| 1 | /* BreakIterator.java -- Breaks text into elements
|
|---|
| 2 | Copyright (C) 1998, 1999, 2001 Free Software Foundation, Inc.
|
|---|
| 3 |
|
|---|
| 4 | This file is part of GNU Classpath.
|
|---|
| 5 |
|
|---|
| 6 | GNU Classpath is free software; you can redistribute it and/or modify
|
|---|
| 7 | it under the terms of the GNU General Public License as published by
|
|---|
| 8 | the Free Software Foundation; either version 2, or (at your option)
|
|---|
| 9 | any later version.
|
|---|
| 10 |
|
|---|
| 11 | GNU Classpath is distributed in the hope that it will be useful, but
|
|---|
| 12 | WITHOUT ANY WARRANTY; without even the implied warranty of
|
|---|
| 13 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|---|
| 14 | General Public License for more details.
|
|---|
| 15 |
|
|---|
| 16 | You should have received a copy of the GNU General Public License
|
|---|
| 17 | along with GNU Classpath; see the file COPYING. If not, write to the
|
|---|
| 18 | Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
|
|---|
| 19 | 02111-1307 USA.
|
|---|
| 20 |
|
|---|
| 21 | Linking this library statically or dynamically with other modules is
|
|---|
| 22 | making a combined work based on this library. Thus, the terms and
|
|---|
| 23 | conditions of the GNU General Public License cover the whole
|
|---|
| 24 | combination.
|
|---|
| 25 |
|
|---|
| 26 | As a special exception, the copyright holders of this library give you
|
|---|
| 27 | permission to link this library with independent modules to produce an
|
|---|
| 28 | executable, regardless of the license terms of these independent
|
|---|
| 29 | modules, and to copy and distribute the resulting executable under
|
|---|
| 30 | terms of your choice, provided that you also meet, for each linked
|
|---|
| 31 | independent module, the terms and conditions of the license of that
|
|---|
| 32 | module. An independent module is a module which is not derived from
|
|---|
| 33 | or based on this library. If you modify this library, you may extend
|
|---|
| 34 | this exception to your version of the library, but you are not
|
|---|
| 35 | obligated to do so. If you do not wish to do so, delete this
|
|---|
| 36 | exception statement from your version. */
|
|---|
| 37 |
|
|---|
| 38 |
|
|---|
| 39 | package java.text;
|
|---|
| 40 |
|
|---|
| 41 | import java.util.Locale;
|
|---|
| 42 | import java.util.MissingResourceException;
|
|---|
| 43 | import java.util.ResourceBundle;
|
|---|
| 44 |
|
|---|
| 45 | /**
|
|---|
| 46 | * This class iterates over text elements such as words, lines, sentences,
|
|---|
| 47 | * and characters. It can only iterate over one of these text elements at
|
|---|
| 48 | * a time. An instance of this class configured for the desired iteration
|
|---|
| 49 | * type is created by calling one of the static factory methods, not
|
|---|
| 50 | * by directly calling a constructor.
|
|---|
| 51 | *
|
|---|
| 52 | * @author Tom Tromey <[email protected]>
|
|---|
| 53 | * @author Aaron M. Renn ([email protected])
|
|---|
| 54 | * @date March 19, 1999
|
|---|
| 55 | */
|
|---|
| 56 | /* Written using "Java Class Libraries", 2nd edition, plus online
|
|---|
| 57 | * API docs for JDK 1.2 beta from http://www.javasoft.com.
|
|---|
| 58 | * Status: Believed complete and correct to 1.1.
|
|---|
| 59 | */
|
|---|
| 60 | public abstract class BreakIterator implements Cloneable
|
|---|
| 61 | {
|
|---|
| 62 | /**
|
|---|
| 63 | * This value is returned by the <code>next()</code> and
|
|---|
| 64 | * <code>previous</code> in order to indicate that the end of the
|
|---|
| 65 | * text has been reached.
|
|---|
| 66 | */
|
|---|
| 67 | // The value was discovered by writing a test program.
|
|---|
| 68 | public static final int DONE = -1;
|
|---|
| 69 |
|
|---|
| 70 | /**
|
|---|
| 71 | * This method initializes a new instance of <code>BreakIterator</code>.
|
|---|
| 72 | * This protected constructor is available to subclasses as a default
|
|---|
| 73 | * no-arg superclass constructor.
|
|---|
| 74 | */
|
|---|
| 75 | protected BreakIterator ()
|
|---|
| 76 | {
|
|---|
| 77 | }
|
|---|
| 78 |
|
|---|
| 79 | /**
|
|---|
| 80 | * This method returns the index of the current text element boundary.
