-
Notifications
You must be signed in to change notification settings - Fork 113
/
slre.c
437 lines (372 loc) · 14.1 KB
/
slre.c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
/*
* Copyright (c) 2004-2013 Sergey Lyubka <[email protected]>
* Copyright (c) 2013 Cesanta Software Limited
* All rights reserved
*
* This library is dual-licensed: you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2 as
* published by the Free Software Foundation. For the terms of this
* license, see <http://www.gnu.org/licenses/>.
*
* You are free to use this library under the terms of the GNU General
* Public License, but WITHOUT ANY WARRANTY; without even the implied
* warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
* See the GNU General Public License for more details.
*
* Alternatively, you can license this library under a commercial
* license, as set out in <http://cesanta.com/products.html>.
*/
#include <stdio.h>
#include <ctype.h>
#include <string.h>
#include "slre.h"
#define MAX_BRANCHES 100
#define MAX_BRACKETS 100
#define FAIL_IF(condition, error_code) if (condition) return (error_code)
#ifndef ARRAY_SIZE
#define ARRAY_SIZE(ar) (sizeof(ar) / sizeof((ar)[0]))
#endif
#ifdef SLRE_DEBUG
#define DBG(x) printf x
#else
#define DBG(x)
#endif
struct bracket_pair {
const char *ptr; /* Points to the first char after '(' in regex */
int len; /* Length of the text between '(' and ')' */
int branches; /* Index in the branches array for this pair */
int num_branches; /* Number of '|' in this bracket pair */
};
struct branch {
int bracket_index; /* index for 'struct bracket_pair brackets' */
/* array defined below */
const char *schlong; /* points to the '|' character in the regex */
};
struct regex_info {
/*
* Describes all bracket pairs in the regular expression.
* First entry is always present, and grabs the whole regex.
*/
struct bracket_pair brackets[MAX_BRACKETS];
int num_brackets;
/*
* Describes alternations ('|' operators) in the regular expression.
* Each branch falls into a specific branch pair.
*/
struct branch branches[MAX_BRANCHES];
int num_branches;
/* Array of captures provided by the user */
struct slre_cap *caps;
int num_caps;
/* E.g. SLRE_IGNORE_CASE. See enum below */
int flags;
};
static int is_metacharacter(const unsigned char *s) {
static const char *metacharacters = "^$().[]*+?|\\Ssdbfnrtv";
return strchr(metacharacters, *s) != NULL;
}
static int op_len(const char *re) {
return re[0] == '\\' && re[1] == 'x' ? 4 : re[0] == '\\' ? 2 : 1;
}
static int set_len(const char *re, int re_len) {
int len = 0;
while (len < re_len && re[len] != ']') {
len += op_len(re + len);
}
return len <= re_len ? len + 1 : -1;
}
static int get_op_len(const char *re, int re_len) {
return re[0] == '[' ? set_len(re + 1, re_len - 1) + 1 : op_len(re);
}
static int is_quantifier(const char *re) {
return re[0] == '*' || re[0] == '+' || re[0] == '?';
}
static int toi(int x) {
return isdigit(x) ? x - '0' : x - 'W';
}
static int hextoi(const unsigned char *s) {
return (toi(tolower(s[0])) << 4) | toi(tolower(s[1]));
}
static int match_op(const unsigned char *re, const unsigned char *s,
struct regex_info *info) {
int result = 0;
switch (*re) {
case '\\':
/* Metacharacters */
switch (re[1]) {
case 'S': FAIL_IF(isspace(*s), SLRE_NO_MATCH); result++; break;
case 's': FAIL_IF(!isspace(*s), SLRE_NO_MATCH); result++; break;
