001 /* Matcher.java -- Instance of a regular expression applied to a char sequence.
002 Copyright (C) 2002, 2004, 2006 Free Software Foundation, Inc.
003
004 This file is part of GNU Classpath.
005
006 GNU Classpath is free software; you can redistribute it and/or modify
007 it under the terms of the GNU General Public License as published by
008 the Free Software Foundation; either version 2, or (at your option)
009 any later version.
010
011 GNU Classpath is distributed in the hope that it will be useful, but
012 WITHOUT ANY WARRANTY; without even the implied warranty of
013 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
014 General Public License for more details.
015
016 You should have received a copy of the GNU General Public License
017 along with GNU Classpath; see the file COPYING. If not, write to the
018 Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
019 02110-1301 USA.
020
021 Linking this library statically or dynamically with other modules is
022 making a combined work based on this library. Thus, the terms and
023 conditions of the GNU General Public License cover the whole
024 combination.
025
026 As a special exception, the copyright holders of this library give you
027 permission to link this library with independent modules to produce an
028 executable, regardless of the license terms of these independent
029 modules, and to copy and distribute the resulting executable under
030 terms of your choice, provided that you also meet, for each linked
031 independent module, the terms and conditions of the license of that
032 module. An independent module is a module which is not derived from
033 or based on this library. If you modify this library, you may extend
034 this exception to your version of the library, but you are not
035 obligated to do so. If you do not wish to do so, delete this
036 exception statement from your version. */
037
038
039 package java.util.regex;
040
041 import gnu.java.util.regex.CharIndexed;
042 import gnu.java.util.regex.RE;
043 import gnu.java.util.regex.REMatch;
044
045 /**
046 * Instance of a regular expression applied to a char sequence.
047 *
048 * @since 1.4
049 */
050 public final class Matcher implements MatchResult
051 {
052 private Pattern pattern;
053 private CharSequence input;
054 // We use CharIndexed as an input object to the getMatch method in order
055 // that /\G/ (the end of the previous match) may work. The information
056 // of the previous match is stored in the CharIndexed object.
057 private CharIndexed inputCharIndexed;
058 private int position;
059 private int appendPosition;
060 private REMatch match;
061
062 Matcher(Pattern pattern, CharSequence input)
063 {
064 this.pattern = pattern;
065 this.input = input;
066 this.inputCharIndexed = RE.makeCharIndexed(input, 0);
067 }
068
069 /**
070 * @param sb The target string buffer
071 * @param replacement The replacement string
072 *
073 * @exception IllegalStateException If no match has yet been attempted,
074 * or if the previous match operation failed
075 * @exception IndexOutOfBoundsException If the replacement string refers
076 * to a capturing group that does not exist in the pattern
077 */
078 public Matcher appendReplacement (StringBuffer sb, String replacement)
079 throws IllegalStateException
080 {
081 assertMatchOp();
082 sb.append(input.subSequence(appendPosition,
083 match.getStartIndex()).toString());
084 sb.append(RE.getReplacement(replacement, match,
085 RE.REG_REPLACE_USE_BACKSLASHESCAPE));
086 appendPosition = match.getEndIndex();
087 return this;
088 }
089
090 /**
091 * @param sb The target string buffer
092 */
093 public StringBuffer appendTail (StringBuffer sb)
094 {
095 sb.append(input.subSequence(appendPosition, input.length()).toString());
096 return sb;
097 }
098
099 /**
100 * @exception IllegalStateException If no match has yet been attempted,
101 * or if the previous match operation failed
102 */
103 public int end ()
104 throws IllegalStateException
105 {
106 assertMatchOp();
107 return match.getEndIndex();
108 }
109
110 /**
111 * @param group The index of a capturing group in this matcher's pattern
112 *
113 * @exception IllegalStateException If no match has yet been attempted,
114 * or if the previous match operation failed
115 * @exception IndexOutOfBoundsException If the replacement string refers
116 * to a capturing group that does not exist in the pattern
117 */
118 public int end (int group)
119 throws IllegalStateException
120 {
121 assertMatchOp();
122 return match.getEndIndex(group);
123 }
124
125 public boolean find ()
126 {
127 boolean first = (match == null);
128 match = pattern.getRE().getMatch(inputCharIndexed, position);
129 if (match != null)
130 {
131 int endIndex = match.getEndIndex();
132 // Are we stuck at the same position?
133 if (!first && endIndex == position)
134 {
135 match = null;
136 // Not at the end of the input yet?
