001: /*
002: * gnu/regexp/REMatch.java
003: * Copyright (C) 1998-2001 Wes Biggs
004: *
005: * This library is free software; you can redistribute it and/or modify
006: * it under the terms of the GNU Lesser General Public License as published
007: * by the Free Software Foundation; either version 2.1 of the License, or
008: * (at your option) any later version.
009: *
010: * This library is distributed in the hope that it will be useful,
011: * but WITHOUT ANY WARRANTY; without even the implied warranty of
012: * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
013: * GNU Lesser General Public License for more details.
014: *
015: * You should have received a copy of the GNU Lesser General Public License
016: * along with this program; if not, write to the Free Software
017: * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
018: */
019:
020: package gnu.regexp;
021:
022: import java.io.Serializable;
023:
024: /**
025: * An instance of this class represents a match
026: * completed by a gnu.regexp matching function. It can be used
027: * to obtain relevant information about the location of a match
028: * or submatch.
029: *
030: * @author <A HREF="mailto:wes@cacas.org">Wes Biggs</A>
031: */
032: public final class REMatch implements Serializable, Cloneable {
033: private String matchedText;
034:
035: // These variables are package scope for fast access within the engine
036: int eflags; // execution flags this match was made using
037:
038: // Offset in source text where match was tried. This is zero-based;
039: // the actual position in the source text is given by (offset + anchor).
040: int offset;
041:
042: // Anchor position refers to the index into the source input
043: // at which the matching operation began.
044: // This is also useful for the ANCHORINDEX option.
045: int anchor;
046:
047: // Package scope; used by RE.
048: int index; // used while matching to mark current match position in input
049: int[] start; // start positions (relative to offset) for each (sub)exp.
050: int[] end; // end positions for the same
051: REMatch next; // other possibility (to avoid having to use arrays)
052:
053: public Object clone() {
054: try {
055: REMatch copy = (REMatch) super .clone();
056: copy.next = null;
057:
058: // Jun 25 2003 3:39 PM PG
059: // Work around problem with jikes 1.18 and IBM Java 1.4.1.
060: //copy.start = (int[]) start.clone();
061: //copy.end = (int[]) end.clone();
062: copy.start = new int[start.length];
063: System.arraycopy(start, 0, copy.start, 0, start.length);
064: copy.end = new int[end.length];
065: System.arraycopy(end, 0, copy.end, 0, end.length);
066:
067: return copy;
068: } catch (CloneNotSupportedException e) {
069: throw new Error(); // doesn't happen
070: }
071: }
072:
073: void assignFrom(REMatch other) {
074: start = other.start;
075: end = other.end;
076: index = other.index;
077: // need to deep clone?
078: next = other.next;
079: }
080:
081: REMatch(int subs, int anchor, int eflags) {
082: start = new int[subs + 1];
083: end = new int[subs + 1];
084: this .anchor = anchor;
085: this .eflags = eflags;
086: clear(anchor);
087: }
088:
089: void finish(CharIndexed text) {
090: start[0] = 0;
091: StringBuffer sb = new StringBuffer();
092: int i;
093: for (i = 0; i < end[0]; i++)
094: sb.append(text.charAt(i));
095: matchedText = sb.toString();
096: for (i = 0; i < start.length; i++) {
097: // If any subexpressions didn't terminate, they don't count
098: // TODO check if this code ever gets hit
099: if ((start[i] == -1) ^ (end[i] == -1)) {
100: start[i] = -1;
101: end[i] = -1;
102: }
103: }
104: next = null; // cut off alternates
105: }
106:
107: /** Clears the current match and moves the offset to the new index. */
108: void clear(int index) {
109: offset = index;
110: this .index = 0;
111: for (int i = 0; i < start.length; i++) {
112: start[i] = end[i] = -1;
113: }
114: next = null; // cut off alternates
115: }
116:
117: /**
118: * Returns the string matching the pattern. This makes it convenient
119: * to write code like the following:
120: * <P>
121: * <code>
122: * REMatch myMatch = myExpression.getMatch(myString);<br>
123: * if (myMatch != null) System.out.println("Regexp found: "+myMatch);
124: * </code>
125: */
126: public String toString() {
127: return matchedText;
128: }
129:
130: /**
131: * Returns the index within the input text where the match in its entirety
132: * began.
