util.c revision 211364 
1/*	$OpenBSD: util.c,v 1.39 2010/07/02 22:18:03 tedu Exp $	*/
2
3/*-
4 * Copyright (c) 1999 James Howard and Dag-Erling Co�dan Sm�rgrav
5 * Copyright (C) 2008-2010 Gabor Kovesdan <gabor@FreeBSD.org>
6 * All rights reserved.
7 *
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
10 * are met:
11 * 1. Redistributions of source code must retain the above copyright
12 *    notice, this list of conditions and the following disclaimer.
13 * 2. Redistributions in binary form must reproduce the above copyright
14 *    notice, this list of conditions and the following disclaimer in the
15 *    documentation and/or other materials provided with the distribution.
16 *
17 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
18 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20 * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
21 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27 * SUCH DAMAGE.
28 */
29
30#include <sys/cdefs.h>
31__FBSDID("$FreeBSD: head/usr.bin/grep/util.c 211364 2010-08-15 22:15:04Z gabor $");
32
33#include <sys/stat.h>
34#include <sys/types.h>
35
36#include <ctype.h>
37#include <err.h>
38#include <errno.h>
39#include <fnmatch.h>
40#include <fts.h>
41#include <libgen.h>
42#include <stdbool.h>
43#include <stdio.h>
44#include <stdlib.h>
45#include <string.h>
46#include <unistd.h>
47#include <wchar.h>
48#include <wctype.h>
49
50#include "grep.h"
51
52static int	 linesqueued;
53static int	 procline(struct str *l, int);
54
55bool
56file_matching(const char *fname)
57{
58	bool ret;
59
60	ret = finclude ? false : true;
61
62	for (unsigned int i = 0; i < fpatterns; ++i) {
63		if (fnmatch(fpattern[i].pat,
64		    fname, 0) == 0 || fnmatch(fpattern[i].pat,
65		    basename(fname), 0) == 0) {
66			if (fpattern[i].mode == EXCL_PAT)
67				return (false);
68			else
69				ret = true;
70		}
71	}
72	return (ret);
73}
74
75static inline bool
76dir_matching(const char *dname)
77{
78	bool ret;
79
80	ret = dinclude ? false : true;
81
82	for (unsigned int i = 0; i < dpatterns; ++i) {
83		if (dname != NULL &&
84		    fnmatch(dname, dpattern[i].pat, 0) == 0) {
85			if (dpattern[i].mode == EXCL_PAT)
86				return (false);
87			else
88				ret = true;
89		}
90	}
91	return (ret);
92}
93
94/*
95 * Processes a directory when a recursive search is performed with
96 * the -R option.  Each appropriate file is passed to procfile().
97 */
98int
99grep_tree(char **argv)
100{
101	FTS *fts;
102	FTSENT *p;
103	char *d, *dir = NULL;
104	int c, fts_flags;
105	bool ok;
106
107	c = fts_flags = 0;
108
109	switch(linkbehave) {
110	case LINK_EXPLICIT:
111		fts_flags = FTS_COMFOLLOW;
112		break;
113	case LINK_SKIP:
114		fts_flags = FTS_PHYSICAL;
115		break;
116	default:
117		fts_flags = FTS_LOGICAL;
118
119	}
120
121	fts_flags |= FTS_NOSTAT | FTS_NOCHDIR;
122
123	if (!(fts = fts_open(argv, fts_flags, NULL)))
124		err(2, "fts_open");
125	while ((p = fts_read(fts)) != NULL) {
126		switch (p->fts_info) {
127		case FTS_DNR:
128			/* FALLTHROUGH */
129		case FTS_ERR:
130			errx(2, "%s: %s", p->fts_path, strerror(p->fts_errno));
131			break;
132		case FTS_D:
133			/* FALLTHROUGH */
134		case FTS_DP:
135			break;
136		case FTS_DC:
137			/* Print a warning for recursive directory loop */
138			warnx("warning: %s: recursive directory loop",
139				p->fts_path);
140			break;
141		default:
142			/* Check for file exclusion/inclusion */
143			ok = true;
144			if (dexclude || dinclude) {
145				if ((d = strrchr(p->fts_path, '/')) != NULL) {
146					dir = grep_malloc(sizeof(char) *
147					    (d - p->fts_path + 1));
148					memcpy(dir, p->fts_path,
149					    d - p->fts_path);
150					dir[d - p->fts_path] = '\0';
151				}
152				ok = dir_matching(dir);
153				free(dir);
154				dir = NULL;
155			}
156			if (fexclude || finclude)
157				ok &= file_matching(p->fts_path);
158
159			if (ok)
160				c += procfile(p->fts_path);
161			break;
162		}
163	}
164
165	fts_close(fts);
166	return (c);
167}
168
169/*
170 * Opens a file and processes it.  Each file is processed line-by-line
171 * passing the lines to procline().
