Mercurial > hg > mpdl-group

comparison software/mpdl-services/mpiwg-mpdl-lt/src/de/mpg/mpiwg/berlin/mpdl/lt/text/norm/Normalizer.java @ 19:4a3641ae14d2

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Erstellung
author Josef Willenborg <jwillenborg@mpiwg-berlin.mpg.de>
date Wed, 09 Nov 2011 15:32:05 +0100
parents
children 7d6d969b10cf
comparison
equal deleted inserted replaced
18:dc5e9fcb3fdc 19:4a3641ae14d2
1 package de.mpg.mpiwg.berlin.mpdl.lt.text.norm;
2
3 import java.io.IOException;
4 import java.io.StringReader;
5 import java.util.ArrayList;
6
7 import de.mpg.mpiwg.berlin.mpdl.exception.ApplicationException;
8 import de.mpg.mpiwg.berlin.mpdl.lt.general.Language;
9 import de.mpg.mpiwg.berlin.mpdl.lt.text.norm.lang.MpdlNormalizerLexAR;
10 import de.mpg.mpiwg.berlin.mpdl.lt.text.norm.lang.MpdlNormalizerLexDE;
11 import de.mpg.mpiwg.berlin.mpdl.lt.text.norm.lang.MpdlNormalizerLexEL;
12 import de.mpg.mpiwg.berlin.mpdl.lt.text.norm.lang.MpdlNormalizerLexEN;
13 import de.mpg.mpiwg.berlin.mpdl.lt.text.norm.lang.MpdlNormalizerLexFR;
14 import de.mpg.mpiwg.berlin.mpdl.lt.text.norm.lang.MpdlNormalizerLexIT;
15 import de.mpg.mpiwg.berlin.mpdl.lt.text.norm.lang.MpdlNormalizerLexLA;
16 import de.mpg.mpiwg.berlin.mpdl.lt.text.norm.lang.MpdlNormalizerLexNL;
17 import de.mpg.mpiwg.berlin.mpdl.lt.text.norm.lang.MpdlNormalizerLexZH;
18 import de.mpg.mpiwg.berlin.mpdl.lt.text.reg.Regularization;
19 import de.mpg.mpiwg.berlin.mpdl.lt.text.reg.RegularizationManager;
20
21 public class Normalizer {
22 public static int DISPLAY = 1; // normalization in DISPLAY mode
23 public static int DICTIONARY = 2; // normalization in DICTIONARY mode
24 public static int SEARCH = 3; // normalization in SEARCH mode; never used so far in indexing because it does not support the morph. lexicons such as CELEX (e.g. eingeschränkt would not be stemmed to eingeschraenkt)
25 private int normMode = DISPLAY; // Default e.g. for indexing and querying
26 private String[] normFunctions = {"norm"}; // default is to use the norm function
27 private String language;
28 private int[] offsets;
29
30 public Normalizer(String[] normFunctions, String lang) {
31 this.normFunctions = normFunctions;
32 String language = Language.getInstance().getLanguageId(lang);
33 this.language = language;
34 }
35
36 public Normalizer(String language) {
37 this.language = language;
38 }
39
40 public String getLanguage() {
41 return language;
42 }
43
44 public void setNormMode(int normMode) {
45 this.normMode = normMode;
46 }
47
48 /**
49 * Applies the normalization rules in <code>language</code> to
50 * <code>s</code>, without offset tracking.
51 *
52 * @param s source string
53 * @return normalized string
54 */
55 public String normalize(String s) throws ApplicationException {
56 String normStr = s;
57 if (useSpecialNormFunction())
58 normStr = removeSpecialNWDMarks(normStr);
59 if (useRegFunction()) {
60 // try to regularize the string to the norm form over predefined regularizations
61 RegularizationManager regManager = RegularizationManager.getInstance();
62 ArrayList<Regularization> regs = regManager.findRegsByOrig(language, s);
63 if (regs != null && regs.size() > 0) {
64 Regularization reg = regs.get(0); // only one: the first one
65 String regNormStr = reg.getNorm();
66 normStr = regNormStr;
67 }
68 }
69 if (useNormFunction()) {
70 // normalize the string by string replacements
71 if (normMode == DICTIONARY) {
72 normStr = normalize(normStr, DICTIONARY);
73 } else if (normMode == DISPLAY) {
74 normStr = normalize(normStr, DISPLAY);
75 } else if (normMode == SEARCH) {
76 normStr = normalize(normStr, SEARCH);
77 }
78 }
79 if (useSpecialNormFunction())
80 normStr = insertSpecialNWDMarks(normStr);
81 return normStr;
82 }
83
84 private boolean useRegFunction() {
85 boolean useReg = false;
86 for (int i=0; i< normFunctions.length; i++) {
87 String function = normFunctions[i];
88 if (function.equals("reg"))
89 return true;
90 }
91 return useReg;
92 }
93
94 private boolean useNormFunction() {
95 boolean useNorm = false;
96 for (int i=0; i< normFunctions.length; i++) {
97 String function = normFunctions[i];
98 if (function.equals("norm") || function.equals("specialNorm"))
99 return true;
100 }
101 return useNorm;
102 }
103
104 private boolean useSpecialNormFunction() {
105 boolean useNorm = false;
106 for (int i=0; i< normFunctions.length; i++) {
107 String function = normFunctions[i];
108 if (function.equals("specialNorm"))
109 return true;
110 }
111 return useNorm;
112 }
113
114 private String normalize(String s, int mode) {
115 String inputStr = s;
116 StringReader strReader = new StringReader(inputStr + "\n");
117 String retStr = "";
118 String token = "";
119 try {
120 if (Language.getInstance().isLatin(language)) {
121 MpdlNormalizerLexLA mpdlNormalizerLex = new MpdlNormalizerLexLA(strReader);
122 if (mode == DISPLAY)
123 mpdlNormalizerLex.yybegin(MpdlNormalizerLexLA.DISP);
124 else if (mode == DICTIONARY)
125 mpdlNormalizerLex.yybegin(MpdlNormalizerLexLA.DICT);
126 else if (mode == SEARCH)
127 mpdlNormalizerLex.yybegin(MpdlNormalizerLexLA.SEARCH);
128 while (token != null) {
129 token = mpdlNormalizerLex.yylex();
130 if (token != null)
131 retStr += token;
132 }
133 } else if (Language.getInstance().isArabic(language)) {
134 MpdlNormalizerLexAR mpdlNormalizerLex = new MpdlNormalizerLexAR(strReader);
135 if (mode == DISPLAY)
136 mpdlNormalizerLex.yybegin(MpdlNormalizerLexAR.DISP);
137 else if (mode == DICTIONARY)
138 mpdlNormalizerLex.yybegin(MpdlNormalizerLexAR.DICT);
139 else if (mode == SEARCH)
140 mpdlNormalizerLex.yybegin(MpdlNormalizerLexAR.SEARCH);
141 while (token != null) {
142 token = mpdlNormalizerLex.yylex();
143 if (token != null)
144 retStr += token;
145 }
146 } else if (Language.getInstance().isGerman(language)) {
147 MpdlNormalizerLexDE mpdlNormalizerLex = new MpdlNormalizerLexDE(strReader);
148 if (mode == DISPLAY)
149 mpdlNormalizerLex.yybegin(MpdlNormalizerLexDE.DISP);
150 else if (mode == DICTIONARY)
151 mpdlNormalizerLex.yybegin(MpdlNormalizerLexDE.CELEX);
152 else if (mode == SEARCH)
153 mpdlNormalizerLex.yybegin(MpdlNormalizerLexDE.SEARCH);
154 while (token != null) {
155 token = mpdlNormalizerLex.yylex();
156 if (token != null)
157 retStr += token;
158 }
159 } else if (Language.getInstance().isGreek(language)) {
160 MpdlNormalizerLexEL mpdlNormalizerLex = new MpdlNormalizerLexEL(strReader);
161 if (mode == DISPLAY)
