diff options
Diffstat (limited to 'usr.bin/grep/util.c')
| -rw-r--r-- | usr.bin/grep/util.c | 856 |
1 files changed, 856 insertions, 0 deletions
diff --git a/usr.bin/grep/util.c b/usr.bin/grep/util.c new file mode 100644 index 000000000000..5b40405852b3 --- /dev/null +++ b/usr.bin/grep/util.c @@ -0,0 +1,856 @@ +/* $NetBSD: util.c,v 1.9 2011/02/27 17:33:37 joerg Exp $ */ +/* $OpenBSD: util.c,v 1.39 2010/07/02 22:18:03 tedu Exp $ */ + +/*- + * SPDX-License-Identifier: BSD-2-Clause + * + * Copyright (c) 1999 James Howard and Dag-Erling Smørgrav + * Copyright (C) 2008-2010 Gabor Kovesdan <gabor@FreeBSD.org> + * Copyright (C) 2017 Kyle Evans <kevans@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. + */ + +#include <sys/cdefs.h> +#include <sys/stat.h> +#include <sys/types.h> + +#include <ctype.h> +#include <err.h> +#include <errno.h> +#include <fnmatch.h> +#include <fts.h> +#include <libgen.h> +#include <stdbool.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <unistd.h> +#include <wchar.h> +#include <wctype.h> + +#include "grep.h" + +static bool first_match = true; + +/* + * Match printing context + */ +struct mprintc { + long long tail; /* Number of trailing lines to record */ + int last_outed; /* Number of lines since last output */ + bool doctx; /* Printing context? */ + bool printmatch; /* Printing matches? */ + bool same_file; /* Same file as previously printed? */ +}; + +static void procmatch_match(struct mprintc *mc, struct parsec *pc); +static void procmatch_nomatch(struct mprintc *mc, struct parsec *pc); +static bool procmatches(struct mprintc *mc, struct parsec *pc, bool matched); +#ifdef WITH_INTERNAL_NOSPEC +static int litexec(const struct pat *pat, const char *string, + size_t nmatch, regmatch_t pmatch[]); +#endif +static bool procline(struct parsec *pc); +static bool printline(struct parsec *pc, int sep, size_t *last_out); +static void printline_metadata(struct str *line, int sep); + +bool +file_matching(const char *fname) +{ + char *fname_base, *fname_buf; + bool ret; + + ret = finclude ? false : true; + fname_buf = strdup(fname); + if (fname_buf == NULL) + err(2, "strdup"); + fname_base = basename(fname_buf); + + for (unsigned int i = 0; i < fpatterns; ++i) { + if (fnmatch(fpattern[i].pat, fname, 0) == 0 || + fnmatch(fpattern[i].pat, fname_base, 0) == 0) + /* + * The last pattern matched wins exclusion/inclusion + * rights, so we can't reasonably bail out early here. + */ + ret = (fpattern[i].mode != EXCL_PAT); + } + free(fname_buf); + return (ret); +} + +static inline bool +dir_matching(const char *dname) +{ + bool ret; + + ret = dinclude ? false : true; + + for (unsigned int i = 0; i < dpatterns; ++i) { + if (dname != NULL && fnmatch(dpattern[i].pat, dname, 0) == 0) + /* + * The last pattern matched wins exclusion/inclusion + * rights, so we can't reasonably bail out early here. + */ + ret = (dpattern[i].mode != EXCL_PAT); + } + return (ret); +} + +/* + * Processes a directory when a recursive search is performed with + * the -R option. Each appropriate file is passed to procfile(). + */ +bool +grep_tree(char **argv) +{ + FTS *fts; + FTSENT *p; + int fts_flags; + bool matched, ok; + const char *wd[] = { ".", NULL }; + + matched = false; + + /* This switch effectively initializes 'fts_flags' */ + switch(linkbehave) { + case LINK_EXPLICIT: + fts_flags = FTS_COMFOLLOW | FTS_PHYSICAL; + break; + case LINK_SKIP: + fts_flags = FTS_PHYSICAL; + break; + default: + fts_flags = FTS_LOGICAL | FTS_NOSTAT; + } + + fts_flags |= FTS_NOCHDIR; + + fts = fts_open((argv[0] == NULL) ? + __DECONST(char * const *, wd) : argv, fts_flags, NULL); + if (fts == NULL) + err(2, "fts_open"); + while (errno = 0, (p = fts_read(fts)) != NULL) { + switch (p->fts_info) { + case FTS_DNR: + case FTS_ERR: + case FTS_NS: + file_err = true; + if(!sflag) + warnc(p->fts_errno, "%s", p->fts_path); + break; + case FTS_D: + if (dexclude || dinclude) + if (!dir_matching(p->fts_name) || + !dir_matching(p->fts_path)) + fts_set(fts, p, FTS_SKIP); + break; + case FTS_DC: + /* Print a warning for recursive directory loop */ + warnx("warning: %s: recursive directory loop", + p->fts_path); + break; + case FTS_DP: + break; + case FTS_SL: + /* + * Skip symlinks for LINK_EXPLICIT and + * LINK_SKIP. Note that due to FTS_COMFOLLOW, + * symlinks on the command line are followed + * for LINK_EXPLICIT and not reported as + * symlinks. + */ + break; + default: + /* Check for file exclusion/inclusion */ + ok = true; + if (fexclude || finclude) + ok &= file_matching(p->fts_path); + + if (ok && procfile(p->fts_path)) + matched = true; + break; + } + } + if (errno != 0) + err(2, "fts_read"); + + fts_close(fts); + return (matched); +} + +static void +procmatch_match(struct mprintc *mc, struct parsec *pc) +{ + + if (mc->doctx) { + if (!first_match && (!mc->same_file || mc->last_outed > 0)) + printf("--\n"); + if (Bflag > 0) + printqueue(); + mc->tail = Aflag; + } + + /* Print the matching line, but only if not quiet/binary */ + if (mc->printmatch) { + size_t last_out; + bool terminated; + + last_out = 0; + terminated = printline(pc, ':', &last_out); + while (pc->matchidx >= MAX_MATCHES) { + /* Reset matchidx and try again */ + pc->matchidx = 0; + if (procline(pc) == !vflag) + terminated = printline(pc, ':', &last_out); + else + break; + } + + /* + * The above loop processes the entire line as long as we keep + * hitting the maximum match count. At this point, we know + * that there's nothing left to be printed and can terminate the + * line. + */ + if (!terminated) + printline(pc, ':', &last_out); + + first_match = false; + mc->same_file = true; + mc->last_outed = 0; + } +} + +static void +procmatch_nomatch(struct mprintc *mc, struct parsec *pc) +{ + + /* Deal with any -A context as needed */ + if (mc->tail > 0) { + grep_printline(&pc->ln, '-'); + mc->tail--; + if (Bflag > 0) + clearqueue(); + } else if (Bflag == 0 || (Bflag > 0 && enqueue(&pc->ln))) + /* + * Enqueue non-matching lines for -B context. If we're not + * actually doing -B context or if the enqueue resulted in a + * line being rotated out, then go ahead and increment + * last_outed to signify a gap between context/match. + */ + ++mc->last_outed; +} + +/* + * Process any matches in the current parsing context, return a boolean + * indicating whether we should halt any further processing or not. 'true' to + * continue processing, 'false' to halt. + */ +static bool +procmatches(struct mprintc *mc, struct parsec *pc, bool matched) +{ + + if (mflag && mcount <= 0) { + /* + * We already hit our match count, but we need to keep dumping + * lines until we've lost our tail. + */ + grep_printline(&pc->ln, '-'); + mc->tail--; + return (mc->tail != 0); + } + + /* + * XXX TODO: This should loop over pc->matches and handle things on a + * line-by-line basis, setting up a `struct str` as needed. + */ + /* Deal with