Annotation of src/usr.bin/grep/grep.c, Revision 1.31
1.31 ! otto 1: /* $OpenBSD: grep.c,v 1.30 2004/09/28 20:51:15 jmc 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>
1.1 deraadt 32:
1.18 millert 33: #include <ctype.h>
1.3 deraadt 34: #include <err.h>
35: #include <errno.h>
36: #include <getopt.h>
37: #include <regex.h>
1.1 deraadt 38: #include <stdio.h>
39: #include <stdlib.h>
1.3 deraadt 40: #include <string.h>
1.1 deraadt 41: #include <unistd.h>
42:
1.3 deraadt 43: #include "grep.h"
44:
45: /* Flags passed to regcomp() and regexec() */
46: int cflags;
47: int eflags = REG_STARTEND;
48:
49: int matchall; /* shortcut */
50: int patterns, pattern_sz;
51: char **pattern;
52: regex_t *r_pattern;
1.13 tedu 53: fastgrep_t *fg_pattern;
1.3 deraadt 54:
55: /* For regex errors */
56: char re_error[RE_ERROR_BUF + 1];
57:
58: /* Command-line flags */
59: int Aflag; /* -A x: print x lines trailing each match */
60: int Bflag; /* -B x: print x lines leading each match */
61: int Eflag; /* -E: interpret pattern as extended regexp */
62: int Fflag; /* -F: interpret pattern as list of fixed strings */
63: int Gflag; /* -G: interpret pattern as basic regexp */
64: int Hflag; /* -H: if -R, follow explicitly listed symlinks */
65: int Lflag; /* -L: only show names of files with no matches */
66: int Pflag; /* -P: if -R, no symlinks are followed */
67: int Rflag; /* -R: recursively search directory trees */
68: int Sflag; /* -S: if -R, follow all symlinks */
69: int Vflag; /* -V: display version information */
1.5 deraadt 70: #ifndef NOZ
1.3 deraadt 71: int Zflag; /* -Z: decompress input before processing */
1.5 deraadt 72: #endif
1.3 deraadt 73: int bflag; /* -b: show block numbers for each match */
74: int cflag; /* -c: only show a count of matching lines */
75: int hflag; /* -h: don't print filename headers */
76: int iflag; /* -i: ignore case */
77: int lflag; /* -l: only show names of files with matches */
78: int nflag; /* -n: show line numbers in front of matching lines */
79: int oflag; /* -o: always print file name */
80: int qflag; /* -q: quiet mode (don't output anything) */
81: int sflag; /* -s: silent mode (ignore errors) */
82: int vflag; /* -v: only show non-matching lines */
83: int wflag; /* -w: pattern must start and end on word boundaries */
84: int xflag; /* -x: pattern must match entire line */
1.27 otto 85: int lbflag; /* --line-buffered */
1.3 deraadt 86:
1.9 tedu 87: int binbehave = BIN_FILE_BIN;
88:
89: enum {
90: BIN_OPT = CHAR_MAX + 1,
91: HELP_OPT,
1.27 otto 92: MMAP_OPT,
93: LINEBUF_OPT
1.9 tedu 94: };
95:
1.3 deraadt 96: /* Housekeeping */
1.10 tedu 97: int first; /* flag whether or not this is our first match */
1.3 deraadt 98: int tail; /* lines left to print */
99: int lead; /* number of lines in leading context queue */
100:
1.8 tedu 101: extern char *__progname;
1.3 deraadt 102:
103: static void
104: usage(void)
105: {
1.5 deraadt 106: fprintf(stderr,
107: #ifdef NOZ
1.30 jmc 108: "usage: %s [-abcEFGHhIiLlnoPqRSsUVvwx] [-A num] [-B num] [-C[num]]\n"
1.5 deraadt 109: #else
1.30 jmc 110: "usage: %s [-abcEFGHhIiLlnoPqRSsUVvwxZ] [-A num] [-B num] [-C[num]]\n"
1.5 deraadt 111: #endif
1.30 jmc 112: "\t[-e pattern] [-f file] [--binary-files=value] [--context[=num]]\n"
