Annotation of src/usr.bin/grep/grep.c, Revision 1.23
1.23 ! millert 1: /* $OpenBSD: grep.c,v 1.22 2003/07/16 19:08:21 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
1.9 tedu 108: "usage: %s [-[AB] num] [-CEFGHILPRSUVabchilnoqsvwx]"
1.5 deraadt 109: #else
1.9 tedu 110: "usage: %s [-[AB] num] [-CEFGHILPRSUVZabchilnoqsvwx]"
1.5 deraadt 111: #endif
1.8 tedu 112: " [-e pattern] [-f file] [file ...]\n", __progname);
1.3 deraadt 113: exit(2);
114: }
115:
1.5 deraadt 116: #ifdef NOZ
1.9 tedu 117: static char *optstr = "0123456789A:B:CEFGHILPSRUVabce:f:hilnoqrsuvwxy";
1.5 deraadt 118: #else
1.9 tedu 119: static char *optstr = "0123456789A:B:CEFGHILPSRUVZabce:f:hilnoqrsuvwxy";
1.5 deraadt 120: #endif
1.3 deraadt 121:
1.5 deraadt 122: struct option long_options[] =
1.3 deraadt 123: {
1.19 deraadt 124: {"binary-files", required_argument, NULL, BIN_OPT},
125: {"help", no_argument, NULL, HELP_OPT},
126: {"mmap", no_argument, NULL, MMAP_OPT},
127: {"after-context", required_argument, NULL, 'A'},
128: {"before-context", required_argument, NULL, 'B'},
129: {"context", optional_argument, NULL, 'C'},
130: {"devices", required_argument, NULL, 'D'},
131: {"extended-regexp", no_argument, NULL, 'E'},
132: {"fixed-strings", no_argument, NULL, 'F'},
133: {"basic-regexp", no_argument, NULL, 'G'},
134: {"binary", no_argument, NULL, 'U'},
135: {"version", no_argument, NULL, 'V'},
136: {"text", no_argument, NULL, 'a'},
137: {"byte-offset", no_argument, NULL, 'b'},
138: {"count", no_argument, NULL, 'c'},
139: {"regexp", required_argument, NULL, 'e'},
140: {"file", required_argument, NULL, 'f'},
141: {"no-filename", no_argument, NULL, 'h'},
142: {"ignore-case", no_argument, NULL, 'i'},
143: {"files-without-match", no_argument, NULL, 'L'},
144: {"files-with-matches", no_argument, NULL, 'l'},
145: {"line-number", no_argument, NULL, 'n'},
146: {"quiet", no_argument, NULL, 'q'},
147: {"silent", no_argument, NULL, 'q'},
148: {"recursive", no_argument, NULL, 'r'},
149: {"no-messages", no_argument, NULL, 's'},
150: {"revert-match", no_argument, NULL, 'v'},
151: {"word-regexp", no_argument, NULL, 'w'},
152: {"line-regexp", no_argument, NULL, 'x'},
153: {"unix-byte-offsets", no_argument, NULL, 'u'},
1.5 deraadt 154: #ifndef NOZ
1.19 deraadt 155: {"decompress", no_argument, NULL, 'Z'},
1.5 deraadt 156: #endif
1.19 deraadt 157: {NULL, no_argument, NULL, 0}
1.3 deraadt 158: };
159:
1.1 deraadt 160:
1.3 deraadt 161: static void
162: add_pattern(char *pat, size_t len)
163: {
164: if (len == 0 || matchall) {
165: matchall = 1;
166: return;
167: }
168: if (patterns == pattern_sz) {
169: pattern_sz *= 2;
1.22 millert 170: pattern = grep_realloc(pattern, ++pattern_sz * sizeof(*pattern));
1.3 deraadt 171: }
1.11 tedu 172: if (pat[len - 1] == '\n')
1.3 deraadt 173: --len;
1.11 tedu 174: pattern[patterns] = grep_malloc(len + 1);
1.15 tedu 175: /* pat may not be NUL-terminated */
176: memcpy(pattern[patterns], pat, len);
1.14 tedu 177: pattern[patterns][len] = '\0';
1.3 deraadt 178: ++patterns;
1.13 tedu 179:
180: if (len > maxPatternLen)
181: maxPatternLen = len;
1.3 deraadt 182: }
1.1 deraadt 183:
1.3 deraadt 184: static void
185: read_patterns(char *fn)
186: {
187: FILE *f;
188: char *line;
189: size_t len;
190: int nl;