|
|---|
| 81 | *
|
|---|
| 82 | * @return The current text boundary.
|
|---|
| 83 | */
|
|---|
| 84 | public abstract int current ();
|
|---|
| 85 |
|
|---|
| 86 | /**
|
|---|
| 87 | * This method returns the first text element boundary in the text being
|
|---|
| 88 | * iterated over.
|
|---|
| 89 | *
|
|---|
| 90 | * @return The first text boundary.
|
|---|
| 91 | */
|
|---|
| 92 | public abstract int first ();
|
|---|
| 93 |
|
|---|
| 94 | /**
|
|---|
| 95 | * This methdod returns the offset of the text element boundary following
|
|---|
| 96 | * the specified offset.
|
|---|
| 97 | *
|
|---|
| 98 | * @param offset The text index from which to find the next text boundary.
|
|---|
| 99 | *
|
|---|
| 100 | * @param The next text boundary following the specified index.
|
|---|
| 101 | */
|
|---|
| 102 | public abstract int following (int pos);
|
|---|
| 103 |
|
|---|
| 104 | /**
|
|---|
| 105 | * This method returns a list of locales for which instances of
|
|---|
| 106 | * <code>BreakIterator</code> are available.
|
|---|
| 107 | *
|
|---|
| 108 | * @return A list of available locales
|
|---|
| 109 | */
|
|---|
| 110 | public static synchronized Locale[] getAvailableLocales ()
|
|---|
| 111 | {
|
|---|
| 112 | Locale[] l = new Locale[1];
|
|---|
| 113 | l[0] = Locale.US;
|
|---|
| 114 | return l;
|
|---|
| 115 | }
|
|---|
| 116 |
|
|---|
| 117 | private static BreakIterator getInstance (String type, Locale loc)
|
|---|
| 118 | {
|
|---|
| 119 | String className;
|
|---|
| 120 | try
|
|---|
| 121 | {
|
|---|
| 122 | ResourceBundle res
|
|---|
| 123 | = ResourceBundle.getBundle("gnu.java.locale.LocaleInformation",
|
|---|
| 124 | loc);
|
|---|
| 125 | className = res.getString(type);
|
|---|
| 126 | }
|
|---|
| 127 | catch (MissingResourceException x)
|
|---|
| 128 | {
|
|---|
| 129 | return null;
|
|---|
| 130 | }
|
|---|
| 131 | try
|
|---|
| 132 | {
|
|---|
| 133 | Class k = Class.forName(className);
|
|---|
| 134 | return (BreakIterator) k.newInstance();
|
|---|
| 135 | }
|
|---|
| 136 | catch (ClassNotFoundException x1)
|
|---|
| 137 | {
|
|---|
| 138 | return null;
|
|---|
| 139 | }
|
|---|
| 140 | catch (InstantiationException x2)
|
|---|
| 141 | {
|
|---|
| 142 | return null;
|
|---|
| 143 | }
|
|---|
| 144 | catch (IllegalAccessException x3)
|
|---|
| 145 | {
|
|---|
| 146 | return null;
|
|---|
| 147 | }
|
|---|
| 148 | }
|
|---|
| 149 |
|
|---|
| 150 | /**
|
|---|
| 151 | * This method returns an instance of <code>BreakIterator</code> that will
|
|---|
| 152 | * iterate over characters as defined in the default locale.
|
|---|
| 153 | *
|
|---|
| 154 | * @return A <code>BreakIterator</code> instance for the default locale.
|
|---|
| 155 | */
|
|---|
| 156 | public static BreakIterator getCharacterInstance ()
|
|---|
| 157 | {
|
|---|
| 158 | return getCharacterInstance (Locale.getDefault());
|
|---|
| 159 | }
|
|---|
| 160 |
|
|---|
| 161 | /**
|
|---|
| 162 | * This method returns an instance of <code>BreakIterator</code> that will
|
|---|
| 163 | * iterate over characters as defined in the specified locale. If the
|
|---|
| 164 | * desired locale is not available, the default locale is used.
|
|---|
| 165 | *
|
|---|
| 166 | * @param locale The desired locale.
|
|---|
| 167 | *
|
|---|
| 168 | * @return A <code>BreakIterator</code> instance for the default locale.
|
|---|
| 169 | */
|
|---|
| 170 | public static BreakIterator getCharacterInstance (Locale loc)
|
|---|
| 171 | {
|
|---|
| 172 | BreakIterator r = getInstance ("CharacterIterator", loc);
|
|---|
| 173 | if (r == null)
|
|---|
| 174 | r = new gnu.java.text.CharacterBreakIterator ();
|
|---|
| 175 | return r;
|
|---|
| 176 | }
|
|---|
| 177 |
|
|---|
| 178 | /**
|
|---|
| 179 | * This method returns an instance of <code>BreakIterator</code> that will
|
|---|
| 180 | * iterate over line breaks as defined in the default locale.