any -B context or context separators */ + if (matched) { + procmatch_match(mc, pc); + + /* Count the matches if we have a match limit */ + if (mflag) { + /* XXX TODO: Decrement by number of matched lines */ + mcount -= 1; + if (mcount <= 0) + return (mc->tail != 0); + } + } else if (mc->doctx) + procmatch_nomatch(mc, pc); + + return (true); +} + +/* + * Opens a file and processes it. Each file is processed line-by-line + * passing the lines to procline(). + */ +bool +procfile(const char *fn) +{ + struct parsec pc; + struct mprintc mc; + struct file *f; + struct stat sb; + mode_t s; + int lines; + bool line_matched; + + if (strcmp(fn, "-") == 0) { + fn = label != NULL ? label : errstr[1]; + f = grep_open(NULL); + } else { + if (stat(fn, &sb) == 0) { + /* Check if we need to process the file */ + s = sb.st_mode & S_IFMT; + if (dirbehave == DIR_SKIP && s == S_IFDIR) + return (false); + if (devbehave == DEV_SKIP && (s == S_IFIFO || + s == S_IFCHR || s == S_IFBLK || s == S_IFSOCK)) + return (false); + } + f = grep_open(fn); + } + if (f == NULL) { + file_err = true; + if (!sflag) + warn("%s", fn); + return (false); + } + + pc.ln.file = grep_strdup(fn); + pc.ln.line_no = 0; + pc.ln.len = 0; + pc.ln.boff = 0; + pc.ln.off = -1; + pc.binary = f->binary; + pc.cntlines = false; + memset(&mc, 0, sizeof(mc)); + mc.printmatch = true; + if ((pc.binary && binbehave == BINFILE_BIN) || cflag || qflag || + lflag || Lflag) + mc.printmatch = false; + if (mc.printmatch && (Aflag != 0 || Bflag != 0)) + mc.doctx = true; + if (mc.printmatch && (Aflag != 0 || Bflag != 0 || mflag || nflag)) + pc.cntlines = true; + mcount = mlimit; + + for (lines = 0; lines == 0 || !(lflag || qflag); ) { + /* + * XXX TODO: We need to revisit this in a chunking world. We're + * not going to be doing per-line statistics because of the + * overhead involved. procmatches can figure that stuff out as + * needed. */ + /* Reset per-line statistics */ + pc.printed = 0; + pc.matchidx = 0; + pc.lnstart = 0; + pc.ln.boff = 0; + pc.ln.off += pc.ln.len + 1; + /* XXX TODO: Grab a chunk */ + if ((pc.ln.dat = grep_fgetln(f, &pc)) == NULL || + pc.ln.len == 0) + break; + + if (pc.ln.len > 0 && pc.ln.dat[pc.ln.len - 1] == fileeol) + --pc.ln.len; + pc.ln.line_no++; + + /* Return if we need to skip a binary file */ + if (pc.binary && binbehave == BINFILE_SKIP) { + grep_close(f); + free(pc.ln.file); + free(f); + return (0); + } + + if (mflag && mcount <= 0) { + /* + * Short-circuit, already hit match count and now we're + * just picking up any remaining pieces. + */ + if (!procmatches(&mc, &pc, false)) + break; + continue; + } + line_matched = procline(&pc) == !vflag; + if (line_matched) + ++lines; + + /* Halt processing if we hit our match limit */ + if (!procmatches(&mc, &pc, line_matched)) + break; + } + if (Bflag > 0) + clearqueue(); + grep_close(f); + + if (cflag && !qflag) { + if (!hflag) + printf("%s:", pc.ln.file); + printf("%u\n", lines); + } + if (lflag && !qflag && lines != 0) + printf("%s%c", fn, nullflag ? 0 : '\n'); + if (Lflag && !qflag && lines == 0) + printf("%s%c", fn, nullflag ? 