1 /* $OpenBSD: grep.c,v 1.65 2020/07/23 20:19:27 martijn Exp $ */ 2 3 /*- 4 * Copyright (c) 1999 James Howard and Dag-Erling Co�dan Sm�rgrav 5 * All rights reserved. 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 29 #include <sys/types.h> 30 #include <sys/stat.h> 31 #include <sys/queue.h> 32 33 #include <ctype.h> 34 #include <err.h> 35 #include <errno.h> 36 #include <getopt.h> 37 #include <regex.h> 38 #include <stdio.h> 39 #include <stdlib.h> 40 #include <string.h> 41 #include <unistd.h> 42 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; 53 fastgrep_t *fg_pattern; 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 Hflag; /* -H: always print filename header */ 64 int Lflag; /* -L: only show names of files with no matches */ 65 int Rflag; /* -R: recursively search directory trees */ 66 int Zflag; /* -Z: decompress input before processing */ 67 int bflag; /* -b: show block numbers for each match */ 68 int cflag; /* -c: only show a count of matching lines */ 69 int hflag; /* -h: don't print filename headers */ 70 int iflag; /* -i: ignore case */ 71 int lflag; /* -l: only show names of files with matches */ 72 int mflag; /* -m x: stop reading the files after x matches */ 73 long long mcount; /* count for -m */ 74 long long mlimit; /* requested value for -m */ 75 int nflag; /* -n: show line numbers in front of matching lines */ 76 int oflag; /* -o: print each match */ 77 int qflag; /* -q: quiet mode (don't output anything) */ 78 int sflag; /* -s: silent mode (ignore errors) */ 79 int vflag; /* -v: only show non-matching lines */ 80 int wflag; /* -w: pattern must start and end on word boundaries */ 81 int xflag; /* -x: pattern must match entire line */ 82 int lbflag; /* --line-buffered */ 83 const char *labelname; /* --label=name */ 84 85 int binbehave = BIN_FILE_BIN; 86 87 enum { 88 BIN_OPT = CHAR_MAX + 1, 89 HELP_OPT, 90 MMAP_OPT, 91 LINEBUF_OPT, 92 LABEL_OPT, 93 }; 94 95 /* Housekeeping */ 96 int first; /* flag whether or not this is our first match */ 97 int tail; /* lines left to print */ 98 int file_err; /* file reading error */ 99 100 struct patfile { 101 const char *pf_file; 102 SLIST_ENTRY(patfile) pf_next; 103 }; 104 SLIST_HEAD(, patfile) patfilelh; 105 106 extern char *__progname; 107 108 static void 109 usage(void) 110 { 111 fprintf(stderr, 112 #ifdef NOZ 113 "usage: %s [-abcEFGHhIiLlnoqRsUVvwx] [-A num] [-B num] [-C[num]]" 114 #else 115 "usage: %s [-abcEFGHhIiLlnoqRsUVvwxZ] [-A num] [-B num] [-C[num]]" 116 #endif 117 " [-e pattern]\n" 118 "\t[-f file] [-m num] [--binary-files=value] [--context[=num]]\n" 119 "\t[--label=name] [--line-buffered] [pattern] [file ...]