Annotation of src/usr.bin/grep/grep.c, Revision 1.40
1.40 ! tedu 1: /* $OpenBSD: grep.c,v 1.39 2007/09/02 15:19:32 deraadt Exp $ */
1.6 deraadt 2:
1.1 deraadt 3: /*-
1.3 deraadt 4: * Copyright (c) 1999 James Howard and Dag-Erling Coïdan Smørgrav
5: * All rights reserved.
1.1 deraadt 6: *
7: * Redistribution and use in source and binary forms, with or without
8: * modification, are permitted provided that the following conditions
9: * are met:
10: * 1. Redistributions of source code must retain the above copyright
11: * notice, this list of conditions and the following disclaimer.
12: * 2. Redistributions in binary form must reproduce the above copyright
13: * notice, this list of conditions and the following disclaimer in the
14: * documentation and/or other materials provided with the distribution.
15: *
16: * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
17: * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
18: * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
19: * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
20: * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
21: * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
22: * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
23: * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
24: * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
25: * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
26: * SUCH DAMAGE.
27: */
28:
1.3 deraadt 29: #include <sys/types.h>
1.9 tedu 30: #include <sys/limits.h>
1.3 deraadt 31: #include <sys/stat.h>
1.33 jaredy 32: #include <sys/queue.h>
1.1 deraadt 33:
1.18 millert 34: #include <ctype.h>
1.3 deraadt 35: #include <err.h>
36: #include <errno.h>
37: #include <getopt.h>
38: #include <regex.h>
1.1 deraadt 39: #include <stdio.h>
40: #include <stdlib.h>
1.3 deraadt 41: #include <string.h>
1.1 deraadt 42: #include <unistd.h>
43:
1.3 deraadt 44: #include "grep.h"
45:
46: /* Flags passed to regcomp() and regexec() */
47: int cflags;
48: int eflags = REG_STARTEND;
49:
50: int matchall; /* shortcut */
51: int patterns, pattern_sz;
52: char **pattern;
53: regex_t *r_pattern;
1.13 tedu 54: fastgrep_t *fg_pattern;
1.3 deraadt 55:
56: /* For regex errors */
57: char re_error[RE_ERROR_BUF + 1];
58:
59: /* Command-line flags */
60: int Aflag; /* -A x: print x lines trailing each match */
61: int Bflag; /* -B x: print x lines leading each match */
62: int Eflag; /* -E: interpret pattern as extended regexp */
63: int Fflag; /* -F: interpret pattern as list of fixed strings */
64: int Gflag; /* -G: interpret pattern as basic regexp */
65: int Lflag; /* -L: only show names of files with no matches */
66: int Rflag; /* -R: recursively search directory trees */
1.5 deraadt 67: #ifndef NOZ
1.3 deraadt 68: int Zflag; /* -Z: decompress input before processing */
1.5 deraadt 69: #endif
1.3 deraadt 70: int bflag; /* -b: show block numbers for each match */
71: int cflag; /* -c: only show a count of matching lines */
