1 /* $NetBSD: util.c,v 1.13 2011/04/18 23:22:42 joerg Exp $ */ 2 /* $FreeBSD: head/usr.bin/grep/util.c 211496 2010-08-19 09:28:59Z des $ */ 3 /* $OpenBSD: util.c,v 1.39 2010/07/02 22:18:03 tedu Exp $ */ 4 5 /*- 6 * Copyright (c) 1999 James Howard and Dag-Erling Coïdan Smørgrav 7 * Copyright (C) 2008-2010 Gabor Kovesdan <gabor@FreeBSD.org> 8 * All rights reserved. 9 * 10 * Redistribution and use in source and binary forms, with or without 11 * modification, are permitted provided that the following conditions 12 * are met: 13 * 1. Redistributions of source code must retain the above copyright 14 * notice, this list of conditions and the following disclaimer. 15 * 2. Redistributions in binary form must reproduce the above copyright 16 * notice, this list of conditions and the following disclaimer in the 17 * documentation and/or other materials provided with the distribution. 18 * 19 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND 20 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 21 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 22 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE 23 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 24 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 25 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 26 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 27 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 28 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 29 * SUCH DAMAGE. 30 */ 31 32 #if HAVE_NBTOOL_CONFIG_H 33 #include "nbtool_config.h" 34 #endif 35 36 #include <sys/cdefs.h> 37 __RCSID("$NetBSD: util.c,v 1.13 2011/04/18 23:22:42 joerg Exp $"); 38 39 #include <sys/stat.h> 40 #include <sys/types.h> 41 42 #include <ctype.h> 43 #include <err.h> 44 #include <errno.h> 45 #include <fnmatch.h> 46 #include <fts.h> 47 #include <libgen.h> 48 #include <stdbool.h> 49 #include <stdio.h> 50 #include <stdlib.h> 51 #include <string.h> 52 #include <unistd.h> 53 #include <wchar.h> 54 #include <wctype.h> 55 56 #include "grep.h" 57 58 static bool first, first_global = true; 59 static unsigned long long since_printed; 60 61 static int procline(struct str *l, int); 62 63 bool 64 file_matching(const char *fname) 65 { 66 char *fname_base, *fname_copy; 67 unsigned int i; 68 bool ret; 69 70 ret = finclude ? false : true; 71 fname_copy = grep_strdup(fname); 72 fname_base = basename(fname_copy); 73 74 for (i = 0; i < fpatterns; ++i) { 75 if (fnmatch(fpattern[i].pat, fname, 0) == 0 || 76 fnmatch(fpattern[i].pat, fname_base, 0) == 0) { 77 if (fpattern[i].mode == EXCL_PAT) 78 return (false); 79 else 80 ret = true; 81 } 82 } 83 free(fname_copy); 84 return (ret); 85 } 86 87 static inline bool 88 dir_matching(const char *dname) 89 { 90 unsigned int i; 91 bool ret; 92 93 ret = dinclude ? false : true; 94 95 for (i = 0; i < dpatterns; ++i) { 96 if (dname != NULL && 97 fnmatch(dname, dpattern[i].pat, 0) == 0) { 98 if (dpattern[i].mode == EXCL_PAT) 99 return (false); 100 else 101 ret = true; 102 } 103 } 104 return (ret); 105 } 106 107 /* 108 * Processes a directory when a recursive search is performed with 109 * the -R option. Each appropriate file is passed to procfile(). 110 */ 111 int 112 grep_tree(char **argv) 113 { 114 FTS *fts; 115 FTSENT *p; 116 char *d, *dir = NULL; 117 int c, fts_flags; 118 bool ok; 119 120 c = fts_flags = 0; 121 122 switch(linkbehave) { 123 case LINK_EXPLICIT: 124 fts_flags = FTS_COMFOLLOW; 125 break; 126 case LINK_SKIP: 127 fts_flags = FTS_PHYSICAL; 128 break; 129 default: 130 fts_flags = FTS_LOGICAL; 131 132 } 133 134 fts_flags |= FTS_NOSTAT | FTS_NOCHDIR; 135 136 if (!