|
|---|
| 181 | *
|
|---|
| 182 | * @return A <code>BreakIterator</code> instance for the default locale.
|
|---|
| 183 | */
|
|---|
| 184 | public static BreakIterator getLineInstance ()
|
|---|
| 185 | {
|
|---|
| 186 | return getLineInstance (Locale.getDefault());
|
|---|
| 187 | }
|
|---|
| 188 |
|
|---|
| 189 | /**
|
|---|
| 190 | * This method returns an instance of <code>BreakIterator</code> that will
|
|---|
| 191 | * iterate over line breaks as defined in the specified locale. If the
|
|---|
| 192 | * desired locale is not available, the default locale is used.
|
|---|
| 193 | *
|
|---|
| 194 | * @param locale The desired locale.
|
|---|
| 195 | *
|
|---|
| 196 | * @return A <code>BreakIterator</code> instance for the default locale.
|
|---|
| 197 | */
|
|---|
| 198 | public static BreakIterator getLineInstance (Locale loc)
|
|---|
| 199 | {
|
|---|
| 200 | BreakIterator r = getInstance ("LineIterator", loc);
|
|---|
| 201 | if (r == null)
|
|---|
| 202 | r = new gnu.java.text.LineBreakIterator ();
|
|---|
| 203 | return r;
|
|---|
| 204 | }
|
|---|
| 205 |
|
|---|
| 206 | /**
|
|---|
| 207 | * This method returns an instance of <code>BreakIterator</code> that will
|
|---|
| 208 | * iterate over sentences as defined in the default locale.
|
|---|
| 209 | *
|
|---|
| 210 | * @return A <code>BreakIterator</code> instance for the default locale.
|
|---|
| 211 | */
|
|---|
| 212 | public static BreakIterator getSentenceInstance ()
|
|---|
| 213 | {
|
|---|
| 214 | return getSentenceInstance (Locale.getDefault());
|
|---|
| 215 | }
|
|---|
| 216 |
|
|---|
| 217 | /**
|
|---|
| 218 | * This method returns an instance of <code>BreakIterator</code> that will
|
|---|
| 219 | * iterate over sentences as defined in the specified locale. If the
|
|---|
| 220 | * desired locale is not available, the default locale is used.
|
|---|
| 221 | *
|
|---|
| 222 | * @param locale The desired locale.
|
|---|
| 223 | *
|
|---|
| 224 | * @return A <code>BreakIterator</code> instance for the default locale.
|
|---|
| 225 | */
|
|---|
| 226 | public static BreakIterator getSentenceInstance (Locale loc)
|
|---|
| 227 | {
|
|---|
| 228 | BreakIterator r = getInstance ("SentenceIterator", loc);
|
|---|
| 229 | if (r == null)
|
|---|
| 230 | r = new gnu.java.text.SentenceBreakIterator ();
|
|---|
| 231 | return r;
|
|---|
| 232 | }
|
|---|
| 233 |
|
|---|
| 234 | /**
|
|---|
| 235 | * This method returns the text this object is iterating over as a
|
|---|
| 236 | * <code>CharacterIterator</code>.
|
|---|
| 237 | *
|
|---|
| 238 | * @param The text being iterated over.
|
|---|
| 239 | */
|
|---|
| 240 | public abstract CharacterIterator getText ();
|
|---|
| 241 |
|
|---|
| 242 | /**
|
|---|
| 243 | * This method returns an instance of <code>BreakIterator</code> that will
|
|---|
| 244 | * iterate over words as defined in the default locale.
|
|---|
| 245 | *
|
|---|
| 246 | * @return A <code>BreakIterator</code> instance for the default locale.
|
|---|
| 247 | */
|
|---|
| 248 | public static BreakIterator getWordInstance ()
|
|---|
| 249 | {
|
|---|
| 250 | return getWordInstance (Locale.getDefault());
|
|---|
| 251 | }
|
|---|
| 252 |
|
|---|
| 253 | /**
|
|---|
| 254 | * This method returns an instance of <code>BreakIterator</code> that will
|
|---|
| 255 | * iterate over words as defined in the specified locale. If the
|
|---|
| 256 | * desired locale is not available, the default locale is used.
|
|---|
| 257 | *
|
|---|
| 258 | * @param locale The desired locale.