case 'd': FAIL_IF(!isdigit(*s), SLRE_NO_MATCH); result++; break;
case 'b': FAIL_IF(*s != '\b', SLRE_NO_MATCH); result++; break;
case 'f': FAIL_IF(*s != '\f', SLRE_NO_MATCH); result++; break;
case 'n': FAIL_IF(*s != '\n', SLRE_NO_MATCH); result++; break;
case 'r': FAIL_IF(*s != '\r', SLRE_NO_MATCH); result++; break;
case 't': FAIL_IF(*s != '\t', SLRE_NO_MATCH); result++; break;
case 'v': FAIL_IF(*s != '\v', SLRE_NO_MATCH); result++; break;
case 'x':
/* Match byte, \xHH where HH is hexadecimal byte representaion */
FAIL_IF(hextoi(re + 2) != *s, SLRE_NO_MATCH);
result++;
break;
default:
/* Valid metacharacter check is done in bar() */
FAIL_IF(re[1] != s[0], SLRE_NO_MATCH);
result++;
break;
}
break;
case '|': FAIL_IF(1, SLRE_INTERNAL_ERROR); break;
case '$': FAIL_IF(1, SLRE_NO_MATCH); break;
case '.': result++; break;
default:
if (info->flags & SLRE_IGNORE_CASE) {
FAIL_IF(tolower(*re) != tolower(*s), SLRE_NO_MATCH);
} else {
FAIL_IF(*re != *s, SLRE_NO_MATCH);
}
result++;
break;
}
return result;
}
static int match_set(const char *re, int re_len, const char *s,
struct regex_info *info) {
int len = 0, result = -1, invert = re[0] == '^';
if (invert) re++, re_len--;
while (len <= re_len && re[len] != ']' && result <= 0) {
/* Support character range */
if (re[len] != '-' && re[len + 1] == '-' && re[len + 2] != ']' &&
re[len + 2] != '\0') {
result = info->flags & SLRE_IGNORE_CASE ?
tolower(*s) >= tolower(re[len]) && tolower(*s) <= tolower(re[len + 2]) :
*s >= re[len] && *s <= re[len + 2];
len += 3;
} else {
result = match_op((const unsigned char *) re + len, (const unsigned char *) s, info);
len += op_len(re + len);
}
}
return (!invert && result > 0) || (invert && result <= 0) ? 1 : -1;
}
static int doh(const char *s, int s_len, struct regex_info *info, int bi);
static int bar(const char *re, int re_len, const char *s, int s_len,
struct regex_info *info, int bi) {
/* i is offset in re, j is offset in s, bi is brackets index */
int i, j, n, step;
for (i = j = 0; i < re_len && j <= s_len; i += step) {
/* Handle quantifiers. Get the length of the chunk. */
step = re[i] == '(' ? info->brackets[bi + 1].len + 2 :
get_op_len(re + i, re_len - i);
DBG(("%s [%.*s] [%.*s] re_len=%d step=%d i=%d j=%d\n", __func__,
re_len - i, re + i, s_len - j, s + j, re_len, step, i, j));
FAIL_IF(is_quantifier(&re[i]), SLRE_UNEXPECTED_QUANTIFIER);
FAIL_IF(step <= 0, SLRE_INVALID_CHARACTER_SET);
if (i + step < re_len && is_quantifier(re + i + step)) {
DBG(("QUANTIFIER: [%.*s]%c [%.*s]\n", step, re + i,
re[i + step], s_len - j, s + j));
if (re[i + step] == '?') {
int result = bar(re + i, step, s + j, s_len - j, info, bi);
j += result > 0 ? result : 0;
i++;
} else if (re[i + step] == '+' || re[i + step] == '*') {
int j2 = j, nj = j, n1, n2 = -1, ni, non_greedy = 0;
/* Points to the regexp code after the quantifier */
ni = i + step + 1;
if (ni < re_len && re[ni] == '?') {
non_greedy = 1;
ni++;
}
do {
if ((n1 = bar(re + i, step, s + j2, s_len - j2, info, bi)) > 0) {
j2 += n1;
}
if (re[i + step] == '+' && n1 < 0) break;
if (ni >= re_len) {
/* After quantifier, there is nothing */
nj = j2;
} else if ((n2 = bar(re + ni, re_len - ni, s + j2,
s_len - j2, info, bi)) >= 0) {
/* Regex after quantifier matched */
nj = j2 + n2;
}
if (nj > j && non_greedy) break;
} while (n1 > 0);
/*
* Even if we found one or more pattern, this branch will be executed,
* changing the next captures.