(fts = fts_open(argv, fts_flags, NULL))) 137 err(2, "fts_open"); 138 while ((p = fts_read(fts)) != NULL) { 139 switch (p->fts_info) { 140 case FTS_DNR: 141 /* FALLTHROUGH */ 142 case FTS_ERR: 143 errx(2, "%s: %s", p->fts_path, strerror(p->fts_errno)); 144 break; 145 case FTS_D: 146 /* FALLTHROUGH */ 147 case FTS_DP: 148 break; 149 case FTS_DC: 150 /* Print a warning for recursive directory loop */ 151 warnx("warning: %s: recursive directory loop", 152 p->fts_path); 153 break; 154 default: 155 /* Check for file exclusion/inclusion */ 156 ok = true; 157 if (dexclude || dinclude) { 158 if ((d = strrchr(p->fts_path, '/')) != NULL) { 159 dir = grep_malloc(sizeof(char) * 160 (d - p->fts_path + 1)); 161 memcpy(dir, p->fts_path, 162 d - p->fts_path); 163 dir[d - p->fts_path] = '\0'; 164 } 165 ok = dir_matching(dir); 166 free(dir); 167 dir = NULL; 168 } 169 if (fexclude || finclude) 170 ok &= file_matching(p->fts_path); 171 172 if (ok) 173 c += procfile(p->fts_path); 174 break; 175 } 176 } 177 178 fts_close(fts); 179 return (c); 180 } 181 182 /* 183 * Opens a file and processes it. Each file is processed line-by-line 184 * passing the lines to procline(). 185 */ 186 int 187 procfile(const char *fn) 188 { 189 struct file *f; 190 struct stat sb; 191 struct str ln; 192 mode_t s; 193 int c, t; 194 195 if (mflag && (mcount <= 0)) 196 return (0); 197 198 if (strcmp(fn, "-") == 0) { 199 fn = label != NULL ? label : getstr(1); 200 f = grep_open(NULL); 201 } else { 202 if (!stat(fn, &sb)) { 203 /* Check if we need to process the file */ 204 s = sb.st_mode & S_IFMT; 205 if (s == S_IFDIR && dirbehave == DIR_SKIP) 206 return (0); 207 if ((s == S_IFIFO || s == S_IFCHR || s == S_IFBLK 208 || s == S_IFSOCK) && devbehave == DEV_SKIP) 209 return (0); 210 } 211 f = grep_open(fn); 212 } 213 if (f == NULL) { 214 if (!sflag) 215 warn("%s", fn); 216 if (errno == ENOENT) 217 notfound = true; 218 return (0); 219 } 220 221 ln.file = grep_malloc(strlen(fn) + 1); 222 strcpy(ln.file, fn); 223 ln.line_no = 0; 224 ln.len = 0; 225 tail = 0; 226 ln.off = -1; 227 228 for (first = true, c = 0; c == 0 || !(lflag || qflag); ) { 229 ln.off += ln.len + 1; 230 if ((ln.dat = grep_fgetln(f, &ln.len)) == NULL || ln.len == 0) { 231 if (ln.line_no == 0 && matchall) 232 exit(0); 233 else 234 break; 235 } 236 if (ln.len > 0 && ln.dat[ln.len - 1] == line_sep) 237 --ln.len; 238 ln.line_no++; 239 240 /* Return if we need to skip a binary file */ 241 if (f->binary && binbehave == BINFILE_SKIP) { 242 grep_close(f); 243 free(ln.file); 244 free(f); 245 return (0); 246 } 247 /* Process the file line-by-line */ 248 t = procline(&ln, f->binary); 249 c += t; 250 251 /* Count the matches if we have a match limit */ 252 if (mflag) { 253 mcount -= t; 254 if (mcount <= 0) 255 break; 256 } 257 } 258 if (Bflag > 0) 259 clearqueue(); 260 grep_close(f); 261 262 if (cflag) { 263 if (!hflag) 264 printf("%s:", ln.file); 265 printf("%u%c", c, line_sep); 266 } 267 if (lflag && !qflag && c != 0) 268 printf("%s%c", fn, line_sep); 269 if (Lflag && !qflag && c == 0) 270 printf("%s%c", fn, line_sep); 271 if (c && !cflag && !lflag && !Lflag && 272 binbehave == BINFILE_BIN && f->binary && !qflag) 273 printf(getstr(8), fn); 274 275 free(ln.file); 276 free(f); 277 return (c); 278 } 279 280 #define iswword(x) (iswalnum((x)) || (x) == L'_') 281 282 /* 283 * Processes a line comparing it with the specified patterns. Each pattern 284 * is looped to be compared along with the full string, saving each and every 285 * match, which is necessary to colorize the output and to count the 286 * matches. The matching lines are passed to printline() to display the 287 * appropriate output. 288 */ 289 static int 290 procline(struct str *l, int nottext) 291 { 292 regmatch_t matches[MAX_LINE_MATCHES]; 293 regmatch_t pmatch; 294 size_t st = 0; 295 unsigned int i; 296 int c = 0, m = 0, r = 0; 297 298 if (!matchall) { 299 /* Loop to process the whole line */ 300 while (st <= l->len) { 301 pmatch.rm_so = st; 302 pmatch.rm_eo = l->len; 303 304 /* Loop to compare with all the patterns */ 305 for (i = 0; i < patterns; i++) { 306 /* 307 * XXX: grep_search() is a workaround for speed up and should be 308 * removed in the future. See fastgrep.c. 309 */ 310 if (fg_pattern[i].pattern) { 311 r = grep_search(&fg_pattern[i], 312 (unsigned char *)l->dat, 313 l->len, &pmatch); 314 r = (r == 0) ? 