * Creates an instance of the Metaphone encoder */ public Metaphone() { super(); } /** * Find the metaphone value of a String. This is similar to the * soundex algorithm, but better at finding similar sounding words. * All input is converted to upper case. * Limitations: Input format is expected to be a single ASCII word * with only characters in the A - Z range, no punctuation or numbers. * * @param txt String to find the metaphone code for * @return A metaphone code corresponding to the String supplied */ public String metaphone(final String txt) { boolean hard = false; int txtLength; if (txt == null || (txtLength = txt.length()) == 0) { return ""; } // single character is itself if (txtLength == 1) { return txt.toUpperCase(java.util.Locale.ENGLISH); } final char[] inwd = txt.toUpperCase(java.util.Locale.ENGLISH).toCharArray(); final StringBuilder local = new StringBuilder(40); // manipulate final StringBuilder code = new StringBuilder(10); // output // handle initial 2 characters exceptions switch(inwd[0]) { case 'K': case 'G': case 'P': /* looking for KN, etc*/ if (inwd[1] == 'N') { local.append(inwd, 1, inwd.length - 1); } else { local.append(inwd); } break; case 'A': /* looking for AE */ if (inwd[1] == 'E') { local.append(inwd, 1, inwd.length - 1); } else { local.append(inwd); } break; case 'W': /* looking for WR or WH */ if (inwd[1] == 'R') { // WR -> R local.append(inwd, 1, inwd.length - 1); break; } if (inwd[1] == 'H') { local.append(inwd, 1, inwd.length - 1); local.setCharAt(0, 'W'); // WH -> W } else { local.append(inwd); } break; case 'X': /* initial X becomes S */ inwd[0] = 'S'; local.append(inwd); break; default: local.append(inwd); } // now local has working string with initials fixed final int wdsz = local.length(); int n = 0; while (code.length() < this.getMaxCodeLen() && n < wdsz ) { // max code size of 4 works well final char symb = local.charAt(n); // remove duplicate letters except C if (symb != 'C' && isPreviousChar( local, n, symb ) ) { n++; } else { // not dup switch(symb) { case 'A': case 'E': case 'I': case 'O': case 'U': if (n == 0) { code.append(symb); } break; // only use vowel if leading char case 'B': if ( isPreviousChar(local, n, 'M') && isLastChar(wdsz, n) ) { // B is silent if word ends in MB break; } code.append(symb); break; case 'C': // lots of C special cases /* discard if SCI, SCE or SCY */ if ( isPreviousChar(local, n, 'S') && !isLastChar(wdsz, n) && FRONTV.indexOf(local.charAt(n + 1)) >= 0 ) { break; } if (regionMatch(local, n, "CIA")) { // "CIA" -> X code.append('X'); break; } if (!isLastChar(wdsz, n) && FRONTV.indexOf(local.charAt(n + 1)) >= 0) { code.append('S'); break; // CI,CE,CY -> S } if (isPreviousChar(local, n, 'S') && isNextChar(local, n, 'H') ) { // SCH->sk code.append('K'); break; } if (isNextChar(local, n, 'H')) { // detect CH if (n == 0 && wdsz >= 3 && isVowel(local,2) ) { // CH consonant -> K consonant code.append('K'); } else { code.append('X'); // CHvowel -> X } } else { code.append('K'); } break; case 'D': if (!isLastChar(wdsz, n + 1) && isNextChar(local, n, 'G') && FRONTV.indexOf(local.charAt(n + 2)) >= 0) { // DGE DGI DGY -> J code.append('J'); n += 2; } else { code.append('T'); } break; case 'G': // GH silent at end or before consonant if (isLastChar(wdsz, n + 1) && isNextChar(local, n, 'H')) { break; } if (!isLastChar(wdsz, n + 1) && isNextChar(local,n,'H') && !isVowel(local,n+2)) { break; } if (n > 0 && ( regionMatch(local, n, "GN") || regionMatch(local, n, "GNED") ) ) { break; // silent G } if (isPreviousChar(local, n, 'G')) { // NOTE: Given that duplicated chars are removed, I don't see how this can ever be true hard = true; } else { hard = false; } if (!isLastChar(wdsz, n) && FRONTV.indexOf(local.charAt(n + 1)) >= 0 && !hard) { code.append('J'); } else { code.append('K'); } break; case 'H': if (isLastChar(wdsz, n)) { break; // terminal H } if (n > 0 && VARSON.indexOf(local.charAt(n - 1)) >= 0) { break; } if (isVowel(local,n+1)) { code.append('H'); // Hvowel } break; case 'F': case 'J': case 'L': case 'M': case 'N': case 'R': code.append(symb); break; case 'K': if (n > 0) { // not initial if (!isPreviousChar(local, n, 'C')) { code.append(symb); } } else { code.append(symb); // initial K } break; case 'P': if (isNextChar(local,n,'H')) { // PH -> F code.append('F'); } else { code.append(symb); } break; case 'Q': code.append('K'); break; case 'S': if (regionMatch(local,n,"SH") || regionMatch(local,n,"SIO") || regionMatch(local,n,"SIA")) { code.append('X'); } else { code.append('S'); } break; case 'T': if (regionMatch(local,n,"TIA") || regionMatch(local,n,"TIO")) { code.append('X'); break; } if (regionMatch(local,n,"TCH")) { // Silent if in "TCH" break; } // substitute numeral 0 for TH (resembles theta after all) if (regionMatch(local,n,"TH")) { code.append('0'); } else { code.append('T'); } break; case 'V': code.append('F'); break; case 'W': case 'Y': // silent if not followed by vowel if (!isLastChar(wdsz,n) && isVowel(local,n+1)) { code.append(symb); } break; case 'X': code.append('K'); code.append('S'); break; case 'Z': code.append('S'); break; default: // do nothing break; } // end switch n++; } // end else from symb != 'C' if (code.length() > this.getMaxCodeLen()) { code.setLength(this.getMaxCodeLen()); } } return code.toString(); } private boolean isVowel(final StringBuilder string, final int index) { return VOWELS.indexOf(string.charAt(index)) >= 0; } private boolean isPreviousChar(final StringBuilder string, final int index, final char c) { boolean matches = false; if( index > 0 && index < string.length() ) { matches = string.charAt(index - 1) == c; } return matches; } private boolean isNextChar(final StringBuilder string, final int index, final char c) { boolean matches = false; if( index >= 0 && index < string.length() - 1 ) { matches = string.charAt(index + 1) == c; } return matches; } private boolean regionMatch(final StringBuilder string, final int index, final String test) { boolean matches = false; if( index >= 0 && index + test.length() - 1 < string.length() ) { final String substring = string.substring( index, index + test.length()); matches = substring.equals( test ); } return matches; } private boolean isLastChar(final int wdsz, final int n) { return n + 1 == wdsz; } /** * Encodes an Object using the metaphone algorithm. This method * is provided in order to satisfy the requirements of the * Encoder interface, and will throw an EncoderException if the * supplied object is not of type java.lang.String. * * @param obj Object to encode * @return An object (or type java.lang.String) containing the * metaphone code which corresponds to the String supplied. * @throws EncoderException if the parameter supplied is not * of type java.lang.String */ @Override public Object encode(final Object obj) throws EncoderException { if (!(obj instanceof String)) { throw new EncoderException("Parameter supplied to Metaphone encode is not of type java.lang.String"); } return metaphone((String) obj); } /** * Encodes a String using the Metaphone algorithm. * * @param str String object to encode * @return The metaphone code corresponding to the String supplied */ @Override public String encode(final String str) { return metaphone(str); } /** * Tests is the metaphones of two strings are identical. * * @param str1 First of two strings to compare * @param str2 Second of two strings to compare * @return {@code true} if the metaphones of these strings are identical, * {@code false} otherwise. */ public boolean isMetaphoneEqual(final String str1, final String str2) { return metaphone(str1).equals(metaphone(str2)); } /** * Returns the maxCodeLen. * @return int */ public int getMaxCodeLen() { return this.maxCodeLen; } /** * Sets the maxCodeLen. * @param maxCodeLen The maxCodeLen to set */ public void setMaxCodeLen(final int maxCodeLen) { this.maxCodeLen = maxCodeLen; } }





© 2015 - 2024 Weber Informatics LLC | Privacy Policy