113: "\t[--line-buffered] [pattern] [file ...]\n", __progname);
1.3 deraadt 114: exit(2);
115: }
116:
1.5 deraadt 117: #ifdef NOZ
1.9 tedu 118: static char *optstr = "0123456789A:B:CEFGHILPSRUVabce:f:hilnoqrsuvwxy";
1.5 deraadt 119: #else
1.9 tedu 120: static char *optstr = "0123456789A:B:CEFGHILPSRUVZabce:f:hilnoqrsuvwxy";
1.5 deraadt 121: #endif
1.3 deraadt 122:
1.5 deraadt 123: struct option long_options[] =
1.3 deraadt 124: {
1.19 deraadt 125: {"binary-files", required_argument, NULL, BIN_OPT},
126: {"help", no_argument, NULL, HELP_OPT},
127: {"mmap", no_argument, NULL, MMAP_OPT},
1.27 otto 128: {"line-buffered", no_argument, NULL, LINEBUF_OPT},
1.19 deraadt 129: {"after-context", required_argument, NULL, 'A'},
130: {"before-context", required_argument, NULL, 'B'},
131: {"context", optional_argument, NULL, 'C'},
132: {"devices", required_argument, NULL, 'D'},
133: {"extended-regexp", no_argument, NULL, 'E'},
134: {"fixed-strings", no_argument, NULL, 'F'},
135: {"basic-regexp", no_argument, NULL, 'G'},
136: {"binary", no_argument, NULL, 'U'},
137: {"version", no_argument, NULL, 'V'},
138: {"text", no_argument, NULL, 'a'},
139: {"byte-offset", no_argument, NULL, 'b'},
140: {"count", no_argument, NULL, 'c'},
141: {"regexp", required_argument, NULL, 'e'},
142: {"file", required_argument, NULL, 'f'},
143: {"no-filename", no_argument, NULL, 'h'},
144: {"ignore-case", no_argument, NULL, 'i'},
145: {"files-without-match", no_argument, NULL, 'L'},
146: {"files-with-matches", no_argument, NULL, 'l'},
147: {"line-number", no_argument, NULL, 'n'},
148: {"quiet", no_argument, NULL, 'q'},
149: {"silent", no_argument, NULL, 'q'},
150: {"recursive", no_argument, NULL, 'r'},
151: {"no-messages", no_argument, NULL, 's'},
152: {"revert-match", no_argument, NULL, 'v'},
153: {"word-regexp", no_argument, NULL, 'w'},
154: {"line-regexp", no_argument, NULL, 'x'},
155: {"unix-byte-offsets", no_argument, NULL, 'u'},
1.5 deraadt 156: #ifndef NOZ
1.19 deraadt 157: {"decompress", no_argument, NULL, 'Z'},
1.5 deraadt 158: #endif
1.19 deraadt 159: {NULL, no_argument, NULL, 0}
1.3 deraadt 160: };
161:
1.1 deraadt 162:
1.3 deraadt 163: static void
164: add_pattern(char *pat, size_t len)
165: {
166: if (len == 0 || matchall) {
167: matchall = 1;
168: return;
169: }
170: if (patterns == pattern_sz) {
171: pattern_sz *= 2;
1.22 millert 172: pattern = grep_realloc(pattern, ++pattern_sz * sizeof(*pattern));
1.3 deraadt 173: }
1.11 tedu 174: if (pat[len - 1] == '\n')
1.3 deraadt 175: --len;
1.15 tedu 176: /* pat may not be NUL-terminated */
1.28 millert 177: if (wflag && !Fflag) {
1.26 millert 178: int bol = 0, eol = 0;
179: if (pat[0] == '^')
180: bol = 1;
181: if (pat[len - 1] == '$')
182: eol = 1;
1.25 millert 183: pattern[patterns] = grep_malloc(len + 15);
1.26 millert 184: snprintf(pattern[patterns], len + 15, "%s[[:<:]]%.*s[[:>:]]%s",
185: bol ? "^" : "", (int)len - bol - eol, pat + bol,
186: eol ? "$" : "");
187: len += 14;
1.25 millert 188: } else {
189: pattern[patterns] = grep_malloc(len + 1);
190: memcpy(pattern[patterns], pat, len);
191: pattern[patterns][len] = '\0';
192: }
1.3 deraadt 193: ++patterns;
194: }
1.1 deraadt 195:
1.3 deraadt 196: static void
197: read_patterns(char *fn)
198: {
199: FILE *f;