162 mpdlNormalizerLex.yybegin(MpdlNormalizerLexEL.DISP);
163 else if (mode == DICTIONARY)
164 mpdlNormalizerLex.yybegin(MpdlNormalizerLexEL.SIGMA);
165 else if (mode == SEARCH)
166 mpdlNormalizerLex.yybegin(MpdlNormalizerLexEL.SEARCH);
167 while (token != null) {
168 token = mpdlNormalizerLex.yylex();
169 if (token != null)
170 retStr += token;
171 }
172 } else if (Language.getInstance().isEnglish(language)) {
173 MpdlNormalizerLexEN mpdlNormalizerLex = new MpdlNormalizerLexEN(strReader);
174 if (mode == DISPLAY)
175 mpdlNormalizerLex.yybegin(MpdlNormalizerLexEN.DISP);
176 else if (mode == DICTIONARY)
177 mpdlNormalizerLex.yybegin(MpdlNormalizerLexEN.DICT);
178 else if (mode == SEARCH)
179 mpdlNormalizerLex.yybegin(MpdlNormalizerLexEN.SEARCH);
180 while (token != null) {
181 token = mpdlNormalizerLex.yylex();
182 if (token != null)
183 retStr += token;
184 }
185 } else if (Language.getInstance().isFrench(language)) {
186 MpdlNormalizerLexFR mpdlNormalizerLex = new MpdlNormalizerLexFR(strReader);
187 if (mode == DISPLAY)
188 mpdlNormalizerLex.yybegin(MpdlNormalizerLexFR.DISP);
189 else if (mode == DICTIONARY)
190 mpdlNormalizerLex.yybegin(MpdlNormalizerLexFR.DICT_ASCII);
191 else if (mode == SEARCH)
192 mpdlNormalizerLex.yybegin(MpdlNormalizerLexFR.SEARCH);
193 while (token != null) {
194 token = mpdlNormalizerLex.yylex();
195 if (token != null)
196 retStr += token;
197 }
198 } else if (Language.getInstance().isItalian(language)) {
199 MpdlNormalizerLexIT mpdlNormalizerLex = new MpdlNormalizerLexIT(strReader);
200 if (mode == DISPLAY)
201 mpdlNormalizerLex.yybegin(MpdlNormalizerLexIT.DISP);
202 else if (mode == DICTIONARY)
203 mpdlNormalizerLex.yybegin(MpdlNormalizerLexIT.DICT);
204 else if (mode == SEARCH)
205 mpdlNormalizerLex.yybegin(MpdlNormalizerLexIT.SEARCH);
206 while (token != null) {
207 token = mpdlNormalizerLex.yylex();
208 if (token != null)
209 retStr += token;
210 }
211 } else if (Language.getInstance().isDutch(language)) {
212 MpdlNormalizerLexNL mpdlNormalizerLex = new MpdlNormalizerLexNL(strReader);
213 if (mode == DISPLAY)
214 mpdlNormalizerLex.yybegin(MpdlNormalizerLexNL.DISP);
215 else if (mode == DICTIONARY)
216 mpdlNormalizerLex.yybegin(MpdlNormalizerLexNL.DICT);
217 else if (mode == SEARCH)
218 mpdlNormalizerLex.yybegin(MpdlNormalizerLexNL.SEARCH);
219 while (token != null) {
220 token = mpdlNormalizerLex.yylex();
221 if (token != null)
222 retStr += token;
223 }
224 } else if (Language.getInstance().isChinese(language)) {
225 MpdlNormalizerLexZH mpdlNormalizerLex = new MpdlNormalizerLexZH(strReader);
226 if (mode == DISPLAY)
227 mpdlNormalizerLex.yybegin(MpdlNormalizerLexZH.DISP);
228 else if (mode == DICTIONARY)
229 mpdlNormalizerLex.yybegin(MpdlNormalizerLexZH.DICT);
230 else if (mode == SEARCH)
231 mpdlNormalizerLex.yybegin(MpdlNormalizerLexZH.SEARCH);
232 while (token != null) {
233 token = mpdlNormalizerLex.yylex();
234 if (token != null)
235 retStr += token;
236 }
237 } else {
238 retStr = s; // return the string unchanged
239 }
240 } catch (IOException e ) {
241 // nothing cause IOException is not needed for a StringReader
242 }
243 return retStr;
244 }
245
246
247 // used only in XmlTokenizerContentHandler // TODO make it better
248 private String removeSpecialNWDMarks(String inputString) {
249 String COMPLEX_ELEMENT_NWD_MARK = new Character('\u2424').toString(); // not word delimiting element
250 String cleanedWord = inputString;
251 boolean startsWithNWDMark = cleanedWord.startsWith(COMPLEX_ELEMENT_NWD_MARK);
252 if (startsWithNWDMark)
253 cleanedWord = cleanedWord.substring(1);
254 int countNWDMarks = cleanedWord.length() - cleanedWord.replaceAll(COMPLEX_ELEMENT_NWD_MARK, "").length();
255 if (countNWDMarks > 1)
256 cleanedWord = cleanedWord.replaceAll(COMPLEX_ELEMENT_NWD_MARK + "+", COMPLEX_ELEMENT_NWD_MARK);
257 // boolean notHyphenPlusNWD = cleanedWord.matches(".*[^-]+" + COMPLEX_ELEMENT_NWD_MARK + "+.*"); // e.g. "praebi ta"
258 // if (notHyphenPlusNWD)
259 // cleanedWord = cleanedWord.replaceAll("([^-]+)" + COMPLEX_ELEMENT_NWD_MARK + "+", "$1-" + COMPLEX_ELEMENT_NWD_MARK); // e.g. "praebi ta" is replaced by "praebi- ta"
260 cleanedWord = cleanedWord.replaceAll(COMPLEX_ELEMENT_NWD_MARK, " ");
261 return cleanedWord;
262 }
263
264 private String insertSpecialNWDMarks(String inputString) {
265 String COMPLEX_ELEMENT_NWD_MARK = new Character('\u2424').toString(); // not word delimiting element
266 String retStr = inputString;
267 boolean startsWithNWDMark = retStr.startsWith(COMPLEX_ELEMENT_NWD_MARK);
268 int countNWDMarks = retStr.length() - retStr.replaceAll(COMPLEX_ELEMENT_NWD_MARK, "").length();
269 retStr = retStr.replaceAll(" ", COMPLEX_ELEMENT_NWD_MARK);
270 // if (notHyphenPlusNWD)
271 // normalizedWordStr = normalizedWordStr.replaceAll("-" + COMPLEX_ELEMENT_NWD_MARK, COMPLEX_ELEMENT_NWD_MARK); // e.g. "praebi- ta" is replaced by "praebi ta"
272 if (countNWDMarks > 1) {
273 String nwdStr = "";
274 for (int i=0; i<countNWDMarks; i++)
275 nwdStr += COMPLEX_ELEMENT_NWD_MARK;
276 retStr = retStr.replaceAll(COMPLEX_ELEMENT_NWD_MARK, nwdStr);
277 }
278 if (startsWithNWDMark)
279 retStr = COMPLEX_ELEMENT_NWD_MARK + retStr;
280 return retStr;
281 }
282
283 /**
284 * Old code from Arboreal (Malcolm Hyman)
285 * Applies the normalization rules in <code>language</code> to
286 * <code>s</code>, with offset tracking.<p>
287 *
288 * <strong>WARNING:</strong>
289 * Arboreal will not work properly if a normalization substitution
290 * replaces a source character with more than two target characters!
291 * This is simply a BUG, and should be fixed. Fortunately, however,
292 * one does not often need such a replacement.<p>
293 *
294 * @param s source string
295 * @param offsets character offset table
296 * @return normalized string
297 */
298 private String normalize4Lexica(String s, int[] offsets) {
299 this.offsets = offsets;
300 if (language.equals("la") || language.equals("lat")) {
301 StringBuffer buf = new StringBuffer();
302 int n = 0;
303 for (int i = 0; i < s.length(); i++) {
304 char c = s.charAt(i);
305 String replace = new String();
306 switch (c) {
307 case 'j': replace = "i"; break;
308 case 'v': replace = "u"; break;
309 /*
310 * Linguistic note: /u/ and /v/ are rarely phonemic
311 * in Latin, as in alui 's/he nourished' vs.
312 * alvi 'of a belly', volui 's/he wished' or 'it rolled'
313 * vs. volvi 'to be rolled', (in)seruit 's/he joined
314 * together' vs. (in)servit 's/he serves'.
