Annotation of src/usr.bin/grep/grep.c, Revision 1.26
1.26 ! millert 1: /* $OpenBSD: grep.c,v 1.25 2004/01/25 21:36:00 millert 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 */
85:
1.9 tedu 86: int binbehave = BIN_FILE_BIN;
87:
88: enum {
89: BIN_OPT = CHAR_MAX + 1,
90: HELP_OPT,
91: MMAP_OPT
92: };
93:
1.3 deraadt 94: /* Housekeeping */
1.10 tedu 95: int first; /* flag whether or not this is our first match */
1.3 deraadt 96: int tail; /* lines left to print */
97: int lead; /* number of lines in leading context queue */
1.13 tedu 98: int boleol; /* At least one pattern has a bol or eol */
1.17 millert 99: size_t maxPatternLen; /* Longest length of all patterns */
1.3 deraadt 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
1.24 mcbride 108: "usage: %s [-AB num] [-CEFGHILPRSUVabchilnoqsvwx]\n"
1.5 deraadt 109: #else
1.24 mcbride 110: "usage: %s [-AB num] [-CEFGHILPRSUVZabchilnoqsvwx]\n"
1.5 deraadt 111: #endif
1.24 mcbride 112: "\t[--context[=num]] [--binary-files=value]\n"
113: "\t[-e pattern] [-f file] [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},
128: {"after-context", required_argument, NULL, 'A'},
129: {"before-context", required_argument, NULL, 'B'},
130: {"context", optional_argument, NULL, 'C'},
131: {"devices", required_argument, NULL, 'D'},
132: {"extended-regexp", no_argument, NULL, 'E'},
133: {"fixed-strings", no_argument, NULL, 'F'},
134: {"basic-regexp", no_argument, NULL, 'G'},
135: {"binary", no_argument, NULL, 'U'},
136: {"version", no_argument, NULL, 'V'},
137: {"text", no_argument, NULL, 'a'},
138: {"byte-offset", no_argument, NULL, 'b'},
139: {"count", no_argument, NULL, 'c'},
140: {"regexp", required_argument, NULL, 'e'},
141: {"file", required_argument, NULL, 'f'},
142: {"no-filename", no_argument, NULL, 'h'},
143: {"ignore-case", no_argument, NULL, 'i'},
144: {"files-without-match", no_argument, NULL, 'L'},
145: {"files-with-matches", no_argument, NULL, 'l'},
146: {"line-number", no_argument, NULL, 'n'},
147: {"quiet", no_argument, NULL, 'q'},
148: {"silent", no_argument, NULL, 'q'},
149: {"recursive", no_argument, NULL, 'r'},
150: {"no-messages", no_argument, NULL, 's'},
151: {"revert-match", no_argument, NULL, 'v'},
152: {"word-regexp", no_argument, NULL, 'w'},
153: {"line-regexp", no_argument, NULL, 'x'},
154: {"unix-byte-offsets", no_argument, NULL, 'u'},
1.5 deraadt 155: #ifndef NOZ
1.19 deraadt 156: {"decompress", no_argument, NULL, 'Z'},
1.5 deraadt 157: #endif
1.19 deraadt 158: {NULL, no_argument, NULL, 0}
1.3 deraadt 159: };
160:
1.1 deraadt 161:
1.3 deraadt 162: static void
163: add_pattern(char *pat, size_t len)
164: {
165: if (len == 0 || matchall) {
166: matchall = 1;
167: return;
168: }
169: if (patterns == pattern_sz) {
170: pattern_sz *= 2;
1.22 millert 171: pattern = grep_realloc(pattern, ++pattern_sz * sizeof(*pattern));
1.3 deraadt 172: }
1.11 tedu 173: if (pat[len - 1] == '\n')
1.3 deraadt 174: --len;
1.15 tedu 175: /* pat may not be NUL-terminated */
1.25 millert 176: if (wflag) {
1.26 ! millert 177: int bol = 0, eol = 0;
! 178: if (pat[0] == '^')
! 179: bol = 1;
! 180: if (pat[len - 1] == '$')
! 181: eol = 1;
1.25 millert 182: pattern[patterns] = grep_malloc(len + 15);
1.26 ! millert 183: snprintf(pattern[patterns], len + 15, "%s[[:<:]]%.*s[[:>:]]%s",
! 184: bol ? "^" : "", (int)len - bol - eol, pat + bol,
! 185: eol ? "$" : "");
! 186: len += 14;
1.25 millert 187: } else {
188: pattern[patterns] = grep_malloc(len + 1);