0 : '\n'); + if (lines != 0 && !cflag && !lflag && !Lflag && + binbehave == BINFILE_BIN && f->binary && !qflag) + printf(errstr[7], fn); + + free(pc.ln.file); + free(f); + return (lines != 0); +} + +#ifdef WITH_INTERNAL_NOSPEC +/* + * Internal implementation of literal string search within a string, modeled + * after regexec(3), for use when the regex(3) implementation doesn't offer + * either REG_NOSPEC or REG_LITERAL. This does not apply in the default FreeBSD + * config, but in other scenarios such as building against libgnuregex or on + * some non-FreeBSD OSes. + */ +static int +litexec(const struct pat *pat, const char *string, size_t nmatch, + regmatch_t pmatch[]) +{ + char *(*strstr_fn)(const char *, const char *); + char *sub, *subject; + const char *search; + size_t idx, n, ofs, stringlen; + + if (cflags & REG_ICASE) + strstr_fn = strcasestr; + else + strstr_fn = strstr; + idx = 0; + ofs = pmatch[0].rm_so; + stringlen = pmatch[0].rm_eo; + if (ofs >= stringlen) + return (REG_NOMATCH); + subject = strndup(string, stringlen); + if (subject == NULL) + return (REG_ESPACE); + for (n = 0; ofs < stringlen;) { + search = (subject + ofs); + if ((unsigned long)pat->len > strlen(search)) + break; + sub = strstr_fn(search, pat->pat); + /* + * Ignoring the empty string possibility due to context: grep optimizes + * for empty patterns and will never reach this point. + */ + if (sub == NULL) + break; + ++n; + /* Fill in pmatch if necessary */ + if (nmatch > 0) { + pmatch[idx].rm_so = ofs + (sub - search); + pmatch[idx].rm_eo = pmatch[idx].rm_so + pat->len; + if (++idx == nmatch) + break; + ofs = pmatch[idx].rm_so + 1; + } else + /* We only needed to know if we match or not */ + break; + } + free(subject); + if (n > 0 && nmatch > 0) + for (n = idx; n < nmatch; ++n) + pmatch[n].rm_so = pmatch[n].rm_eo = -1; + + return (n > 0 ? 0 : REG_NOMATCH); +} +#endif /* WITH_INTERNAL_NOSPEC */ + +#define iswword(x) (iswalnum((x)) || (x) == L'_') + +/* + * Processes a line comparing it with the specified patterns. Each pattern + * is looped to be compared along with the full string, saving each and every + * match, which is necessary to colorize the output and to count the + * matches. The matching lines are passed to printline() to display the + * appropriate output. + */ +static bool +procline(struct parsec *pc) +{ + regmatch_t pmatch, lastmatch, chkmatch; + wchar_t wbegin, wend; + size_t st, nst; + unsigned int i; + int r = 0, leflags = eflags; + size_t startm = 0, matchidx; + unsigned int retry; + bool lastmatched, matched; + + matchidx = pc->matchidx; + + /* Null pattern shortcuts. */ + if (matchall) { + if (xflag && pc->ln.len == 0) { + /* Matches empty lines (-x). */ + return (true); + } else if (!wflag && !xflag) { + /* Matches every line (no -w or -x). */ + return (true); + } + + /* + * If we only have the NULL pattern, whether we match or not + * depends on if we got here with -w or -x. If either is set, + * the answer is no. If we have other patterns, we'll defer + * to them. + */ + if (patterns == 0) { + return (!(wflag || xflag)); + } + } else if (patterns == 0) { + /* Pattern file with no patterns. */ + return (false); + } + + matched = false; + st = pc->lnstart; + nst = 0; + /* Initialize to avoid a false positive warning from GCC. */ + lastmatch.rm_so = lastmatch.rm_eo = 0; + + /* Loop to process the whole line */ + while (st <= pc->ln.len) { + lastmatched = false; + startm = matchidx; + retry = 0; + if (st > 0 && pc->ln.dat[st - 1] != fileeol) + leflags |= REG_NOTBOL; + /* Loop to compare with all the patterns */ + for (i = 0; i < patterns; i++) { + pmatch.rm_so = st; + pmatch.rm_eo = pc->ln.len; +#ifdef WITH_INTERNAL_NOSPEC + if (grepbehave == GREP_FIXED) + r = litexec(&pattern[i], pc->ln.dat, 