191:
192: if ((f = fopen(fn, "r")) == NULL)
1.20 millert 193: err(2, "%s", fn);
1.3 deraadt 194: nl = 0;
195: while ((line = fgetln(f, &len)) != NULL) {
196: if (*line == '\n') {
197: ++nl;
198: continue;
199: }
200: if (nl) {
201: matchall = 1;
202: break;
203: }
204: nl = 0;
205: add_pattern(line, len);
206: }
207: if (ferror(f))
1.20 millert 208: err(2, "%s", fn);
1.3 deraadt 209: fclose(f);
210: }
1.1 deraadt 211:
1.13 tedu 212: static void
213: free_patterns(void)
214: {
215: int i;
216:
217: for (i = 0; i < patterns; i++) {
218: if (fg_pattern[i].pattern)
219: free(fg_pattern[i].pattern);
220: else
221: regfree(&r_pattern[i]);
222: free(pattern[i]);
223: }
224:
225: free(fg_pattern);
226: free(r_pattern);
227: free(pattern);
228: }
229:
1.1 deraadt 230: int
1.3 deraadt 231: main(int argc, char *argv[])
1.1 deraadt 232: {
1.23 ! millert 233: int c, lastc, prevoptind, newarg, i;
1.21 millert 234: long l;
235: char *ep;
1.3 deraadt 236:
1.16 tedu 237: switch (__progname[0]) {
238: case 'e':
239: Eflag++;
240: break;
241: case 'f':
242: Fflag++;
243: break;
244: case 'g':
245: Gflag++;
246: break;
247: #ifndef NOZ
248: case 'z':
249: Zflag++;
250: switch(__progname[1]) {
251: case 'e':
252: Eflag++;
253: break;
254: case 'f':
255: Fflag++;
256: break;
257: case 'g':
258: Gflag++;
259: break;
260: }
261: break;
262: #endif
263: }
264:
1.18 millert 265: lastc = '\0';
1.23 ! millert 266: newarg = 1;
! 267: prevoptind = 1;
1.5 deraadt 268: while ((c = getopt_long(argc, argv, optstr,
1.17 millert 269: long_options, NULL)) != -1) {
1.3 deraadt 270: switch (c) {
271: case '0': case '1': case '2': case '3': case '4':
272: case '5': case '6': case '7': case '8': case '9':
1.23 ! millert 273: if (newarg || !isdigit(lastc))
! 274: Aflag = 0;
! 275: else if (Aflag > INT_MAX / 10)
! 276: errx(2, "context out of range");
! 277: Aflag = Bflag = (Aflag * 10) + (c - '0');
1.3 deraadt 278: break;
279: case 'A':
280: case 'B':
1.21 millert 281: l = strtol(optarg, &ep, 10);
282: if (ep == optarg || *ep != '\0' ||
283: l <= 0 || l >= INT_MAX)
284: errx(2, "context out of range");
285: if (c == 'A')
286: Aflag = (int)l;
287: else
288: Bflag = (int)l;
1.3 deraadt 289: break;
290: case 'C':
1.5 deraadt 291: if (optarg == NULL)
1.3 deraadt 292: Aflag = Bflag = 2;
1.21 millert 293: else {
294: l = strtol(optarg, &ep, 10);
295: if (ep == optarg || *ep != '\0' ||
296: l <= 0 || l >= INT_MAX)
297: errx(2, "context out of range");
298: Aflag = Bflag = (int)l;
299: }
1.1 deraadt 300: break;
301: case 'E':
1.16 tedu 302: Fflag = Gflag = 0;
1.3 deraadt 303: Eflag++;
1.1 deraadt 304: break;
305: case 'F':
1.16 tedu 306: Eflag = Gflag = 0;
1.3 deraadt 307: Fflag++;
308: break;
309: case 'G':
1.16 tedu 310: Eflag = Fflag = 0;
1.3 deraadt 311: Gflag++;
1.1 deraadt 312: break;
313: case 'H':
1.3 deraadt 314: Hflag++;
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 'P':
1.3 deraadt 324: Pflag++;
325: break;
326: case 'S':
327: Sflag++;
1.1 deraadt 328: break;
329: case 'R':
1.3 deraadt 330: case 'r':
331: Rflag++;
332: oflag++;
333: break;
334: case 'U':
1.9 tedu 335: binbehave = BIN_FILE_BIN;
1.3 deraadt 336: break;
337: case 'V':
338: fprintf(stderr, "grep version %u.%u\n", VER_MAJ, VER_MIN);
1.9 tedu 339: exit(0);