189: memcpy(pattern[patterns], pat, len);
190: pattern[patterns][len] = '\0';
191: }
1.3 deraadt 192: ++patterns;
1.13 tedu 193:
194: if (len > maxPatternLen)
195: maxPatternLen = len;
1.3 deraadt 196: }
1.1 deraadt 197:
1.3 deraadt 198: static void
199: read_patterns(char *fn)
200: {
201: FILE *f;
202: char *line;
203: size_t len;
204: int nl;
205:
206: if ((f = fopen(fn, "r")) == NULL)
1.20 millert 207: err(2, "%s", fn);
1.3 deraadt 208: nl = 0;
209: while ((line = fgetln(f, &len)) != NULL) {
210: if (*line == '\n') {
211: ++nl;
212: continue;
213: }
214: if (nl) {
215: matchall = 1;
216: break;
217: }
218: nl = 0;
219: add_pattern(line, len);
220: }
221: if (ferror(f))
1.20 millert 222: err(2, "%s", fn);
1.3 deraadt 223: fclose(f);
224: }
1.1 deraadt 225:
1.13 tedu 226: static void
227: free_patterns(void)
228: {
229: int i;
230:
231: for (i = 0; i < patterns; i++) {
232: if (fg_pattern[i].pattern)
233: free(fg_pattern[i].pattern);
234: else
235: regfree(&r_pattern[i]);
236: free(pattern[i]);
237: }
238:
239: free(fg_pattern);
240: free(r_pattern);
241: free(pattern);
242: }
243:
1.1 deraadt 244: int
1.3 deraadt 245: main(int argc, char *argv[])
1.1 deraadt 246: {
1.23 millert 247: int c, lastc, prevoptind, newarg, i;
1.21 millert 248: long l;
249: char *ep;
1.3 deraadt 250:
1.16 tedu 251: switch (__progname[0]) {
252: case 'e':
253: Eflag++;
254: break;
255: case 'f':
256: Fflag++;
257: break;
258: case 'g':
259: Gflag++;
260: break;
261: #ifndef NOZ
262: case 'z':
263: Zflag++;
264: switch(__progname[1]) {
265: case 'e':
266: Eflag++;
267: break;
268: case 'f':
269: Fflag++;
270: break;
271: case 'g':
272: Gflag++;
273: break;
274: }
275: break;
276: #endif
277: }
278:
1.18 millert 279: lastc = '\0';
1.23 millert 280: newarg = 1;
281: prevoptind = 1;
1.5 deraadt 282: while ((c = getopt_long(argc, argv, optstr,
1.17 millert 283: long_options, NULL)) != -1) {
1.3 deraadt 284: switch (c) {
285: case '0': case '1': case '2': case '3': case '4':
286: case '5': case '6': case '7': case '8': case '9':
1.23 millert 287: if (newarg || !isdigit(lastc))
288: Aflag = 0;
289: else if (Aflag > INT_MAX / 10)
290: errx(2, "context out of range");
291: Aflag = Bflag = (Aflag * 10) + (c - '0');
1.3 deraadt 292: break;
293: case 'A':
294: case 'B':
1.21 millert 295: l = strtol(optarg, &ep, 10);
296: if (ep == optarg || *ep != '\0' ||
297: l <= 0 || l >= INT_MAX)
298: errx(2, "context out of range");
299: if (c == 'A')
300: Aflag = (int)l;
301: else
302: Bflag = (int)l;
1.3 deraadt 303: break;
304: case 'C':
1.5 deraadt 305: if (optarg == NULL)
1.3 deraadt 306: Aflag = Bflag = 2;
1.21 millert 307: else {
308: l = strtol(optarg, &ep, 10);
309: if (ep == optarg || *ep != '\0' ||
310: l <= 0 || l >= INT_MAX)
311: errx(2, "context out of range");
312: Aflag = Bflag = (int)l;
313: }
1.1 deraadt 314: break;
315: case 'E':
1.16 tedu 316: Fflag = Gflag = 0;
1.3 deraadt 317: Eflag++;
1.1 deraadt 318: break;
319: case 'F':
1.16 tedu 320: Eflag = Gflag = 0;
1.3 deraadt 321: Fflag++;
322: break;
323: case 'G':
1.16 tedu 324: Eflag = Fflag = 0;
1.3 deraadt 325: Gflag++;
1.1 deraadt 326: break;
327: case 'H':
1.3 deraadt 328: Hflag++;
1.1 deraadt 329: break;
1.9 tedu 330: case 'I':
331: binbehave = BIN_FILE_SKIP;
332: break;
1.1 deraadt 333: case 'L':
1.3 deraadt 334: lflag = 0;
335: Lflag = qflag = 1;
1.1 deraadt 336: break;
337: case 'P':
1.3 deraadt 338: Pflag++;
339: break;
340: case 'S':
341: Sflag++;
1.1 deraadt 342: break;
343: case 'R':