315 */
316 case 'q':
317 if ((i < s.length() - 1) && (s.charAt(i + 1) == ';'))
318 replace = "qu";
319 else
320 replace = "q";
321 break;
322 case ';':
323 if ((i > 0) && (s.charAt(i - 1) == 'q'))
324 replace = "e";
325 else if ((i == 0) || ! Character.isLetter(s.charAt(i - 1)))
326 replace = ";";
327 else
328 replace = "";
329 break;
330 case '\u0300': replace = ""; break; // COMBINING GRAVE ACCENT
331 case '\u0301': replace = ""; break; // COMBINING ACCUTE ACCENT
332 case '\u0302': replace = ""; break; // COMBINING CIRCUMFLEX ACCENT
333
334 case '\u00c0': replace = "A"; break; // LATIN CAPITAL LETTER A GRAVE
335 case '\u00c1': replace = "A"; break; // LATIN CAPITAL LETTER A ACUTE
336 case '\u00c2': replace = "A"; break; // LATIN CAPITAL LETTER A CIRCUMFLEX
337 case '\u00c4': replace = "A"; break; // LATIN CAPITAL LETTER A DIAERESIS
338 case '\u00c6': replace = "Ae"; break; // LATIN CAPITAL LETTER A E
339 case '\u00c7': replace = "C"; break; // LATIN CAPITAL LETTER C CEDILLA
340 case '\u00c8': replace = "E"; break; // LATIN CAPITAL LETTER E GRAVE
341 case '\u00c9': replace = "E"; break; // LATIN CAPITAL LETTER E ACUTE
342 case '\u00ca': replace = "E"; break; // LATIN CAPITAL LETTER E CIRCUMFLEX
343 case '\u00cb': replace = "E"; break; // LATIN CAPITAL LETTER E DIAERESIS
344 case '\u00cc': replace = "I"; break; // LATIN CAPITAL LETTER I GRAVE;
345 case '\u00cd': replace = "I"; break; // LATIN CAPITAL LETTER I ACUTE
346 case '\u00ce': replace = "I"; break; // LATIN CAPITAL LETTER I CIRCUMFLEX
347 case '\u00cf': replace = "I"; break; // LATIN CAPITAL LETTER I DIAERESIS
348 case '\u00d2': replace = "O"; break; // LATIN CAPITAL LETTER O GRAVE
349 case '\u00d3': replace = "O"; break; // LATIN CAPITAL LETTER O ACUTE
350 case '\u00d4': replace = "O"; break; // LATIN CAPITAL LETTER O CIRCUMFLEX
351 case '\u00d6': replace = "O"; break; // LATIN CAPITAL LETTER O DIAERESIS
352 case '\u00d9': replace = "U"; break; // LATIN CAPITAL LETTER U GRAVE
353 case '\u00da': replace = "U"; break; // LATIN CAPITAL LETTER U ACUTE
354 case '\u00db': replace = "U"; break; // LATIN CAPITAL LETTER U CIRCUMFLEX
355 case '\u00dc': replace = "U"; break; // LATIN CAPITAL LETTER U DIAERESIS
356 case '\u00e0': replace = "a"; break; // LATIN SMALL LETTER A GRAVE
357 case '\u00e1': replace = "a"; break; // LATIN SMALL LETTER A ACUTE
358 case '\u00e2': replace = "a"; break; // LATIN SMALL LETTER A CIRCUMFLEX
359 case '\u00e4': replace = "a"; break; // LATIN SMALL LETTER A DIAERESIS
360 case '\u00e6': replace = "ae"; break; // LATIN SMALL LETTER A E
361 case '\u00e7': replace = "c"; break; // LATIN SMALL LETTER C CEDILLA
362 case '\u00e8': replace = "e"; break; // LATIN SMALL LETTER E GRAVE
363 case '\u00e9': replace = "e"; break; // LATIN SMALL LETTER E ACUTE
364 case '\u00ea': replace = "e"; break; // LATIN SMALL LETTER E CIRCUMFLEX
365 case '\u00eb': replace = "e"; break; // LATIN SMALL LETTER E DIAERESIS
366 case '\u00ec': replace = "i"; break; // LATIN SMALL LETTER I GRAVE
367 case '\u00ed': replace = "i"; break; // LATIN SMALL LETTER I ACUTE
368 case '\u00ee': replace = "i"; break; // LATIN SMALL LETTER I CIRCUMFLEX
369 case '\u00ef': replace = "i"; break; // LATIN SMALL LETTER I DIAERESIS
370 case '\u00f2': replace = "o"; break; // LATIN SMALL LETTER O GRAVE
371 case '\u00f3': replace = "o"; break; // LATIN SMALL LETTER O ACUTE
372 case '\u00f4': replace = "o"; break; // LATIN SMALL LETTER O CIRCUMFLEX
373 case '\u00f6': replace = "o"; break; // LATIN SMALL LETTER O DIAERESIS
374 case '\u00f9': replace = "u"; break; // LATIN SMALL LETTER U GRAVE
375 case '\u00fa': replace = "u"; break; // LATIN SMALL LETTER U ACUTE
376 case '\u00fb': replace = "u"; break; // LATIN SMALL LETTER U CIRCUMFLEX
377 case '\u00fc': replace = "u"; break; // LATIN SMALL LETTER U DIAERESIS
378 case '\u0100': replace = "A"; break; // LATIN CAPITAL LETTER A MACRON
379 case '\u0101': replace = "a"; break; // LATIN SMALL LETTER A MACRON
380 case '\u0102': replace = "A"; break; // LATIN CAPITAL LETTER A BREVE
381 case '\u0103': replace = "a"; break; // LATIN SMALL LETTER A BREVE
382 case '\u0112': replace = "E"; break; // LATIN CAPITAL LETTER E MACRON
383 case '\u0113': replace = "e"; break; // LATIN SMALL LETTER E MACRON
384 case '\u0114': replace = "E"; break; // LATIN CAPITAL LETTER E BREVE
385 case '\u0115': replace = "e"; break; // LATIN SMALL LETTER E BREVE
386 case '\u0118': replace = "Ae"; break; // LATIN CAPITAL LETTER E OGONEK
387 case '\u0119': replace = "ae"; break; // LATIN SMALL LETTER E OGONEK
388 case '\u012a': replace = "I"; break; // LATIN CAPITAL LETTER I MACRON
389 case '\u012b': replace = "i"; break; // LATIN SMALL LETTER I MACRON
390 case '\u012c': replace = "I"; break; // LATIN CAPITAL LETTER I BREVE
391 case '\u012d': replace = "i"; break; // LATIN SMALL LETTER I BREVE
392 case '\u014c': replace = "O"; break; // LATIN CAPITAL LETTER O MACRON
393 case '\u014d': replace = "o"; break; // LATIN SMALL LETTER O MACRON
394 case '\u014e': replace = "O"; break; // LATIN CAPITAL LETTER O BREVE
395 case '\u014f': replace = "o"; break; // LATIN SMALL LETTER O BREVE
396 case '\u0152': replace = "Oe"; break; // LATIN CAPITAL LETTER O E
397 case '\u0153': replace = "oe"; break; // LATIN SMALL LETTER O E
398 case '\u016a': replace = "U"; break; // LATIN CAPITAL LETTER U MACRON
399 case '\u016b': replace = "u"; break; // LATIN SMALL LETTER U MACRON
400 case '\u016c': replace = "U"; break; // LATIN CAPITAL LETTER U BREVE
401 case '\u016d': replace = "u"; break; // LATIN SMALL LETTER U BREVE
402 case '\u017f': replace = "s"; break; // LATIN SMALL LETTER LONG S
403 case '\u00df': replace = "ss"; break; // LATIN SMALL LETTER SHARP S
404 case '\u00ad': break; // soft hyphen
405 // new in MPDL project by J. Willenborg
406 case '\u1e14': replace = "e"; break; // LATIN ... LETTER E WITH ...
407 case '\u1e15': replace = "e"; break; // LATIN ... LETTER E WITH ...
408 case '\u1e16': replace = "e"; break; // LATIN ... LETTER E WITH ...
409 case '\u1e17': replace = "e"; break; // LATIN ... LETTER E WITH ...
410 case '\u1e18': replace = "e"; break; // LATIN ... LETTER E WITH ...
411 case '\u1e19': replace = "e"; break; // LATIN ... LETTER E WITH ...
412 case '\u1e1a': replace = "e"; break; // LATIN ... LETTER E WITH ...
413 case '\u1e1b': replace = "e"; break; // LATIN ... LETTER E WITH ...
414 case '\u1e1c': replace = "e"; break; // LATIN ... LETTER E WITH ...
415 case '\u1e1d': replace = "e"; break; // LATIN ... LETTER E WITH ...
416 case '\u1eb8': replace = "e"; break; // LATIN ... LETTER E WITH ...
417 case '\u1eb9': replace = "e"; break; // LATIN ... LETTER E WITH ...
418 case '\u1eba': replace = "e"; break; // LATIN ... LETTER E WITH ...
419 case '\u1ebb': replace = "e"; break; // LATIN ... LETTER E WITH ...
420 case '\u1ebc': replace = "e"; break; // LATIN ... LETTER E WITH ...
421 case '\u1ebd': replace = "e"; break; // LATIN ... LETTER E WITH ...
422 case '\u1ebe': replace = "e"; break; // LATIN ... LETTER E WITH ...
423 case '\u1ebf': replace = "e"; break; // LATIN ... LETTER E WITH ...
424 case '\u1ec0': replace = "e"; break; // LATIN ... LETTER E WITH ...
425 case '\u1ec1': replace = "e"; break; // LATIN ... LETTER E WITH ...
426 case '\u1ec2': replace = "e"; break; // LATIN ... LETTER E WITH ...
427 case '\u1ec3': replace = "e"; break; // LATIN ... LETTER E WITH ...
428 case '\u1ec4': replace = "e"; break; // LATIN ... LETTER E WITH ...
429 case '\u1ec5': replace = "e"; break; // LATIN ... LETTER E WITH ...
430 case '\u1ec6': replace = "e"; break; // LATIN ... LETTER E WITH ...
