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 lnstart; /* Position in line */
67 size_t matchidx; /* Latest match index */
68 int printed; /* Metadata printed? */
69 bool binary; /* Binary file? */
73 static int procline(struct parsec *pc);
74 static void printline(struct parsec *pc, int sep);
75 static void printline_metadata(struct str *line, int sep);
78 file_matching(const char *fname)
80 char *fname_base, *fname_buf;
83 ret = finclude ? false : true;
84 fname_buf = strdup(fname);
85 if (fname_buf == NULL)
87 fname_base = basename(fname_buf);
89 for (unsigned int i = 0; i < fpatterns; ++i) {
90 if (fnmatch(fpattern[i].pat, fname, 0) == 0 ||
91 fnmatch(fpattern[i].pat, fname_base, 0) == 0) {
92 if (fpattern[i].mode == EXCL_PAT) {
104 dir_matching(const char *dname)
108 ret = dinclude ? false : true;
110 for (unsigned int i = 0; i < dpatterns; ++i) {
112 fnmatch(dpattern[i].pat, dname, 0) == 0) {
113 if (dpattern[i].mode == EXCL_PAT)
123 * Processes a directory when a recursive search is performed with
124 * the -R option. Each appropriate file is passed to procfile().
127 grep_tree(char **argv)
133 const char *wd[] = { ".", NULL };
139 fts_flags = FTS_COMFOLLOW;
142 fts_flags = FTS_PHYSICAL;
145 fts_flags = FTS_LOGICAL;
149 fts_flags |= FTS_NOSTAT | FTS_NOCHDIR;
151 fts = fts_open((argv[0] == NULL) ?
152 __DECONST(char * const *, wd) : argv, fts_flags, NULL);
155 while ((p = fts_read(fts)) != NULL) {
156 switch (p->fts_info) {
162 warnx("%s: %s", p->fts_path, strerror(p->fts_errno));
167 if (dexclude || dinclude)
168 if (!dir_matching(p->fts_name) ||
169 !dir_matching(p->fts_path))
170 fts_set(fts, p, FTS_SKIP);
173 /* Print a warning for recursive directory loop */
174 warnx("warning: %s: recursive directory loop",
178 /* Check for file exclusion/inclusion */
180 if (fexclude || finclude)
181 ok &= file_matching(p->fts_path);
184 c += procfile(p->fts_path);
194 * Opens a file and processes it. Each file is processed line-by-line
195 * passing the lines to procline().
198 procfile(const char *fn)
206 int c, last_outed, t;
207 bool doctx, printmatch, same_file;
209 if (strcmp(fn, "-") == 0) {
210 fn = label != NULL ? label : getstr(1);
213 if (!stat(fn, &sb)) {
214 /* Check if we need to process the file */
215 s = sb.st_mode & S_IFMT;
216 if (s == S_IFDIR && dirbehave == DIR_SKIP)
218 if ((s == S_IFIFO || s == S_IFCHR || s == S_IFBLK
219 || s == S_IFSOCK) && devbehave == DEV_SKIP)
233 pc.ln.file = grep_malloc(strlen(fn) + 1);
234 strcpy(pc.ln.file, fn);
239 pc.binary = f->binary;
246 if ((pc.binary && binbehave == BINFILE_BIN) || cflag || qflag ||
249 if (printmatch && (Aflag != 0 || Bflag != 0))
253 for (c = 0; c == 0 || !(lflag || qflag); ) {
254 /* Reset per-line statistics */
259 pc.ln.off += pc.ln.len + 1;
260 if ((pc.ln.dat = grep_fgetln(f, &pc.ln.len)) == NULL ||
262 if (pc.ln.line_no == 0 && matchall)
264 * An empty file with an empty pattern and the
265 * -w flag does not match
267 exit(matchall && wflag ? 1 : 0);
272 if (pc.ln.len > 0 && pc.ln.dat[pc.ln.len - 1] == fileeol)
276 /* Return if we need to skip a binary file */
277 if (pc.binary && binbehave == BINFILE_SKIP) {
284 if ((t = procline(&pc)) == 0)
287 /* Deal with any -B context or context separators */
288 if (t == 0 && doctx) {
289 if (!first_match && (!same_file || last_outed > 0))
295 /* Print the matching line, but only if not quiet/binary */
296 if (t == 0 && printmatch) {
298 while (pc.matchidx >= MAX_LINE_MATCHES) {
299 /* Reset matchidx and try again */
301 if (procline(&pc) == 0)
310 if (t != 0 && doctx) {
311 /* Deal with any -A context */
313 grep_printline(&pc.ln, '-');
319 * Enqueue non-matching lines for -B context.
