3840
|
1 /**
|
|
2 * @test
|
|
3 * @bug 7070134
|
|
4 * @summary Hotspot crashes with sigsegv from PorterStemmer
|
|
5 *
|
|
6 * @run shell Test7070134.sh
|
|
7 */
|
|
8
|
|
9 /*
|
|
10
|
|
11 Porter stemmer in Java. The original paper is in
|
|
12
|
|
13 Porter, 1980, An algorithm for suffix stripping, Program, Vol. 14,
|
|
14 no. 3, pp 130-137,
|
|
15
|
6046
|
16 http://www.tartarus.org/~martin/PorterStemmer
|
|
17
|
|
18 The software is completely free for any purpose, unless notes at the head
|
|
19 of the program text indicates otherwise (which is rare). In any case,
|
|
20 the notes about licensing are never more restrictive than the BSD License.
|
|
21
|
|
22 In every case where the software is not written by me (Martin Porter),
|
|
23 this licensing arrangement has been endorsed by the contributor, and it is
|
|
24 therefore unnecessary to ask the contributor again to confirm it.
|
|
25
|
|
26 I have not asked any contributors (or their employers, if they have them)
|
|
27 for proofs that they have the right to distribute their software in this way.
|
3840
|
28
|
|
29 History:
|
|
30
|
|
31 Release 1
|
|
32
|
|
33 Bug 1 (reported by Gonzalo Parra 16/10/99) fixed as marked below.
|
|
34 The words 'aed', 'eed', 'oed' leave k at 'a' for step 3, and b[k-1]
|
|
35 is then out outside the bounds of b.
|
|
36
|
|
37 Release 2
|
|
38
|
|
39 Similarly,
|
|
40
|
|
41 Bug 2 (reported by Steve Dyrdahl 22/2/00) fixed as marked below.
|
|
42 'ion' by itself leaves j = -1 in the test for 'ion' in step 5, and
|
|
43 b[j] is then outside the bounds of b.
|
|
44
|
|
45 Release 3
|
|
46
|
|
47 Considerably revised 4/9/00 in the light of many helpful suggestions
|
|
48 from Brian Goetz of Quiotix Corporation (brian@quiotix.com).
|
|
49
|
|
50 Release 4
|
|
51
|
|
52 */
|
|
53
|
|
54 import java.io.*;
|
|
55
|
|
56 /**
|
|
57 * Stemmer, implementing the Porter Stemming Algorithm
|
|
58 *
|
|
59 * The Stemmer class transforms a word into its root form. The input
|
|
60 * word can be provided a character at time (by calling add()), or at once
|
|
61 * by calling one of the various stem(something) methods.
|
|
62 */
|
|
63
|
|
64 class Stemmer
|
|
65 { private char[] b;
|
|
66 private int i, /* offset into b */
|
|
67 i_end, /* offset to end of stemmed word */
|
|
68 j, k;
|
|
69 private static final int INC = 50;
|
|
70 /* unit of size whereby b is increased */
|
|
71 public Stemmer()
|
|
72 { b = new char[INC];
|
|
73 i = 0;
|
|
74 i_end = 0;
|
|
75 }
|
|
76
|
|
77 /**
|
|
78 * Add a character to the word being stemmed. When you are finished
|
|
79 * adding characters, you can call stem(void) to stem the word.
|
|
80 */
|
|
81
|
|
82 public void add(char ch)
|
|
83 { if (i == b.length)
|
|
84 { char[] new_b = new char[i+INC];
|
|
85 for (int c = 0; c < i; c++) new_b[c] = b[c];
|
|
86 b = new_b;
|
|
87 }
|
|
88 b[i++] = ch;
|
|
89 }
|
|
90
|
|
91
|
|
92 /** Adds wLen characters to the word being stemmed contained in a portion
|
|
93 * of a char[] array. This is like repeated calls of add(char ch), but
|
|
94 * faster.
|
|
95 */
|
|
96
|
|
97 public void add(char[] w, int wLen)
|
|
98 { if (i+wLen >= b.length)
|
|
99 { char[] new_b = new char[i+wLen+INC];
|
|
100 for (int c = 0; c < i; c++) new_b[c] = b[c];
|
|
101 b = new_b;
|
|
102 }
|
|
103 for (int c = 0; c < wLen; c++) b[i++] = w[c];
|
|
104 }
|
|
105
|
|
106 /**
|
|
107 * After a word has been stemmed, it can be retrieved by toString(),
|
|
108 * or a reference to the internal buffer can be retrieved by getResultBuffer
|
|
109 * and getResultLength (which is generally more efficient.)
|
|
110 */
|
|
111 public String toString() { return new String(b,0,i_end); }
|
|
112
|
|
113 /**
|
|
114 * Returns the length of the word resulting from the stemming process.
