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 */
017package org.apache.commons.codec.language;
018
019import java.util.Locale;
020
021import org.apache.commons.codec.EncoderException;
022import org.apache.commons.codec.StringEncoder;
023
024/**
025 * Match Rating Approach Phonetic Algorithm Developed by <CITE>Western Airlines</CITE> in 1977.
026 *
027 * This class is immutable and thread-safe.
028 *
029 * @see <a href="http://en.wikipedia.org/wiki/Match_rating_approach">Wikipedia - Match Rating Approach</a>
030 * @since 1.8
031 */
032public class MatchRatingApproachEncoder implements StringEncoder {
033
034    private static final String SPACE = " ";
035
036    private static final String EMPTY = "";
037
038    /**
039     * Constants used mainly for the min rating value.
040     */
041    private static final int ONE = 1, TWO = 2, THREE = 3, FOUR = 4, FIVE = 5, SIX = 6, SEVEN = 7,
042                             ELEVEN = 11, TWELVE = 12;
043
044    /**
045     * The plain letter equivalent of the accented letters.
046     */
047    private static final String PLAIN_ASCII = "AaEeIiOoUu" + // grave
048            "AaEeIiOoUuYy" + // acute
049            "AaEeIiOoUuYy" + // circumflex
050            "AaOoNn" + // tilde
051            "AaEeIiOoUuYy" + // umlaut
052            "Aa" + // ring
053            "Cc" + // cedilla
054            "OoUu"; // double acute
055
056    /**
057     * Unicode characters corresponding to various accented letters. For example: \u00DA is U acute etc...
058     */
059    private static final String UNICODE = "\u00C0\u00E0\u00C8\u00E8\u00CC\u00EC\u00D2\u00F2\u00D9\u00F9" +
060            "\u00C1\u00E1\u00C9\u00E9\u00CD\u00ED\u00D3\u00F3\u00DA\u00FA\u00DD\u00FD" +
061            "\u00C2\u00E2\u00CA\u00EA\u00CE\u00EE\u00D4\u00F4\u00DB\u00FB\u0176\u0177" +
062            "\u00C3\u00E3\u00D5\u00F5\u00D1\u00F1" +
063            "\u00C4\u00E4\u00CB\u00EB\u00CF\u00EF\u00D6\u00F6\u00DC\u00FC\u0178\u00FF" +
064            "\u00C5\u00E5" + "\u00C7\u00E7" + "\u0150\u0151\u0170\u0171";
065
066    private static final String[] DOUBLE_CONSONANT =
067            new String[] { "BB", "CC", "DD", "FF", "GG", "HH", "JJ", "KK", "LL", "MM", "NN", "PP", "QQ", "RR", "SS",
068                           "TT", "VV", "WW", "XX", "YY", "ZZ" };
069
070    /**
071     * Cleans up a name: 1. Upper-cases everything 2. Removes some common punctuation 3. Removes accents 4. Removes any
072     * spaces.
073     *
074     * <h2>API Usage</h2>
075     * <p>
076     * Consider this method private, it is package protected for unit testing only.
077     * </p>
078     *
079     * @param name
080     *            The name to be cleaned
081     * @return The cleaned name
082     */
083    String cleanName(final String name) {
084        String upperName = name.toUpperCase(Locale.ENGLISH);
085
086        final String[] charsToTrim = { "\\-", "[&]", "\\'", "\\.", "[\\,]" };
087        for (final String str : charsToTrim) {
088            upperName = upperName.replaceAll(str, EMPTY);
089        }
090
091        upperName = removeAccents(upperName);
092        upperName = upperName.replaceAll("\\s+", EMPTY);
093
094        return upperName;
095    }
096
097    /**
098     * Encodes an Object using the Match Rating Approach algorithm. Method is here to satisfy the requirements of the
099     * Encoder interface Throws an EncoderException if input object is not of type java.lang.String.
100     *
101     * @param pObject
102     *            Object to encode
103     * @return An object (or type java.lang.String) containing the Match Rating Approach code which corresponds to the
104     *         String supplied.
105     * @throws EncoderException
106     *             if the parameter supplied is not of type java.lang.String
107     */
108    @Override
109    public final Object encode(final Object pObject) throws EncoderException {
110        if (!(pObject instanceof String)) {
111            throw new EncoderException(
112                    "Parameter supplied to Match Rating Approach encoder is not of type java.lang.String");
113        }
114        return encode((String) pObject);
115    }
116
117    /**
118     * Encodes a String using the Match Rating Approach (MRA) algorithm.
119     *
120     * @param name
121     *            String object to encode
122     * @return The MRA code corresponding to the String supplied
123     */
124    @Override
125    public final String encode(String name) {
126        // Bulletproof for trivial input - NINO
127        if (name == null || EMPTY.equalsIgnoreCase(name) || SPACE.equalsIgnoreCase(name) || name.length() == 1) {
128            return EMPTY;
129        }
130
131        // Preprocessing
132        name = cleanName(name);
133
134        // BEGIN: Actual encoding part of the algorithm...
