001 /*
002 * Licensed to the Apache Software Foundation (ASF) under one or more
003 * contributor license agreements. See the NOTICE file distributed with
004 * this work for additional information regarding copyright ownership.
005 * The ASF licenses this file to You under the Apache License, Version 2.0
006 * (the "License"); you may not use this file except in compliance with
007 * the License. You may obtain a copy of the License at
008 *
009 * http://www.apache.org/licenses/LICENSE-2.0
010 *
011 * Unless required by applicable law or agreed to in writing, software
012 * distributed under the License is distributed on an "AS IS" BASIS,
013 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
014 * See the License for the specific language governing permissions and
015 * limitations under the License.
016 */
017
018 package org.apache.commons.codec.language;
019
020 import org.apache.commons.codec.EncoderException;
021 import org.apache.commons.codec.StringEncoder;
022
023 /**
024 * Encodes a string into a Soundex value. Soundex is an encoding used to relate similar names, but can also be used as a
025 * general purpose scheme to find word with similar phonemes.
026 *
027 * This class is thread-safe.
028 * Although not strictly immutable, the {@link #maxLength} field is not actually used.
029 *
030 * @version $Id: Soundex.html 889935 2013-12-11 05:05:13Z ggregory $
031 */
032 public class Soundex implements StringEncoder {
033
034 /**
035 * This is a default mapping of the 26 letters used in US English. A value of <code>0</code> for a letter position
036 * means do not encode.
037 * <p>
038 * (This constant is provided as both an implementation convenience and to allow Javadoc to pick
039 * up the value for the constant values page.)
040 * </p>
041 *
042 * @see #US_ENGLISH_MAPPING
043 */
044 public static final String US_ENGLISH_MAPPING_STRING = "01230120022455012623010202";
045
046 /**
047 * This is a default mapping of the 26 letters used in US English. A value of <code>0</code> for a letter position
048 * means do not encode.
049 *
050 * @see Soundex#Soundex(char[])
051 */
052 private static final char[] US_ENGLISH_MAPPING = US_ENGLISH_MAPPING_STRING.toCharArray();
053
054 /**
055 * An instance of Soundex using the US_ENGLISH_MAPPING mapping.
056 *
057 * @see #US_ENGLISH_MAPPING
058 */
059 public static final Soundex US_ENGLISH = new Soundex();
060
061 /**
062 * The maximum length of a Soundex code - Soundex codes are only four characters by definition.
063 *
064 * @deprecated This feature is not needed since the encoding size must be constant. Will be removed in 2.0.
065 */
066 @Deprecated
067 private int maxLength = 4;
068
069 /**
070 * Every letter of the alphabet is "mapped" to a numerical value. This char array holds the values to which each
071 * letter is mapped. This implementation contains a default map for US_ENGLISH
072 */
073 private final char[] soundexMapping;
074
075 /**
076 * Creates an instance using US_ENGLISH_MAPPING
077 *
078 * @see Soundex#Soundex(char[])
079 * @see Soundex#US_ENGLISH_MAPPING
080 */
081 public Soundex() {
082 this.soundexMapping = US_ENGLISH_MAPPING;
083 }
084
085 /**
086 * Creates a soundex instance using the given mapping. This constructor can be used to provide an internationalized
087 * mapping for a non-Western character set.
088 *
089 * Every letter of the alphabet is "mapped" to a numerical value. This char array holds the values to which each
090 * letter is mapped. This implementation contains a default map for US_ENGLISH
091 *
092 * @param mapping
093 * Mapping array to use when finding the corresponding code for a given character
094 */
095 public Soundex(char[] mapping) {
096 this.soundexMapping = new char[mapping.length];
097 System.arraycopy(mapping, 0, this.soundexMapping, 0, mapping.length);
098 }
099
100 /**
101 * Creates a refined soundex instance using a custom mapping. This constructor can be used to customize the mapping,
102 * and/or possibly provide an internationalized mapping for a non-Western character set.
103 *
104 * @param mapping
105 * Mapping string to use when finding the corresponding code for a given character
106 * @since 1.4
107 */
108 public Soundex(String mapping) {
109 this.soundexMapping = mapping.toCharArray();
110 }
111
112 /**
113 * Encodes the Strings and returns the number of characters in the two encoded Strings that are the same. This
114 * return value ranges from 0 through 4: 0 indicates little or no similarity, and 4 indicates strong similarity or
115 * identical values.
116 *
117 * @param s1
118 * A String that will be encoded and compared.
119 * @param s2
120 * A String that will be encoded and compared.
121 * @return The number of characters in the two encoded Strings that are the same from 0 to 4.