\n", 120 __progname); 121 exit(2); 122 } 123 124 #ifdef NOZ 125 static const char optstr[] = "0123456789A:B:CEFGHILRUVabce:f:hilm:noqrsuvwxy"; 126 #else 127 static const char optstr[] = "0123456789A:B:CEFGHILRUVZabce:f:hilm:noqrsuvwxy"; 128 #endif 129 130 static const struct option long_options[] = 131 { 132 {"binary-files", required_argument, NULL, BIN_OPT}, 133 {"help", no_argument, NULL, HELP_OPT}, 134 {"mmap", no_argument, NULL, MMAP_OPT}, 135 {"label", required_argument, NULL, LABEL_OPT}, 136 {"line-buffered", no_argument, NULL, LINEBUF_OPT}, 137 {"after-context", required_argument, NULL, 'A'}, 138 {"before-context", required_argument, NULL, 'B'}, 139 {"context", optional_argument, NULL, 'C'}, 140 {"devices", required_argument, NULL, 'D'}, 141 {"extended-regexp", no_argument, NULL, 'E'}, 142 {"fixed-strings", no_argument, NULL, 'F'}, 143 {"basic-regexp", no_argument, NULL, 'G'}, 144 {"with-filename", no_argument, NULL, 'H'}, 145 {"binary", no_argument, NULL, 'U'}, 146 {"version", no_argument, NULL, 'V'}, 147 {"text", no_argument, NULL, 'a'}, 148 {"byte-offset", no_argument, NULL, 'b'}, 149 {"count", no_argument, NULL, 'c'}, 150 {"regexp", required_argument, NULL, 'e'}, 151 {"file", required_argument, NULL, 'f'}, 152 {"no-filename", no_argument, NULL, 'h'}, 153 {"ignore-case", no_argument, NULL, 'i'}, 154 {"files-without-match", no_argument, NULL, 'L'}, 155 {"files-with-matches", no_argument, NULL, 'l'}, 156 {"max-count", required_argument, NULL, 'm'}, 157 {"line-number", no_argument, NULL, 'n'}, 158 {"quiet", no_argument, NULL, 'q'}, 159 {"silent", no_argument, NULL, 'q'}, 160 {"recursive", no_argument, NULL, 'r'}, 161 {"no-messages", no_argument, NULL, 's'}, 162 {"revert-match", no_argument, NULL, 'v'}, 163 {"word-regexp", no_argument, NULL, 'w'}, 164 {"line-regexp", no_argument, NULL, 'x'}, 165 {"unix-byte-offsets", no_argument, NULL, 'u'}, 166 #ifndef NOZ 167 {"decompress", no_argument, NULL, 'Z'}, 168 #endif 169 {NULL, no_argument, NULL, 0} 170 }; 171 172 173 static void 174 add_pattern(char *pat, size_t len) 175 { 176 if (!xflag && (len == 0 || matchall)) { 177 matchall = 1; 178 return; 179 } 180 if (patterns == pattern_sz) { 181 pattern_sz *= 2; 182 pattern = grep_reallocarray(pattern, ++pattern_sz, sizeof(*pattern)); 183 } 184 if (len > 0 && pat[len - 1] == '\n') 185 --len; 186 /* pat may not be NUL-terminated */ 187 if (wflag && !Fflag) { 188 int bol = 0, eol = 0, extra; 189 if (pat[0] == '^') 190 bol = 1; 191 if (len > 0 && pat[len - 1] == '$') 192 eol = 1; 193 extra = Eflag ? 2 : 4; 194 pattern[patterns] = grep_malloc(len + 15 + extra); 195 snprintf(pattern[patterns], len + 15 + extra, 196 "%s[[:<:]]%s%.*s%s[[:>:]]%s", 197 bol ? "^" : "", 198 Eflag ? "(" : "\\(", 199 (int)len - bol - eol, pat + bol, 200 Eflag ? ")" : "\\)", 201 eol ? "$" : ""); 202 len += 14 + extra; 203 } else { 204 pattern[patterns] = grep_malloc(len + 1); 205 memcpy(pattern[patterns], pat, len); 206 pattern[patterns][len] = '\0'; 207 } 208 ++patterns; 209 } 210 211 static void 212 add_patterns(char *pats) 213 { 214 char *nl; 215 216 while ((nl = strchr(pats, '\n')) != NULL) { 217 add_pattern(pats, nl - pats); 218 pats = nl + 1; 219 } 220 add_pattern(pats, strlen(pats)); 221 } 222 223 static void 224 read_patterns(const char *fn) 225 { 226 FILE *f; 227 char *line; 228 ssize_t len; 229 size_t linesize; 230 231 if ((f = fopen(fn, "r")) == NULL) 232 err(2, "%s", fn); 233 line = NULL; 234 linesize = 0; 235 while ((len = getline(&line, &linesize, f)) != -1) 236 add_pattern(line, *line == '\n' ? 