431 case '\u1ec7': replace = "e"; break; // LATIN ... LETTER E WITH ...
432 // by Malcolm
433 case '\u2329': break; // BRA
434 case '\u232a': break; // KET
435 default: replace += c; break;
436 }
437 buf.append(replace);
438 // update offsets if replacement is a different length
439 if (offsets != null) {
440 int r = replace.length();
441 if (r == 0)
442 this.offsets = arrayKill(this.offsets, i - n);
443 else if (r == 2)
444 this.offsets = arrayInsert(this.offsets, i - n + 1, this.offsets[i - n], r - 1);
445 n += 1 - r;
446 }
447 }
448 return buf.toString();
449 } else if (language.equals("it")) {
450 // new Mpdl code: added by J. Willenborg: some of Malcolms code did not work without errors so it has to be taken away, also all latin stuff is imported
451 StringBuffer buf = new StringBuffer();
452 int n = 0;
453 for (int i = 0; i < s.length(); i++) {
454 char c = s.charAt(i);
455 String replace = new String();
456 switch (c) {
457 case '\u00c0': replace = "A"; break; // LATIN CAPITAL LETTER A GRAVE
458 case '\u00c1': replace = "A"; break; // LATIN CAPITAL LETTER A ACUTE
459 case '\u00c2': replace = "A"; break; // LATIN CAPITAL LETTER A CIRCUMFLEX
460 case '\u00c4': replace = "A"; break; // LATIN CAPITAL LETTER A DIAERESIS
461 case '\u00c6': replace = "Ae"; break; // LATIN CAPITAL LETTER A E
462 case '\u00c7': replace = "C"; break; // LATIN CAPITAL LETTER C CEDILLA
463 case '\u00c8': replace = "E"; break; // LATIN CAPITAL LETTER E GRAVE
464 case '\u00c9': replace = "E"; break; // LATIN CAPITAL LETTER E ACUTE
465 case '\u00ca': replace = "E"; break; // LATIN CAPITAL LETTER E CIRCUMFLEX
466 case '\u00cb': replace = "E"; break; // LATIN CAPITAL LETTER E DIAERESIS
467 case '\u00cc': replace = "I"; break; // LATIN CAPITAL LETTER I GRAVE;
468 case '\u00cd': replace = "I"; break; // LATIN CAPITAL LETTER I ACUTE
469 case '\u00ce': replace = "I"; break; // LATIN CAPITAL LETTER I CIRCUMFLEX
470 case '\u00cf': replace = "I"; break; // LATIN CAPITAL LETTER I DIAERESIS
471 case '\u00d2': replace = "O"; break; // LATIN CAPITAL LETTER O GRAVE
472 case '\u00d3': replace = "O"; break; // LATIN CAPITAL LETTER O ACUTE
473 case '\u00d4': replace = "O"; break; // LATIN CAPITAL LETTER O CIRCUMFLEX
474 case '\u00d6': replace = "O"; break; // LATIN CAPITAL LETTER O DIAERESIS
475 case '\u00d9': replace = "U"; break; // LATIN CAPITAL LETTER U GRAVE
476 case '\u00da': replace = "U"; break; // LATIN CAPITAL LETTER U ACUTE
477 case '\u00db': replace = "U"; break; // LATIN CAPITAL LETTER U CIRCUMFLEX
478 case '\u00dc': replace = "U"; break; // LATIN CAPITAL LETTER U DIAERESIS
479 case '\u00e0': replace = "a"; break; // LATIN SMALL LETTER A GRAVE
480 case '\u00e1': replace = "a"; break; // LATIN SMALL LETTER A ACUTE
481 case '\u00e2': replace = "a"; break; // LATIN SMALL LETTER A CIRCUMFLEX
482 case '\u00e4': replace = "a"; break; // LATIN SMALL LETTER A DIAERESIS
483 case '\u00e6': replace = "ae"; break; // LATIN SMALL LETTER A E
484 case '\u00e7': replace = "c"; break; // LATIN SMALL LETTER C CEDILLA
485 case '\u00e8': replace = "e"; break; // LATIN SMALL LETTER E GRAVE
486 case '\u00e9': replace = "e"; break; // LATIN SMALL LETTER E ACUTE
487 case '\u00ea': replace = "e"; break; // LATIN SMALL LETTER E CIRCUMFLEX
488 case '\u00eb': replace = "e"; break; // LATIN SMALL LETTER E DIAERESIS
489 case '\u00ec': replace = "i"; break; // LATIN SMALL LETTER I GRAVE
490 case '\u00ed': replace = "i"; break; // LATIN SMALL LETTER I ACUTE
491 case '\u00ee': replace = "i"; break; // LATIN SMALL LETTER I CIRCUMFLEX
492 case '\u00ef': replace = "i"; break; // LATIN SMALL LETTER I DIAERESIS
493 case '\u00f2': replace = "o"; break; // LATIN SMALL LETTER O GRAVE
494 case '\u00f3': replace = "o"; break; // LATIN SMALL LETTER O ACUTE
495 case '\u00f4': replace = "o"; break; // LATIN SMALL LETTER O CIRCUMFLEX
496 case '\u00f6': replace = "o"; break; // LATIN SMALL LETTER O DIAERESIS
497 case '\u00f9': replace = "u"; break; // LATIN SMALL LETTER U GRAVE
498 case '\u00fa': replace = "u"; break; // LATIN SMALL LETTER U ACUTE
499 case '\u00fb': replace = "u"; break; // LATIN SMALL LETTER U CIRCUMFLEX
500 case '\u00fc': replace = "u"; break; // LATIN SMALL LETTER U DIAERESIS
501 case '\u0100': replace = "A"; break; // LATIN CAPITAL LETTER A MACRON
502 case '\u0101': replace = "a"; break; // LATIN SMALL LETTER A MACRON
503 case '\u0102': replace = "A"; break; // LATIN CAPITAL LETTER A BREVE
504 case '\u0103': replace = "a"; break; // LATIN SMALL LETTER A BREVE
505 case '\u0112': replace = "E"; break; // LATIN CAPITAL LETTER E MACRON
506 case '\u0113': replace = "e"; break; // LATIN SMALL LETTER E MACRON
507 case '\u0114': replace = "E"; break; // LATIN CAPITAL LETTER E BREVE
508 case '\u0115': replace = "e"; break; // LATIN SMALL LETTER E BREVE
509 case '\u0118': replace = "Ae"; break; // LATIN CAPITAL LETTER E OGONEK
510 case '\u0119': replace = "ae"; break; // LATIN SMALL LETTER E OGONEK
511 case '\u012a': replace = "I"; break; // LATIN CAPITAL LETTER I MACRON
512 case '\u012b': replace = "i"; break; // LATIN SMALL LETTER I MACRON
513 case '\u012c': replace = "I"; break; // LATIN CAPITAL LETTER I BREVE
514 case '\u012d': replace = "i"; break; // LATIN SMALL LETTER I BREVE
515 case '\u014c': replace = "O"; break; // LATIN CAPITAL LETTER O MACRON
516 case '\u014d': replace = "o"; break; // LATIN SMALL LETTER O MACRON
517 case '\u014e': replace = "O"; break; // LATIN CAPITAL LETTER O BREVE
518 case '\u014f': replace = "o"; break; // LATIN SMALL LETTER O BREVE
519 case '\u0152': replace = "Oe"; break; // LATIN CAPITAL LETTER O E
520 case '\u0153': replace = "oe"; break; // LATIN SMALL LETTER O E
521 case '\u016a': replace = "U"; break; // LATIN CAPITAL LETTER U MACRON
522 case '\u016b': replace = "u"; break; // LATIN SMALL LETTER U MACRON
523 case '\u016c': replace = "U"; break; // LATIN CAPITAL LETTER U BREVE
524 case '\u016d': replace = "u"; break; // LATIN SMALL LETTER U BREVE
525 case '\u017f': replace = "s"; break; // LATIN SMALL LETTER LONG S
526 case '\u00df': replace = "ss"; break; // LATIN SMALL LETTER SHARP S
527 // new in MPDL project by J. Willenborg
528 case '\u1e8d': replace = "e"; break; // LATIN SMALL LETTER E WITH TILDE
529 default: replace += c; break;
530 }
531 buf.append(replace);
532 // update offsets if replacement is a different length
533 if (offsets != null) {
534 int r = replace.length();
535 if (r == 0) this.offsets = arrayKill(this.offsets, i - n);
536 else if (r == 2)
537 this.offsets = arrayInsert(this.offsets, i - n + 1, this.offsets[i - n], r - 1);
538 n += 1 - r;
539 }
540 }
541 return buf.toString();
542 // new Mpdl code: added by J. Willenborg: most of the latin replacements also in english
543 } else if (language.equals("en")) {
544 StringBuffer buf = new StringBuffer();
545 int n = 0;
546 for (int i = 0; i < s.length(); i++) {
547 char c = s.charAt(i);
548 String replace = new String();
549 switch (c) {
550 case '\u0300': replace = ""; break; // COMBINING GRAVE ACCENT
551 case '\u0301': replace = ""; break; // COMBINING ACCUTE ACCENT
552 case '\u0302': replace = ""; break; // COMBINING CIRCUMFLEX ACCENT
553
554 case '\u00c0': replace = "A"; break; // LATIN CAPITAL LETTER A GRAVE
555 case '\u00c1': replace = "A"; break; // LATIN CAPITAL LETTER A ACUTE
556 case '\u00c2': replace = "A"; break; // LATIN CAPITAL LETTER A CIRCUMFLEX
557 case '\u00c4': replace = "A"; break; // LATIN CAPITAL LETTER A DIAERESIS
558 case '\u00c6': replace = "Ae"; break; // LATIN CAPITAL LETTER A E
559 case '\u00c7': replace = "C"; break; // LATIN CAPITAL LETTER C CEDILLA
560 case '\u00c8': replace = "E"; break; // LATIN CAPITAL LETTER E GRAVE
561 case '\u00c9': replace = "E"; break; // LATIN CAPITAL LETTER E ACUTE
562 case '\u00ca': replace = "E"; break; // LATIN CAPITAL LETTER E CIRCUMFLEX
563 case '\u00cb': replace = "E"; break; // LATIN CAPITAL LETTER E DIAERESIS