172 */
173int
174procfile(const char *fn)
175{
176	struct file *f;
177	struct stat sb;
178	struct str ln;
179	mode_t s;
180	int c, t;
181
182	if (mflag && (mcount <= 0))
183		return (0);
184
185	if (strcmp(fn, "-") == 0) {
186		fn = label != NULL ? label : getstr(1);
187		f = grep_stdin_open();
188	} else {
189		if (!stat(fn, &sb)) {
190			/* Check if we need to process the file */
191			s = sb.st_mode & S_IFMT;
192			if (s == S_IFDIR && dirbehave == DIR_SKIP)
193				return (0);
194			if ((s == S_IFIFO || s == S_IFCHR || s == S_IFBLK
195				|| s == S_IFSOCK) && devbehave == DEV_SKIP)
196					return (0);
197		}
198		f = grep_open(fn);
199	}
200	if (f == NULL) {
201		if (!sflag)
202			warn("%s", fn);
203		if (errno == ENOENT)
204			notfound = true;
205		return (0);
206	}
207
208	ln.file = grep_malloc(strlen(fn) + 1);
209	strcpy(ln.file, fn);
210	ln.line_no = 0;
211	ln.len = 0;
212	linesqueued = 0;
213	tail = 0;
214	ln.off = -1;
215
216	for (c = 0;  c == 0 || !(lflag || qflag); ) {
217		ln.off += ln.len + 1;
218		if ((ln.dat = grep_fgetln(f, &ln.len)) == NULL) {
219			if (ln.line_no == 0 && matchall)
220				exit(0);
221			else
222				break;
223		}
224		if (ln.len > 0 && ln.dat[ln.len - 1] == '\n')
225			--ln.len;
226		ln.line_no++;
227
228		/* Return if we need to skip a binary file */
229		if (f->binary && binbehave == BINFILE_SKIP) {
230			grep_close(f);
231			free(ln.file);
232			free(f);
233			return (0);
234		}
235		/* Process the file line-by-line */
236		if ((t = procline(&ln, f->binary)) == 0 && Bflag > 0) {
237			enqueue(&ln);
238			linesqueued++;
239		}
240		c += t;
241
242		/* Count the matches if we have a match limit */
243		if (mflag) {
244			mcount -= t;
245			if (mcount <= 0)
246				break;
247		}
248	}
249	if (Bflag > 0)
250		clearqueue();
251	grep_close(f);
252
253	if (cflag) {
254		if (!hflag)
255			printf("%s:", ln.file);
256		printf("%u\n", c);
257	}
258	if (lflag && !qflag && c != 0)
259		printf("%s\n", fn);
260	if (Lflag && !qflag && c == 0)
261		printf("%s\n", fn);
262	if (c && !cflag && !lflag && !Lflag &&
263	    binbehave == BINFILE_BIN && f->binary && !qflag)
264		printf(getstr(8), fn);
265
266	free(ln.file);
267	free(f);
268	return (c);
269}
270
271#define iswword(x)	(iswalnum((x)) || (x) == L'_')
272
273/*
274 * Processes a line comparing it with the specified patterns.  Each pattern
275 * is looped to be compared along with the full string, saving each and every
276 * match, which is necessary to colorize the output and to count the
277 * matches.  The matching lines are passed to printline() to display the
278 * appropriate output.
279 */
280static inline int
281procline(struct str *l, int nottext)
282{
283	regmatch_t matches[MAX_LINE_MATCHES];
284	regmatch_t pmatch;
285	size_t st = 0;
286	unsigned int i;
287	int c = 0, m = 0, r = 0;
288
289	if (!matchall) {
290		/* Loop to process the whole line */
291		while (st <= l->len) {
292			pmatch.rm_so = st;
293			pmatch.rm_eo = l->len;
294
295			/* Loop to compare with all the patterns */
296			for (i = 0; i < patterns; i++) {
297/*
298 * XXX: grep_search() is a workaround for speed up and should be
299 * removed in the future.  See fastgrep.c.
