1 |
14 |
jlechner |
/* gnu/regexp/RETokenRepeated.java
|
2 |
|
|
Copyright (C) 1998-2001, 2004 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., 51 Franklin Street, Fifth Floor, Boston, MA
|
19 |
|
|
02110-1301 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 gnu.regexp;
|
40 |
|
|
|
41 |
|
|
import java.util.Vector;
|
42 |
|
|
|
43 |
|
|
final class RETokenRepeated extends REToken {
|
44 |
|
|
private REToken token;
|
45 |
|
|
private int min,max;
|
46 |
|
|
private boolean stingy;
|
47 |
|
|
private boolean possessive;
|
48 |
|
|
private boolean alwaysEmpty; // Special case of {0}
|
49 |
|
|
|
50 |
|
|
RETokenRepeated(int subIndex, REToken token, int min, int max) {
|
51 |
|
|
super(subIndex);
|
52 |
|
|
this.token = token;
|
53 |
|
|
this.min = min;
|
54 |
|
|
this.max = max;
|
55 |
|
|
alwaysEmpty = (min == 0 && max == 0);
|
56 |
|
|
}
|
57 |
|
|
|
58 |
|
|
/** Sets the minimal matching mode to true. */
|
59 |
|
|
void makeStingy() {
|
60 |
|
|
stingy = true;
|
61 |
|
|
}
|
62 |
|
|
|
63 |
|
|
/** Queries if this token has minimal matching enabled. */
|
64 |
|
|
boolean isStingy() {
|
65 |
|
|
return stingy;
|
66 |
|
|
}
|
67 |
|
|
|
68 |
|
|
/** Sets possessive matching mode to true. */
|
69 |
|
|
void makePossessive() {
|
70 |
|
|
possessive = true;
|
71 |
|
|
}
|
72 |
|
|
|
73 |
|
|
/** Queries if this token has possessive matching enabled. */
|
74 |
|
|
boolean isPossessive() {
|
75 |
|
|
return possessive;
|
76 |
|
|
}
|
77 |
|
|
|
78 |
|
|
/**
|
79 |
|
|
* The minimum length of a repeated token is the minimum length
|
80 |
|
|
* of the token multiplied by the minimum number of times it must
|
81 |
|
|
* match.
|
82 |
|
|
*/
|
83 |
|
|
int getMinimumLength() {
|
84 |
|
|
return (min * token.getMinimumLength());
|
85 |
|
|
}
|
86 |
|
|
|
87 |
|
|
int getMaximumLength() {
|
88 |
|
|
if (max == Integer.MAX_VALUE) return Integer.MAX_VALUE;
|
89 |
|
|
int tmax = token.getMaximumLength();
|
90 |
|
|
if (tmax == Integer.MAX_VALUE) return tmax;
|
91 |
|
|
return (max * tmax);
|
92 |
|
|
}
|
93 |
|
|
|
94 |
|
|
boolean stopMatchingIfSatisfied = true;
|
95 |
|
|
|
96 |
|
|
private static REMatch findDoables(REToken tk,
|
97 |
|
|
CharIndexed input, REMatch mymatch) {
|
98 |
|
|
|
99 |
|
|
REMatch.REMatchList doables = new REMatch.REMatchList();
|
100 |
|
|
|
101 |
|
|
// try next repeat at all possible positions
|
102 |
|
|
for (REMatch current = mymatch;
|
103 |
|
|
current != null; current = current.next) {
|
104 |
|
|
REMatch recurrent = (REMatch) current.clone();
|
105 |
|
|
int origin = recurrent.index;
|
106 |
|
|
tk = (REToken) tk.clone();
|
107 |
|
|
tk.next = tk.uncle = null;
|
108 |
|
|
if (tk.match(input, recurrent)) {
|
109 |
|
|
if (recurrent.index == origin) recurrent.empty = true;
|
110 |
|
|
// add all items in current to doables array
|
111 |
|
|
doables.addTail(recurrent);
|
112 |
|
|
}
|
113 |
|
|
}
|
114 |
|
|
return doables.head;
|
115 |
|
|
}
|
116 |
|
|
|
117 |
|
|
// We do need to save every possible point, but the number of clone()
|
118 |
|
|
// invocations here is really a killer for performance on non-stingy
|
119 |
|
|
// repeat operators. I'm open to suggestions...