564 case '\u00cc': replace = "I"; break; // LATIN CAPITAL LETTER I GRAVE;
565 case '\u00cd': replace = "I"; break; // LATIN CAPITAL LETTER I ACUTE
566 case '\u00ce': replace = "I"; break; // LATIN CAPITAL LETTER I CIRCUMFLEX
567 case '\u00cf': replace = "I"; break; // LATIN CAPITAL LETTER I DIAERESIS
568 case '\u00d2': replace = "O"; break; // LATIN CAPITAL LETTER O GRAVE
569 case '\u00d3': replace = "O"; break; // LATIN CAPITAL LETTER O ACUTE
570 case '\u00d4': replace = "O"; break; // LATIN CAPITAL LETTER O CIRCUMFLEX
571 case '\u00d6': replace = "O"; break; // LATIN CAPITAL LETTER O DIAERESIS
572 case '\u00d9': replace = "U"; break; // LATIN CAPITAL LETTER U GRAVE
573 case '\u00da': replace = "U"; break; // LATIN CAPITAL LETTER U ACUTE
574 case '\u00db': replace = "U"; break; // LATIN CAPITAL LETTER U CIRCUMFLEX
575 case '\u00dc': replace = "U"; break; // LATIN CAPITAL LETTER U DIAERESIS
576 case '\u00e0': replace = "a"; break; // LATIN SMALL LETTER A GRAVE
577 case '\u00e1': replace = "a"; break; // LATIN SMALL LETTER A ACUTE
578 case '\u00e2': replace = "a"; break; // LATIN SMALL LETTER A CIRCUMFLEX
579 case '\u00e4': replace = "a"; break; // LATIN SMALL LETTER A DIAERESIS
580 case '\u00e6': replace = "ae"; break; // LATIN SMALL LETTER A E
581 case '\u00e7': replace = "c"; break; // LATIN SMALL LETTER C CEDILLA
582 case '\u00e8': replace = "e"; break; // LATIN SMALL LETTER E GRAVE
583 case '\u00e9': replace = "e"; break; // LATIN SMALL LETTER E ACUTE
584 case '\u00ea': replace = "e"; break; // LATIN SMALL LETTER E CIRCUMFLEX
585 case '\u00eb': replace = "e"; break; // LATIN SMALL LETTER E DIAERESIS
586 case '\u00ec': replace = "i"; break; // LATIN SMALL LETTER I GRAVE
587 case '\u00ed': replace = "i"; break; // LATIN SMALL LETTER I ACUTE
588 case '\u00ee': replace = "i"; break; // LATIN SMALL LETTER I CIRCUMFLEX
589 case '\u00ef': replace = "i"; break; // LATIN SMALL LETTER I DIAERESIS
590 case '\u00f2': replace = "o"; break; // LATIN SMALL LETTER O GRAVE
591 case '\u00f3': replace = "o"; break; // LATIN SMALL LETTER O ACUTE
592 case '\u00f4': replace = "o"; break; // LATIN SMALL LETTER O CIRCUMFLEX
593 case '\u00f6': replace = "o"; break; // LATIN SMALL LETTER O DIAERESIS
594 case '\u00f9': replace = "u"; break; // LATIN SMALL LETTER U GRAVE
595 case '\u00fa': replace = "u"; break; // LATIN SMALL LETTER U ACUTE
596 case '\u00fb': replace = "u"; break; // LATIN SMALL LETTER U CIRCUMFLEX
597 case '\u00fc': replace = "u"; break; // LATIN SMALL LETTER U DIAERESIS
598 case '\u0100': replace = "A"; break; // LATIN CAPITAL LETTER A MACRON
599 case '\u0101': replace = "a"; break; // LATIN SMALL LETTER A MACRON
600 case '\u0102': replace = "A"; break; // LATIN CAPITAL LETTER A BREVE
601 case '\u0103': replace = "a"; break; // LATIN SMALL LETTER A BREVE
602 case '\u0112': replace = "E"; break; // LATIN CAPITAL LETTER E MACRON
603 case '\u0113': replace = "e"; break; // LATIN SMALL LETTER E MACRON
604 case '\u0114': replace = "E"; break; // LATIN CAPITAL LETTER E BREVE
605 case '\u0115': replace = "e"; break; // LATIN SMALL LETTER E BREVE
606 case '\u0118': replace = "Ae"; break; // LATIN CAPITAL LETTER E OGONEK
607 case '\u0119': replace = "ae"; break; // LATIN SMALL LETTER E OGONEK
608 case '\u012a': replace = "I"; break; // LATIN CAPITAL LETTER I MACRON
609 case '\u012b': replace = "i"; break; // LATIN SMALL LETTER I MACRON
610 case '\u012c': replace = "I"; break; // LATIN CAPITAL LETTER I BREVE
611 case '\u012d': replace = "i"; break; // LATIN SMALL LETTER I BREVE
612 case '\u014c': replace = "O"; break; // LATIN CAPITAL LETTER O MACRON
613 case '\u014d': replace = "o"; break; // LATIN SMALL LETTER O MACRON
614 case '\u014e': replace = "O"; break; // LATIN CAPITAL LETTER O BREVE
615 case '\u014f': replace = "o"; break; // LATIN SMALL LETTER O BREVE
616 case '\u0152': replace = "Oe"; break; // LATIN CAPITAL LETTER O E
617 case '\u0153': replace = "oe"; break; // LATIN SMALL LETTER O E
618 case '\u016a': replace = "U"; break; // LATIN CAPITAL LETTER U MACRON
619 case '\u016b': replace = "u"; break; // LATIN SMALL LETTER U MACRON
620 case '\u016c': replace = "U"; break; // LATIN CAPITAL LETTER U BREVE
621 case '\u016d': replace = "u"; break; // LATIN SMALL LETTER U BREVE
622 case '\u017f': replace = "s"; break; // LATIN SMALL LETTER LONG S
623 case '\u00df': replace = "ss"; break; // LATIN SMALL LETTER SHARP S
624 // new in MPDL project by J. Willenborg
625 case '\u1e8d': replace = "e"; break; // LATIN SMALL LETTER E WITH TILDE
626 // by Malcolm
627 case '\u00ad': break; // soft hyphen
628 case '\u2329': break; // BRA
629 case '\u232a': break; // KET
630 default: replace += c; break;
631 }
632 buf.append(replace);
633 // update offsets if replacement is a different length
634 if (offsets != null) {
635 int r = replace.length();
636 if (r == 0)
637 this.offsets = arrayKill(this.offsets, i - n);
638 else if (r == 2)
639 this.offsets = arrayInsert(this.offsets, i - n + 1, this.offsets[i - n], r - 1);
640 n += 1 - r;
641 }
642 }
643 return buf.toString();
644 } else if (language.equals("fr")) {
645 // new Mpdl code: added by J. Willenborg: some of Malcolms code did not work without errors so it has to be taken away
646 StringBuffer buf = new StringBuffer();
647 int n = 0;
648 for (int i = 0; i < s.length(); i++) {
649 char c = s.charAt(i);
650 String replace = new String();
651 switch (c) {
652 case '\u00e6': replace = "ae"; break; // LATIN SMALL LETTER A E
653 case '\u017f': replace = "s"; break; // LATIN SMALL LETTER LONG S
654 case '\u00df': replace = "ss"; break; // LATIN SMALL LETTER SHARP S
655 case '\u00ad': break; // soft hyphen
656 case '-': break;
657 default: replace += c; break;
658 }
659 buf.append(replace);
660 // update offsets if replacement is a different length
661 if (offsets != null) {
662 int r = replace.length();
663 if (r == 0)
664 this.offsets = arrayKill(this.offsets, i - n);
665 else if (r == 2)
666 this.offsets = arrayInsert(this.offsets, i - n + 1, this.offsets[i - n], r - 1);
667 n += 1 - r;
668 }
669 }
670 return buf.toString();
671 } else if (language.equals("de")) {
672 StringBuffer buf = new StringBuffer();
673 int n = 0;
674 for (int i = 0; i < s.length(); i++) {
675 char c = s.charAt(i);
676 String replace = new String();
677 switch (c) {
678 case '\u00c4': replace = "Ae"; break;
679 case '\u00d6': replace = "Oe"; break;
680 case '\u00dc': replace = "Ue"; break;
681 case '\u00df': replace = "ss"; break;
682 case '\u00e4': replace = "ae"; break;
683 case '\u00f6': replace = "oe"; break;
684 case '\u00fc': replace = "ue"; break;
685 case '\u00ad': break; // soft hyphen
686 case '\u00e9': replace = "e"; break;
687 // new in MPDL project by J. Willenborg
688 case '\u017f': replace = "s"; break; // LATIN SMALL LETTER LONG S
689 // case '-': break;
690 default: replace += c; break;
691 }
692 buf.append(replace);
693 // update offsets if replacement is a different length
694 if (offsets != null) {
695 int r = replace.length();
696 if (r == 0)
697 this.offsets = arrayKill(this.offsets, i - n);
698 else if (r == 2)
699 this.offsets = arrayInsert(this.offsets, i - n + 1, this.offsets[i - n], r - 1);
700 n += 1 - r;
701 }
702 }
703 return buf.toString();
704 } else if (language.equals("zh")) {
705 StringBuffer buf = new StringBuffer();
706 int n = 0;
707 for (int i = 0; i < s.length(); i++) {
708 char c = s.charAt(i);
709 String replace = new String();
710 switch (c) {
711 case '\u00b9': replace = "1"; break;
712 case '\u00b2': replace = "2"; break;
713 case '\u00b3': replace = "3"; break;
714 case '\u2074': replace = "4"; break;
715 case '\u2075': replace = "5"; break;
716 // original by Malcolm Hyman: with the following replacements
717 // case '\u3000': replace = " "; break;
718 // case '\u3001': replace = ","; break;
719 // case '\u3002': replace = "."; break;
720 // case '\u200b': break; // BREAKS EVERYTHING!