137 if (position < input.length() - 1)
138 {
139 position++;
140 return find(position);
141 }
142 else
143 return false;
144 }
145 position = endIndex;
146 return true;
147 }
148 return false;
149 }
150
151 /**
152 * @param start The index to start the new pattern matching
153 *
154 * @exception IndexOutOfBoundsException If the replacement string refers
155 * to a capturing group that does not exist in the pattern
156 */
157 public boolean find (int start)
158 {
159 match = pattern.getRE().getMatch(inputCharIndexed, start);
160 if (match != null)
161 {
162 position = match.getEndIndex();
163 return true;
164 }
165 return false;
166 }
167
168 /**
169 * @exception IllegalStateException If no match has yet been attempted,
170 * or if the previous match operation failed
171 */
172 public String group ()
173 {
174 assertMatchOp();
175 return match.toString();
176 }
177
178 /**
179 * @param group The index of a capturing group in this matcher's pattern
180 *
181 * @exception IllegalStateException If no match has yet been attempted,
182 * or if the previous match operation failed
183 * @exception IndexOutOfBoundsException If the replacement string refers
184 * to a capturing group that does not exist in the pattern
185 */
186 public String group (int group)
187 throws IllegalStateException
188 {
189 assertMatchOp();
190 return match.toString(group);
191 }
192
193 /**
194 * @param replacement The replacement string
195 */
196 public String replaceFirst (String replacement)
197 {
198 reset();
199 // Semantics might not quite match
200 return pattern.getRE().substitute(input, replacement, position,
201 RE.REG_REPLACE_USE_BACKSLASHESCAPE);
202 }
203
204 /**
205 * @param replacement The replacement string
206 */
207 public String replaceAll (String replacement)
208 {
209 reset();
210 return pattern.getRE().substituteAll(input, replacement, position,
211 RE.REG_REPLACE_USE_BACKSLASHESCAPE);
212 }
213
214 public int groupCount ()
215 {
216 return pattern.getRE().getNumSubs();
217 }
218
219 public boolean lookingAt ()
220 {
221 match = pattern.getRE().getMatch(inputCharIndexed, 0, RE.REG_FIX_STARTING_POSITION, null);
222 if (match != null)
223 {
224 if (match.getStartIndex() == 0)
225 {
226 position = match.getEndIndex();
227 return true;
228 }
229 match = null;
230 }
231 return false;
232 }
233
234 /**
235 * Attempts to match the entire input sequence against the pattern.
236 *
237 * If the match succeeds then more information can be obtained via the
238 * start, end, and group methods.
239 *
240 * @see #start()
241 * @see #end()
242 * @see #group()
243 */
244 public boolean matches ()
245 {
246 match = pattern.getRE().getMatch(inputCharIndexed, 0, RE.REG_TRY_ENTIRE_MATCH|RE.REG_FIX_STARTING_POSITION, null);
247 if (match != null)
248 {
249 if (match.getStartIndex() == 0)
250 {
251 position = match.getEndIndex();
252 if (position == input.length())
253 return true;
254 }
255 match = null;
256 }
257 return false;
258 }
259
260 /**
261 * Returns the Pattern that is interpreted by this Matcher
262 */
263 public Pattern pattern ()
264 {
265 return pattern;
266 }
267
268 public Matcher reset ()
269 {
270 position = 0;
271 match = null;
272 return this;
273 }
274
275 /**
276 * @param input The new input character sequence
277 */
278 public Matcher reset (CharSequence input)
279 {
280 this.input = input;
281 this.inputCharIndexed = RE.makeCharIndexed(input, 0);
282 return reset();
283 }
284
285 /**
286 * @returns the index of a capturing group in this matcher's pattern
287 *
288 * @exception IllegalStateException If no match has yet been attempted,
289 * or if the previous match operation failed
290 */
291 public int start ()
292 throws IllegalStateException
293 {
294 assertMatchOp();
295 return match.getStartIndex();
296 }
297
298 /**
299 * @param group The index of a capturing group in this matcher's pattern
300 *
301 * @exception IllegalStateException If no match has yet been attempted,
302 * or if the previous match operation failed
303 * @exception IndexOutOfBoundsException If the replacement string refers
304 * to a capturing group that does not exist in the pattern
305 */
306 public int start (int group)
307 throws IllegalStateException
308 {
309 assertMatchOp();
310 return match.getStartIndex(group);
311 }
312
313 /**
314 * @return True if and only if the matcher hit the end of input.
315 */
316 public boolean hitEnd()
317 {
318 return inputCharIndexed.hitEnd();
319 }
320
321 /**
322 * @return A string expression of this matcher.
323 */
324 public String toString()
325 {
326 StringBuilder sb = new StringBuilder();
327 sb.append(this.getClass().getName())
328 .append("[pattern=").append(pattern.pattern())
329 .append(" region=").append("0").append(",").append(input.length())
330 .append(" lastmatch=").append(match == null ? "" : match.toString())
331 .append("]");
332 return sb.toString();
333 }
334
335 private void assertMatchOp()
336 {
337 if (match == null) throw new IllegalStateException();
338 }
339 }