|
120 |
|
|
|
121 |
|
|
// Hypothetical question: can you have a RE that matches 1 times,
|
122 |
|
|
// 3 times, 5 times, but not 2 times or 4 times? Does having
|
123 |
|
|
// the subexpression back-reference operator allow that?
|
124 |
|
|
|
125 |
|
|
boolean match(CharIndexed input, REMatch mymatch) {
|
126 |
|
|
// Possible positions for the next repeat to match at
|
127 |
|
|
REMatch newMatch = mymatch;
|
128 |
|
|
|
129 |
|
|
// {0} needs some special treatment.
|
130 |
|
|
if (alwaysEmpty) {
|
131 |
|
|
REMatch result = matchRest(input, newMatch);
|
132 |
|
|
if (result != null) {
|
133 |
|
|
mymatch.assignFrom(result);
|
134 |
|
|
return true;
|
135 |
|
|
}
|
136 |
|
|
else {
|
137 |
|
|
return false;
|
138 |
|
|
}
|
139 |
|
|
}
|
140 |
|
|
|
141 |
|
|
// number of times we've matched so far
|
142 |
|
|
int numRepeats = 0;
|
143 |
|
|
|
144 |
|
|
REMatch doables;
|
145 |
|
|
int lastIndex = mymatch.index;
|
146 |
|
|
boolean emptyMatchFound = false;
|
147 |
|
|
|
148 |
|
|
while (numRepeats < min) {
|
149 |
|
|
doables = findDoables(token, input, newMatch);
|
150 |
|
|
|
151 |
|
|
// if none of the possibilities worked out,
|
152 |
|
|
// it means that minimum number of repeats could not be found.
|
153 |
|
|
if (doables == null) return false;
|
154 |
|
|
|
155 |
|
|
// reassign where the next repeat can match
|
156 |
|
|
newMatch = doables;
|
157 |
|
|
|
158 |
|
|
// increment how many repeats we've successfully found
|
159 |
|
|
++numRepeats;
|
160 |
|
|
|
161 |
|
|
if (newMatch.empty) {
|
162 |
|
|
numRepeats = min;
|
163 |
|
|
emptyMatchFound = true;
|
164 |
|
|
break;
|
165 |
|
|
}
|
166 |
|
|
lastIndex = newMatch.index;
|
167 |
|
|
}
|
168 |
|
|
|
169 |
|
|
Vector positions = new Vector();
|
170 |
|
|
|
171 |
|
|
while (numRepeats <= max) {
|
172 |
|
|
// We want to check something like
|
173 |
|
|
// if (stingy)
|
174 |
|
|
// and neglect the further matching. But experience tells
|
175 |
|
|
// such neglection may cause incomplete matching.
|
176 |
|
|
// For example, if we neglect the seemingly unnecessay
|
177 |
|
|
// matching, /^(b+?|a){1,2}?c/ cannot match "bbc".
|
178 |
|
|
// On the other hand, if we do not stop the unnecessary
|
179 |
|
|
// matching, /(([a-c])b*?\2)*/ matches "ababbbcbc"
|
180 |
|
|
// entirely when we wan to find only "ababb".
|
181 |
|
|
// In order to make regression tests pass, we do as we did.
|
182 |
|
|
if (stopMatchingIfSatisfied && stingy) {
|
183 |
|
|
REMatch results = matchRest(input, newMatch);
|
184 |
|
|
if (results != null) {
|
185 |
|
|
mymatch.assignFrom(results);
|
186 |
|
|
return true;
|
187 |
|
|
}
|
188 |
|
|
}
|
189 |
|
|
positions.add(newMatch);
|
190 |
|
|
if (emptyMatchFound) break;
|
191 |
|
|
|
192 |
|
|
doables = findDoables(token, input, newMatch);
|
193 |
|
|
if (doables == null) break;
|
194 |
|
|
|
195 |
|
|
// doables.index == lastIndex occurs either
|
196 |
|
|
// (1) when an empty string was the longest
|
197 |
|
|
// that matched this token.
|
198 |
|
|
// or
|
199 |
|
|
// (2) when the same string matches this token many times.
|
200 |
|
|
// For example, "acbab" itself matches "a.*b" and
|
201 |
|
|
// its substrings "acb" and "ab" also match.