1.3 deraadt 344: case 'r':
345: Rflag++;
346: oflag++;
347: break;
348: case 'U':
1.9 tedu 349: binbehave = BIN_FILE_BIN;
1.3 deraadt 350: break;
351: case 'V':
352: fprintf(stderr, "grep version %u.%u\n", VER_MAJ, VER_MIN);
1.9 tedu 353: exit(0);
1.1 deraadt 354: break;
1.5 deraadt 355: #ifndef NOZ
1.3 deraadt 356: case 'Z':
357: Zflag++;
1.1 deraadt 358: break;
1.5 deraadt 359: #endif
1.1 deraadt 360: case 'a':
1.9 tedu 361: binbehave = BIN_FILE_TEXT;
1.1 deraadt 362: break;
363: case 'b':
1.3 deraadt 364: bflag = 1;
1.1 deraadt 365: break;
366: case 'c':
1.3 deraadt 367: cflag = 1;
1.1 deraadt 368: break;
369: case 'e':
1.3 deraadt 370: add_pattern(optarg, strlen(optarg));
1.1 deraadt 371: break;
372: case 'f':
1.3 deraadt 373: read_patterns(optarg);
1.1 deraadt 374: break;
375: case 'h':
1.3 deraadt 376: oflag = 0;
377: hflag = 1;
1.1 deraadt 378: break;
379: case 'i':
1.3 deraadt 380: case 'y':
1.13 tedu 381: iflag = 1;
1.1 deraadt 382: cflags |= REG_ICASE;
383: break;
384: case 'l':
1.3 deraadt 385: Lflag = 0;
386: lflag = qflag = 1;
1.1 deraadt 387: break;
388: case 'n':
1.3 deraadt 389: nflag = 1;
390: break;
391: case 'o':
392: hflag = 0;
393: oflag = 1;
1.1 deraadt 394: break;
395: case 'q':
1.3 deraadt 396: qflag = 1;
1.1 deraadt 397: break;
398: case 's':
1.3 deraadt 399: sflag = 1;
1.1 deraadt 400: break;
401: case 'v':
1.3 deraadt 402: vflag = 1;
1.1 deraadt 403: break;
404: case 'w':
1.3 deraadt 405: wflag = 1;
1.1 deraadt 406: break;
407: case 'x':
1.3 deraadt 408: xflag = 1;
1.1 deraadt 409: break;
1.9 tedu 410: case BIN_OPT:
411: if (strcmp("binary", optarg) == 0)
412: binbehave = BIN_FILE_BIN;
413: else if (strcmp("without-match", optarg) == 0)
414: binbehave = BIN_FILE_SKIP;
415: else if (strcmp("text", optarg) == 0)
416: binbehave = BIN_FILE_TEXT;
417: else
418: errx(2, "Unknown binary-files option");
419: break;
420: case 'u':
421: case MMAP_OPT:
422: /* default, compatibility */
423: break;
424: case HELP_OPT:
1.1 deraadt 425: default:
426: usage();
427: }
1.18 millert 428: lastc = c;
1.23 millert 429: newarg = optind != prevoptind;
1.18 millert 430: prevoptind = optind;
1.1 deraadt 431: }
1.3 deraadt 432: argc -= optind;
433: argv += optind;
434:
435: if (argc == 0 && patterns == 0)
1.1 deraadt 436: usage();
437:
1.3 deraadt 438: if (patterns == 0) {
439: add_pattern(*argv, strlen(*argv));
440: --argc;
441: ++argv;
442: }
1.5 deraadt 443:
1.16 tedu 444: if (Eflag)
445: cflags |= REG_EXTENDED;
446: else if (Fflag)
447: cflags |= REG_NOSPEC;
1.13 tedu 448: fg_pattern = grep_malloc(patterns * sizeof(*fg_pattern));
1.22 millert 449: r_pattern = grep_malloc(patterns * sizeof(*r_pattern));
1.3 deraadt 450: for (i = 0; i < patterns; ++i) {
1.13 tedu 451: /* Check if cheating is allowed */
452: if (fastcomp(&fg_pattern[i], pattern[i])) {
453: /* Fall back to full regex library */
454: if ((c = regcomp(&r_pattern[i], pattern[i], cflags))) {
455: regerror(c, &r_pattern[i], re_error,
456: RE_ERROR_BUF);
1.20 millert 457: errx(2, "%s", re_error);
1.13 tedu 458: }
1.1 deraadt 459: }
460: }
461:
1.3 deraadt 462: if ((argc == 0 || argc == 1) && !oflag)
463: hflag = 1;
1.1 deraadt 464:
1.3 deraadt 465: if (argc == 0)
466: exit(!procfile(NULL));
1.5 deraadt 467:
1.3 deraadt 468: if (Rflag)
469: c = grep_tree(argv);
1.1 deraadt 470: else
1.3 deraadt 471: for (c = 0; argc--; ++argv)
472: c += procfile(*argv);
1.13 tedu 473:
474: free_patterns();
1.1 deraadt 475:
1.3 deraadt 476: exit(!c);
1.1 deraadt 477: }