1 /* $NetBSD: util.c,v 1.9 2011/02/27 17:33:37 joerg Exp $ */
3 /* $OpenBSD: util.c,v 1.39 2010/07/02 22:18:03 tedu Exp $ */
6 * Copyright (c) 1999 James Howard and Dag-Erling Coïdan Smørgrav
7 * Copyright (C) 2008-2010 Gabor Kovesdan <gabor@FreeBSD.org>
10 * Redistribution and use in source and binary forms, with or without
11 * modification, are permitted provided that the following conditions
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.
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
32 #include <sys/cdefs.h>
33 __FBSDID("$FreeBSD$");
36 #include <sys/types.h>
52 #ifndef WITHOUT_FASTMATCH
53 #include "fastmatch.h"
57 static bool first_match = true;
60 * Parsing context; used to hold things like matches made and
64 regmatch_t matches[MAX_LINE_MATCHES]; /* Matches made */
65 struct str ln; /* Current line */
66 size_t matchidx; /* Latest used match index */
67 bool binary; /* Binary file? */
71 static int procline(struct parsec *pc);
72 static void printline(struct parsec *pc, int sep);
73 static void printline_metadata(struct str *line, int sep);
76 file_matching(const char *fname)
78 char *fname_base, *fname_buf;
81 ret = finclude ? false : true;
82 fname_buf = strdup(fname);
83 if (fname_buf == NULL)
85 fname_base = basename(fname_buf);
87 for (unsigned int i = 0; i < fpatterns; ++i) {
88 if (fnmatch(fpattern[i].pat, fname, 0) == 0 ||
89 fnmatch(fpattern[i].pat, fname_base, 0) == 0) {
90 if (fpattern[i].mode == EXCL_PAT) {
102 dir_matching(const char *dname)
106 ret = dinclude ? false : true;
108 for (unsigned int i = 0; i < dpatterns; ++i) {
110 fnmatch(dpattern[i].pat, dname, 0) == 0) {
111 if (dpattern[i].mode == EXCL_PAT)
121 * Processes a directory when a recursive search is performed with
122 * the -R option. Each appropriate file is passed to procfile().
125 grep_tree(char **argv)
131 const char *wd[] = { ".", NULL };
137 fts_flags = FTS_COMFOLLOW;
140 fts_flags = FTS_PHYSICAL;
143 fts_flags = FTS_LOGICAL;
147 fts_flags |= FTS_NOSTAT | FTS_NOCHDIR;
149 fts = fts_open((argv[0] == NULL) ?
150 __DECONST(char * const *, wd) : argv, fts_flags, NULL);
153 while ((p = fts_read(fts)) != NULL) {
154 switch (p->fts_info) {
160 warnx("%s: %s", p->fts_path, strerror(p->fts_errno));
165 if (dexclude || dinclude)
166 if (!dir_matching(p->fts_name) ||
167 !dir_matching(p->fts_path))
168 fts_set(fts, p, FTS_SKIP);
171 /* Print a warning for recursive directory loop */
172 warnx("warning: %s: recursive directory loop",
176 /* Check for file exclusion/inclusion */
178 if (fexclude || finclude)
179 ok &= file_matching(p->fts_path);
182 c += procfile(p->fts_path);
192 * Opens a file and processes it. Each file is processed line-by-line
193 * passing the lines to procline().
