summaryrefslogtreecommitdiff
path: root/usr.bin/grep/fastgrep.c
blob: 2fcd864e2d7616a0b2d800ba228dd75f7a1ef91f (plain) (blame)
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
/*	$OpenBSD: util.c,v 1.36 2007/10/02 17:59:18 otto Exp $	*/
/*	$FreeBSD: head/usr.bin/grep/fastgrep.c 211496 2010-08-19 09:28:59Z des $ */

/*-
 * Copyright (c) 1999 James Howard and Dag-Erling Coïdan Smørgrav
 * Copyright (C) 2008 Gabor Kovesdan <gabor@FreeBSD.org>
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 *
 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 */

/*
 * XXX: This file is a speed up for grep to cover the defects of the
 * regex library.  These optimizations should practically be implemented
 * there keeping this code clean.  This is a future TODO, but for the
 * meantime, we need to use this workaround.
 */

#if HAVE_NBTOOL_CONFIG_H
#include "nbtool_config.h"
#endif

#include <sys/cdefs.h>
__RCSID("$NetBSD: fastgrep.c,v 1.5 2011/04/18 03:27:40 joerg Exp $");

#include <limits.h>
#include <stdbool.h>
#include <stdlib.h>
#include <string.h>
#include <wchar.h>
#include <wctype.h>

#include "grep.h"

static inline int	grep_cmp(const unsigned char *, const unsigned char *, size_t);
static inline void	grep_revstr(unsigned char *, int);

void
fgrepcomp(fastgrep_t *fg, const char *pat)
{
	unsigned int i;

	/* Initialize. */
	fg->len = strlen(pat);
	fg->bol = false;
	fg->eol = false;
	fg->reversed = false;

	fg->pattern = (unsigned char *)grep_strdup(pat);

	/* Preprocess pattern. */
	for (i = 0; i <= UCHAR_MAX; i++)
		fg->qsBc[i] = fg->len;
	for (i = 1; i < fg->len; i++)
		fg->qsBc[fg->pattern[i]] = fg->len - i;
}

/*
 * Returns: -1 on failure, 0 on success
 */
int
fastcomp(fastgrep_t *fg, const char *pat)
{
	unsigned int i;
	int firstHalfDot = -1;
	int firstLastHalfDot = -1;
	int hasDot = 0;
	int lastHalfDot = 0;
	int shiftPatternLen;

	/* Initialize. */
	fg->len = strlen(pat);
	fg->bol = false;
	fg->eol = false;
	fg->reversed = false;
	fg->word = wflag;

	/* Remove end-of-line character ('$'). */
	if (fg->len > 0 && pat[fg->len - 1] == '$') {
		fg->eol = true;
		fg->len--;
	}

	/* Remove beginning-of-line character ('^'). */
	if (pat[0] == '^') {
		fg->bol = true;
		fg->len--;
		pat++;
	}

	if (fg->len >= 14 &&
	    memcmp(pat, "[[:<:]]", 7) == 0 &&
	    memcmp(pat + fg->len - 7, "[[:>:]]", 7) == 0) {
		fg->len -= 14;
		pat += 7;
		/* Word boundary is handled separately in util.c */
		fg->word = true;
	}

	/*
	 * pat has been adjusted earlier to not include '^', '$' or
	 * the word match character classes at the beginning and ending
	 * of the string respectively.
	 */
	fg->pattern = grep_malloc(fg->len + 1);
	memcpy(fg->pattern, pat, fg->len);
	fg->pattern[fg->len] = '\0';

	/* Look for ways to cheat...er...avoid the full regex engine. */
	for (i = 0; i < fg->len; i++) {
		/* Can still cheat? */
		if (fg->pattern[i] == '.') {
			hasDot = i;
			if (i < fg->len / 2) {
				if (firstHalfDot < 0)
					/* Closest dot to the beginning */
					firstHalfDot = i;
			} else {
				/* Closest dot to the end of the pattern. */
				lastHalfDot = i;
				if (firstLastHalfDot < 0)
					firstLastHalfDot = i;
			}
		} else {
			/* Free memory and let others know this is empty. */
			free(fg->pattern);
			fg->pattern = NULL;
			return (-1);
		}
	}

	/*
	 * Determine if a reverse search would be faster based on the placement
	 * of the dots.
	 */
	if ((!(lflag || cflag)) && ((!(fg->bol || fg->eol)) &&
	    ((lastHalfDot) && ((firstHalfDot < 0) ||
	    ((fg->len - (lastHalfDot + 1)) < (size_t)firstHalfDot)))) &&
	    !oflag && !color) {
		fg->reversed = true;
		hasDot = fg->len - (firstHalfDot < 0 ?
		    firstLastHalfDot : firstHalfDot) - 1;
		grep_revstr(fg->pattern, fg->len);
	}