0 : len); 237 if (ferror(f)) 238 err(2, "%s", fn); 239 fclose(f); 240 free(line); 241 } 242 243 int 244 main(int argc, char *argv[]) 245 { 246 int c, lastc, prevoptind, newarg, i, needpattern, exprs, expr_sz; 247 struct patfile *patfile, *pf_next; 248 long l; 249 char **expr; 250 const char *errstr; 251 252 if (pledge("stdio rpath", NULL) == -1) 253 err(2, "pledge"); 254 255 SLIST_INIT(&patfilelh); 256 switch (__progname[0]) { 257 case 'e': 258 Eflag = 1; 259 break; 260 case 'f': 261 Fflag = 1; 262 break; 263 #ifndef NOZ 264 case 'z': 265 Zflag = 1; 266 switch(__progname[1]) { 267 case 'e': 268 Eflag = 1; 269 break; 270 case 'f': 271 Fflag = 1; 272 break; 273 } 274 break; 275 #endif 276 } 277 278 lastc = '\0'; 279 newarg = 1; 280 prevoptind = 1; 281 needpattern = 1; 282 expr_sz = exprs = 0; 283 expr = NULL; 284 while ((c = getopt_long(argc, argv, optstr, 285 long_options, NULL)) != -1) { 286 switch (c) { 287 case '0': case '1': case '2': case '3': case '4': 288 case '5': case '6': case '7': case '8': case '9': 289 if (newarg || !isdigit(lastc)) 290 Aflag = 0; 291 else if (Aflag > INT_MAX / 10) 292 errx(2, "context out of range"); 293 Aflag = Bflag = (Aflag * 10) + (c - '0'); 294 break; 295 case 'A': 296 case 'B': 297 l = strtonum(optarg, 1, INT_MAX, &errstr); 298 if (errstr != NULL) 299 errx(2, "context %s", errstr); 300 if (c == 'A') 301 Aflag = (int)l; 302 else 303 Bflag = (int)l; 304 break; 305 case 'C': 306 if (optarg == NULL) 307 Aflag = Bflag = 2; 308 else { 309 l = strtonum(optarg, 1, INT_MAX, &errstr); 310 if (errstr != NULL) 311 errx(2, "context %s", errstr); 312 Aflag = Bflag = (int)l; 313 } 314 break; 315 case 'E': 316 Fflag = 0; 317 Eflag = 1; 318 break; 319 case 'F': 320 Eflag = 0; 321 Fflag = 1; 322 break; 323 case 'G': 324 Eflag = Fflag = 0; 325 break; 326 case 'H': 327 Hflag = 1; 328 break; 329 case 'I': 330 binbehave = BIN_FILE_SKIP; 331 break; 332 case 'L': 333 lflag = 0; 334 Lflag = qflag = 1; 335 break; 336 case 'R': 337 case 'r': 338 Rflag = 1; 339 break; 340 case 'U': 341 binbehave = BIN_FILE_BIN; 342 break; 343 case 'V': 344 fprintf(stderr, "grep version %u.%u\n", VER_MAJ, VER_MIN); 345 exit(0); 346 break; 347 #ifndef NOZ 348 case 'Z': 349 Zflag = 1; 350 break; 351 #endif 352 case 'a': 353 binbehave = BIN_FILE_TEXT; 354 break; 355 case 'b': 356 bflag = 1; 357 break; 358 case 'c': 359 cflag = 1; 360 break; 361 case 'e': 362 /* defer adding of expressions until all arguments are parsed */ 363 if (exprs == expr_sz) { 364 expr_sz *= 2; 365 expr = grep_reallocarray(expr, ++expr_sz, 366 sizeof(*expr)); 367 } 368 needpattern = 0; 369 expr[exprs] = optarg; 370 ++exprs; 371 break; 372 case 'f': 373 patfile = grep_malloc(sizeof(*patfile)); 374 patfile->pf_file = optarg; 375 SLIST_INSERT_HEAD(&patfilelh, patfile, pf_next); 376 needpattern = 0; 377 break; 378 case 'h': 379 hflag = 1; 380 break; 