196 procfile(const char *fn)
204 int c, last_outed, t;
205 bool doctx, printmatch, same_file;
207 if (strcmp(fn, "-") == 0) {
208 fn = label != NULL ? label : getstr(1);
211 if (!stat(fn, &sb)) {
212 /* Check if we need to process the file */
213 s = sb.st_mode & S_IFMT;
214 if (s == S_IFDIR && dirbehave == DIR_SKIP)
216 if ((s == S_IFIFO || s == S_IFCHR || s == S_IFBLK
217 || s == S_IFSOCK) && devbehave == DEV_SKIP)
231 pc.ln.file = grep_malloc(strlen(fn) + 1);
232 strcpy(pc.ln.file, fn);
236 pc.binary = f->binary;
242 if ((pc.binary && binbehave == BINFILE_BIN) || cflag || qflag ||
245 if (printmatch && (Aflag != 0 || Bflag != 0))
249 for (c = 0; c == 0 || !(lflag || qflag); ) {
250 /* Reset match count for every line processed */
252 pc.ln.off += pc.ln.len + 1;
253 if ((pc.ln.dat = grep_fgetln(f, &pc.ln.len)) == NULL ||
255 if (pc.ln.line_no == 0 && matchall)
257 * An empty file with an empty pattern and the
258 * -w flag does not match
260 exit(matchall && wflag ? 1 : 0);
265 if (pc.ln.len > 0 && pc.ln.dat[pc.ln.len - 1] == fileeol)
269 /* Return if we need to skip a binary file */
270 if (pc.binary && binbehave == BINFILE_SKIP) {
277 if ((t = procline(&pc)) == 0)
280 /* Deal with any -B context or context separators */
281 if (t == 0 && doctx) {
282 if (!first_match && (!same_file || last_outed > 0))
288 /* Print the matching line, but only if not quiet/binary */
289 if (t == 0 && printmatch) {
295 if (t != 0 && doctx) {
296 /* Deal with any -A context */
304 * Enqueue non-matching lines for -B context.
305 * If we're not actually doing -B context or if
306 * the enqueue resulted in a line being rotated
307 * out, then go ahead and increment last_outed
308 * to signify a gap between context/match.
310 if (Bflag == 0 || (Bflag > 0 && enqueue(ln)))
315 /* Count the matches if we have a match limit */
316 if (t == 0 && mflag) {
318 if (mflag && mcount <= 0)
329 printf("%s:", pc.ln.file);
332 if (lflag && !qflag && c != 0)
333 printf("%s%c", fn, nullflag ? 0 : '\n');
334 if (Lflag && !qflag && c == 0)
335 printf("%s%c", fn, nullflag ? 0 : '\n');
336 if (c && !cflag && !lflag && !Lflag &&
337 binbehave == BINFILE_BIN && f->binary && !qflag)
338 printf(getstr(8), fn);
345 #define iswword(x) (iswalnum((x)) || (x) == L'_')
348 * Processes a line comparing it with the specified patterns. Each pattern
349 * is looped to be compared along with the full string, saving each and every
350 * match, which is necessary to colorize the output and to count the
351 * matches. The matching lines are passed to printline() to display the
352 * appropriate output.
355 procline(struct parsec *pc)
357 regmatch_t pmatch, lastmatch, chkmatch;
358 wchar_t wbegin, wend;
359 size_t st = 0, nst = 0;
361 int c = 0, r = 0, lastmatches = 0, leflags = eflags;
362 size_t startm = 0, matchidx;
365 matchidx = pc->matchidx;
367 /* Special case: empty pattern with -w flag, check first character */
368 if (matchall && wflag) {
372 if (sscanf(&pc->ln.dat[0], "%lc", &wend) != 1 || iswword(wend))
379 /* Initialize to avoid a false positive warning from GCC. */
380 lastmatch.rm_so = lastmatch.rm_eo = 0;
382 /* Loop to process the whole line */
383 while (st <= pc->ln.len) {
388 leflags |= REG_NOTBOL;
389 /* Loop to compare with all the patterns */
390 for (i = 0; i < patterns; i++) {
392 pmatch.rm_eo = pc->ln.len;
393 #ifndef WITHOUT_FASTMATCH
394 if (fg_pattern[i].pattern)
395 r = fastexec(&fg_pattern[i],
396 pc->ln.dat, 1, &pmatch, leflags);
399 r = regexec(&r_pattern[i], pc->ln.dat, 1,
403 /* Check for full match */
404 if (xflag && (pmatch.rm_so != 0 ||
405 (size_t)pmatch.rm_eo != pc->ln.len))
407 /* Check for whole word match */
408 #ifndef WITHOUT_FASTMATCH
409 if (wflag || fg_pattern[i].word) {
413 wbegin = wend = L' ';
414 if (pmatch.rm_so != 0 &&
415 sscanf(&pc->ln.dat[pmatch.rm_so - 1],
416 "%lc", &wbegin) != 1)
418 else if ((size_t)pmatch.rm_eo !=
420 sscanf(&pc->ln.dat[pmatch.rm_eo],
423 else if (iswword(wbegin) ||
427 * If we're doing whole word matching and we
428 * matched once, then we should try the pattern
429 * again after advancing just past the start of
430 * the earliest match. This allows the pattern
431 * to match later on in the line and possibly
432 * still match a whole word.
