1 /* $NetBSD: util.c,v 1.9 2011/02/27 17:33:37 joerg Exp $ */ 2 /* $FreeBSD$ */ 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 #include <sys/cdefs.h> 33 __FBSDID("$FreeBSD$"); 34 35 #include <sys/stat.h> 36 #include <sys/types.h> 37 38 #include <ctype.h> 39 #include <err.h> 40 #include <errno.h> 41 #include <fnmatch.h> 42 #include <fts.h> 43 #include <libgen.h> 44 #include <stdbool.h> 45 #include <stdio.h> 46 #include <stdlib.h> 47 #include <string.h> 48 #include <unistd.h> 49 #include <wchar.h> 50 #include <wctype.h> 51 52 #include "grep.h" 53 54 static int linesqueued; 55 static int procline(struct str *l, int); 56 57 bool 58 file_matching(const char *fname) 59 { 60 char *fname_base; 61 bool ret; 62 63 ret = finclude ? false : true; 64 fname_base = basename(fname); 65 66 for (unsigned int i = 0; i < fpatterns; ++i) { 67 if (fnmatch(fpattern[i].pat, fname, 0) == 0 || 68 fnmatch(fpattern[i].pat, fname_base, 0) == 0) { 69 if (fpattern[i].mode == EXCL_PAT) 70 return (false); 71 else 72 ret = true; 73 } 74 } 75 return (ret); 76 } 77 78 static inline bool 79 dir_matching(const char *dname) 80 { 81 bool ret; 82 83 ret = dinclude ? false : true; 84 85 for (unsigned int i = 0; i < dpatterns; ++i) { 86 if (dname != NULL && 87 fnmatch(dpattern[i].pat, dname, 0) == 0) { 88 if (dpattern[i].mode == EXCL_PAT) 89 return (false); 90 else 91 ret = true; 92 } 93 } 94 return (ret); 95 } 96 97 /* 98 * Processes a directory when a recursive search is performed with 99 * the -R option. Each appropriate file is passed to procfile(). 100 */ 101 int 102 grep_tree(char **argv) 103 { 104 FTS *fts; 105 FTSENT *p; 106 int c, fts_flags; 107 bool ok; 108 109 c = fts_flags = 0; 110 111 switch(linkbehave) { 112 case LINK_EXPLICIT: 113 fts_flags = FTS_COMFOLLOW; 114 break; 115 case LINK_SKIP: 116 fts_flags = FTS_PHYSICAL; 117 break; 118 default: 119 fts_flags = FTS_LOGICAL; 120 121 } 122 123 fts_flags |= FTS_NOSTAT | FTS_NOCHDIR; 124 125 if (!(fts = fts_open(argv, fts_flags, NULL))) 126 err(2, "fts_open"); 127 while ((p = fts_read(fts)) != NULL) { 128 switch (p->fts_info) { 129 case FTS_DNR: 130 /* FALLTHROUGH */ 131 case FTS_ERR: 132 errx(2, "%s: %s", p->fts_path, strerror(p->fts_errno)); 133 break; 134 case FTS_D: 135 /* FALLTHROUGH */ 136 case FTS_DP: 137 if (dexclude || dinclude) 138 if (!dir_matching(p->fts_name) || 139 !dir_matching(p->fts_path)) 140 fts_set(fts, p, FTS_SKIP); 141 break; 142 case FTS_DC: 143 /* Print a warning for recursive directory loop */ 144 warnx("warning: %s: recursive directory loop", 145 p->fts_path); 146 break; 147 default: 148 /* Check for file exclusion/inclusion */ 149 ok = true; 150 if (fexclude || finclude) 151 ok &= file_matching(p->fts_path); 152 153 if (ok) 154 c += procfile(p->fts_path); 155 break; 156 } 157 } 158 159 fts_close(fts); 160 return (c); 161 } 162 163 /* 164 * Opens a file and processes it. Each file is processed line-by-line 165 * passing the lines to procline(). 