72: int hflag; /* -h: don't print filename headers */
73: int iflag; /* -i: ignore case */
74: int lflag; /* -l: only show names of files with matches */
75: int nflag; /* -n: show line numbers in front of matching lines */
76: int qflag; /* -q: quiet mode (don't output anything) */
77: int sflag; /* -s: silent mode (ignore errors) */
78: int vflag; /* -v: only show non-matching lines */
79: int wflag; /* -w: pattern must start and end on word boundaries */
80: int xflag; /* -x: pattern must match entire line */
1.27 otto 81: int lbflag; /* --line-buffered */
1.3 deraadt 82:
1.9 tedu 83: int binbehave = BIN_FILE_BIN;
84:
85: enum {
86: BIN_OPT = CHAR_MAX + 1,
87: HELP_OPT,
1.27 otto 88: MMAP_OPT,
89: LINEBUF_OPT
1.9 tedu 90: };
91:
1.3 deraadt 92: /* Housekeeping */
1.10 tedu 93: int first; /* flag whether or not this is our first match */
1.3 deraadt 94: int tail; /* lines left to print */
95:
1.33 jaredy 96: struct patfile {
97: const char *pf_file;
98: SLIST_ENTRY(patfile) pf_next;
99: };
100: SLIST_HEAD(, patfile) patfilelh;
101:
1.8 tedu 102: extern char *__progname;
1.3 deraadt 103:
104: static void
105: usage(void)
106: {
1.5 deraadt 107: fprintf(stderr,
108: #ifdef NOZ
1.40 ! tedu 109: "usage: %s [-abcEFGhIiLlnqRsUVvwx] [-A num] [-B num] [-C[num]]\n"
1.5 deraadt 110: #else
1.40 ! tedu 111: "usage: %s [-abcEFGhIiLlnqRsUVvwxZ] [-A num] [-B num] [-C[num]]\n"
1.5 deraadt 112: #endif
1.30 jmc 113: "\t[-e pattern] [-f file] [--binary-files=value] [--context[=num]]\n"
114: "\t[--line-buffered] [pattern] [file ...]\n", __progname);
1.3 deraadt 115: exit(2);
116: }
117:
1.5 deraadt 118: #ifdef NOZ
1.40 ! tedu 119: static char *optstr = "0123456789A:B:CEFGILRUVabce:f:hilnqrsuvwxy";
1.5 deraadt 120: #else
1.40 ! tedu 121: static char *optstr = "0123456789A:B:CEFGILRUVZabce:f:hilnqrsuvwxy";
1.5 deraadt 122: #endif
1.3 deraadt 123:
1.5 deraadt 124: struct option long_options[] =
1.3 deraadt 125: {
1.19 deraadt 126: {"binary-files", required_argument, NULL, BIN_OPT},
127: {"help", no_argument, NULL, HELP_OPT},
128: {"mmap", no_argument, NULL, MMAP_OPT},
1.27 otto 129: {"line-buffered", no_argument, NULL, LINEBUF_OPT},
1.19 deraadt 130: {"after-context", required_argument, NULL, 'A'},
131: {"before-context", required_argument, NULL, 'B'},
132: {"context", optional_argument, NULL, 'C'},
133: {"devices", required_argument, NULL, 'D'},
134: {"extended-regexp", no_argument, NULL, 'E'},
135: {"fixed-strings", no_argument, NULL, 'F'},
136: {"basic-regexp", no_argument, NULL, 'G'},
137: {"binary", no_argument, NULL, 'U'},
138: {"version", no_argument, NULL, 'V'},
139: {"text", no_argument, NULL, 'a'},
140: {"byte-offset", no_argument, NULL, 'b'},
141: {"count", no_argument, NULL, 'c'},
142: {"regexp", required_argument, NULL, 'e'},
143: {"file", required_argument, NULL, 'f'},
144: {"no-filename", no_argument, NULL, 'h'},
145: {"ignore-case", no_argument, NULL, 'i'},
146: {"files-without-match", no_argument, NULL, 'L'},
