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