1/*	$NetBSD: util.c,v 1.17 2013/01/21 03:24:43 msaitoh Exp $	*/
2/*	$FreeBSD: head/usr.bin/grep/util.c 211496 2010-08-19 09:28:59Z des $	*/
3/*	$OpenBSD: util.c,v 1.39 2010/07/02 22:18:03 tedu Exp $	*/
4
5/*-
6 * Copyright (c) 1999 James Howard and Dag-Erling Coïdan Smørgrav
7 * Copyright (C) 2008-2010 Gabor Kovesdan <gabor@FreeBSD.org>
8 * All rights reserved.
9 *
10 * Redistribution and use in source and binary forms, with or without
11 * modification, are permitted provided that the following conditions
12 * are met:
13 * 1. Redistributions of source code must retain the above copyright
14 *    notice, this list of conditions and the following disclaimer.
15 * 2. Redistributions in binary form must reproduce the above copyright
16 *    notice, this list of conditions and the following disclaimer in the
17 *    documentation and/or other materials provided with the distribution.
18 *
19 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
20 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
21 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
22 * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
23 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
24 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
25 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
26 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
27 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
28 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
29 * SUCH DAMAGE.
30 */
31
32#if HAVE_NBTOOL_CONFIG_H
33#include "nbtool_config.h"
34#endif
35
36#include <sys/cdefs.h>
37__RCSID("$NetBSD: util.c,v 1.17 2013/01/21 03:24:43 msaitoh Exp $");
38
39#include <sys/stat.h>
40#include <sys/types.h>
41
42#include <ctype.h>
43#include <err.h>
44#include <errno.h>
45#include <fnmatch.h>
46#include <fts.h>
47#include <libgen.h>
48#include <stdbool.h>
49#include <stdio.h>
50#include <stdlib.h>
51#include <string.h>
52#include <unistd.h>
53#include <wchar.h>
54#include <wctype.h>
55
56#include "grep.h"
57
58static bool	 first, first_global = true;
59static unsigned long long since_printed;
60
61static int	 procline(struct str *l, int);
62
63bool
64file_matching(const char *fname)
65{
66	char *fname_base, *fname_copy;
67	unsigned int i;
68	bool ret;
69
70	ret = finclude ? false : true;
71	fname_copy = grep_strdup(fname);
72	fname_base = basename(fname_copy);
73
74	for (i = 0; i < fpatterns; ++i) {
75		if (fnmatch(fpattern[i].pat, fname, 0) == 0 ||
76		    fnmatch(fpattern[i].pat, fname_base, 0) == 0) {
77			if (fpattern[i].mode == EXCL_PAT) {
78				free(fname_copy);
79				return (false);
80			} else
81				ret = true;
82		}
83	}
84	free(fname_copy);
85	return (ret);
86}
87
88static inline bool
89dir_matching(const char *dname)
90{
91	unsigned int i;
92	bool ret;
93
94	ret = dinclude ? false : true;
95
96	for (i = 0; i < dpatterns; ++i) {
97		if (dname != NULL &&
98		    fnmatch(dname, dpattern[i].pat, 0) == 0) {
99			if (dpattern[i].mode == EXCL_PAT)
100				return (false);
101			else
102				ret = true;
103		}
104	}
105	return (ret);
106}
107
108/*
109 * Processes a directory when a recursive search is performed with
110 * the -R option.  Each appropriate file is passed to procfile().
111 */
112int
113grep_tree(char **argv)
114{
115	FTS *fts;
116	FTSENT *p;
117	char *d, *dir = NULL;
118	int c, fts_flags;
119	bool ok;
120
121	c = fts_flags = 0;
122
123	switch(linkbehave) {
124	case LINK_EXPLICIT:
125		fts_flags = FTS_COMFOLLOW;
126		break;
127	case LINK_SKIP:
128		fts_flags = FTS_PHYSICAL;
129		break;
130	default:
131		fts_flags = FTS_LOGICAL;
132
133	}
134
135	fts_flags |= FTS_NOSTAT | FTS_NOCHDIR;
136
137	if (!(fts = fts_open(argv, fts_flags, NULL)))
138		err(2, "fts_open");
139	while ((p = fts_read(fts)) != NULL) {
140		switch (p->fts_info) {
141		case FTS_DNR:
142			/* FALLTHROUGH */
143		case FTS_ERR:
144			errx(2, "%s: %s", p->fts_path, strerror(p->fts_errno));
145			break;
146		case FTS_D:
147			/* FALLTHROUGH */
148		case FTS_DP:
149			break;
150		case FTS_DC:
151			/* Print a warning for recursive directory loop */
152			warnx("warning: %s: recursive directory loop",
153				p->fts_path);
154			break;
155		default:
156			/* Check for file exclusion/inclusion */
157			ok = true;
158			if (dexclude || dinclude) {
159				if ((d = strrchr(p->fts_path, '/')) != NULL) {
160					dir = grep_malloc(sizeof(char) *
161					    (d - p->fts_path + 1));
162					memcpy(dir, p->fts_path,
163					    d - p->fts_path);
164					dir[d - p->fts_path] = '\0';
165				}
166				ok = dir_matching(dir);
167				free(dir);
168				dir = NULL;
169			}
170			if (fexclude || finclude)
171				ok &= file_matching(p->fts_path);
172
173			if (ok)
174				c += procfile(p->fts_path);
175			break;
176		}
177	}
178
179	fts_close(fts);
180	return (c);
181}
182
183/*
184 * Opens a file and processes it.  Each file is processed line-by-line
185 * passing the lines to procline().