135        // 1. Delete all vowels unless the vowel begins the word
136        name = removeVowels(name);
137
138        // 2. Remove second consonant from any double consonant
139        name = removeDoubleConsonants(name);
140
141        // 3. Reduce codex to 6 letters by joining the first 3 and last 3 letters
142        name = getFirst3Last3(name);
143
144        return name;
145    }
146
147    /**
148     * Gets the first and last 3 letters of a name (if &gt; 6 characters) Else just returns the name.
149     *
150     * <h2>API Usage</h2>
151     * <p>
152     * Consider this method private, it is package protected for unit testing only.
153     * </p>
154     *
155     * @param name
156     *            The string to get the substrings from
157     * @return Annexed first and last 3 letters of input word.
158     */
159    String getFirst3Last3(final String name) {
160        final int nameLength = name.length();
161
162        if (nameLength > SIX) {
163            final String firstThree = name.substring(0, THREE);
164            final String lastThree = name.substring(nameLength - THREE, nameLength);
165            return firstThree + lastThree;
166        }
167        return name;
168    }
169
170    /**
171     * Obtains the min rating of the length sum of the 2 names. In essence the larger the sum length the smaller the
172     * min rating. Values strictly from documentation.
173     *
174     * <h2>API Usage</h2>
175     * <p>
176     * Consider this method private, it is package protected for unit testing only.
177     * </p>
178     *
179     * @param sumLength
180     *            The length of 2 strings sent down
181     * @return The min rating value
182     */
183    int getMinRating(final int sumLength) {
184        int minRating = 0;
185
186        if (sumLength <= FOUR) {
187            minRating = FIVE;
188        } else if (sumLength <= SEVEN) { // aready know it is at least 5
189            minRating = FOUR;
190        } else if (sumLength <= ELEVEN) { // aready know it is at least 8
191            minRating = THREE;
192        } else if (sumLength == TWELVE) {
193            minRating = TWO;
194        } else {
195            minRating = ONE; // docs said little here.
196        }
197
198        return minRating;
199    }
200
201    /**
202     * Determines if two names are homophonous via Match Rating Approach (MRA) algorithm. It should be noted that the
203     * strings are cleaned in the same way as {@link #encode(String)}.
204     *
205     * @param name1
206     *            First of the 2 strings (names) to compare
207     * @param name2
208     *            Second of the 2 names to compare
209     * @return <code>true</code> if the encodings are identical <code>false</code> otherwise.
210     */
211    public boolean isEncodeEquals(String name1, String name2) {
212        // Bulletproof for trivial input - NINO
213        if (name1 == null || EMPTY.equalsIgnoreCase(name1) || SPACE.equalsIgnoreCase(name1)) {
214            return false;
215        } else if (name2 == null || EMPTY.equalsIgnoreCase(name2) || SPACE.equalsIgnoreCase(name2)) {
216            return false;
217        } else if (name1.length() == 1 || name2.length() == 1) {
218            return false;
219        } else if (name1.equalsIgnoreCase(name2)) {
220            return true;
221        }
222
223        // Preprocessing
224        name1 = cleanName(name1);
225        name2 = cleanName(name2);
226
227        // Actual MRA Algorithm
228
229        // 1. Remove vowels
230        name1 = removeVowels(name1);
231        name2 = removeVowels(name2);
232
233        // 2. Remove double consonants
234        name1 = removeDoubleConsonants(name1);
235        name2 = removeDoubleConsonants(name2);
236
237        // 3. Reduce down to 3 letters
238        name1 = getFirst3Last3(name1);
239        name2 = getFirst3Last3(name2);
240
241        // 4. Check for length difference - if 3 or greater then no similarity
242        // comparison is done
243        if (Math.abs(name1.length() - name2.length()) >= THREE) {
244            return false;
245        }
246
247        // 5. Obtain the minimum rating value by calculating the length sum of the
248        // encoded Strings and sending it down.
249        final int sumLength = Math.abs(name1.length() + name2.length());
250        int minRating = 0;
251        minRating = getMinRating(sumLength);
252
253        // 6. Process the encoded Strings from left to right and remove any
254        // identical characters found from both Strings respectively.
255        final int count = leftToRightThenRightToLeftProcessing(name1, name2);
256
257        // 7. Each PNI item that has a similarity rating equal to or greater than
258        // the min is considered to be a good candidate match
259        return count >= minRating;
260
261    }
262
263    /**
264     * Determines if a letter is a vowel.
265     *
266     * <h2>API Usage</h2>
267     * <p>
268     * Consider this method private, it is package protected for unit testing only.