200: char *line;
201: size_t len;
202: int nl;
203:
204: if ((f = fopen(fn, "r")) == NULL)
1.20 millert 205: err(2, "%s", fn);
1.3 deraadt 206: nl = 0;
207: while ((line = fgetln(f, &len)) != NULL) {
208: if (*line == '\n') {
209: ++nl;
210: continue;
211: }
212: if (nl) {
213: matchall = 1;
214: break;
215: }
216: nl = 0;
217: add_pattern(line, len);
218: }
219: if (ferror(f))
1.20 millert 220: err(2, "%s", fn);
1.3 deraadt 221: fclose(f);
222: }
1.1 deraadt 223:
224: int
1.3 deraadt 225: main(int argc, char *argv[])
1.1 deraadt 226: {
1.23 millert 227: int c, lastc, prevoptind, newarg, i;
1.21 millert 228: long l;
229: char *ep;
1.3 deraadt 230:
1.16 tedu 231: switch (__progname[0]) {
232: case 'e':
233: Eflag++;
234: break;
235: case 'f':
236: Fflag++;
237: break;
238: case 'g':
239: Gflag++;
240: break;
241: #ifndef NOZ
242: case 'z':
243: Zflag++;
244: switch(__progname[1]) {
245: case 'e':
246: Eflag++;
247: break;
248: case 'f':
249: Fflag++;
250: break;
251: case 'g':
252: Gflag++;
253: break;
254: }
255: break;
256: #endif
257: }
258:
1.18 millert 259: lastc = '\0';
1.23 millert 260: newarg = 1;
261: prevoptind = 1;
1.5 deraadt 262: while ((c = getopt_long(argc, argv, optstr,
1.17 millert 263: long_options, NULL)) != -1) {
1.3 deraadt 264: switch (c) {
265: case '0': case '1': case '2': case '3': case '4':
266: case '5': case '6': case '7': case '8': case '9':
1.23 millert 267: if (newarg || !isdigit(lastc))
268: Aflag = 0;
269: else if (Aflag > INT_MAX / 10)
270: errx(2, "context out of range");
271: Aflag = Bflag = (Aflag * 10) + (c - '0');
1.3 deraadt 272: break;
273: case 'A':
274: case 'B':
1.21 millert 275: l = strtol(optarg, &ep, 10);
276: if (ep == optarg || *ep != '\0' ||
277: l <= 0 || l >= INT_MAX)
278: errx(2, "context out of range");
279: if (c == 'A')
280: Aflag = (int)l;
281: else
282: Bflag = (int)l;
1.3 deraadt 283: break;
284: case 'C':
1.5 deraadt 285: if (optarg == NULL)
1.3 deraadt 286: Aflag = Bflag = 2;
1.21 millert 287: else {
288: l = strtol(optarg, &ep, 10);
289: if (ep == optarg || *ep != '\0' ||
290: l <= 0 || l >= INT_MAX)
291: errx(2, "context out of range");
292: Aflag = Bflag = (int)l;
293: }
1.1 deraadt 294: break;
295: case 'E':
1.16 tedu 296: Fflag = Gflag = 0;
1.3 deraadt 297: Eflag++;
1.1 deraadt 298: break;
299: case 'F':
1.16 tedu 300: Eflag = Gflag = 0;
1.3 deraadt 301: Fflag++;
302: break;
303: case 'G':
1.16 tedu 304: Eflag = Fflag = 0;
1.3 deraadt 305: Gflag++;
1.1 deraadt 306: break;
307: case 'H':
1.3 deraadt 308: Hflag++;
1.1 deraadt 309: break;
1.9 tedu 310: case 'I':
311: binbehave = BIN_FILE_SKIP;
312: break;
1.1 deraadt 313: case 'L':
1.3 deraadt 314: lflag = 0;
315: Lflag = qflag = 1;
1.1 deraadt 316: break;
317: case 'P':
1.3 deraadt 318: Pflag++;
319: break;
320: case 'S':
321: Sflag++;
1.1 deraadt 322: break;
323: case 'R':
1.3 deraadt 324: case 'r':
325: Rflag++;
326: oflag++;
327: break;
328: case 'U':
1.9 tedu 329: binbehave = BIN_FILE_BIN;
1.3 deraadt 330: break;
331: case 'V':
332: fprintf(stderr, "grep version %u.%u\n", VER_MAJ, VER_MIN);
1.9 tedu 333: exit(0);
1.1 deraadt 334: break;
1.5 deraadt 335: #ifndef NOZ
1.3 deraadt 336: case 'Z':
337: Zflag++;
1.1 deraadt 338: break;
1.5 deraadt 339: #endif
1.1 deraadt 340: case 'a':