147: {"files-with-matches", no_argument, NULL, 'l'},
148: {"line-number", no_argument, NULL, 'n'},
149: {"quiet", no_argument, NULL, 'q'},
150: {"silent", no_argument, NULL, 'q'},
151: {"recursive", no_argument, NULL, 'r'},
152: {"no-messages", no_argument, NULL, 's'},
153: {"revert-match", no_argument, NULL, 'v'},
154: {"word-regexp", no_argument, NULL, 'w'},
155: {"line-regexp", no_argument, NULL, 'x'},
156: {"unix-byte-offsets", no_argument, NULL, 'u'},
1.5 deraadt 157: #ifndef NOZ
1.19 deraadt 158: {"decompress", no_argument, NULL, 'Z'},
1.5 deraadt 159: #endif
1.19 deraadt 160: {NULL, no_argument, NULL, 0}
1.3 deraadt 161: };
162:
1.1 deraadt 163:
1.3 deraadt 164: static void
165: add_pattern(char *pat, size_t len)
166: {
1.36 jaredy 167: if (!xflag && (len == 0 || matchall)) {
1.3 deraadt 168: matchall = 1;
169: return;
170: }
171: if (patterns == pattern_sz) {
172: pattern_sz *= 2;
1.22 millert 173: pattern = grep_realloc(pattern, ++pattern_sz * sizeof(*pattern));
1.3 deraadt 174: }
1.37 ray 175: if (len > 0 && pat[len - 1] == '\n')
1.3 deraadt 176: --len;
1.15 tedu 177: /* pat may not be NUL-terminated */
1.28 millert 178: if (wflag && !Fflag) {
1.32 otto 179: int bol = 0, eol = 0, extra;
1.26 millert 180: if (pat[0] == '^')
181: bol = 1;
1.36 jaredy 182: if (len > 0 && pat[len - 1] == '$')
1.26 millert 183: eol = 1;
1.32 otto 184: extra = Eflag ? 2 : 4;
185: pattern[patterns] = grep_malloc(len + 15 + extra);
186: snprintf(pattern[patterns], len + 15 + extra,
187: "%s[[:<:]]%s%.*s%s[[:>:]]%s",
188: bol ? "^" : "",
189: Eflag ? "(" : "\\(",
190: (int)len - bol - eol, pat + bol,
191: Eflag ? ")" : "\\)",
1.26 millert 192: eol ? "$" : "");
1.32 otto 193: len += 14 + extra;
1.25 millert 194: } else {
195: pattern[patterns] = grep_malloc(len + 1);
196: memcpy(pattern[patterns], pat, len);
197: pattern[patterns][len] = '\0';
198: }
1.3 deraadt 199: ++patterns;
200: }
1.1 deraadt 201:
1.3 deraadt 202: static void
1.35 otto 203: add_patterns(char *pats)
204: {
205: char *nl;
206:
207: while ((nl = strchr(pats, '\n')) != NULL) {
208: add_pattern(pats, nl - pats);
209: pats = nl + 1;
210: }
211: add_pattern(pats, strlen(pats));
212: }
213:
214: static void
1.33 jaredy 215: read_patterns(const char *fn)
1.3 deraadt 216: {
217: FILE *f;
218: char *line;
219: size_t len;
220:
221: if ((f = fopen(fn, "r")) == NULL)
1.20 millert 222: err(2, "%s", fn);
1.38 kili 223: while ((line = fgetln(f, &len)) != NULL)
224: add_pattern(line, *line == '\n' ? 0 : len);
1.3 deraadt 225: if (ferror(f))
1.20 millert 226: err(2, "%s", fn);
1.3 deraadt 227: fclose(f);
228: }
1.1 deraadt 229:
230: int
1.3 deraadt 231: main(int argc, char *argv[])
1.1 deraadt 232: {
1.38 kili 233: int c, lastc, prevoptind, newarg, i, needpattern;
1.33 jaredy 234: struct patfile *patfile, *pf_next;
1.21 millert 235: long l;
236: char *ep;
1.3 deraadt 237:
1.33 jaredy 238: SLIST_INIT(&patfilelh);
1.16 tedu 239: switch (__progname[0]) {
240: case 'e':
241: Eflag++;