0 : REG_NOMATCH; 315 st = pmatch.rm_eo; 316 } else { 317 r = regexec(&r_pattern[i], l->dat, 1, 318 &pmatch, eflags); 319 r = (r == 0) ? 0 : REG_NOMATCH; 320 st = pmatch.rm_eo; 321 } 322 if (r == REG_NOMATCH) 323 continue; 324 /* Check for full match */ 325 if (r == 0 && xflag) 326 if (pmatch.rm_so != 0 || 327 (size_t)pmatch.rm_eo != l->len) 328 r = REG_NOMATCH; 329 /* Check for whole word match */ 330 if (r == 0 && fg_pattern[i].word && 331 pmatch.rm_so != 0) { 332 wint_t wbegin, wend; 333 334 wbegin = wend = L' '; 335 if (pmatch.rm_so != 0 && 336 sscanf(&l->dat[pmatch.rm_so - 1], 337 "%lc", &wbegin) != 1) 338 r = REG_NOMATCH; 339 else if ((size_t)pmatch.rm_eo != l->len && 340 sscanf(&l->dat[pmatch.rm_eo], 341 "%lc", &wend) != 1) 342 r = REG_NOMATCH; 343 else if (iswword(wbegin) || iswword(wend)) 344 r = REG_NOMATCH; 345 } 346 if (r == 0) { 347 if (m == 0) 348 c++; 349 if (m < MAX_LINE_MATCHES) 350 matches[m++] = pmatch; 351 /* matches - skip further patterns */ 352 if ((color != NULL && !oflag) || qflag || lflag) 353 break; 354 } 355 } 356 357 if (vflag) { 358 c = !c; 359 break; 360 } 361 /* One pass if we are not recording matches */ 362 if ((color != NULL && !oflag) || qflag || lflag) 363 break; 364 365 if (st == (size_t)pmatch.rm_so) 366 break; /* No matches */ 367 } 368 } else 369 c = !vflag; 370 371 if (c && binbehave == BINFILE_BIN && nottext) 372 return (c); /* Binary file */ 373 374 /* Dealing with the context */ 375 if ((tail || c) && !cflag && !qflag && !lflag && !Lflag) { 376 if (c) { 377 if ((Aflag || Bflag) && !first_global && 378 (first || since_printed > Bflag)) 379 printf("--\n"); 380 tail = Aflag; 381 if (Bflag > 0) 382 printqueue(); 383 printline(l, ':', matches, m); 384 } else { 385 printline(l, '-', matches, m); 386 tail--; 387 } 388 first = false; 389 first_global = false; 390 since_printed = 0; 391 } else { 392 if (Bflag) 393 enqueue(l); 394 since_printed++; 395 } 396 return (c); 397 } 398 399 /* 400 * Safe malloc() for internal use. 401 */ 402 void * 403 grep_malloc(size_t size) 404 { 405 void *ptr; 406 407 if ((ptr = malloc(size)) == NULL) 408 err(2, "malloc"); 409 return (ptr); 410 } 411 412 /* 413 * Safe calloc() for internal use. 414 */ 415 void * 416 grep_calloc(size_t nmemb, size_t size) 417 { 418 void *ptr; 419 420 if ((ptr = calloc(nmemb, size)) == NULL) 421 err(2, "calloc"); 422 return (ptr); 423 } 424 425 /* 426 * Safe realloc() for internal use. 427 */ 428 void * 429 grep_realloc(void *ptr, size_t size) 430 { 431 432 if ((ptr = realloc(ptr, size)) == NULL) 433 err(2, "realloc"); 434 return (ptr); 435 } 436 437 /* 438 * Safe strdup() for internal use. 439 */ 440 char * 441 grep_strdup(const char *str) 442 { 443 char *ret; 444 445 if ((ret = strdup(str)) == NULL) 446 err(2, "strdup"); 447 return (ret); 448 } 449 450 /* 451 * Prints a matching line according to the command line options. 452 */ 453 void 454 printline(struct str *line, int sep, regmatch_t *matches, int m) 455 { 456 size_t a = 0; 457 int i, n = 0; 458 459 if (!hflag) { 460 if (nullflag == 0) 461 fputs(line->file, stdout); 462 else { 463 printf("%s", line->file); 464 putchar(0); 465 } 466 ++n; 467 } 468 if (nflag) { 469 if (n > 0) 470 putchar(sep); 471 printf("%d", line->line_no); 472 ++n; 473 } 474 if (bflag) { 475 if (n > 0) 476 putchar(sep); 477 printf("%lld", (long long)line->off); 478 ++n; 479 } 480 if (n) 481 putchar(sep); 482 /* --color and -o */ 483 if ((oflag || color) && m > 0) { 484 for (i = 0; i < m; i++) { 485 if (!oflag) 486 fwrite(line->dat + a, matches[i].rm_so - a, 1, 487 stdout); 488 if (color) 489 fprintf(stdout, "\33[%sm\33[K", color); 490 491 fwrite(line->dat + matches[i].rm_so, 492 matches[i].rm_eo - matches[i].rm_so, 1, 493 stdout); 494 if (color) 495 fprintf(stdout, "\33[m\33[K"); 496 a = matches[i].rm_eo; 497 if (oflag) 498 putchar('\n'); 499 } 500 if (!oflag) { 501 if (line->len - a > 0) 502 fwrite(line->dat + a, line->len - a, 1, stdout); 503 putchar(line_sep); 504 } 505 } else { 506 fwrite(line->dat, line->len, 1, stdout); 507 putchar(line_sep); 508 } 509 } 510