166 */ 167 int 168 procfile(const char *fn) 169 { 170 struct file *f; 171 struct stat sb; 172 struct str ln; 173 mode_t s; 174 int c, t; 175 176 if (mflag && (mcount <= 0)) 177 return (0); 178 179 if (strcmp(fn, "-") == 0) { 180 fn = label != NULL ? label : getstr(1); 181 f = grep_open(NULL); 182 } else { 183 if (!stat(fn, &sb)) { 184 /* Check if we need to process the file */ 185 s = sb.st_mode & S_IFMT; 186 if (s == S_IFDIR && dirbehave == DIR_SKIP) 187 return (0); 188 if ((s == S_IFIFO || s == S_IFCHR || s == S_IFBLK 189 || s == S_IFSOCK) && devbehave == DEV_SKIP) 190 return (0); 191 } 192 f = grep_open(fn); 193 } 194 if (f == NULL) { 195 if (!sflag) 196 warn("%s", fn); 197 if (errno == ENOENT) 198 notfound = true; 199 return (0); 200 } 201 202 ln.file = grep_malloc(strlen(fn) + 1); 203 strcpy(ln.file, fn); 204 ln.line_no = 0; 205 ln.len = 0; 206 linesqueued = 0; 207 tail = 0; 208 ln.off = -1; 209 210 for (c = 0; c == 0 || !(lflag || qflag); ) { 211 ln.off += ln.len + 1; 212 if ((ln.dat = grep_fgetln(f, &ln.len)) == NULL || ln.len == 0) { 213 if (ln.line_no == 0 && matchall) 214 exit(0); 215 else 216 break; 217 } 218 if (ln.len > 0 && ln.dat[ln.len - 1] == '\n') 219 --ln.len; 220 ln.line_no++; 221 222 /* Return if we need to skip a binary file */ 223 if (f->binary && binbehave == BINFILE_SKIP) { 224 grep_close(f); 225 free(ln.file); 226 free(f); 227 return (0); 228 } 229 /* Process the file line-by-line */ 230 if ((t = procline(&ln, f->binary)) == 0 && Bflag > 0) { 231 enqueue(&ln); 232 linesqueued++; 233 } 234 c += t; 235 236 /* Count the matches if we have a match limit */ 237 if (mflag) { 238 mcount -= t; 239 if (mcount <= 0) 240 break; 241 } 242 } 243 if (Bflag > 0) 244 clearqueue(); 245 grep_close(f); 246 247 if (cflag) { 248 if (!hflag) 249 printf("%s:", ln.file); 250 printf("%u\n", c); 251 } 252 if (lflag && !qflag && c != 0) 253 printf("%s\n", fn); 254 if (Lflag && !qflag && c == 0) 255 printf("%s\n", fn); 256 if (c && !cflag && !lflag && !Lflag && 257 binbehave == BINFILE_BIN && f->binary && !qflag) 258 printf(getstr(8), fn); 259 260 free(ln.file); 261 free(f); 262 return (c); 263 } 264 265 #define iswword(x) (iswalnum((x)) || (x) == L'_') 266 267 /* 268 * Processes a line comparing it with the specified patterns. Each pattern 269 * is looped to be compared along with the full string, saving each and every 270 * match, which is necessary to colorize the output and to count the 271 * matches. The matching lines are passed to printline() to display the 272 * appropriate output. 273 */ 274 static int 275 procline(struct str *l, int nottext) 276 { 277 regmatch_t matches[MAX_LINE_MATCHES]; 278 regmatch_t pmatch; 279 size_t st = 0; 280 unsigned int i; 281 int c = 0, m = 0, r = 0; 282 283 if (!matchall) { 284 /* Loop to process the whole line */ 285 while (st <= l->len) { 286 pmatch.rm_so = st; 287 pmatch.rm_eo = l->len; 288 289 /* Loop to compare with all the patterns */ 290 for (i = 0; i < patterns; i++) { 291 /* 292 * XXX: grep_search() is a workaround for speed up and should be 293 * removed in the future. See fastgrep.c. 294 */ 295 if (fg_pattern[i].pattern) 296 r = grep_search(&fg_pattern[i], 297 (unsigned char *)l->dat, 298 l->len, &pmatch); 299 else 300 r = regexec(&r_pattern[i], l->dat, 1, 301 &pmatch, eflags); 302 r = (r == 0) ? 0 : REG_NOMATCH; 303 st = (cflags & REG_NOSUB) 304 ? (size_t)l->len 305 : (size_t)pmatch.rm_eo; 306 if (r == REG_NOMATCH) 307 continue; 308 /* Check for full match */ 309 if (r == 0 && xflag) 310 if (pmatch.rm_so != 0 || 311 (size_t)pmatch.rm_eo != l->len) 312 r = REG_NOMATCH; 313 /* Check for whole word match */ 314 if (r == 0 && (wflag || fg_pattern[i].word)) { 315 wint_t wbegin, wend; 316 317 wbegin = wend = L' '; 318 if (pmatch.rm_so != 0 && 319 sscanf(&l->dat[pmatch.rm_so - 1], 320 "%lc", &wbegin) != 1) 321 r = REG_NOMATCH; 322 else if ((size_t)pmatch.rm_eo != 323 l->len && 324 sscanf(&l->dat[pmatch.rm_eo], 325 "%lc", &wend) != 1) 326 r = REG_NOMATCH; 327 else if (iswword(wbegin) || 328 iswword(wend)) 329 r = REG_NOMATCH; 330 } 331 if (r == 0) { 332 if (m == 0) 333 c++; 334 if (m < MAX_LINE_MATCHES) 335 matches[m++] = pmatch; 336 /* matches - skip further patterns */ 337 if ((color == NULL && !oflag) || 338 qflag || lflag) 339 break; 340 } 341 } 342 343 if (vflag) { 344 c = !c; 345 break; 346 } 347 /* One pass if we are not recording matches */ 348 if ((color == NULL && !oflag) || qflag || lflag) 349 break; 350 351 if (st == (size_t)pmatch.rm_so) 352 break; /* No matches */ 353 } 354 } else 355 c = !vflag; 356 357 if (c && binbehave == BINFILE_BIN && nottext) 358 return (c); /* Binary file */ 359 360 /* Dealing with the context */ 361 if ((tail || c) && !cflag && !qflag && !lflag && !Lflag) { 362 if (c) { 363 if (!first && !prev && !tail && Aflag) 364 printf("--\n"); 365 tail = Aflag; 366 if (Bflag > 0) { 367 if (!first && !prev) 368 printf("--\n"); 369 printqueue(); 370 } 371 linesqueued = 0; 372 printline(l, ':', matches, m); 373 } else { 374 printline(l, '-', matches, m); 375 tail--; 376 } 377 } 378 379 if (c) { 380 prev = true; 381 first = false; 382 } else 383 prev = false; 384 385 return (c); 386 } 387 388 /* 389 * Safe malloc() for internal use. 390 */ 391 void * 392 grep_malloc(size_t size) 393 { 394 void *ptr; 395 396 if ((ptr = malloc(size)) == NULL) 397 err(2, "malloc"); 398 return (ptr); 399 } 400 401 /* 402 * Safe calloc() for internal use. 403 */ 404 void * 405 grep_calloc(size_t nmemb, size_t size) 406 { 407 void *ptr; 408 409 if ((ptr = calloc(nmemb, size)) == NULL) 410 err(2, "calloc"); 411 return (ptr); 412 } 413 414 /* 415 * Safe realloc() for internal use. 416 */ 417 void * 418 grep_realloc(void *ptr, size_t size) 419 { 420 421 if ((ptr = realloc(ptr, size)) == NULL) 422 err(2, "realloc"); 423 return (ptr); 424 } 425 426 /* 427 * Safe strdup() for internal use. 428 */ 429 char * 430 grep_strdup(const char *str) 431 { 432 char *ret; 433 434 if ((ret = strdup(str)) == NULL) 435 err(2, "strdup"); 436 return (ret); 437 } 438 439 /* 440 * Prints a matching line according to the command line options. 441 */ 442 void 443 printline(struct str *line, int sep, regmatch_t *matches, int m) 444 { 445 size_t a = 0; 446 int i, n = 0; 447 448 if (!hflag) { 449 if (nullflag == 0) 450 fputs(line->file, stdout); 451 else { 452 printf("%s", line->file); 453 putchar(0); 454 } 455 ++n; 456 } 457 if (nflag) { 458 if (n > 0) 459 putchar(sep); 460 printf("%d", line->line_no); 461 ++n; 462 } 463 if (bflag) { 464 if (n > 0) 465 putchar(sep); 466 printf("%lld", (long long)line->off); 467 ++n; 468 } 469 if (n) 470 putchar(sep); 471 /* --color and -o */ 472 if ((oflag || color) && m > 0) { 473 for (i = 0; i < m; i++) { 474 if (!oflag) 475 fwrite(line->dat + a, matches[i].rm_so - a, 1, 476 stdout); 477 if (color) 478 fprintf(stdout, "\33[%sm\33[K", color); 479 480 fwrite(line->dat + matches[i].rm_so, 481 matches[i].rm_eo - matches[i].rm_so, 1, 482 stdout); 483 if (color) 484 fprintf(stdout, "\33[m\33[K"); 485 a = matches[i].rm_eo; 486 if (oflag) 487 putchar('\n'); 488 } 489 if (!oflag) { 490 if (line->len - a > 0) 491 fwrite(line->dat + a, line->len - a, 1, stdout); 492 putchar('\n'); 493 } 494 } else { 495 fwrite(line->dat, line->len, 1, stdout); 496 putchar('\n'); 497 } 498 } 499