|
|---|
| 259 | *
|
|---|
| 260 | * @return A <code>BreakIterator</code> instance for the default locale.
|
|---|
| 261 | */
|
|---|
| 262 | public static BreakIterator getWordInstance (Locale loc)
|
|---|
| 263 | {
|
|---|
| 264 | BreakIterator r = getInstance ("WordIterator", loc);
|
|---|
| 265 | if (r == null)
|
|---|
| 266 | r = new gnu.java.text.WordBreakIterator ();
|
|---|
| 267 | return r;
|
|---|
| 268 | }
|
|---|
| 269 |
|
|---|
| 270 | /**
|
|---|
| 271 | * This method tests whether or not the specified position is a text
|
|---|
| 272 | * element boundary.
|
|---|
| 273 | *
|
|---|
| 274 | * @param offset The text position to test.
|
|---|
| 275 | *
|
|---|
| 276 | * @return <code>true</code> if the position is a boundary,
|
|---|
| 277 | * <code>false</code> otherwise.
|
|---|
| 278 | */
|
|---|
| 279 | public boolean isBoundary (int pos)
|
|---|
| 280 | {
|
|---|
| 281 | if (pos == 0)
|
|---|
| 282 | return true;
|
|---|
| 283 | return following (pos - 1) == pos;
|
|---|
| 284 | }
|
|---|
| 285 |
|
|---|
| 286 | /**
|
|---|
| 287 | * This method returns the last text element boundary in the text being
|
|---|
| 288 | * iterated over.
|
|---|
| 289 | *
|
|---|
| 290 | * @return The last text boundary.
|
|---|
| 291 | */
|
|---|
| 292 | public abstract int last ();
|
|---|
| 293 |
|
|---|
| 294 | /**
|
|---|
| 295 | * This method returns the text element boundary following the current
|
|---|
| 296 | * text position.
|
|---|
| 297 | *
|
|---|
| 298 | * @return The next text boundary.
|
|---|
| 299 | */
|
|---|
| 300 | public abstract int next ();
|
|---|
| 301 |
|
|---|
| 302 | /**
|
|---|
| 303 | * This method returns the n'th text element boundary following the current
|
|---|
| 304 | * text position.
|
|---|
| 305 | *
|
|---|
| 306 | * @param n The number of text element boundaries to skip.
|
|---|
| 307 | *
|
|---|
| 308 | * @return The next text boundary.
|
|---|
| 309 | */
|
|---|
| 310 | public abstract int next (int n);
|
|---|
| 311 |
|
|---|
| 312 | /**
|
|---|
| 313 | * This methdod returns the offset of the text element boundary preceding
|
|---|
| 314 | * the specified offset.
|
|---|
| 315 | *
|
|---|
| 316 | * @param offset The text index from which to find the preceding
|
|---|
| 317 | * text boundary.
|
|---|
| 318 | *
|
|---|
| 319 | * @returns The next text boundary preceding the specified index.
|
|---|
| 320 | */
|
|---|
| 321 | public int preceding (int pos)
|
|---|
| 322 | {
|
|---|
| 323 | if (following (pos) == DONE)
|
|---|
| 324 | last ();
|
|---|
| 325 | while (previous () >= pos)
|
|---|
| 326 | ;
|
|---|
| 327 | return current ();
|
|---|
| 328 | }
|
|---|
| 329 |
|
|---|
| 330 | /**
|
|---|
| 331 | * This method returns the text element boundary preceding the current
|
|---|
| 332 | * text position.
|
|---|
| 333 | *
|
|---|
| 334 | * @return The previous text boundary.
|
|---|
| 335 | */
|
|---|
| 336 | public abstract int previous ();
|
|---|
| 337 |
|
|---|
| 338 | /**
|
|---|
| 339 | * This method sets the text string to iterate over.
|
|---|
| 340 | *
|
|---|
| 341 | * @param str The <code>String</code> to iterate over.
|
|---|
| 342 | */
|
|---|
| 343 | public void setText (String newText)
|
|---|
| 344 | {
|
|---|
| 345 | setText (new StringCharacterIterator (newText));
|
|---|
| 346 | }
|
|---|
| 347 |
|
|---|
| 348 | /**
|
|---|
| 349 | * This method sets the text to iterate over from the specified
|
|---|
| 350 | * <code>CharacterIterator</code>.
|
|---|
| 351 | *
|
|---|
| 352 | * @param ci The desired <code>CharacterIterator</code>.
|
|---|
| 353 | */
|
|---|
| 354 | public abstract void setText (CharacterIterator newText);
|
|---|
| 355 | }
|
|---|