186 */
187int
188procfile(const char *fn)
189{
190	struct file *f;
191	struct stat sb;
192	struct str ln;
193	mode_t s;
194	int c, t;
195
196	if (mflag && (mcount <= 0))
197		return (0);
198
199	if (strcmp(fn, "-") == 0) {
200		fn = label != NULL ? label : getstr(1);
201		f = grep_open(NULL);
202	} else {
203		if (!stat(fn, &sb)) {
204			/* Check if we need to process the file */
205			s = sb.st_mode & S_IFMT;
206			if (s == S_IFDIR && dirbehave == DIR_SKIP)
207				return (0);
208			if ((s == S_IFIFO || s == S_IFCHR || s == S_IFBLK
209				|| s == S_IFSOCK) && devbehave == DEV_SKIP)
210					return (0);
211		}
212		f = grep_open(fn);
213	}
214	if (f == NULL) {
215		if (!sflag)
216			warn("%s", fn);
217		if (errno == ENOENT)
218			notfound = true;
219		return (0);
220	}
221
222	ln.file = grep_malloc(strlen(fn) + 1);
223	strcpy(ln.file, fn);
224	ln.line_no = 0;
225	ln.len = 0;
226	tail = 0;
227	ln.off = -1;
228
229	for (first = true, c = 0;  c == 0 || !(lflag || qflag); ) {
230		ln.off += ln.len + 1;
231		if ((ln.dat = grep_fgetln(f, &ln.len)) == NULL || ln.len == 0)
232			break;
233		if (ln.len > 0 && ln.dat[ln.len - 1] == line_sep)
234			--ln.len;
235		ln.line_no++;
236
237		/* Return if we need to skip a binary file */
238		if (f->binary && binbehave == BINFILE_SKIP) {
239			grep_close(f);
240			free(ln.file);
241			free(f);
242			return (0);
243		}
244		/* Process the file line-by-line */
245		t = procline(&ln, f->binary);
246		c += t;
247
248		/* Count the matches if we have a match limit */
249		if (mflag) {
250			mcount -= t;
251			if (mcount <= 0)
252				break;
253		}
254	}
255	if (Bflag > 0)
256		clearqueue();
257	grep_close(f);
258
259	if (cflag) {
260		if (!hflag)
261			printf("%s:", ln.file);
262		printf("%u%c", c, line_sep);
263	}
264	if (lflag && !qflag && c != 0)
265		printf("%s%c", fn, line_sep);
266	if (Lflag && !qflag && c == 0)
267		printf("%s%c", fn, line_sep);
268	if (c && !cflag && !lflag && !Lflag &&
269	    binbehave == BINFILE_BIN && f->binary && !qflag)
270		printf(getstr(8), fn);
271
272	free(ln.file);
273	free(f);
274	return (c);
275}
276
277#define iswword(x)	(iswalnum((x)) || (x) == L'_')
278
279/*
280 * Processes a line comparing it with the specified patterns.  Each pattern
281 * is looped to be compared along with the full string, saving each and every
282 * match, which is necessary to colorize the output and to count the
283 * matches.  The matching lines are passed to printline() to display the
284 * appropriate output.
285 */
286static int
287procline(struct str *l, int nottext)
288{
289	regmatch_t matches[MAX_LINE_MATCHES];
290	regmatch_t pmatch;
291	size_t st = 0;
292	unsigned int i;
293	int c = 0, m = 0, r = 0;
294
295	/* Loop to process the whole line */
296	while (st <= l->len) {
297		pmatch.rm_so = st;
298		pmatch.rm_eo = l->len;
299
300		/* Loop to compare with all the patterns */
301		for (i = 0; i < patterns; i++) {
302/*
303 * XXX: grep_search() is a workaround for speed up and should be
304 * removed in the future.  See fastgrep.c.