|
202 |
|
|
// In this case, we do not have to go further until
|
203 |
|
|
// numRepeats == max because the more numRepeats grows,
|
204 |
|
|
// the shorter the substring matching this token becomes.
|
205 |
|
|
// So the previous succesful match must have bee the best
|
206 |
|
|
// match. But this is not necessarily the case if stingy.
|
207 |
|
|
if (doables.index == lastIndex) {
|
208 |
|
|
if (doables.empty) {
|
209 |
|
|
emptyMatchFound = true;
|
210 |
|
|
}
|
211 |
|
|
else {
|
212 |
|
|
if (!stingy) break;
|
213 |
|
|
}
|
214 |
|
|
}
|
215 |
|
|
numRepeats++;
|
216 |
|
|
newMatch = doables;
|
217 |
|
|
lastIndex = newMatch.index;
|
218 |
|
|
}
|
219 |
|
|
|
220 |
|
|
// We're greedy, but ease off until a true match is found.
|
221 |
|
|
// At this point we've either got too many or just the right amount.
|
222 |
|
|
// See if this numRepeats works with the rest of the regexp.
|
223 |
|
|
|
224 |
|
|
REMatch.REMatchList allResults = new REMatch.REMatchList();
|
225 |
|
|
|
226 |
|
|
int posCount = positions.size();
|
227 |
|
|
int posIndex = (stingy ? 0 : posCount - 1);
|
228 |
|
|
|
229 |
|
|
while (posCount-- > 0) {
|
230 |
|
|
REMatch m = (REMatch) positions.elementAt(posIndex);
|
231 |
|
|
if (stingy) posIndex++; else posIndex--;
|
232 |
|
|
|
233 |
|
|
REMatch results = matchRest(input, m);
|
234 |
|
|
if (results != null) {
|
235 |
|
|
// Order these from longest to shortest
|
236 |
|
|
// Start by assuming longest (more repeats)
|
237 |
|
|
// If stingy the order is shortest to longest.
|
238 |
|
|
allResults.addTail(results);
|
239 |
|
|
}
|
240 |
|
|
else {
|
241 |
|
|
if (possessive) break;
|
242 |
|
|
}
|
243 |
|
|
}
|
244 |
|
|
|
245 |
|
|
if (allResults.head != null) {
|
246 |
|
|
mymatch.assignFrom(allResults.head); // does this get all?
|
247 |
|
|
return true;
|
248 |
|
|
}
|
249 |
|
|
// If we fall out, no matches.
|
250 |
|
|
return false;
|
251 |
|
|
}
|
252 |
|
|
|
253 |
|
|
private REMatch matchRest(CharIndexed input, final REMatch newMatch) {
|
254 |
|
|
REMatch current, single;
|
255 |
|
|
REMatch.REMatchList doneIndex = new REMatch.REMatchList();
|
256 |
|
|
// Test all possible matches for this number of repeats
|
257 |
|
|
for (current = newMatch; current != null; current = current.next) {
|
258 |
|
|
// clone() separates a single match from the chain
|
259 |
|
|
single = (REMatch) current.clone();
|
260 |
|
|
if (next(input, single)) {
|
261 |
|
|
// chain results to doneIndex
|
262 |
|
|
doneIndex.addTail(single);
|
263 |
|
|
}
|
264 |
|
|
}
|
265 |
|
|
return doneIndex.head;
|
266 |
|
|
}
|
267 |
|
|
|
268 |
|
|
void dump(StringBuffer os) {
|
269 |
|
|
os.append("(?:");
|
270 |
|
|
token.dumpAll(os);
|
271 |
|
|
os.append(')');
|
272 |
|
|
if ((max == Integer.MAX_VALUE) && (min <= 1))
|
273 |
|
|
os.append( (min == 0) ? '*' : '+' );
|
274 |
|
|
else if ((min == 0) && (max == 1))
|
275 |
|
|
os.append('?');
|
276 |
|
|
else {
|
277 |
|
|
os.append('{').append(min);
|
278 |
|
|
if (max > min) {
|
279 |
|
|
os.append(',');
|
280 |
|
|
if (max != Integer.MAX_VALUE) os.append(max);
|
281 |
|
|
}
|
282 |
|
|
os.append('}');
|
283 |
|
|
}
|
284 |
|
|
if (stingy) os.append('?');
|
285 |
|
|
}
|
286 |
|
|
}
|