|
|
115 */
|
|
116 public int getResultLength() { return i_end; }
|
|
117
|
|
118 /**
|
|
119 * Returns a reference to a character buffer containing the results of
|
|
120 * the stemming process. You also need to consult getResultLength()
|
|
121 * to determine the length of the result.
|
|
122 */
|
|
123 public char[] getResultBuffer() { return b; }
|
|
124
|
|
125 /* cons(i) is true <=> b[i] is a consonant. */
|
|
126
|
|
127 private final boolean cons(int i)
|
|
128 { switch (b[i])
|
|
129 { case 'a': case 'e': case 'i': case 'o': case 'u': return false;
|
|
130 case 'y': return (i==0) ? true : !cons(i-1);
|
|
131 default: return true;
|
|
132 }
|
|
133 }
|
|
134
|
|
135 /* m() measures the number of consonant sequences between 0 and j. if c is
|
|
136 a consonant sequence and v a vowel sequence, and <..> indicates arbitrary
|
|
137 presence,
|
|
138
|
|
139 <c><v> gives 0
|
|
140 <c>vc<v> gives 1
|
|
141 <c>vcvc<v> gives 2
|
|
142 <c>vcvcvc<v> gives 3
|
|
143 ....
|
|
144 */
|
|
145
|
|
146 private final int m()
|
|
147 { int n = 0;
|
|
148 int i = 0;
|
|
149 while(true)
|
|
150 { if (i > j) return n;
|
|
151 if (! cons(i)) break; i++;
|
|
152 }
|
|
153 i++;
|
|
154 while(true)
|
|
155 { while(true)
|
|
156 { if (i > j) return n;
|
|
157 if (cons(i)) break;
|
|
158 i++;
|
|
159 }
|
|
160 i++;
|
|
161 n++;
|
|
162 while(true)
|
|
163 { if (i > j) return n;
|
|
164 if (! cons(i)) break;
|
|
165 i++;
|
|
166 }
|
|
167 i++;
|
|
168 }
|
|
169 }
|
|
170
|
|
171 /* vowelinstem() is true <=> 0,...j contains a vowel */
|
|
172
|
|
173 private final boolean vowelinstem()
|
|
174 { int i; for (i = 0; i <= j; i++) if (! cons(i)) return true;
|
|
175 return false;
|
|
176 }
|
|
177
|
|
178 /* doublec(j) is true <=> j,(j-1) contain a double consonant. */
|
|
179
|
|
180 private final boolean doublec(int j)
|
|
181 { if (j < 1) return false;
|
|
182 if (b[j] != b[j-1]) return false;
|
|
183 return cons(j);
|
|
184 }
|
|
185
|
|
186 /* cvc(i) is true <=> i-2,i-1,i has the form consonant - vowel - consonant
|
|
187 and also if the second c is not w,x or y. this is used when trying to
|
|
188 restore an e at the end of a short word. e.g.
|
|
189
|
|
190 cav(e), lov(e), hop(e), crim(e), but
|
|
191 snow, box, tray.
|
|
192
|
|
193 */
|
|
194
|
|
195 private final boolean cvc(int i)
|
|
196 { if (i < 2 || !cons(i) || cons(i-1) || !cons(i-2)) return false;
|
|
197 { int ch = b[i];
|
|
198 if (ch == 'w' || ch == 'x' || ch == 'y') return false;
|
|
199 }
|
|
200 return true;
|
|
201 }
|
|
202
|
|
203 private final boolean ends(String s)
|
|
204 { int l = s.length();
|
|
205 int o = k-l+1;
|
|
206 if (o < 0) return false;
|
|
207 for (int i = 0; i < l; i++) if (b[o+i] != s.charAt(i)) return false;
|
|
208 j = k-l;
|
|
209 return true;
|
|
210 }
|
|
211
|
|
212 /* setto(s) sets (j+1),...k to the characters in the string s, readjusting
|
|
213 k. */
|
|
214
|
|
215 private final void setto(String s)
|
|
216 { int l = s.length();
|
|
217 int o = j+1;
|
|
218 for (int i = 0; i < l; i++) b[o+i] = s.charAt(i);
|
|
219 k = j+l;
|
|
220 }
|
|
221
|
|
222 /* r(s) is used further down. */
|
|
223
|
|
224 private final void r(String s) { if (m() > 0) setto(s); }
|
|
225
|
|
226 /* step1() gets rid of plurals and -ed or -ing. e.g.