122 *
123 * @see SoundexUtils#difference(StringEncoder,String,String)
124 * @see <a href="http://msdn.microsoft.com/library/default.asp?url=/library/en-us/tsqlref/ts_de-dz_8co5.asp"> MS
125 * T-SQL DIFFERENCE </a>
126 *
127 * @throws EncoderException
128 * if an error occurs encoding one of the strings
129 * @since 1.3
130 */
131 public int difference(String s1, String s2) throws EncoderException {
132 return SoundexUtils.difference(this, s1, s2);
133 }
134
135 /**
136 * Encodes an Object using the soundex algorithm. This method is provided in order to satisfy the requirements of
137 * the Encoder interface, and will throw an EncoderException if the supplied object is not of type java.lang.String.
138 *
139 * @param obj
140 * Object to encode
141 * @return An object (or type java.lang.String) containing the soundex code which corresponds to the String
142 * supplied.
143 * @throws EncoderException
144 * if the parameter supplied is not of type java.lang.String
145 * @throws IllegalArgumentException
146 * if a character is not mapped
147 */
148 @Override
149 public Object encode(Object obj) throws EncoderException {
150 if (!(obj instanceof String)) {
151 throw new EncoderException("Parameter supplied to Soundex encode is not of type java.lang.String");
152 }
153 return soundex((String) obj);
154 }
155
156 /**
157 * Encodes a String using the soundex algorithm.
158 *
159 * @param str
160 * A String object to encode
161 * @return A Soundex code corresponding to the String supplied
162 * @throws IllegalArgumentException
163 * if a character is not mapped
164 */
165 @Override
166 public String encode(String str) {
167 return soundex(str);
168 }
169
170 /**
171 * Used internally by the SoundEx algorithm.
172 *
173 * Consonants from the same code group separated by W or H are treated as one.
174 *
175 * @param str
176 * the cleaned working string to encode (in upper case).
177 * @param index
178 * the character position to encode
179 * @return Mapping code for a particular character
180 * @throws IllegalArgumentException
181 * if the character is not mapped
182 */
183 private char getMappingCode(String str, int index) {
184 // map() throws IllegalArgumentException
185 char mappedChar = this.map(str.charAt(index));
186 // HW rule check
187 if (index > 1 && mappedChar != '0') {
188 char hwChar = str.charAt(index - 1);
189 if ('H' == hwChar || 'W' == hwChar) {
190 char preHWChar = str.charAt(index - 2);
191 char firstCode = this.map(preHWChar);
192 if (firstCode == mappedChar || 'H' == preHWChar || 'W' == preHWChar) {
193 return 0;
194 }
195 }
196 }
197 return mappedChar;
198 }
199
200 /**
201 * Returns the maxLength. Standard Soundex
202 *
203 * @deprecated This feature is not needed since the encoding size must be constant. Will be removed in 2.0.
204 * @return int
205 */
206 @Deprecated
207 public int getMaxLength() {
208 return this.maxLength;
209 }
210
211 /**
212 * Returns the soundex mapping.
213 *
214 * @return soundexMapping.
215 */
216 private char[] getSoundexMapping() {
217 return this.soundexMapping;
218 }
219
220 /**
221 * Maps the given upper-case character to its Soundex code.
222 *
223 * @param ch
224 * An upper-case character.
225 * @return A Soundex code.
226 * @throws IllegalArgumentException
227 * Thrown if <code>ch</code> is not mapped.
228 */
229 private char map(char ch) {
230 int index = ch - 'A';
231 if (index < 0 || index >= this.getSoundexMapping().length) {
232 throw new IllegalArgumentException("The character is not mapped: " + ch);
233 }
234 return this.getSoundexMapping()[index];
235 }
236
237 /**
238 * Sets the maxLength.
239 *
240 * @deprecated This feature is not needed since the encoding size must be constant. Will be removed in 2.0.
241 * @param maxLength
242 * The maxLength to set
243 */
244 @Deprecated
245 public void setMaxLength(int maxLength) {
246 this.maxLength = maxLength;
247 }
248
249 /**
250 * Retrieves the Soundex code for a given String object.
251 *
252 * @param str
253 * String to encode using the Soundex algorithm
254 * @return A soundex code for the String supplied
255 * @throws IllegalArgumentException
256 * if a character is not mapped
257 */
258 public String soundex(String str) {
259 if (str == null) {
260 return null;
261 }
262 str = SoundexUtils.clean(str);
263 if (str.length() == 0) {
264 return str;
265 }
266 char out[] = {'0', '0', '0', '0'};
267 char last, mapped;
268 int incount = 1, count = 1;
269 out[0] = str.charAt(0);
270 // getMappingCode() throws IllegalArgumentException
271 last = getMappingCode(str, 0);
272 while (incount < str.length() && count < out.length) {
273 mapped = getMappingCode(str, incount++);
274 if (mapped != 0) {
275 if (mapped != '0' && mapped != last) {
276 out[count++] = mapped;
277 }
278 last = mapped;
279 }
280 }
281 return new String(out);
282 }
283
284 }