721 default: replace += c; break;
722 }
723 buf.append(replace);
724 // update offsets if replacement is a different length
725 if (offsets != null) {
726 int r = replace.length();
727 if (r == 0)
728 this.offsets = arrayKill(this.offsets, i - n);
729 else if (r == 2)
730 this.offsets = arrayInsert(this.offsets, i - n + 1, this.offsets[i - n], r - 1);
731 n += 1 - r;
732 }
733 }
734 return buf.toString();
735 } else if (language.equals("akk") ||
736 language.equals("qam") ||
737 language.equals("qpc") ||
738 language.equals("elx") ||
739 language.equals("sux") ||
740 language.equals("hit") ||
741 language.equals("qhu") ||
742 language.equals("peo") ||
743 language.equals("uga") ||
744 language.equals("ura") ||
745 language.equals("qcu")) {
746 StringBuffer buf = new StringBuffer();
747 int n = 0;
748 char last = '\u0000';
749 for (int i = 0; i < s.length(); i++) {
750 char c = s.charAt(i);
751 c = Character.toLowerCase(c);
752 String replace = new String();
753 switch (c) {
754 case '{': replace += "-"; break;
755 case '}': replace += "-"; break;
756 // These are from PSD::ATF::Unicode by Steve Tinney
757 case '\u0161': replace += "sz"; break;
758 case '\u1e63': replace += "s,"; break;
759 case '\u1e6d': replace += "t,"; break;
760 case '\u014b': replace += "j"; break;
761 case '\u015b': replace += "s'"; break;
762 case '\u2080': replace += "0"; break;
763 case '\u2081': replace += "1"; break;
764 case '\u2082': replace += "2"; break;
765 case '\u2083': replace += "3"; break;
766 case '\u2084': replace += "4"; break;
767 case '\u2085': replace += "5"; break;
768 case '\u2086': replace += "6"; break;
769 case '\u2087': replace += "7"; break;
770 case '\u2088': replace += "8"; break;
771 case '\u2089': replace += "9"; break;
772
773 case 'c': // shin (except where used as modifier)
774 if ((i > 0) && ((last == '~') || (last == '@')))
775 replace += "c";
776 else replace += "sz";
777 break;
778 default: replace += c; break;
779 }
780 // suppress grapheme boundary before or after word boundary
781 if (replace.equals("-")) {
782 if ((i + 1 == s.length()) || (s.charAt(i + 1) == ' ') || (i == 0) || (s.charAt(i - 1) == ' '))
783 replace = "";
784 }
785 last = c;
786 buf.append(replace);
787 // update offsets if replacement is a different length
788 if (offsets != null) {
789 int r = replace.length();
790 if (r == 0)
791 this.offsets = arrayKill(this.offsets, i - n);
792 else if (r == 2)
793 this.offsets = arrayInsert(this.offsets, i - n + 1, this.offsets[i - n], r - 1);
794 n += 1 - r;
795 }
796 }
797 return buf.toString();
798 } else if (language.equals("el") || language.equals("grc")) {
799 StringBuffer buf = new StringBuffer();
800 int n = 0;
801 for (int i = 0; i < s.length(); i++) {
802 char c = s.charAt(i);
803 String replace = new String();
804 switch (c) {
805 case '\u03c2': replace = "\u03c3"; break; // GREEK SMALL LETTER FINAL SIGMA
806 case '<': break;
807 case '>': break;
808 case '[': break;
809 case ']': break;
810 case '1': break;
811 case '2': break;
812 case '\u03ac': replace = "\u1f71"; break;
813 case '\u03ad': replace = "\u1f73"; break;
814 case '\u03ae': replace = "\u1f75"; break;
815 case '\u03af': replace = "\u1f77"; break;
816 case '\u03cc': replace = "\u1f79"; break;
817 case '\u03cd': replace = "\u1f7b"; break;
818 case '\u03ce': replace = "\u1f7d"; break;
819 case '-': break; // same treatment as soft hyphen
820 case '\u00ad': break; // soft hyphen
821 default: replace += c; break;
822 }
823 buf.append(replace);
824 // update offsets if replacement is a different length
825 if (offsets != null) {
826 int r = replace.length();
827 if (r == 0)
828 this.offsets = arrayKill(this.offsets, i - n);
829 else if (r == 2)
830 this.offsets = arrayInsert(this.offsets, i - n + 1, this.offsets[i - n], r - 1);
831 n += 1 - r;
832 }
833 }
834 return buf.toString();
835 } else if (language.equals("el_atonic")) {
836 StringBuffer buf = new StringBuffer();
837 int n = 0;
838 for (int i = 0; i < s.length(); i++) {
839 char c = s.charAt(i);
840 String replace = new String();
841 switch (c) {
842 case '\u03c2': replace = "\u03c3"; break; // GREEK SMALL LETTER FINAL SIGMA
843 // map characters with diacritics to their plain equivalent
844 // cf. <code>BetaCode.java</code>
845 case '\u03aa': replace = "\u0399"; break;
846 case '\u03ab': replace = "\u03a5"; break;
847 case '\u03ac': replace = "\u0381"; break;
848 case '\u03ad': replace = "\u0385"; break;
849 case '\u03ae': replace = "\u0387"; break;
850 case '\u03af': replace = "\u0389"; break;
851 case '\u03ca': replace = "\u03b9"; break;
852 case '\u03cb': replace = "\u03c5"; break;
853 case '\u03cc': replace = "\u03bf"; break;
854 case '\u03cd': replace = "\u03c5"; break;
855 case '\u03ce': replace = "\u03c9"; break;
856 case '\u1f00': replace = "\u03b1"; break;
857 case '\u1f01': replace = "\u03b1"; break;
858 case '\u1f02': replace = "\u03b1"; break;
859 case '\u1f03': replace = "\u03b1"; break;
860 case '\u1f04': replace = "\u03b1"; break;
861 case '\u1f05': replace = "\u03b1"; break;
862 case '\u1f06': replace = "\u03b1"; break;
863 case '\u1f07': replace = "\u03b1"; break;
864 case '\u1f08': replace = "\u0391"; break;
865 case '\u1f09': replace = "\u0391"; break;
866 case '\u1f0a': replace = "\u0391"; break;
867 case '\u1f0b': replace = "\u0391"; break;
868 case '\u1f0c': replace = "\u0391"; break;
869 case '\u1f0d': replace = "\u0391"; break;
870 case '\u1f0e': replace = "\u0391"; break;
871 case '\u1f0f': replace = "\u0391"; break;
872 case '\u1f10': replace = "\u03b5"; break;
873 case '\u1f11': replace = "\u03b5"; break;
874 case '\u1f12': replace = "\u03b5"; break;
875 case '\u1f13': replace = "\u03b5"; break;
876 case '\u1f14': replace = "\u03b5"; break;
877 case '\u1f15': replace = "\u03b5"; break;
878 case '\u1f18': replace = "\u0395"; break;
879 case '\u1f19': replace = "\u0395"; break;
880 case '\u1f1a': replace = "\u0395"; break;
881 case '\u1f1b': replace = "\u0395"; break;
882 case '\u1f1c': replace = "\u0395"; break;
883 case '\u1f1d': replace = "\u0395"; break;
884 case '\u1f20': replace = "\u03b7"; break;
885 case '\u1f21': replace = "\u03b7"; break;
886 case '\u1f22': replace = "\u03b7"; break;
887 case '\u1f23': replace = "\u03b7"; break;
888 case '\u1f24': replace = "\u03b7"; break;
889 case '\u1f25': replace = "\u03b7"; break;
890 case '\u1f26': replace = "\u03b7"; break;
891 case '\u1f27': replace = "\u03b7"; break;
892 case '\u1f28': replace = "\u0397"; break;