|
|
227
|
|
228 caresses -> caress
|
|
229 ponies -> poni
|
|
230 ties -> ti
|
|
231 caress -> caress
|
|
232 cats -> cat
|
|
233
|
|
234 feed -> feed
|
|
235 agreed -> agree
|
|
236 disabled -> disable
|
|
237
|
|
238 matting -> mat
|
|
239 mating -> mate
|
|
240 meeting -> meet
|
|
241 milling -> mill
|
|
242 messing -> mess
|
|
243
|
|
244 meetings -> meet
|
|
245
|
|
246 */
|
|
247
|
|
248 private final void step1()
|
|
249 { if (b[k] == 's')
|
|
250 { if (ends("sses")) k -= 2; else
|
|
251 if (ends("ies")) setto("i"); else
|
|
252 if (b[k-1] != 's') k--;
|
|
253 }
|
|
254 if (ends("eed")) { if (m() > 0) k--; } else
|
|
255 if ((ends("ed") || ends("ing")) && vowelinstem())
|
|
256 { k = j;
|
|
257 if (ends("at")) setto("ate"); else
|
|
258 if (ends("bl")) setto("ble"); else
|
|
259 if (ends("iz")) setto("ize"); else
|
|
260 if (doublec(k))
|
|
261 { k--;
|
|
262 { int ch = b[k];
|
|
263 if (ch == 'l' || ch == 's' || ch == 'z') k++;
|
|
264 }
|
|
265 }
|
|
266 else if (m() == 1 && cvc(k)) setto("e");
|
|
267 }
|
|
268 }
|
|
269
|
|
270 /* step2() turns terminal y to i when there is another vowel in the stem. */
|
|
271
|
|
272 private final void step2() { if (ends("y") && vowelinstem()) b[k] = 'i'; }
|
|
273
|
|
274 /* step3() maps double suffices to single ones. so -ization ( = -ize plus
|
|
275 -ation) maps to -ize etc. note that the string before the suffix must give
|
|
276 m() > 0. */
|
|
277
|
|
278 private final void step3() { if (k == 0) return; /* For Bug 1 */ switch (b[k-1])
|
|
279 {
|
|
280 case 'a': if (ends("ational")) { r("ate"); break; }
|
|
281 if (ends("tional")) { r("tion"); break; }
|
|
282 break;
|
|
283 case 'c': if (ends("enci")) { r("ence"); break; }
|
|
284 if (ends("anci")) { r("ance"); break; }
|
|
285 break;
|
|
286 case 'e': if (ends("izer")) { r("ize"); break; }
|
|
287 break;
|
|
288 case 'l': if (ends("bli")) { r("ble"); break; }
|
|
289 if (ends("alli")) { r("al"); break; }
|
|
290 if (ends("entli")) { r("ent"); break; }
|
|
291 if (ends("eli")) { r("e"); break; }
|
|
292 if (ends("ousli")) { r("ous"); break; }
|
|
293 break;
|
|
294 case 'o': if (ends("ization")) { r("ize"); break; }
|
|
295 if (ends("ation")) { r("ate"); break; }
|
|
296 if (ends("ator")) { r("ate"); break; }
|
|
297 break;
|
|
298 case 's': if (ends("alism")) { r("al"); break; }
|
|
299 if (ends("iveness")) { r("ive"); break; }
|
|
300 if (ends("fulness")) { r("ful"); break; }
|
|
301 if (ends("ousness")) { r("ous"); break; }
|
|
302 break;
|
|
303 case 't': if (ends("aliti")) { r("al"); break; }
|
|
304 if (ends("iviti")) { r("ive"); break; }
|
|
305 if (ends("biliti")) { r("ble"); break; }
|
|
306 break;
|
|
307 case 'g': if (ends("logi")) { r("log"); break; }
|
|
308 } }
|
|
309
|
|
310 /* step4() deals with -ic-, -full, -ness etc. similar strategy to step3. */
|
|
311
|
|
312 private final void step4() { switch (b[k])
|
|
313 {
|
|
314 case 'e': if (ends("icate")) { r("ic"); break; }
|
|
315 if (ends("ative")) { r(""); break; }
|
|
316 if (ends("alize")) { r("al"); break; }
|
|
317 break;
|
|
318 case 'i': if (ends("iciti")) { r("ic"); break; }
|
|
319 break;
|
|
320 case 'l': if (ends("ical")) { r("ic"); break; }
|
|
321 if (ends("ful")) { r(""); break; }
|
|
322 break;
|
|
323 case 's': if (ends("ness")) { r(""); break; }
|
|
324 break;
|
|
325 } }
|
|
326
|
|
327 /* step5() takes off -ant, -ence etc., in context <c>vcvc<v>. */
|
|
328
|
|
329 private final void step5()
|
|
330 { if (k == 0) return; /* for Bug 1 */ switch (b[k-1])
|
|
331 { case 'a': if (ends("al")) break; return;
|
|
332 case 'c': if (ends("ance")) break;