	/*
	 * Normal Quick Search would require a shift based on the position the
	 * next character after the comparison is within the pattern.  With
	 * wildcards, the position of the last dot effects the maximum shift
	 * distance.
	 * The closer to the end the wild card is the slower the search.  A
	 * reverse version of this algorithm would be useful for wildcards near
	 * the end of the string.
	 *
	 * Examples:
	 * Pattern	Max shift
	 * -------	---------
	 * this		5
	 * .his		4
	 * t.is		3
	 * th.s		2
	 * thi.		1
	 */

	/* Adjust the shift based on location of the last dot ('.'). */
	shiftPatternLen = fg->len - hasDot;

	/* Preprocess pattern. */
	for (i = 0; i <= (signed)UCHAR_MAX; i++)
		fg->qsBc[i] = shiftPatternLen;
	for (i = hasDot + 1; i < fg->len; i++) {
		fg->qsBc[fg->pattern[i]] = fg->len - i;
	}

	/*
	 * Put pattern back to normal after pre-processing to allow for easy
	 * comparisons later.
	 */
	if (fg->reversed)
		grep_revstr(fg->pattern, fg->len);

	return (0);
}

int
grep_search(fastgrep_t *fg, const unsigned char *data, size_t len, regmatch_t *pmatch)
{
	unsigned int j;
	int ret = REG_NOMATCH;

	if (pmatch->rm_so == (ssize_t)len)
		return (ret);

	if (fg->bol && pmatch->rm_so != 0) {
		pmatch->rm_so = len;
		pmatch->rm_eo = len;
		return (ret);
	}

	/* No point in going farther if we do not have enough data. */
	if (len < fg->len)
		return (ret);

	/* Only try once at the beginning or ending of the line. */
	if (fg->bol || fg->eol) {
		/* Simple text comparison. */
		/* Verify data is >= pattern length before searching on it. */
		if (len >= fg->len) {
			/* Determine where in data to start search at. */
			j = fg->eol ? len - fg->len : 0;
			if (!((fg->bol && fg->eol) && (len != fg->len)))
				if (grep_cmp(fg->pattern, data + j,
				    fg->len) == -1) {
					pmatch->rm_so = j;
					pmatch->rm_eo = j + fg->len;
						ret = 0;
				}
		}
	} else if (fg->reversed) {
		/* Quick Search algorithm. */
		j = len;
		do {
			if (grep_cmp(fg->pattern, data + j - fg->len,
				fg->len) == -1) {
				pmatch->rm_so = j - fg->len;
				pmatch->rm_eo = j;
				ret = 0;
				break;
			}
			/* Shift if within bounds, otherwise, we are done. */
			if (j == fg->len)
				break;
			j -= fg->qsBc[data[j - fg->len - 1]];
		} while (j >= fg->len);
	} else {
		/* Quick Search algorithm. */
		j = pmatch->rm_so;
		do {
			if (grep_cmp(fg->pattern, data + j, fg->len) == -1) {
				pmatch->rm_so = j;
				pmatch->rm_eo = j + fg->len;
				ret = 0;
				break;
			}

			/* Shift if within bounds, otherwise, we are done. */
			if (j + fg->len == len)
				break;
			else
				j += fg->qsBc[data[j + fg->len]];
		} while (j <= (len - fg->len));
	}

	return (ret);
}

/*
 * Returns:	i >= 0 on failure (position that it failed)
 *		-1 on success
 */
static inline int
grep_cmp(const unsigned char *pat, const unsigned char *data, size_t len)
{
	size_t size;
	wchar_t *wdata, *wpat;
	unsigned int i;

	if (iflag) {
		if ((size = mbstowcs(NULL, (const char *)data, 0)) ==
		    ((size_t) - 1))
			return (-1);

		wdata = grep_malloc(size * sizeof(wint_t));

		if (mbstowcs(wdata, (const char *)data, size) ==
		    ((size_t) - 1))
			return (-1);

		if ((size = mbstowcs(NULL, (const char *)pat, 0)) ==
		    ((size_t) - 1))
			return (-1);

		wpat = grep_malloc(size * sizeof(wint_t));

		if (mbstowcs(wpat, (const char *)pat, size) == ((size_t) - 1))
			return (-1);
		for (i = 0; i < len; i++) {
			if ((towlower(wpat[i]) == towlower(wdata[i])) ||
			    ((grepbehave != GREP_FIXED) && wpat[i] == L'.'))
				continue;
			free(wpat);
			free(wdata);
				return (i);
		}
	} else {
		for (i = 0; i < len; i++) {
			if ((pat[i] == data[i]) || ((grepbehave != GREP_FIXED) &&
			    pat[i] == '.'))
				continue;
			return (i);
		}
	}
	return (-1);
}

static inline void
grep_revstr(unsigned char *str, int len)
{
	int i;
	char c;

	for (i = 0; i < len / 2; i++) {
		c = str[i];
		str[i] = str[len - i - 1];
		str[len - i - 1] = c;
	}
}