893 case '\u1f29': replace = "\u0397"; break;
894 case '\u1f2a': replace = "\u0397"; break;
895 case '\u1f2b': replace = "\u0397"; break;
896 case '\u1f2c': replace = "\u0397"; break;
897 case '\u1f2d': replace = "\u0397"; break;
898 case '\u1f2e': replace = "\u0397"; break;
899 case '\u1f2f': replace = "\u0397"; break;
900 case '\u1f30': replace = "\u03b9"; break;
901 case '\u1f31': replace = "\u03b9"; break;
902 case '\u1f32': replace = "\u03b9"; break;
903 case '\u1f33': replace = "\u03b9"; break;
904 case '\u1f34': replace = "\u03b9"; break;
905 case '\u1f35': replace = "\u03b9"; break;
906 case '\u1f36': replace = "\u03b9"; break;
907 case '\u1f37': replace = "\u03b9"; break;
908 case '\u1f38': replace = "\u0399"; break;
909 case '\u1f39': replace = "\u0399"; break;
910 case '\u1f3a': replace = "\u0399"; break;
911 case '\u1f3b': replace = "\u0399"; break;
912 case '\u1f3c': replace = "\u0399"; break;
913 case '\u1f3d': replace = "\u0399"; break;
914 case '\u1f3e': replace = "\u0399"; break;
915 case '\u1f3f': replace = "\u0399"; break;
916 case '\u1f40': replace = "\u03bf"; break;
917 case '\u1f41': replace = "\u03bf"; break;
918 case '\u1f42': replace = "\u03bf"; break;
919 case '\u1f43': replace = "\u03bf"; break;
920 case '\u1f44': replace = "\u03bf"; break;
921 case '\u1f45': replace = "\u03bf"; break;
922 case '\u1f48': replace = "\u039f"; break;
923 case '\u1f49': replace = "\u039f"; break;
924 case '\u1f4a': replace = "\u039f"; break;
925 case '\u1f4b': replace = "\u039f"; break;
926 case '\u1f4c': replace = "\u039f"; break;
927 case '\u1f4d': replace = "\u039f"; break;
928 case '\u1f50': replace = "\u03c5"; break;
929 case '\u1f51': replace = "\u03c5"; break;
930 case '\u1f52': replace = "\u03c5"; break;
931 case '\u1f53': replace = "\u03c5"; break;
932 case '\u1f54': replace = "\u03c5"; break;
933 case '\u1f55': replace = "\u03c5"; break;
934 case '\u1f56': replace = "\u03c5"; break;
935 case '\u1f57': replace = "\u03c5"; break;
936 case '\u1f58': replace = "\u03a5"; break;
937 case '\u1f59': replace = "\u03a5"; break;
938 case '\u1f5a': replace = "\u03a5"; break;
939 case '\u1f5b': replace = "\u03a5"; break;
940 case '\u1f5c': replace = "\u03a5"; break;
941 case '\u1f5d': replace = "\u03a5"; break;
942 case '\u1f5e': replace = "\u03a5"; break;
943 case '\u1f5f': replace = "\u03a5"; break;
944 case '\u1f60': replace = "\u03c9"; break;
945 case '\u1f61': replace = "\u03c9"; break;
946 case '\u1f62': replace = "\u03c9"; break;
947 case '\u1f63': replace = "\u03c9"; break;
948 case '\u1f64': replace = "\u03c9"; break;
949 case '\u1f65': replace = "\u03c9"; break;
950 case '\u1f66': replace = "\u03c9"; break;
951 case '\u1f67': replace = "\u03c9"; break;
952 case '\u1f68': replace = "\u03a9"; break;
953 case '\u1f69': replace = "\u03a9"; break;
954 case '\u1f6a': replace = "\u03a9"; break;
955 case '\u1f6b': replace = "\u03a9"; break;
956 case '\u1f6c': replace = "\u03a9"; break;
957 case '\u1f6d': replace = "\u03a9"; break;
958 case '\u1f6e': replace = "\u03a9"; break;
959 case '\u1f6f': replace = "\u03a9"; break;
960 case '\u1f70': replace = "\u03b1"; break;
961 case '\u1f71': replace = "\u03b1"; break;
962 case '\u1f72': replace = "\u03b5"; break;
963 case '\u1f73': replace = "\u03b5"; break;
964 case '\u1f74': replace = "\u03b7"; break;
965 case '\u1f75': replace = "\u03b7"; break;
966 case '\u1f76': replace = "\u03b9"; break;
967 case '\u1f77': replace = "\u03b9"; break;
968 case '\u1f78': replace = "\u03bf"; break;
969 case '\u1f79': replace = "\u03bf"; break;
970 case '\u1f7a': replace = "\u03c5"; break;
971 case '\u1f7b': replace = "\u03c5"; break;
972 case '\u1f7c': replace = "\u03c9"; break;
973 case '\u1f7d': replace = "\u03c9"; break;
974 case '\u1f80': replace = "\u03b1"; break;
975 case '\u1f81': replace = "\u03b1"; break;
976 case '\u1f82': replace = "\u03b1"; break;
977 case '\u1f83': replace = "\u03b1"; break;
978 case '\u1f84': replace = "\u03b1"; break;
979 case '\u1f85': replace = "\u03b1"; break;
980 case '\u1f86': replace = "\u03b1"; break;
981 case '\u1f87': replace = "\u03b1"; break;
982 case '\u1f88': replace = "\u0391"; break;
983 case '\u1f89': replace = "\u0391"; break;
984 case '\u1f8a': replace = "\u0391"; break;
985 case '\u1f8b': replace = "\u0391"; break;
986 case '\u1f8c': replace = "\u0391"; break;
987 case '\u1f8d': replace = "\u0391"; break;
988 case '\u1f8e': replace = "\u0391"; break;
989 case '\u1f8f': replace = "\u0391"; break;
990 case '\u1f90': replace = "\u03b7"; break;
991 case '\u1f91': replace = "\u03b7"; break;
992 case '\u1f92': replace = "\u03b7"; break;
993 case '\u1f93': replace = "\u03b7"; break;
994 case '\u1f94': replace = "\u03b7"; break;
995 case '\u1f95': replace = "\u03b7"; break;
996 case '\u1f96': replace = "\u03b7"; break;
997 case '\u1f97': replace = "\u03b7"; break;
998 case '\u1f98': replace = "\u0397"; break;
999 case '\u1f99': replace = "\u0397"; break;
1000 case '\u1f9a': replace = "\u0397"; break;
1001 case '\u1f9b': replace = "\u0397"; break;
1002 case '\u1f9c': replace = "\u0397"; break;
1003 case '\u1f9d': replace = "\u0397"; break;
1004 case '\u1f9e': replace = "\u0397"; break;
1005 case '\u1f9f': replace = "\u0397"; break;
1006 case '\u1fa0': replace = "\u03c9"; break;
1007 case '\u1fa1': replace = "\u03c9"; break;
1008 case '\u1fa2': replace = "\u03c9"; break;
1009 case '\u1fa3': replace = "\u03c9"; break;
1010 case '\u1fa4': replace = "\u03c9"; break;
1011 case '\u1fa5': replace = "\u03c9"; break;
1012 case '\u1fa6': replace = "\u03c9"; break;
1013 case '\u1fa7': replace = "\u03c9"; break;
1014 case '\u1fa8': replace = "\u03a9"; break;
1015 case '\u1fa9': replace = "\u03a9"; break;
1016 case '\u1faa': replace = "\u03a9"; break;
1017 case '\u1fab': replace = "\u03a9"; break;
1018 case '\u1fac': replace = "\u03a9"; break;
1019 case '\u1fad': replace = "\u03a9"; break;
1020 case '\u1fae': replace = "\u03a9"; break;
1021 case '\u1faf': replace = "\u03a9"; break;
1022 case '\u1fb2': replace = "\u03b1"; break;
1023 case '\u1fb3': replace = "\u03b1"; break;
1024 case '\u1fb4': replace = "\u03b1"; break;
1025 case '\u1fb6': replace = "\u03b1"; break;
1026 case '\u1fb7': replace = "\u03b1"; break;
1027 case '\u1fba': replace = "\u0391"; break;
1028 case '\u1fbb': replace = "\u0391"; break;
1029 case '\u1fbc': replace = "\u0391"; break;
1030 case '\u1fc2': replace = "\u03b7"; break;
1031 case '\u1fc3': replace = "\u03b7"; break;