|
|
333 if (ends("ence")) break; return;
|
|
334 case 'e': if (ends("er")) break; return;
|
|
335 case 'i': if (ends("ic")) break; return;
|
|
336 case 'l': if (ends("able")) break;
|
|
337 if (ends("ible")) break; return;
|
|
338 case 'n': if (ends("ant")) break;
|
|
339 if (ends("ement")) break;
|
|
340 if (ends("ment")) break;
|
|
341 /* element etc. not stripped before the m */
|
|
342 if (ends("ent")) break; return;
|
|
343 case 'o': if (ends("ion") && j >= 0 && (b[j] == 's' || b[j] == 't')) break;
|
|
344 /* j >= 0 fixes Bug 2 */
|
|
345 if (ends("ou")) break; return;
|
|
346 /* takes care of -ous */
|
|
347 case 's': if (ends("ism")) break; return;
|
|
348 case 't': if (ends("ate")) break;
|
|
349 if (ends("iti")) break; return;
|
|
350 case 'u': if (ends("ous")) break; return;
|
|
351 case 'v': if (ends("ive")) break; return;
|
|
352 case 'z': if (ends("ize")) break; return;
|
|
353 default: return;
|
|
354 }
|
|
355 if (m() > 1) k = j;
|
|
356 }
|
|
357
|
|
358 /* step6() removes a final -e if m() > 1. */
|
|
359
|
|
360 private final void step6()
|
|
361 { j = k;
|
|
362 if (b[k] == 'e')
|
|
363 { int a = m();
|
|
364 if (a > 1 || a == 1 && !cvc(k-1)) k--;
|
|
365 }
|
|
366 if (b[k] == 'l' && doublec(k) && m() > 1) k--;
|
|
367 }
|
|
368
|
|
369 /** Stem the word placed into the Stemmer buffer through calls to add().
|
|
370 * Returns true if the stemming process resulted in a word different
|
|
371 * from the input. You can retrieve the result with
|
|
372 * getResultLength()/getResultBuffer() or toString().
|
|
373 */
|
|
374 public void stem()
|
|
375 { k = i - 1;
|
|
376 if (k > 1) { step1(); step2(); step3(); step4(); step5(); step6(); }
|
|
377 i_end = k+1; i = 0;
|
|
378 }
|
|
379
|
|
380 /** Test program for demonstrating the Stemmer. It reads text from a
|
|
381 * a list of files, stems each word, and writes the result to standard
|
|
382 * output. Note that the word stemmed is expected to be in lower case:
|
|
383 * forcing lower case must be done outside the Stemmer class.
|
|
384 * Usage: Stemmer file-name file-name ...
|
|
385 */
|
|
386 public static void main(String[] args)
|
|
387 {
|
|
388 char[] w = new char[501];
|
|
389 Stemmer s = new Stemmer();
|
|
390 for (int i = 0; i < args.length; i++)
|
|
391 try
|
|
392 {
|
|
393 FileInputStream in = new FileInputStream(args[i]);
|
|
394
|
|
395 try
|
|
396 { while(true)
|
|
397
|
|
398 { int ch = in.read();
|
|
399 if (Character.isLetter((char) ch))
|
|
400 {
|
|
401 int j = 0;
|
|
402 while(true)
|
|
403 { ch = Character.toLowerCase((char) ch);
|
|
404 w[j] = (char) ch;
|
|
405 if (j < 500) j++;
|
|
406 ch = in.read();
|
|
407 if (!Character.isLetter((char) ch))
|
|
408 {
|
|
409 /* to test add(char ch) */
|
|
410 for (int c = 0; c < j; c++) s.add(w[c]);
|
|
411
|
|
412 /* or, to test add(char[] w, int j) */
|
|
413 /* s.add(w, j); */
|
|
414
|
|
415 s.stem();
|
|
416 { String u;
|
|
417
|
|
418 /* and now, to test toString() : */
|
|
419 u = s.toString();
|
|
420
|
|
421 /* to test getResultBuffer(), getResultLength() : */
|
|
422 /* u = new String(s.getResultBuffer(), 0, s.getResultLength()); */
|
|
423
|
|
424 System.out.print(u);
|
|
425 }
|
|
426 break;
|
|
427 }
|
|
428 }
|
|
429 }
|
|
430 if (ch < 0) break;
|
|
431 System.out.print((char)ch);
|
|
432 }
|
|
433 }
|
|
434 catch (IOException e)
|
|
435 { System.out.println("error reading " + args[i]);
|
|
436 break;
|
|
437 }
|
|
438 }
|
|
439 catch (FileNotFoundException e)
|
|
440 { System.out.println("file " + args[i] + " not found");
|
|
441 break;
|
|
442 }
|
|
443 }
|
|
444 }
|