133: */
134: public int getStartIndex() {
135: return offset + start[0];
136: }
137:
138: /**
139: * Returns the index within the input string where the match in
140: * its entirety ends. The return value is the next position after
141: * the end of the string; therefore, a match created by the
142: * following call:
143: *
144: * <P>
145: * <code>REMatch myMatch = myExpression.getMatch(myString);</code>
146: * <P>
147: * can be viewed (given that myMatch is not null) by creating
148: * <P>
149: * <code>String theMatch = myString.substring(myMatch.getStartIndex(),
150: * myMatch.getEndIndex());</code>
151: * <P>
152: * But you can save yourself that work, since the <code>toString()</code>
153: * method (above) does exactly that for you.
154: */
155: public int getEndIndex() {
156: return offset + end[0];
157: }
158:
159: /**
160: * Returns the string matching the given subexpression. The subexpressions
161: * are indexed starting with one, not zero. That is, the subexpression
162: * identified by the first set of parentheses in a regular expression
163: * could be retrieved from an REMatch by calling match.toString(1).
164: *
165: * @param sub Index of the subexpression.
166: */
167: public String toString(int sub) {
168: if ((sub >= start.length) || (start[sub] == -1))
169: return "";
170: return (matchedText.substring(start[sub], end[sub]));
171: }
172:
173: /**
174: * Returns the index within the input string used to generate this match
175: * where subexpression number <i>sub</i> begins, or <code>-1</code> if
176: * the subexpression does not exist. The initial position is zero.
177: *
178: * @param sub Subexpression index
179: * @deprecated Use getStartIndex(int) instead.
180: */
181: public int getSubStartIndex(int sub) {
182: if (sub >= start.length)
183: return -1;
184: int x = start[sub];
185: return (x == -1) ? x : offset + x;
186: }
187:
188: /**
189: * Returns the index within the input string used to generate this match
190: * where subexpression number <i>sub</i> begins, or <code>-1</code> if
191: * the subexpression does not exist. The initial position is zero.
192: *
193: * @param sub Subexpression index
194: * @since gnu.regexp 1.1.0
195: */
196: public int getStartIndex(int sub) {
197: if (sub >= start.length)
198: return -1;
199: int x = start[sub];
200: return (x == -1) ? x : offset + x;
201: }
202:
203: /**
204: * Returns the index within the input string used to generate this match
205: * where subexpression number <i>sub</i> ends, or <code>-1</code> if
206: * the subexpression does not exist. The initial position is zero.
207: *
208: * @param sub Subexpression index
209: * @deprecated Use getEndIndex(int) instead
210: */
211: public int getSubEndIndex(int sub) {
212: if (sub >= start.length)
213: return -1;
214: int x = end[sub];
215: return (x == -1) ? x : offset + x;
216: }
217:
218: /**
219: * Returns the index within the input string used to generate this match
220: * where subexpression number <i>sub</i> ends, or <code>-1</code> if
221: * the subexpression does not exist. The initial position is zero.
222: *
223: * @param sub Subexpression index
224: */
225: public int getEndIndex(int sub) {
226: if (sub >= start.length)
227: return -1;
228: int x = end[sub];
229: return (x == -1) ? x : offset + x;
230: }
231:
232: /**
233: * Substitute the results of this match to create a new string.
234: * This is patterned after PERL, so the tokens to watch out for are
235: * <code>$0</code> through <code>$9</code>. <code>$0</code> matches
236: * the full substring matched; <code>$<i>n</i></code> matches
237: * subexpression number <i>n</i>.
238: *
239: * @param input A string consisting of literals and <code>$<i>n</i></code> tokens.
240: */
241: public String substituteInto(String input) {
242: // a la Perl, $0 is whole thing, $1 - $9 are subexpressions
243: StringBuffer output = new StringBuffer();
244: int pos;
245: for (pos = 0; pos < input.length() - 1; pos++) {
246: if ((input.charAt(pos) == '$')
247: && (Character.isDigit(input.charAt(pos + 1)))) {
248: int val = Character.digit(input.charAt(++pos), 10);
249: if (val < start.length) {
250: output.append(toString(val));
251: }
252: } else
253: output.append(input.charAt(pos));
254: }
255: if (pos < input.length())
256: output.append(input.charAt(pos));
257: return output.toString();
258: }
259: }
|