1.1 deraadt 340: break;
1.5 deraadt 341: #ifndef NOZ
1.3 deraadt 342: case 'Z':
343: Zflag++;
1.1 deraadt 344: break;
1.5 deraadt 345: #endif
1.1 deraadt 346: case 'a':
1.9 tedu 347: binbehave = BIN_FILE_TEXT;
1.1 deraadt 348: break;
349: case 'b':
1.3 deraadt 350: bflag = 1;
1.1 deraadt 351: break;
352: case 'c':
1.3 deraadt 353: cflag = 1;
1.1 deraadt 354: break;
355: case 'e':
1.3 deraadt 356: add_pattern(optarg, strlen(optarg));
1.1 deraadt 357: break;
358: case 'f':
1.3 deraadt 359: read_patterns(optarg);
1.1 deraadt 360: break;
361: case 'h':
1.3 deraadt 362: oflag = 0;
363: hflag = 1;
1.1 deraadt 364: break;
365: case 'i':
1.3 deraadt 366: case 'y':
1.13 tedu 367: iflag = 1;
1.1 deraadt 368: cflags |= REG_ICASE;
369: break;
370: case 'l':
1.3 deraadt 371: Lflag = 0;
372: lflag = qflag = 1;
1.1 deraadt 373: break;
374: case 'n':
1.3 deraadt 375: nflag = 1;
376: break;
377: case 'o':
378: hflag = 0;
379: oflag = 1;
1.1 deraadt 380: break;
381: case 'q':
1.3 deraadt 382: qflag = 1;
1.1 deraadt 383: break;
384: case 's':
1.3 deraadt 385: sflag = 1;
1.1 deraadt 386: break;
387: case 'v':
1.3 deraadt 388: vflag = 1;
1.1 deraadt 389: break;
390: case 'w':
1.3 deraadt 391: wflag = 1;
1.1 deraadt 392: break;
393: case 'x':
1.3 deraadt 394: xflag = 1;
1.1 deraadt 395: break;
1.9 tedu 396: case BIN_OPT:
397: if (strcmp("binary", optarg) == 0)
398: binbehave = BIN_FILE_BIN;
399: else if (strcmp("without-match", optarg) == 0)
400: binbehave = BIN_FILE_SKIP;
401: else if (strcmp("text", optarg) == 0)
402: binbehave = BIN_FILE_TEXT;
403: else
404: errx(2, "Unknown binary-files option");
405: break;
406: case 'u':
407: case MMAP_OPT:
408: /* default, compatibility */
409: break;
410: case HELP_OPT:
1.1 deraadt 411: default:
412: usage();
413: }
1.18 millert 414: lastc = c;
1.23 ! millert 415: newarg = optind != prevoptind;
1.18 millert 416: prevoptind = optind;
1.1 deraadt 417: }
1.3 deraadt 418: argc -= optind;
419: argv += optind;
420:
421: if (argc == 0 && patterns == 0)
1.1 deraadt 422: usage();
423:
1.3 deraadt 424: if (patterns == 0) {
425: add_pattern(*argv, strlen(*argv));
426: --argc;
427: ++argv;
428: }
1.5 deraadt 429:
1.16 tedu 430: if (Eflag)
431: cflags |= REG_EXTENDED;
432: else if (Fflag)
433: cflags |= REG_NOSPEC;
1.13 tedu 434: fg_pattern = grep_malloc(patterns * sizeof(*fg_pattern));
1.22 millert 435: r_pattern = grep_malloc(patterns * sizeof(*r_pattern));
1.3 deraadt 436: for (i = 0; i < patterns; ++i) {
1.13 tedu 437: /* Check if cheating is allowed */
438: if (fastcomp(&fg_pattern[i], pattern[i])) {
439: /* Fall back to full regex library */
440: if ((c = regcomp(&r_pattern[i], pattern[i], cflags))) {
441: regerror(c, &r_pattern[i], re_error,
442: RE_ERROR_BUF);
1.20 millert 443: errx(2, "%s", re_error);
1.13 tedu 444: }
1.1 deraadt 445: }
446: }
447:
1.3 deraadt 448: if ((argc == 0 || argc == 1) && !oflag)
449: hflag = 1;
1.1 deraadt 450:
1.3 deraadt 451: if (argc == 0)
452: exit(!procfile(NULL));
1.5 deraadt 453:
1.3 deraadt 454: if (Rflag)
455: c = grep_tree(argv);
1.1 deraadt 456: else
1.3 deraadt 457: for (c = 0; argc--; ++argv)
458: c += procfile(*argv);
1.13 tedu 459:
460: free_patterns();
1.1 deraadt 461:
1.3 deraadt 462: exit(!c);
1.1 deraadt 463: }