*/
if (n1 < 0 && n2 < 0 && re[i + step] == '*' &&
(n2 = bar(re + ni, re_len - ni, s + j, s_len - j, info, bi)) > 0) {
nj = j + n2;
}
DBG(("STAR/PLUS END: %d %d %d %d %d\n", j, nj, re_len - ni, n1, n2));
FAIL_IF(re[i + step] == '+' && nj == j, SLRE_NO_MATCH);
/* If while loop body above was not executed for the * quantifier, */
/* make sure the rest of the regex matches */
FAIL_IF(nj == j && ni < re_len && n2 < 0, SLRE_NO_MATCH);
/* Returning here cause we've matched the rest of RE already */
return nj;
}
continue;
}
if (re[i] == '[') {
n = match_set(re + i + 1, re_len - (i + 2), s + j, info);
DBG(("SET %.*s [%.*s] -> %d\n", step, re + i, s_len - j, s + j, n));
FAIL_IF(n <= 0, SLRE_NO_MATCH);
j += n;
} else if (re[i] == '(') {
n = SLRE_NO_MATCH;
bi++;
FAIL_IF(bi >= info->num_brackets, SLRE_INTERNAL_ERROR);
DBG(("CAPTURING [%.*s] [%.*s] [%s]\n",
step, re + i, s_len - j, s + j, re + i + step));
if (re_len - (i + step) <= 0) {
/* Nothing follows brackets */
n = doh(s + j, s_len - j, info, bi);
} else {
int j2;
for (j2 = 0; j2 <= s_len - j; j2++) {
if ((n = doh(s + j, s_len - (j + j2), info, bi)) >= 0 &&
bar(re + i + step, re_len - (i + step),
s + j + n, s_len - (j + n), info, bi) >= 0) break;
}
}
DBG(("CAPTURED [%.*s] [%.*s]:%d\n", step, re + i, s_len - j, s + j, n));
FAIL_IF(n < 0, n);
if (info->caps != NULL && n > 0) {
info->caps[bi - 1].ptr = s + j;
info->caps[bi - 1].len = n;
}
j += n;
} else if (re[i] == '^') {
FAIL_IF(j != 0, SLRE_NO_MATCH);
} else if (re[i] == '$') {
FAIL_IF(j != s_len, SLRE_NO_MATCH);
} else {
FAIL_IF(j >= s_len, SLRE_NO_MATCH);
n = match_op((const unsigned char *) (re + i), (const unsigned char *) (s + j), info);
FAIL_IF(n <= 0, n);
j += n;
}
}
return j;
}
/* Process branch points */
static int doh(const char *s, int s_len, struct regex_info *info, int bi) {
const struct bracket_pair *b = &info->brackets[bi];
int i = 0, len, result;
const char *p;
do {
p = i == 0 ? b->ptr : info->branches[b->branches + i - 1].schlong + 1;
len = b->num_branches == 0 ? b->len :
i == b->num_branches ? (int) (b->ptr + b->len - p) :
(int) (info->branches[b->branches + i].schlong - p);
DBG(("%s %d %d [%.*s] [%.*s]\n", __func__, bi, i, len, p, s_len, s));
result = bar(p, len, s, s_len, info, bi);
DBG(("%s <- %d\n", __func__, result));
} while (result <= 0 && i++ < b->num_branches); /* At least 1 iteration */
return result;
}
static int baz(const char *s, int s_len, struct regex_info *info) {
int i, result = -1, is_anchored = info->brackets[0].ptr[0] == '^';
for (i = 0; i <= s_len; i++) {
result = doh(s + i, s_len - i, info, 0);
if (result >= 0) {
result += i;
break;
}
if (is_anchored) break;
}
return result;
}
static void setup_branch_points(struct regex_info *info) {
int i, j;
struct branch tmp;
/* First, sort branches. Must be stable, no qsort. Use bubble algo. */
for (i = 0; i < info->num_branches; i++) {
for (j = i + 1; j < info->num_branches; j++) {
if (info->branches[i].bracket_index > info->branches[j].bracket_index) {
tmp = info->branches[i];
info->branches[i] = info->branches[j];
info->branches[j] = tmp;
}
}
}
/*
* For each bracket, set their branch points. This way, for every bracket
* (i.e. every chunk of regex) we know all branch points before matching.