434 if (r == REG_NOMATCH &&
435 (retry == 0 || pmatch.rm_so + 1 < retry))
436 retry = pmatch.rm_so + 1;
437 if (r == REG_NOMATCH)
448 * Replace previous match if the new one is earlier
449 * and/or longer. This will lead to some amount of
450 * extra work if -o/--color are specified, but it's
451 * worth it from a correctness point of view.
453 if (matchidx > startm) {
454 chkmatch = pc->matches[matchidx - 1];
455 if (pmatch.rm_so < chkmatch.rm_so ||
456 (pmatch.rm_so == chkmatch.rm_so &&
457 (pmatch.rm_eo - pmatch.rm_so) >
458 (chkmatch.rm_eo - chkmatch.rm_so))) {
459 pc->matches[matchidx - 1] = pmatch;
463 /* Advance as normal if not */
464 pc->matches[matchidx++] = pmatch;
467 /* avoid excessive matching - skip further patterns */
468 if ((color == NULL && !oflag) || qflag || lflag ||
469 matchidx >= MAX_LINE_MATCHES)
474 * Advance to just past the start of the earliest match, try
475 * again just in case we still have a chance to match later in
478 if (lastmatches == 0 && retry > 0) {
483 /* One pass if we are not recording matches */
484 if (!wflag && ((color == NULL && !oflag) || qflag || lflag || Lflag))
487 /* If we didn't have any matches or REG_NOSUB set */
488 if (lastmatches == 0 || (cflags & REG_NOSUB))
491 if (lastmatches == 0)
494 else if (st == nst && lastmatch.rm_so == lastmatch.rm_eo)
495 /* Zero-length match -- advance one more so we don't get stuck */
498 /* Advance st based on previous matches */
502 /* Reflect the new matchidx in the context */
503 pc->matchidx = matchidx;
510 * Safe malloc() for internal use.
513 grep_malloc(size_t size)
517 if ((ptr = malloc(size)) == NULL)
523 * Safe calloc() for internal use.
526 grep_calloc(size_t nmemb, size_t size)
530 if ((ptr = calloc(nmemb, size)) == NULL)
536 * Safe realloc() for internal use.
539 grep_realloc(void *ptr, size_t size)
542 if ((ptr = realloc(ptr, size)) == NULL)
548 * Safe strdup() for internal use.
551 grep_strdup(const char *str)
555 if ((ret = strdup(str)) == NULL)
561 * Print an entire line as-is, there are no inline matches to consider. This is
562 * used for printing context.
564 void grep_printline(struct str *line, int sep) {
565 printline_metadata(line, sep);
566 fwrite(line->dat, line->len, 1, stdout);
571 printline_metadata(struct str *line, int sep)
578 fputs(line->file, stdout);
581 printf("%s", line->file);
588 printf("%d", line->line_no);
594 printf("%lld", (long long)line->off);
602 * Prints a matching line according to the command line options.
605 printline(struct parsec *pc, int sep)
611 /* If matchall, everything matches but don't actually print for -o */
612 if (oflag && matchall)
615 matchidx = pc->matchidx;
618 if ((oflag || color) && matchidx > 0) {
619 printline_metadata(&pc->ln, sep);
620 for (i = 0; i < matchidx; i++) {
621 match = pc->matches[i];
622 /* Don't output zero length matches */
623 if (match.rm_so == match.rm_eo)
626 fwrite(pc->ln.dat + a, match.rm_so - a, 1,
629 fprintf(stdout, "\33[%sm\33[K", color);
630 fwrite(pc->ln.dat + match.rm_so,
631 match.rm_eo - match.rm_so, 1, stdout);
633 fprintf(stdout, "\33[m\33[K");
639 if (pc->ln.len - a > 0)
640 fwrite(pc->ln.dat + a, pc->ln.len - a, 1,
645 grep_printline(&pc->ln, sep);