1, &pmatch); + else +#endif + r = regexec(&r_pattern[i], pc->ln.dat, 1, &pmatch, + leflags); + if (r != 0) + continue; + /* Check for full match */ + if (xflag && (pmatch.rm_so != 0 || + (size_t)pmatch.rm_eo != pc->ln.len)) + continue; + /* Check for whole word match */ + if (wflag) { + wbegin = wend = L' '; + if (pmatch.rm_so != 0 && + sscanf(&pc->ln.dat[pmatch.rm_so - 1], + "%lc", &wbegin) != 1) + r = REG_NOMATCH; + else if ((size_t)pmatch.rm_eo != + pc->ln.len && + sscanf(&pc->ln.dat[pmatch.rm_eo], + "%lc", &wend) != 1) + r = REG_NOMATCH; + else if (iswword(wbegin) || + iswword(wend)) + r = REG_NOMATCH; + /* + * If we're doing whole word matching and we + * matched once, then we should try the pattern + * again after advancing just past the start of + * the earliest match. This allows the pattern + * to match later on in the line and possibly + * still match a whole word. + */ + if (r == REG_NOMATCH && + (retry == pc->lnstart || + (unsigned int)pmatch.rm_so + 1 < retry)) + retry = pmatch.rm_so + 1; + if (r == REG_NOMATCH) + continue; + } + lastmatched = true; + lastmatch = pmatch; + + if (matchidx == 0) + matched = true; + + /* + * Replace previous match if the new one is earlier + * and/or longer. This will lead to some amount of + * extra work if -o/--color are specified, but it's + * worth it from a correctness point of view. + */ + if (matchidx > startm) { + chkmatch = pc->matches[matchidx - 1]; + if (pmatch.rm_so < chkmatch.rm_so || + (pmatch.rm_so == chkmatch.rm_so && + (pmatch.rm_eo - pmatch.rm_so) > + (chkmatch.rm_eo - chkmatch.rm_so))) { + pc->matches[matchidx - 1] = pmatch; + nst = pmatch.rm_eo; + } + } else { + /* Advance as normal if not */ + pc->matches[matchidx++] = pmatch; + nst = pmatch.rm_eo; + } + /* avoid excessive matching - skip further patterns */ + if ((color == NULL && !oflag) || qflag || lflag || + matchidx >= MAX_MATCHES) { + pc->lnstart = nst; + lastmatched = false; + break; + } + } + + /* + * Advance to just past the start of the earliest match, try + * again just in case we still have a chance to match later in + * the string. + */ + if (!lastmatched && retry > pc->lnstart) { + st = retry; + continue; + } + + /* XXX TODO: We will need to keep going, since we're chunky */ + /* One pass if we are not recording matches */ + if (!wflag && ((color == NULL && !oflag) || qflag || lflag || Lflag)) + break; + + /* If we didn't have any matches or REG_NOSUB set */ + if (!lastmatched || (cflags & REG_NOSUB)) + nst = pc->ln.len; + + if (!lastmatched) + /* No matches */ + break; + else if (st == nst && lastmatch.rm_so == lastmatch.rm_eo) + /* Zero-length match -- advance one more so we don't get stuck */ + nst++; + + /* Advance st based on previous matches */ + st = nst; + pc->lnstart = st; + } + + /* Reflect the new matchidx in the context */ + pc->matchidx = matchidx; + return matched; +} + +/* + * Safe malloc() for internal use. + */ +void * +grep_malloc(size_t size) +{ + void *ptr; + + if (size == 0) + return (NULL); + if ((ptr = malloc(size)) == NULL) + err(2, "malloc"); + return (ptr); +} + +/* + * Safe calloc() for internal use. + */ +void * +grep_calloc(size_t nmemb, size_t size) +{ + void *ptr; + + if (nmemb == 0 || size == 0) + return (NULL); + if ((ptr = calloc(nmemb, size)) == NULL) + err(2, "calloc"); + return (ptr); +} + +/* + * Safe realloc() for internal use. + */ +void * +grep_realloc(void *ptr, size_t size) +{ + + if ((ptr = realloc(ptr, size)) == NULL) + err(2, "realloc"); + return (ptr); +} + +/* + * Safe strdup() for internal use. + */ +char * +grep_strdup(const char *str) +{ + char *ret; + + if ((ret = strdup(str)) == NULL) + err(2, "strdup"); + return (ret); +} + +/* + * Print an entire line as-is, there are no inline matches to consider. This is + * used for printing context. + */ +void grep_printline(struct str *line, int sep) { + printline_metadata(line, sep); + fwrite(line->dat, line->len, 1, stdout); + putchar(fileeol); + + fflush(stdout); +} + +static void +printline_metadata(struct str *line, int sep) +{ + bool printsep; + + printsep = false; + if (!hflag) { + if (!nullflag) { + fputs(line->file, stdout); + printsep = true; + } else { + printf("%s", line->file); + putchar(0); + } + } + if (nflag) { + if (printsep) + putchar(sep); + printf("%d", line->line_no); + printsep = true; + } + if (bflag) { + if (printsep) + putchar(sep); + printf("%lld", (long long)(line->off + line->boff)); + printsep = true; + } + if (printsep) + putchar(sep); +} + +/* + * Prints a matching line according to the command line options. We need + * *last_out to be populated on entry in case this is just a continuation of + * matches within the same line. + * + * Returns true if the line was terminated, false if it was not. + */ +static bool +printline(struct parsec *pc, int sep, size_t *last_out) +{ + size_t a = *last_out; + size_t i, matchidx; + regmatch_t match; + bool terminated; + + /* + * Nearly all paths below will terminate the line by default, but it is + * avoided in some circumstances in case we don't have the full context + * available here. + */ + terminated = true; + + /* If matchall, everything matches but don't actually print for -o */ + if (oflag && matchall) + return (terminated); + + matchidx = pc->matchidx; + + /* --color and -o */ + if ((oflag || color) && (pc->printed > 0 || matchidx > 0)) { + /* Only print metadata once per line if --color */ + if (!oflag && pc->printed == 0) { + printline_metadata(&pc->ln, sep); + } + for (i = 0; i < matchidx; i++) { + match = pc->matches[i]; + /* Don't output zero length matches */ + if (match.rm_so == match.rm_eo) + continue; + /* + * Metadata is printed on a per-line basis, so every + * match gets file metadata with the -o flag. + */ + if (oflag) { + pc->ln.boff = match.rm_so; + printline_metadata(&pc->ln, sep); + } else { + fwrite(pc->ln.dat + a, match.rm_so - a, 1, + stdout); + } + if (color) + fprintf(stdout, "\33[%sm\33[K", color); + fwrite(pc->ln.dat + match.rm_so, + match.rm_eo - match.rm_so, 1, stdout); + if (color) + fprintf(stdout, "\33[m\33[K"); + a = match.rm_eo; + if (oflag) + putchar('\n'); + } + + /* + * Don't terminate if we reached the match limit; we may have + * other matches on this line to process. + */ + *last_out = a; + if (!oflag && matchidx != MAX_MATCHES) { + if (pc->ln.len - a > 0) { + fwrite(pc->ln.dat + a, pc->ln.len - a, 1, + stdout); + *last_out = pc->ln.len; + } + putchar('\n'); + fflush(stdout); + } else if (!oflag) { + /* + * -o is terminated on every match output, so this + * branch is only designed to capture MAX_MATCHES in a + * line which may be a signal to us for a lack of + * context. The caller will know more and call us again + * to terminate if it needs to. + */ + terminated = false; + } else { + fflush(stdout); + } + } else + grep_printline(&pc->ln, sep); + pc->printed++; + return (terminated); +} |