*/
for (i = j = 0; i < info->num_brackets; i++) {
info->brackets[i].num_branches = 0;
info->brackets[i].branches = j;
while (j < info->num_branches && info->branches[j].bracket_index == i) {
info->brackets[i].num_branches++;
j++;
}
}
}
static int foo(const char *re, int re_len, const char *s, int s_len,
struct regex_info *info) {
int i, step, depth = 0;
/* First bracket captures everything */
info->brackets[0].ptr = re;
info->brackets[0].len = re_len;
info->num_brackets = 1;
/* Make a single pass over regex string, memorize brackets and branches */
for (i = 0; i < re_len; i += step) {
step = get_op_len(re + i, re_len - i);
if (re[i] == '|') {
FAIL_IF(info->num_branches >= (int) ARRAY_SIZE(info->branches),
SLRE_TOO_MANY_BRANCHES);
info->branches[info->num_branches].bracket_index =
info->brackets[info->num_brackets - 1].len == -1 ?
info->num_brackets - 1 : depth;
info->branches[info->num_branches].schlong = &re[i];
info->num_branches++;
} else if (re[i] == '\\') {
FAIL_IF(i >= re_len - 1, SLRE_INVALID_METACHARACTER);
if (re[i + 1] == 'x') {
/* Hex digit specification must follow */
FAIL_IF(re[i + 1] == 'x' && i >= re_len - 3,
SLRE_INVALID_METACHARACTER);
FAIL_IF(re[i + 1] == 'x' && !(isxdigit(re[i + 2]) &&
isxdigit(re[i + 3])), SLRE_INVALID_METACHARACTER);
} else {
FAIL_IF(!is_metacharacter((const unsigned char *) re + i + 1),
SLRE_INVALID_METACHARACTER);
}
} else if (re[i] == '(') {
FAIL_IF(info->num_brackets >= (int) ARRAY_SIZE(info->brackets),
SLRE_TOO_MANY_BRACKETS);
depth++; /* Order is important here. Depth increments first. */
info->brackets[info->num_brackets].ptr = re + i + 1;
info->brackets[info->num_brackets].len = -1;
info->num_brackets++;
FAIL_IF(info->num_caps > 0 && info->num_brackets - 1 > info->num_caps,
SLRE_CAPS_ARRAY_TOO_SMALL);
} else if (re[i] == ')') {
int ind = info->brackets[info->num_brackets - 1].len == -1 ?
info->num_brackets - 1 : depth;
info->brackets[ind].len = (int) (&re[i] - info->brackets[ind].ptr);
DBG(("SETTING BRACKET %d [%.*s]\n",
ind, info->brackets[ind].len, info->brackets[ind].ptr));
depth--;
FAIL_IF(depth < 0, SLRE_UNBALANCED_BRACKETS);
FAIL_IF(i > 0 && re[i - 1] == '(', SLRE_NO_MATCH);
}
}
FAIL_IF(depth != 0, SLRE_UNBALANCED_BRACKETS);
setup_branch_points(info);
return baz(s, s_len, info);
}
int slre_match(const char *regexp, const char *s, int s_len,
struct slre_cap *caps, int num_caps, int flags) {
struct regex_info info;
/* Initialize info structure */
info.flags = flags;
info.num_brackets = info.num_branches = 0;
info.num_caps = num_caps;
info.caps = caps;
DBG(("========================> [%s] [%.*s]\n", regexp, s_len, s));
return foo(regexp, (int) strlen(regexp), s, s_len, &info);
}