300 */
301				if (fg_pattern[i].pattern) {
302					r = grep_search(&fg_pattern[i],
303					    (unsigned char *)l->dat,
304					    l->len, &pmatch);
305					r = (r == 0) ? 0 : REG_NOMATCH;
306					st = pmatch.rm_eo;
307				} else {
308					r = regexec(&r_pattern[i], l->dat, 1,
309					    &pmatch, eflags);
310					r = (r == 0) ? 0 : REG_NOMATCH;
311					st = pmatch.rm_eo;
312				}
313				if (r == REG_NOMATCH)
314					continue;
315				/* Check for full match */
316				if (r == 0 && xflag)
317					if (pmatch.rm_so != 0 ||
318					    (size_t)pmatch.rm_eo != l->len)
319						r = REG_NOMATCH;
320				/* Check for whole word match */
321				if (r == 0 && wflag && pmatch.rm_so != 0) {
322					wint_t wbegin, wend;
323
324					wbegin = wend = L' ';
325					if (pmatch.rm_so != 0 &&
326					    sscanf(&l->dat[pmatch.rm_so - 1],
327					    "%lc", &wbegin) != 1)
328						r = REG_NOMATCH;
329					else if ((size_t)pmatch.rm_eo != l->len &&
330					    sscanf(&l->dat[pmatch.rm_eo],
331					    "%lc", &wend) != 1)
332						r = REG_NOMATCH;
333					else if (iswword(wbegin) || iswword(wend))
334						r = REG_NOMATCH;
335				}
336				if (r == 0) {
337					if (m == 0)
338						c++;
339					if (m < MAX_LINE_MATCHES)
340						matches[m++] = pmatch;
341					/* matches - skip further patterns */
342					if ((color != NULL && !oflag) || qflag || lflag)
343						break;
344				}
345			}
346
347			if (vflag) {
348				c = !c;
349				break;
350			}
351			/* One pass if we are not recording matches */
352			if ((color != NULL && !oflag) || qflag || lflag)
353				break;
354
355			if (st == (size_t)pmatch.rm_so)
356				break; 	/* No matches */
357		}
358	} else
359		c = !vflag;
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 (!first && !prev && !tail && Aflag)
368				printf("--\n");
369			tail = Aflag;
370			if (Bflag > 0) {
371				if (!first && !prev)
372					printf("--\n");
373				printqueue();
374			}
375			linesqueued = 0;
376			printline(l, ':', matches, m);
377		} else {
378			printline(l, '-', matches, m);
379			tail--;
380		}
381	}
382
383	if (c) {
384		prev = true;
385		first = false;
386	} else
387		prev = false;
388
389	return (c);
390}
391
392/*
393 * Safe malloc() for internal use.
394 */
395void *
396grep_malloc(size_t size)
397{
398	void *ptr;
399
400	if ((ptr = malloc(size)) == NULL)
401		err(2, "malloc");
402	return (ptr);
403}
404
405/*
406 * Safe calloc() for internal use.
407 */
408void *
409grep_calloc(size_t nmemb, size_t size)
410{
411	void *ptr;
412
413	if ((ptr = calloc(nmemb, size)) == NULL)
414		err(2, "calloc");
415	return (ptr);
416}
417
418/*
419 * Safe realloc() for internal use.
420 */
421void *
422grep_realloc(void *ptr, size_t size)
423{
424
425	if ((ptr = realloc(ptr, size)) == NULL)
426		err(2, "realloc");
427	return (ptr);
428}
429
430/*
431 * Safe strdup() for internal use.
432 */
433char *
434grep_strdup(const char *str)
435{
436	char *ret;
437
438	if ((ret = strdup(str)) == NULL)
439		err(2, "strdup");
440	return (ret);
441}
442
443/*
444 * Prints a matching line according to the command line options.
445 */
446void
447printline(struct str *line, int sep, regmatch_t *matches, int m)
448{
449	size_t a = 0;
450	int i, n = 0;
451
452	if (!hflag) {
453		if (nullflag == 0)
454			fputs(line->file, stdout);
455		else {
456			printf("%s", line->file);
457			putchar(0);
458		}
459		++n;
460	}
461	if (nflag) {
462		if (n > 0)
463			putchar(sep);
464		printf("%d", line->line_no);
465		++n;
466	}
467	if (bflag) {
468		if (n > 0)
469			putchar(sep);
470		printf("%lld", (long long)line->off);
471		++n;
472	}
473	if (n)
474		putchar(sep);
475	/* --color and -o */
476	if ((oflag || color) && m > 0) {
477		for (i = 0; i < m; i++) {
478			if (!oflag)
479				fwrite(line->dat + a, matches[i].rm_so - a, 1,
480				    stdout);
481			if (color)
482				fprintf(stdout, "\33[%sm\33[K", color);
483
484				fwrite(line->dat + matches[i].rm_so,
485				    matches[i].rm_eo - matches[i].rm_so, 1,
486				    stdout);
487			if (color)
488				fprintf(stdout, "\33[m\33[K");
489			a = matches[i].rm_eo;
490			if (oflag)
491				putchar('\n');
492		}
493		if (!oflag) {
494			if (line->len - a > 0)
495				fwrite(line->dat + a, line->len - a, 1, stdout);
496			putchar('\n');
497		}
498	} else {
499		fwrite(line->dat, line->len, 1, stdout);
500		putchar('\n');
501	}
502}
503