320 * If we're not actually doing -B context or if
321 * the enqueue resulted in a line being rotated
322 * out, then go ahead and increment last_outed
323 * to signify a gap between context/match.
325 if (Bflag == 0 || (Bflag > 0 && enqueue(ln)))
330 /* Count the matches if we have a match limit */
331 if (t == 0 && mflag) {
333 if (mflag && mcount <= 0)
344 printf("%s:", pc.ln.file);
347 if (lflag && !qflag && c != 0)
348 printf("%s%c", fn, nullflag ? 0 : '\n');
349 if (Lflag && !qflag && c == 0)
350 printf("%s%c", fn, nullflag ? 0 : '\n');
351 if (c && !cflag && !lflag && !Lflag &&
352 binbehave == BINFILE_BIN && f->binary && !qflag)
353 printf(getstr(8), fn);
360 #define iswword(x) (iswalnum((x)) || (x) == L'_')
363 * Processes a line comparing it with the specified patterns. Each pattern
364 * is looped to be compared along with the full string, saving each and every
365 * match, which is necessary to colorize the output and to count the
366 * matches. The matching lines are passed to printline() to display the
367 * appropriate output.
370 procline(struct parsec *pc)
372 regmatch_t pmatch, lastmatch, chkmatch;
373 wchar_t wbegin, wend;
376 int c = 0, r = 0, lastmatches = 0, leflags = eflags;
377 size_t startm = 0, matchidx;
380 matchidx = pc->matchidx;
382 /* Special case: empty pattern with -w flag, check first character */
383 if (matchall && wflag) {
387 if (sscanf(&pc->ln.dat[0], "%lc", &wend) != 1 || iswword(wend))
396 /* Initialize to avoid a false positive warning from GCC. */
397 lastmatch.rm_so = lastmatch.rm_eo = 0;
399 /* Loop to process the whole line */
400 while (st <= pc->ln.len) {
405 leflags |= REG_NOTBOL;
406 /* Loop to compare with all the patterns */
407 for (i = 0; i < patterns; i++) {
409 pmatch.rm_eo = pc->ln.len;
410 #ifndef WITHOUT_FASTMATCH
411 if (fg_pattern[i].pattern)
412 r = fastexec(&fg_pattern[i],
413 pc->ln.dat, 1, &pmatch, leflags);
416 r = regexec(&r_pattern[i], pc->ln.dat, 1,
420 /* Check for full match */
421 if (xflag && (pmatch.rm_so != 0 ||
422 (size_t)pmatch.rm_eo != pc->ln.len))
424 /* Check for whole word match */
425 #ifndef WITHOUT_FASTMATCH
426 if (wflag || fg_pattern[i].word) {
430 wbegin = wend = L' ';
431 if (pmatch.rm_so != 0 &&
432 sscanf(&pc->ln.dat[pmatch.rm_so - 1],
433 "%lc", &wbegin) != 1)
435 else if ((size_t)pmatch.rm_eo !=
437 sscanf(&pc->ln.dat[pmatch.rm_eo],
440 else if (iswword(wbegin) ||
444 * If we're doing whole word matching and we
445 * matched once, then we should try the pattern
446 * again after advancing just past the start of
447 * the earliest match. This allows the pattern
448 * to match later on in the line and possibly
449 * still match a whole word.