269     * </p>
270     *
271     * @param letter
272     *            The letter under investiagtion
273     * @return True if a vowel, else false
274     */
275    boolean isVowel(final String letter) {
276        return letter.equalsIgnoreCase("E") || letter.equalsIgnoreCase("A") || letter.equalsIgnoreCase("O") ||
277               letter.equalsIgnoreCase("I") || letter.equalsIgnoreCase("U");
278    }
279
280    /**
281     * Processes the names from left to right (first) then right to left removing identical letters in same positions.
282     * Then subtracts the longer string that remains from 6 and returns this.
283     *
284     * <h2>API Usage</h2>
285     * <p>
286     * Consider this method private, it is package protected for unit testing only.
287     * </p>
288     *
289     * @param name1
290     *            name2
291     * @return the length as above
292     */
293    int leftToRightThenRightToLeftProcessing(final String name1, final String name2) {
294        final char[] name1Char = name1.toCharArray();
295        final char[] name2Char = name2.toCharArray();
296
297        final int name1Size = name1.length() - 1;
298        final int name2Size = name2.length() - 1;
299
300        String name1LtRStart = EMPTY;
301        String name1LtREnd = EMPTY;
302
303        String name2RtLStart = EMPTY;
304        String name2RtLEnd = EMPTY;
305
306        for (int i = 0; i < name1Char.length; i++) {
307            if (i > name2Size) {
308                break;
309            }
310
311            name1LtRStart = name1.substring(i, i + 1);
312            name1LtREnd = name1.substring(name1Size - i, name1Size - i + 1);
313
314            name2RtLStart = name2.substring(i, i + 1);
315            name2RtLEnd = name2.substring(name2Size - i, name2Size - i + 1);
316
317            // Left to right...
318            if (name1LtRStart.equals(name2RtLStart)) {
319                name1Char[i] = ' ';
320                name2Char[i] = ' ';
321            }
322
323            // Right to left...
324            if (name1LtREnd.equals(name2RtLEnd)) {
325                name1Char[name1Size - i] = ' ';
326                name2Char[name2Size - i] = ' ';
327            }
328        }
329
330        // Char arrays -> string & remove extraneous space
331        final String strA = new String(name1Char).replaceAll("\\s+", EMPTY);
332        final String strB = new String(name2Char).replaceAll("\\s+", EMPTY);
333
334        // Final bit - subtract longest string from 6 and return this int value
335        if (strA.length() > strB.length()) {
336            return Math.abs(SIX - strA.length());
337        }
338        return Math.abs(SIX - strB.length());
339    }
340
341    /**
342     * Removes accented letters and replaces with non-accented ascii equivalent Case is preserved.
343     * http://www.codecodex.com/wiki/Remove_accent_from_letters_%28ex_.%C3%A9_to_e%29
344     *
345     * @param accentedWord
346     *            The word that may have accents in it.
347     * @return De-accented word
348     */
349    String removeAccents(final String accentedWord) {
350        if (accentedWord == null) {
351            return null;
352        }
353
354        final StringBuilder sb = new StringBuilder();
355        final int n = accentedWord.length();
356
357        for (int i = 0; i < n; i++) {
358            final char c = accentedWord.charAt(i);
359            final int pos = UNICODE.indexOf(c);
360            if (pos > -1) {
361                sb.append(PLAIN_ASCII.charAt(pos));
362            } else {
363                sb.append(c);
364            }
365        }
366
367        return sb.toString();
368    }
369
370    /**
371     * Replaces any double consonant pair with the single letter equivalent.
372     *
373     * <h2>API Usage</h2>
374     * <p>
375     * Consider this method private, it is package protected for unit testing only.
376     * </p>
377     *
378     * @param name
379     *            String to have double consonants removed
380     * @return Single consonant word
381     */
382    String removeDoubleConsonants(final String name) {
383        String replacedName = name.toUpperCase(Locale.ENGLISH);
384        for (final String dc : DOUBLE_CONSONANT) {
385            if (replacedName.contains(dc)) {
386                final String singleLetter = dc.substring(0, 1);
387                replacedName = replacedName.replace(dc, singleLetter);
388            }
389        }
390        return replacedName;
391    }
392
393    /**
394     * Deletes all vowels unless the vowel begins the word.
395     *
396     * <h2>API Usage</h2>
397     * <p>
398     * Consider this method private, it is package protected for unit testing only.
399     * </p>
400     *
401     * @param name
402     *            The name to have vowels removed
403     * @return De-voweled word
404     */
405    String removeVowels(String name) {
406        // Extract first letter
407        final String firstLetter = name.substring(0, 1);
408
409        name = name.replaceAll("A", EMPTY);
410        name = name.replaceAll("E", EMPTY);
411        name = name.replaceAll("I", EMPTY);
412        name = name.replaceAll("O", EMPTY);
413        name = name.replaceAll("U", EMPTY);
414
415        name = name.replaceAll("\\s{2,}\\b", SPACE);
416
417        // return isVowel(firstLetter) ? (firstLetter + name) : name;
418        if (isVowel(firstLetter)) {
419            return firstLetter + name;
420        }
421        return name;
422    }
423}