381 case 'i': 382 case 'y': 383 iflag = 1; 384 cflags |= REG_ICASE; 385 break; 386 case 'l': 387 Lflag = 0; 388 lflag = qflag = 1; 389 break; 390 case 'm': 391 mflag = 1; 392 mlimit = mcount = strtonum(optarg, 0, LLONG_MAX, 393 &errstr); 394 if (errstr != NULL) 395 errx(2, "invalid max-count %s: %s", 396 optarg, errstr); 397 break; 398 case 'n': 399 nflag = 1; 400 break; 401 case 'o': 402 oflag = 1; 403 break; 404 case 'q': 405 qflag = 1; 406 break; 407 case 's': 408 sflag = 1; 409 break; 410 case 'v': 411 vflag = 1; 412 break; 413 case 'w': 414 wflag = 1; 415 break; 416 case 'x': 417 xflag = 1; 418 break; 419 case BIN_OPT: 420 if (strcmp("binary", optarg) == 0) 421 binbehave = BIN_FILE_BIN; 422 else if (strcmp("without-match", optarg) == 0) 423 binbehave = BIN_FILE_SKIP; 424 else if (strcmp("text", optarg) == 0) 425 binbehave = BIN_FILE_TEXT; 426 else 427 errx(2, "Unknown binary-files option"); 428 break; 429 case 'u': 430 case MMAP_OPT: 431 /* default, compatibility */ 432 break; 433 case LABEL_OPT: 434 labelname = optarg; 435 break; 436 case LINEBUF_OPT: 437 lbflag = 1; 438 break; 439 case HELP_OPT: 440 default: 441 usage(); 442 } 443 lastc = c; 444 newarg = optind != prevoptind; 445 prevoptind = optind; 446 } 447 argc -= optind; 448 argv += optind; 449 450 for (i = 0; i < exprs; i++) 451 add_patterns(expr[i]); 452 free(expr); 453 expr = NULL; 454 455 for (patfile = SLIST_FIRST(&patfilelh); patfile != NULL; 456 patfile = pf_next) { 457 pf_next = SLIST_NEXT(patfile, pf_next); 458 read_patterns(patfile->pf_file); 459 free(patfile); 460 } 461 462 if (argc == 0 && needpattern) 463 usage(); 464 465 if (argc != 0 && needpattern) { 466 add_patterns(*argv); 467 --argc; 468 ++argv; 469 } 470 if (argc == 1 && strcmp(*argv, "-") == 0) { 471 /* stdin */ 472 --argc; 473 ++argv; 474 } 475 476 if (Eflag) 477 cflags |= REG_EXTENDED; 478 if (Fflag) 479 cflags |= REG_NOSPEC; 480 #ifdef SMALL 481 /* Sorry, this won't work */ 482 if (Fflag && wflag) 483 errx(1, "Can't use small fgrep with -w"); 484 #endif 485 fg_pattern = grep_calloc(patterns, sizeof(*fg_pattern)); 486 r_pattern = grep_calloc(patterns, sizeof(*r_pattern)); 487 for (i = 0; i < patterns; ++i) { 488 /* Check if cheating is allowed (always is for fgrep). */ 489 #ifndef SMALL 490 if (Fflag) { 491 fgrepcomp(&fg_pattern[i], pattern[i]); 492 } else 493 #endif 494 { 495 if (fastcomp(&fg_pattern[i], pattern[i])) { 496 /* Fall back to full regex library */ 497 c = regcomp(&r_pattern[i], pattern[i], cflags); 498 if (c != 0) { 499 regerror(c, &r_pattern[i], re_error, 500 RE_ERROR_BUF); 501 errx(2, "%s", re_error); 502 } 503 } 504 } 505 } 506 507 if (lbflag) 508 setvbuf(stdout, NULL, _IOLBF, 0); 509 510 if ((argc == 0 || argc == 1) && !Rflag && !Hflag) 511 hflag = 1; 512 513 if (argc == 0 && !Rflag) 514 exit(!procfile(NULL)); 515 516 if (Rflag) 517 c = grep_tree(argv); 518 else 519 for (c = 0; argc--; ++argv) 520 c |= procfile(*argv); 521 522 exit(c ? (file_err ? (qflag ? 0 : 2) : 0) : (file_err ? 2 : 1)); 523 } 524