451 if (r == REG_NOMATCH &&
452 (retry == pc->lnstart ||
453 pmatch.rm_so + 1 < retry))
454 retry = pmatch.rm_so + 1;
455 if (r == REG_NOMATCH)
465 * Replace previous match if the new one is earlier
466 * and/or longer. This will lead to some amount of
467 * extra work if -o/--color are specified, but it's
468 * worth it from a correctness point of view.
470 if (matchidx > startm) {
471 chkmatch = pc->matches[matchidx - 1];
472 if (pmatch.rm_so < chkmatch.rm_so ||
473 (pmatch.rm_so == chkmatch.rm_so &&
474 (pmatch.rm_eo - pmatch.rm_so) >
475 (chkmatch.rm_eo - chkmatch.rm_so))) {
476 pc->matches[matchidx - 1] = pmatch;
480 /* Advance as normal if not */
481 pc->matches[matchidx++] = pmatch;
484 /* avoid excessive matching - skip further patterns */
485 if ((color == NULL && !oflag) || qflag || lflag ||
486 matchidx >= MAX_LINE_MATCHES) {
494 * Advance to just past the start of the earliest match, try
495 * again just in case we still have a chance to match later in
498 if (lastmatches == 0 && retry > pc->lnstart) {
503 /* One pass if we are not recording matches */
504 if (!wflag && ((color == NULL && !oflag) || qflag || lflag || Lflag))
507 /* If we didn't have any matches or REG_NOSUB set */
508 if (lastmatches == 0 || (cflags & REG_NOSUB))
511 if (lastmatches == 0)
514 else if (st == nst && lastmatch.rm_so == lastmatch.rm_eo)
515 /* Zero-length match -- advance one more so we don't get stuck */
518 /* Advance st based on previous matches */
523 /* Reflect the new matchidx in the context */
524 pc->matchidx = matchidx;
531 * Safe malloc() for internal use.
534 grep_malloc(size_t size)
538 if ((ptr = malloc(size)) == NULL)
544 * Safe calloc() for internal use.
547 grep_calloc(size_t nmemb, size_t size)
551 if ((ptr = calloc(nmemb, size)) == NULL)
557 * Safe realloc() for internal use.
560 grep_realloc(void *ptr, size_t size)
563 if ((ptr = realloc(ptr, size)) == NULL)
569 * Safe strdup() for internal use.
572 grep_strdup(const char *str)
576 if ((ret = strdup(str)) == NULL)
582 * Print an entire line as-is, there are no inline matches to consider. This is
583 * used for printing context.
585 void grep_printline(struct str *line, int sep) {
586 printline_metadata(line, sep);
587 fwrite(line->dat, line->len, 1, stdout);
592 printline_metadata(struct str *line, int sep)
599 fputs(line->file, stdout);
602 printf("%s", line->file);
609 printf("%d", line->line_no);
615 printf("%lld", (long long)(line->off + line->boff));
623 * Prints a matching line according to the command line options.
626 printline(struct parsec *pc, int sep)
632 /* If matchall, everything matches but don't actually print for -o */
633 if (oflag && matchall)
636 matchidx = pc->matchidx;
639 if ((oflag || color) && matchidx > 0) {
640 /* Only print metadata once per line if --color */
641 if (!oflag && pc->printed == 0)
642 printline_metadata(&pc->ln, sep);
643 for (i = 0; i < matchidx; i++) {
644 match = pc->matches[i];
645 /* Don't output zero length matches */
646 if (match.rm_so == match.rm_eo)
649 * Metadata is printed on a per-line basis, so every
650 * match gets file metadata with the -o flag.
653 pc->ln.boff = match.rm_so;
654 printline_metadata(&pc->ln, sep);
656 fwrite(pc->ln.dat + a, match.rm_so - a, 1,
659 fprintf(stdout, "\33[%sm\33[K", color);
660 fwrite(pc->ln.dat + match.rm_so,
661 match.rm_eo - match.rm_so, 1, stdout);
663 fprintf(stdout, "\33[m\33[K");
669 if (pc->ln.len - a > 0)
670 fwrite(pc->ln.dat + a, pc->ln.len - a, 1,
675 grep_printline(&pc->ln, sep);