242: break;
243: case 'f':
244: Fflag++;
245: break;
246: case 'g':
247: Gflag++;
248: break;
249: #ifndef NOZ
250: case 'z':
251: Zflag++;
252: switch(__progname[1]) {
253: case 'e':
254: Eflag++;
255: break;
256: case 'f':
257: Fflag++;
258: break;
259: case 'g':
260: Gflag++;
261: break;
262: }
263: break;
264: #endif
265: }
266:
1.18 millert 267: lastc = '\0';
1.23 millert 268: newarg = 1;
269: prevoptind = 1;
1.38 kili 270: needpattern = 1;
1.5 deraadt 271: while ((c = getopt_long(argc, argv, optstr,
1.17 millert 272: long_options, NULL)) != -1) {
1.3 deraadt 273: switch (c) {
274: case '0': case '1': case '2': case '3': case '4':
275: case '5': case '6': case '7': case '8': case '9':
1.23 millert 276: if (newarg || !isdigit(lastc))
277: Aflag = 0;
278: else if (Aflag > INT_MAX / 10)
279: errx(2, "context out of range");
280: Aflag = Bflag = (Aflag * 10) + (c - '0');
1.3 deraadt 281: break;
282: case 'A':
283: case 'B':
1.21 millert 284: l = strtol(optarg, &ep, 10);
285: if (ep == optarg || *ep != '\0' ||
286: l <= 0 || l >= INT_MAX)
287: errx(2, "context out of range");
288: if (c == 'A')
289: Aflag = (int)l;
290: else
291: Bflag = (int)l;
1.3 deraadt 292: break;
293: case 'C':
1.5 deraadt 294: if (optarg == NULL)
1.3 deraadt 295: Aflag = Bflag = 2;
1.21 millert 296: else {
297: l = strtol(optarg, &ep, 10);
298: if (ep == optarg || *ep != '\0' ||
299: l <= 0 || l >= INT_MAX)
300: errx(2, "context out of range");
301: Aflag = Bflag = (int)l;
302: }
1.1 deraadt 303: break;
304: case 'E':
1.16 tedu 305: Fflag = Gflag = 0;
1.3 deraadt 306: Eflag++;
1.1 deraadt 307: break;
308: case 'F':
1.16 tedu 309: Eflag = Gflag = 0;
1.3 deraadt 310: Fflag++;
311: break;
312: case 'G':
1.16 tedu 313: Eflag = Fflag = 0;
1.3 deraadt 314: Gflag++;
1.1 deraadt 315: break;
1.9 tedu 316: case 'I':
317: binbehave = BIN_FILE_SKIP;
318: break;
1.1 deraadt 319: case 'L':
1.3 deraadt 320: lflag = 0;
321: Lflag = qflag = 1;
1.1 deraadt 322: break;
323: case 'R':
1.3 deraadt 324: case 'r':
325: Rflag++;
326: break;
327: case 'U':
1.9 tedu 328: binbehave = BIN_FILE_BIN;
1.3 deraadt 329: break;
330: case 'V':
331: fprintf(stderr, "grep version %u.%u\n", VER_MAJ, VER_MIN);
1.9 tedu 332: exit(0);
1.1 deraadt 333: break;
1.5 deraadt 334: #ifndef NOZ
1.3 deraadt 335: case 'Z':
336: Zflag++;
1.1 deraadt 337: break;
1.5 deraadt 338: #endif
1.1 deraadt 339: case 'a':
1.9 tedu 340: binbehave = BIN_FILE_TEXT;
1.1 deraadt 341: break;
342: case 'b':
1.3 deraadt 343: bflag = 1;
1.1 deraadt 344: break;
345: case 'c':
1.3 deraadt 346: cflag = 1;
1.1 deraadt 347: break;
348: case 'e':
1.35 otto 349: add_patterns(optarg);
1.38 kili 350: needpattern = 0;
1.1 deraadt 351: break;
352: case 'f':
1.33 jaredy 353: patfile = grep_malloc(sizeof(*patfile));
354: patfile->pf_file = optarg;
355: SLIST_INSERT_HEAD(&patfilelh, patfile, pf_next);
1.38 kili 356: needpattern = 0;
1.1 deraadt 357: break;
358: case 'h':
1.3 deraadt 359: hflag = 1;
1.1 deraadt 360: break;