305 */
306			if (fg_pattern[i].pattern) {
307				r = grep_search(&fg_pattern[i],
308				    (unsigned char *)l->dat,
309				    l->len, &pmatch);
310				r = (r == 0) ? 0 : REG_NOMATCH;
311				st = pmatch.rm_eo;
312			} else {
313				r = regexec(&r_pattern[i], l->dat, 1,
314				    &pmatch, eflags);
315				r = (r == 0) ? 0 : REG_NOMATCH;
316				st = pmatch.rm_eo;
317			}
318			if (r == REG_NOMATCH)
319				continue;
320			/* Check for full match */
321			if (xflag &&
322			    (pmatch.rm_so != 0 ||
323			     (size_t)pmatch.rm_eo != l->len))
324				continue;
325			/* Check for whole word match */
326			if (fg_pattern[i].word && pmatch.rm_so != 0) {
327				wchar_t wbegin, wend;
328
329				wbegin = wend = L' ';
330				if (pmatch.rm_so != 0 &&
331				    sscanf(&l->dat[pmatch.rm_so - 1],
332				    "%lc", &wbegin) != 1)
333					continue;
334				if ((size_t)pmatch.rm_eo != l->len &&
335				    sscanf(&l->dat[pmatch.rm_eo],
336				    "%lc", &wend) != 1)
337					continue;
338				if (iswword(wbegin) || iswword(wend))
339					continue;
340			}
341			c = 1;
342			if (m < MAX_LINE_MATCHES)
343				matches[m++] = pmatch;
344			/* matches - skip further patterns */
345			if ((color != NULL && !oflag) || qflag || lflag)
346				break;
347		}
348
349		if (vflag) {
350			c = !c;
351			break;
352		}
353		/* One pass if we are not recording matches */
354		if ((color != NULL && !oflag) || qflag || lflag)
355			break;
356
357		if (st == (size_t)pmatch.rm_so)
358			break; 	/* No matches */
359	}
360
361	if (c && binbehave == BINFILE_BIN && nottext)
362		return (c); /* Binary file */
363
364	/* Dealing with the context */
365	if ((tail || c) && !cflag && !qflag && !lflag && !Lflag) {
366		if (c) {
367			if ((Aflag || Bflag) && !first_global &&
368			    (first || since_printed > Bflag))
369				printf("--\n");
370			tail = Aflag;
371			if (Bflag > 0)
372				printqueue();
373			printline(l, ':', matches, m);
374		} else {
375			printline(l, '-', matches, m);
376			tail--;
377		}
378		first = false;
379		first_global = false;
380		since_printed = 0;
381	} else {
382		if (Bflag)
383			enqueue(l);
384		since_printed++;
385	}
386	return (c);
387}
388
389/*
390 * Safe malloc() for internal use.
391 */
392void *
393grep_malloc(size_t size)
394{
395	void *ptr;
396
397	if ((ptr = malloc(size)) == NULL)
398		err(2, "malloc");
399	return (ptr);
400}
401
402/*
403 * Safe calloc() for internal use.
404 */
405void *
406grep_calloc(size_t nmemb, size_t size)
407{
408	void *ptr;
409
410	if ((ptr = calloc(nmemb, size)) == NULL)
411		err(2, "calloc");
412	return (ptr);
413}
414
415/*
416 * Safe realloc() for internal use.
417 */
418void *
419grep_realloc(void *ptr, size_t size)
420{
421
422	if ((ptr = realloc(ptr, size)) == NULL)
423		err(2, "realloc");
424	return (ptr);
425}
426
427/*
428 * Safe strdup() for internal use.
429 */
430char *
431grep_strdup(const char *str)
432{
433	char *ret;
434
435	if ((ret = strdup(str)) == NULL)
436		err(2, "strdup");
437	return (ret);
438}
439
440/*
441 * Prints a matching line according to the command line options.
442 */
443void
444printline(struct str *line, int sep, regmatch_t *matches, int m)
445{
446	size_t a = 0;
447	int i, n = 0;
448
449	if (!hflag) {
450		if (nullflag == 0)
451			fputs(line->file, stdout);
452		else {
453			printf("%s", line->file);
454			putchar(0);
455		}
456		++n;
457	}
458	if (nflag) {
459		if (n > 0)
460			putchar(sep);
461		printf("%d", line->line_no);
462		++n;
463	}
464	if (bflag) {
465		if (n > 0)
466			putchar(sep);
467		printf("%lld", (long long)line->off);
468		++n;
469	}
470	if (n)
471		putchar(sep);
472	/* --color and -o */
473	if ((oflag || color) && m > 0) {
474		for (i = 0; i < m; i++) {
475			if (!oflag)
476				fwrite(line->dat + a, matches[i].rm_so - a, 1,
477				    stdout);
478			if (color)
479				fprintf(stdout, "\33[%sm\33[K", color);
480
481				fwrite(line->dat + matches[i].rm_so,
482				    matches[i].rm_eo - matches[i].rm_so, 1,
483				    stdout);
484			if (color)
485				fprintf(stdout, "\33[m\33[K");
486			a = matches[i].rm_eo;
487			if (oflag)
488				putchar('\n');
489		}
490		if (!oflag) {
491			if (line->len - a > 0)
492				fwrite(line->dat + a, line->len - a, 1, stdout);
493			putchar(line_sep);
494		}
495	} else {
496		fwrite(line->dat, line->len, 1, stdout);
497		putchar(line_sep);
498	}
499}
500