1032 case '\u1fc4': replace = "\u03b7"; break;
1033 case '\u1fc6': replace = "\u03b7"; break;
1034 case '\u1fc7': replace = "\u03b7"; break;
1035 case '\u1fca': replace = "\u0397"; break;
1036 case '\u1fcb': replace = "\u0397"; break;
1037 case '\u1fcc': replace = "\u0397"; break;
1038 case '\u1fd2': replace = "\u03b9"; break;
1039 case '\u1fd3': replace = "\u03b9"; break;
1040 case '\u1fd6': replace = "\u03b9"; break;
1041 case '\u1fd7': replace = "\u03b9"; break;
1042 case '\u1fda': replace = "\u0399"; break;
1043 case '\u1fdb': replace = "\u039f"; break;
1044 case '\u1fe2': replace = "\u03c5"; break;
1045 case '\u1fe3': replace = "\u03c5"; break;
1046 case '\u1fe4': replace = "\u03c1"; break;
1047 case '\u1fe5': replace = "\u03c1"; break;
1048 case '\u1fe6': replace = "\u03c5"; break;
1049 case '\u1fe7': replace = "\u03c5"; break;
1050 case '\u1fea': replace = "\u03a5"; break;
1051 case '\u1feb': replace = "\u03a5"; break;
1052 case '\u1fec': replace = "\u03a1"; break;
1053 case '\u1ff2': replace = "\u03c9"; break;
1054 case '\u1ff3': replace = "\u03c9"; break;
1055 case '\u1ff4': replace = "\u03c9"; break;
1056 case '\u1ff6': replace = "\u03c9"; break;
1057 case '\u1ff7': replace = "\u03c9"; break;
1058 case '\u1ff8': replace = "\u039f"; break;
1059 case '\u1ff9': replace = "\u039f"; break;
1060 case '\u1ffa': replace = "\u03a9"; break;
1061 case '\u1ffb': replace = "\u03a9"; break;
1062 case '\u1ffc': replace = "\u03a9"; break;
1063
1064 case '<': break;
1065 case '>': break;
1066 case '-': break; // same treatment as soft hyphen
1067 case '\u00ad': break; // soft hyphen
1068 default: replace += c; break;
1069 }
1070 buf.append(replace);
1071 // update offsets if replacement is a different length
1072 if (offsets != null) {
1073 int r = replace.length();
1074 if (r == 0)
1075 this.offsets = arrayKill(this.offsets, i - n);
1076 else if (r == 2)
1077 this.offsets = arrayInsert(this.offsets, i - n + 1, this.offsets[i - n], r - 1);
1078 n += 1 - r;
1079 }
1080 }
1081 return buf.toString();
1082 } else { // unknown or no language
1083 return s;
1084 }
1085 }
1086
1087 /*
1088 // explicit words
1089 normStr = normStr.replaceAll("aliàs", "alias");
1090 normStr = normStr.replaceAll("hîc", "hic");
1091 normStr = normStr.replaceAll("quòd", "quod");
1092 normStr = normStr.replaceAll("Quòd", "Quod");
1093 normStr = normStr.replaceAll("QVòd", "Quod");
1094 normStr = normStr.replaceAll("Cùmque", "Cumque");
1095 normStr = normStr.replaceAll("aër", "aer");
1096 // ij
1097 normStr = normStr.replaceAll("ij", "ii");
1098 // qu/qv
1099 normStr = normStr.replaceAll("qv", "qu");
1100 // normStr = normStr.replaceAll("qV", "qU");
1101 normStr = normStr.replaceAll("Qv", "Qu");
1102 normStr = normStr.replaceAll("QV", "QU");
1103 // u/v
1104 String vowels = getVowels();
1105 String consonants = getConsonants();
1106 normStr = normStr.replaceAll("([" + vowels + "])([-]*)u([" + vowels +"])", "$1$2v$3"); // vowel + u + vowel --> vowel + v + vowel
1107 normStr = normStr.replaceAll("([" + vowels + "])([-]*)U([" + vowels +"])", "$1$2V$3"); // vowel + U + vowel --> vowel + V + vowel
1108 normStr = normStr.replaceAll("([" + consonants + "])([-]*)v([" + consonants +"])", "$1$2u$3"); // consonant + v + consonant --> consonant + u + consonant
1109 normStr = normStr.replaceAll("([" + consonants + "])([-]*)V([" + consonants +"])", "$1$2U$3"); // consonant + V + consonant --> consonant + U + consonant
1110 normStr = normStr.replaceAll("^v([" + consonants +"])", "u$1"); // v + consonant --> u + consonant
1111 normStr = normStr.replaceAll("^V([" + consonants +"])", "U$1"); // V + consonant --> U + consonant
1112 // end of word: diacritica
1113 normStr = normStr.replaceAll("à$", "a");
1114 normStr = normStr.replaceAll("è$", "e");
1115 normStr = normStr.replaceAll("ò$", "o");
1116 normStr = normStr.replaceAll("àm$", "am");
1117 normStr = normStr.replaceAll("ùm$", "um");
1118 String normStrTmp = normStr;
1119 normStr = "";
1120 for (int i = 0; i < normStrTmp.length(); i++) {
1121 char c = normStrTmp.charAt(i);
1122 String replace = "";
1123 switch (c) {
1124 case 'ſ': replace = "s"; break;
1125 case 'ß': replace = "ss"; break;
1126 case 'æ': replace = "ae"; break;
1127 case 'Æ': replace = "AE"; break;
1128 case 'ę': replace = "ae"; break;
1129 case 'œ': replace = "oe"; break;
1130 default: replace += c; break;
1131 }
1132 normStr = normStr + replace;
1133 }
1134
1135
1136 private String getVowels() {
1137 String retStr = null;
1138 if (Language.getInstance().isItalian(language)) {
1139 retStr = "AEIOUaeiou" +
1140 "\u00c6\u00e6" + // AE ligatures
1141 "\u0152\u0153"; // OE ligatures
1142 } else if (Language.getInstance().isLatin(language)) {
1143 retStr = "AEIOUaeiouÆœęàèòù";
1144 }
1145 return retStr;
1146 }
1147
1148 private String getConsonants() {
1149 String retStr = null;
1150 if (Language.getInstance().isItalian(language)) {
1151 retStr = "BCDFGHKLMNPQRSTVWXZ" +
1152 "bcdfghklmnpqrstvwxz" +
1153 "ſß"; // long/sharp S
1154 } else if (Language.getInstance().isLatin(language)) {
1155 retStr = "BCDFGHKLMNPQRSTVWXZ" +
1156 "bcdfghklmnpqrstvwxz" +
1157 "ſß"; // long/sharp S
1158 }
1159 return retStr;
1160 }
1161
1162
1163
1164
1165
1166 *
1167 *
1168 *
1169 *
1170 */
1171
1172
1173
1174
1175
1176
1177 /**
1178 * Returns a copy of an integer array with the element at
1179 * <code>index</code> removed ("killed").
1180 *
1181 * @param array integer array
1182 * @param index index of element to remove
1183 */
1184 private int[] arrayKill(int[] array, int index) {
1185 int[] newArray = new int[array.length - 1];
1186 System.arraycopy(array, 0, newArray, 0, index);
1187 System.arraycopy(array, index + 1, newArray, index, array.length - index - 1);
1188 return newArray;
1189 }
1190
1191 /**
1192 * Returns a copy of an integer array with <code>count</code> elements
1193 * inserted at <code>index</code>.
1194 *
1195 * @param array integer array
1196 * @param index index to insert new elements
1197 * @param value value to insert into new slots
1198 * @param count number of new slots to insert
1199 */
1200 private int[] arrayInsert(int[] array, int index, int value, int count) {
1201 int[] newArray = new int[array.length + count];
1202 System.arraycopy(array, 0, newArray, 0, index);
1203 for (int i = 0; i < count; i++) newArray[index + i] = value;
1204 System.arraycopy(array, index, newArray, index + count, array.length - index);
1205 return newArray;
1206 }
1207
1208 }