1 /*
2 * Licensed to the Apache Software Foundation (ASF) under one or more
3 * contributor license agreements. See the NOTICE file distributed with
4 * this work for additional information regarding copyright ownership.
5 * The ASF licenses this file to You under the Apache License, Version 2.0
6 * (the "License"); you may not use this file except in compliance with
7 * the License. You may obtain a copy of the License at
8 *
9 * http://www.apache.org/licenses/LICENSE-2.0
10 *
11 * Unless required by applicable law or agreed to in writing, software
12 * distributed under the License is distributed on an "AS IS" BASIS,
13 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
14 * See the License for the specific language governing permissions and
15 * limitations under the License.
16 */
17
18 package org.apache.commons.codec.language;
19
20 /**
21 * Encodes a string into a Caverphone 1.0 value.
22 *
23 * This is an algorithm created by the Caversham Project at the University of Otago. It implements the Caverphone 1.0
24 * algorithm:
25 *
26 * @version $Id: Caverphone1.html 889935 2013-12-11 05:05:13Z ggregory $
27 * @see <a href="http://en.wikipedia.org/wiki/Caverphone">Wikipedia - Caverphone</a>
28 * @see <a href="http://caversham.otago.ac.nz/files/working/ctp060902.pdf">Caverphone 1.0 specification</a>
29 * @since 1.5
30 *
31 * <p>This class is immutable and thread-safe.</p>
32 */
33 public class Caverphone1 extends AbstractCaverphone {
34
35 private static final String SIX_1 = "111111";
36
37 /**
38 * Encodes the given String into a Caverphone value.
39 *
40 * @param source
41 * String the source string
42 * @return A caverphone code for the given String
43 */
44 @Override
45 public String encode(String source) {
46 String txt = source;
47 if (txt == null || txt.length() == 0) {
48 return SIX_1;
49 }
50
51 // 1. Convert to lowercase
52 txt = txt.toLowerCase(java.util.Locale.ENGLISH);
53
54 // 2. Remove anything not A-Z
55 txt = txt.replaceAll("[^a-z]", "");
56
57 // 3. Handle various start options
58 // 2 is a temporary placeholder to indicate a consonant which we are no longer interested in.
59 txt = txt.replaceAll("^cough", "cou2f");
60 txt = txt.replaceAll("^rough", "rou2f");
61 txt = txt.replaceAll("^tough", "tou2f");
62 txt = txt.replaceAll("^enough", "enou2f");
63 txt = txt.replaceAll("^gn", "2n");
64
65 // End
66 txt = txt.replaceAll("mb$", "m2");
67
68 // 4. Handle replacements
69 txt = txt.replaceAll("cq", "2q");
70 txt = txt.replaceAll("ci", "si");
71 txt = txt.replaceAll("ce", "se");
72 txt = txt.replaceAll("cy", "sy");
73 txt = txt.replaceAll("tch", "2ch");
74 txt = txt.replaceAll("c", "k");
75 txt = txt.replaceAll("q", "k");
76 txt = txt.replaceAll("x", "k");
77 txt = txt.replaceAll("v", "f");
78 txt = txt.replaceAll("dg", "2g");
79 txt = txt.replaceAll("tio", "sio");
80 txt = txt.replaceAll("tia", "sia");
81 txt = txt.replaceAll("d", "t");
82 txt = txt.replaceAll("ph", "fh");
83 txt = txt.replaceAll("b", "p");
84 txt = txt.replaceAll("sh", "s2");
85 txt = txt.replaceAll("z", "s");
86 txt = txt.replaceAll("^[aeiou]", "A");
87 // 3 is a temporary placeholder marking a vowel
88 txt = txt.replaceAll("[aeiou]", "3");
89 txt = txt.replaceAll("3gh3", "3kh3");
90 txt = txt.replaceAll("gh", "22");
91 txt = txt.replaceAll("g", "k");
92 txt = txt.replaceAll("s+", "S");
93 txt = txt.replaceAll("t+", "T");
94 txt = txt.replaceAll("p+", "P");
95 txt = txt.replaceAll("k+", "K");
96 txt = txt.replaceAll("f+", "F");
97 txt = txt.replaceAll("m+", "M");
98 txt = txt.replaceAll("n+", "N");
99 txt = txt.replaceAll("w3", "W3");
100 txt = txt.replaceAll("wy", "Wy"); // 1.0 only
101 txt = txt.replaceAll("wh3", "Wh3");
102 txt = txt.replaceAll("why", "Why"); // 1.0 only
103 txt = txt.replaceAll("w", "2");
104 txt = txt.replaceAll("^h", "A");
105 txt = txt.replaceAll("h", "2");
106 txt = txt.replaceAll("r3", "R3");
107 txt = txt.replaceAll("ry", "Ry"); // 1.0 only
108 txt = txt.replaceAll("r", "2");
109 txt = txt.replaceAll("l3", "L3");
110 txt = txt.replaceAll("ly", "Ly"); // 1.0 only
111 txt = txt.replaceAll("l", "2");
112 txt = txt.replaceAll("j", "y"); // 1.0 only
113 txt = txt.replaceAll("y3", "Y3"); // 1.0 only
114 txt = txt.replaceAll("y", "2"); // 1.0 only
115
116 // 5. Handle removals
117 txt = txt.replaceAll("2", "");
118 txt = txt.replaceAll("3", "");
119
120 // 6. put ten 1s on the end
121 txt = txt + SIX_1;
122
123 // 7. take the first six characters as the code
124 return txt.substring(0, SIX_1.length());
125 }
126
127 }