361: case 'i':
1.3 deraadt 362: case 'y':
1.13 tedu 363: iflag = 1;
1.1 deraadt 364: cflags |= REG_ICASE;
365: break;
366: case 'l':
1.3 deraadt 367: Lflag = 0;
368: lflag = qflag = 1;
1.1 deraadt 369: break;
370: case 'n':
1.3 deraadt 371: nflag = 1;
372: break;
1.1 deraadt 373: case 'q':
1.3 deraadt 374: qflag = 1;
1.1 deraadt 375: break;
376: case 's':
1.3 deraadt 377: sflag = 1;
1.1 deraadt 378: break;
379: case 'v':
1.3 deraadt 380: vflag = 1;
1.1 deraadt 381: break;
382: case 'w':
1.3 deraadt 383: wflag = 1;
1.1 deraadt 384: break;
385: case 'x':
1.3 deraadt 386: xflag = 1;
1.1 deraadt 387: break;
1.9 tedu 388: case BIN_OPT:
389: if (strcmp("binary", optarg) == 0)
390: binbehave = BIN_FILE_BIN;
391: else if (strcmp("without-match", optarg) == 0)
392: binbehave = BIN_FILE_SKIP;
393: else if (strcmp("text", optarg) == 0)
394: binbehave = BIN_FILE_TEXT;
395: else
396: errx(2, "Unknown binary-files option");
397: break;
398: case 'u':
399: case MMAP_OPT:
400: /* default, compatibility */
401: break;
1.27 otto 402: case LINEBUF_OPT:
403: lbflag = 1;
404: break;
1.9 tedu 405: case HELP_OPT:
1.1 deraadt 406: default:
407: usage();
408: }
1.18 millert 409: lastc = c;
1.23 millert 410: newarg = optind != prevoptind;
1.18 millert 411: prevoptind = optind;
1.1 deraadt 412: }
1.3 deraadt 413: argc -= optind;
414: argv += optind;
1.33 jaredy 415:
416: for (patfile = SLIST_FIRST(&patfilelh); patfile != NULL;
417: patfile = pf_next) {
418: pf_next = SLIST_NEXT(patfile, pf_next);
419: read_patterns(patfile->pf_file);
420: free(patfile);
421: }
1.3 deraadt 422:
1.38 kili 423: if (argc == 0 && needpattern)
1.1 deraadt 424: usage();
425:
1.38 kili 426: if (argc != 0 && needpattern) {
1.35 otto 427: add_patterns(*argv);
1.3 deraadt 428: --argc;
429: ++argv;
430: }
1.5 deraadt 431:
1.16 tedu 432: if (Eflag)
433: cflags |= REG_EXTENDED;
1.39 deraadt 434: fg_pattern = grep_calloc(patterns, sizeof(*fg_pattern));
435: r_pattern = grep_calloc(patterns, sizeof(*r_pattern));
1.3 deraadt 436: for (i = 0; i < patterns; ++i) {
1.28 millert 437: /* Check if cheating is allowed (always is for fgrep). */
438: if (Fflag) {
439: fgrepcomp(&fg_pattern[i], pattern[i]);
440: } else {
441: if (fastcomp(&fg_pattern[i], pattern[i])) {
442: /* Fall back to full regex library */
443: c = regcomp(&r_pattern[i], pattern[i], cflags);
444: if (c != 0) {
445: regerror(c, &r_pattern[i], re_error,
446: RE_ERROR_BUF);
447: errx(2, "%s", re_error);
448: }
1.13 tedu 449: }
1.1 deraadt 450: }
451: }
1.29 deraadt 452:
1.27 otto 453: if (lbflag)
454: setlinebuf(stdout);
1.1 deraadt 455:
1.40 ! tedu 456: if ((argc == 0 || argc == 1) && !Rflag)
1.3 deraadt 457: hflag = 1;
1.1 deraadt 458:
1.3 deraadt 459: if (argc == 0)
460: exit(!procfile(NULL));
1.5 deraadt 461:
1.3 deraadt 462: if (Rflag)
463: c = grep_tree(argv);
1.1 deraadt 464: else
1.3 deraadt 465: for (c = 0; argc--; ++argv)
466: c += procfile(*argv);
1.1 deraadt 467:
1.3 deraadt 468: exit(!c);
1.1 deraadt 469: }