1.9 tedu 341: binbehave = BIN_FILE_TEXT;
1.1 deraadt 342: break;
343: case 'b':
1.3 deraadt 344: bflag = 1;
1.1 deraadt 345: break;
346: case 'c':
1.3 deraadt 347: cflag = 1;
1.1 deraadt 348: break;
349: case 'e':
1.3 deraadt 350: add_pattern(optarg, strlen(optarg));
1.1 deraadt 351: break;
352: case 'f':
1.3 deraadt 353: read_patterns(optarg);
1.1 deraadt 354: break;
355: case 'h':
1.3 deraadt 356: oflag = 0;
357: hflag = 1;
1.1 deraadt 358: break;
359: case 'i':
1.3 deraadt 360: case 'y':
1.13 tedu 361: iflag = 1;
1.1 deraadt 362: cflags |= REG_ICASE;
363: break;
364: case 'l':
1.3 deraadt 365: Lflag = 0;
366: lflag = qflag = 1;
1.1 deraadt 367: break;
368: case 'n':
1.3 deraadt 369: nflag = 1;
370: break;
371: case 'o':
372: hflag = 0;
373: oflag = 1;
1.1 deraadt 374: break;
375: case 'q':
1.3 deraadt 376: qflag = 1;
1.1 deraadt 377: break;
378: case 's':
1.3 deraadt 379: sflag = 1;
1.1 deraadt 380: break;
381: case 'v':
1.3 deraadt 382: vflag = 1;
1.1 deraadt 383: break;
384: case 'w':
1.3 deraadt 385: wflag = 1;
1.1 deraadt 386: break;
387: case 'x':
1.3 deraadt 388: xflag = 1;
1.1 deraadt 389: break;
1.9 tedu 390: case BIN_OPT:
391: if (strcmp("binary", optarg) == 0)
392: binbehave = BIN_FILE_BIN;
393: else if (strcmp("without-match", optarg) == 0)
394: binbehave = BIN_FILE_SKIP;
395: else if (strcmp("text", optarg) == 0)
396: binbehave = BIN_FILE_TEXT;
397: else
398: errx(2, "Unknown binary-files option");
399: break;
400: case 'u':
401: case MMAP_OPT:
402: /* default, compatibility */
403: break;
1.27 otto 404: case LINEBUF_OPT:
405: lbflag = 1;
406: break;
1.9 tedu 407: case HELP_OPT:
1.1 deraadt 408: default:
409: usage();
410: }
1.18 millert 411: lastc = c;
1.23 millert 412: newarg = optind != prevoptind;
1.18 millert 413: prevoptind = optind;
1.1 deraadt 414: }
1.3 deraadt 415: argc -= optind;
416: argv += optind;
417:
418: if (argc == 0 && patterns == 0)
1.1 deraadt 419: usage();
420:
1.3 deraadt 421: if (patterns == 0) {
422: add_pattern(*argv, strlen(*argv));
423: --argc;
424: ++argv;
425: }
1.5 deraadt 426:
1.16 tedu 427: if (Eflag)
428: cflags |= REG_EXTENDED;
1.13 tedu 429: fg_pattern = grep_malloc(patterns * sizeof(*fg_pattern));
1.22 millert 430: r_pattern = grep_malloc(patterns * sizeof(*r_pattern));
1.3 deraadt 431: for (i = 0; i < patterns; ++i) {
1.28 millert 432: /* Check if cheating is allowed (always is for fgrep). */
433: if (Fflag) {
434: fgrepcomp(&fg_pattern[i], pattern[i]);
435: } else {
436: if (fastcomp(&fg_pattern[i], pattern[i])) {
437: /* Fall back to full regex library */
438: c = regcomp(&r_pattern[i], pattern[i], cflags);
439: if (c != 0) {
440: regerror(c, &r_pattern[i], re_error,
441: RE_ERROR_BUF);
442: errx(2, "%s", re_error);
443: }
1.13 tedu 444: }
1.1 deraadt 445: }
446: }
1.29 deraadt 447:
1.27 otto 448: if (lbflag)
449: setlinebuf(stdout);
1.1 deraadt 450:
1.3 deraadt 451: if ((argc == 0 || argc == 1) && !oflag)
452: hflag = 1;
1.1 deraadt 453:
1.3 deraadt 454: if (argc == 0)
455: exit(!procfile(NULL));
1.5 deraadt 456:
1.3 deraadt 457: if (Rflag)
458: c = grep_tree(argv);
1.1 deraadt 459: else
1.3 deraadt 460: for (c = 0; argc--; ++argv)
461: c += procfile(*argv);
1.1 deraadt 462:
1.3 deraadt 463: exit(!c);
1.1 deraadt 464: }