1 /* $OpenBSD: diff.c,v 1.67 2019/06/28 13:35:00 deraadt Exp $ */ 2 3 /* 4 * Copyright (c) 2003 Todd C. Miller <Todd.Miller@courtesan.com> 5 * 6 * Permission to use, copy, modify, and distribute this software for any 7 * purpose with or without fee is hereby granted, provided that the above 8 * copyright notice and this permission notice appear in all copies. 9 * 10 * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES 11 * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF 12 * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR 13 * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES 14 * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN 15 * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF 16 * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. 17 * 18 * Sponsored in part by the Defense Advanced Research Projects 19 * Agency (DARPA) and Air Force Research Laboratory, Air Force 20 * Materiel Command, USAF, under agreement number F39502-99-1-0512. 21 */ 22 23 #include <sys/cdefs.h> 24 __FBSDID("$FreeBSD$"); 25 26 #include <sys/stat.h> 27 28 #include <ctype.h> 29 #include <err.h> 30 #include <errno.h> 31 #include <getopt.h> 32 #include <stdlib.h> 33 #include <stdio.h> 34 #include <string.h> 35 #include <unistd.h> 36 #include <limits.h> 37 38 #include "diff.h" 39 #include "xmalloc.h" 40 41 bool lflag, Nflag, Pflag, rflag, sflag, Tflag, cflag; 42 bool ignore_file_case, suppress_common, color, noderef; 43 int diff_format, diff_context, status; 44 int tabsize = 8, width = 130; 45 static int colorflag = COLORFLAG_NEVER; 46 char *start, *ifdefname, *diffargs, *label[2]; 47 char *ignore_pats, *most_recent_pat; 48 char *group_format = NULL; 49 const char *add_code, *del_code; 50 struct stat stb1, stb2; 51 struct excludes *excludes_list; 52 regex_t ignore_re, most_recent_re; 53 54 #define OPTIONS "0123456789aBbC:cdD:efF:HhI:iL:lnNPpqrS:sTtU:uwW:X:x:y" 55 enum { 56 OPT_TSIZE = CHAR_MAX + 1, 57 OPT_STRIPCR, 58 OPT_IGN_FN_CASE, 59 OPT_NO_IGN_FN_CASE, 60 OPT_NORMAL, 61 OPT_HORIZON_LINES, 62 OPT_CHANGED_GROUP_FORMAT, 63 OPT_SUPPRESS_COMMON, 64 OPT_COLOR, 65 OPT_NO_DEREFERENCE, 66 }; 67 68 static struct option longopts[] = { 69 { "text", no_argument, 0, 'a' }, 70 { "ignore-space-change", no_argument, 0, 'b' }, 71 { "context", optional_argument, 0, 'C' }, 72 { "ifdef", required_argument, 0, 'D' }, 73 { "minimal", no_argument, 0, 'd' }, 74 { "ed", no_argument, 0, 'e' }, 75 { "forward-ed", no_argument, 0, 'f' }, 76 { "show-function-line", required_argument, 0, 'F' }, 77 { "speed-large-files", no_argument, NULL, 'H' }, 78 { "ignore-blank-lines", no_argument, 0, 'B' }, 79 { "ignore-matching-lines", required_argument, 0, 'I' }, 80 { "ignore-case", no_argument, 0, 'i' }, 81 { "paginate", no_argument, NULL, 'l' }, 82 { "label", required_argument, 0, 'L' }, 83 { "new-file", no_argument, 0, 'N' }, 84 { "rcs", no_argument, 0, 'n' }, 85 { "unidirectional-new-file", no_argument, 0, 'P' }, 86 { "show-c-function", no_argument, 0, 'p' }, 87 { "brief", no_argument, 0, 'q' }, 88 { "recursive", no_argument, 0, 'r' }, 89 { "report-identical-files", no_argument, 0, 's' }, 90 { "starting-file", required_argument, 0, 'S' }, 91 { "expand-tabs", no_argument, 0, 't' }, 92 { "initial-tab", no_argument, 0, 'T' }, 93 { "unified", optional_argument, 0, 'U' }, 94 { "ignore-all-space", no_argument, 0, 'w' }, 95 { "width", required_argument, 0, 'W' }, 96 { "exclude", required_argument, 0, 'x' }, 97 { "exclude-from", required_argument, 0, 'X' }, 98 { "side-by-side", no_argument, NULL, 'y' }, 99 { "ignore-file-name-case", no_argument, NULL, OPT_IGN_FN_CASE }, 100 { "horizon-lines", required_argument, NULL, OPT_HORIZON_LINES }, 101 { "no-dereference", no_argument, NULL, OPT_NO_DEREFERENCE}, 102 { "no-ignore-file-name-case", no_argument, NULL, OPT_NO_IGN_FN_CASE }, 103 { "normal", no_argument, NULL, OPT_NORMAL }, 104 { "strip-trailing-cr", no_argument, NULL, OPT_STRIPCR }, 105 { "tabsize", required_argument, NULL, OPT_TSIZE }, 106 { "changed-group-format", required_argument, NULL, OPT_CHANGED_GROUP_FORMAT}, 107 { "suppress-common-lines", no_argument, NULL, OPT_SUPPRESS_COMMON }, 108 { "color", optional_argument, NULL, OPT_COLOR }, 109 { NULL, 0, 0, '\0'} 110 }; 111 112 static void checked_regcomp(char const *, regex_t *); 113 static void usage(void) __dead2; 114 static void conflicting_format(void) __dead2; 115 static void push_excludes(char *); 116 static void push_ignore_pats(char *); 117 static void read_excludes_file(char *file); 118 static void set_argstr(char **, char **); 119 static char *splice(char *, char *); 120 static bool do_color(void); 121 122 int 123 main(int argc, char **argv) 124 { 125 const char *errstr = NULL; 126 char *ep, **oargv; 127 long l; 128 int ch, dflags, lastch, gotstdin, prevoptind, newarg; 129 130 oargv = argv; 131 gotstdin = 0; 132 dflags = 0; 133 lastch = '\0'; 134 prevoptind = 1; 135 newarg = 1; 136 diff_context = 3; 137 diff_format = D_UNSET; 138 #define FORMAT_MISMATCHED(type) \ 139 (diff_format != D_UNSET && diff_format != (type)) 140 while ((ch = getopt_long(argc, argv, OPTIONS, longopts, NULL)) != -1) { 141 switch (ch) { 142 case '0': case '1': case '2': case '3': case '4': 143 case '5': case '6': case '7': case '8': case '9': 144 if (newarg) 145 usage(); /* disallow -[0-9]+ */ 146 else if (lastch == 'c' || lastch == 'u') 147 diff_context = 0; 148 else if (!isdigit(lastch) || diff_context > INT_MAX / 10) 149 usage(); 150 diff_context = (diff_context * 10) + (ch - '0'); 151 break; 152 case 'a': 153 dflags |= D_FORCEASCII; 154 break; 155 case 'b': 156 dflags |= D_FOLDBLANKS; 157 break; 158 case 'C': 159 case 'c': 160 if (FORMAT_MISMATCHED(D_CONTEXT)) 161 conflicting_format(); 162 cflag = true; 163 diff_format = D_CONTEXT; 164 if (optarg != NULL) { 165 l = strtol(optarg, &ep, 10); 166 if (*ep != '\0' || l < 0 || l >= INT_MAX) 167 usage(); 168 diff_context = (int)l; 169 } 170 break; 171 case 'd': 172 dflags |= D_MINIMAL; 173 break; 174 case 'D': 175 if (FORMAT_MISMATCHED(D_IFDEF)) 176 conflicting_format(); 177 diff_format = D_IFDEF; 178 ifdefname = optarg; 179 break; 180 case 'e': 181 if (FORMAT_MISMATCHED(D_EDIT)) 182 conflicting_format(); 183 diff_format = D_EDIT; 184 break; 185 case 'f': 186 if (FORMAT_MISMATCHED(D_REVERSE)) 187 conflicting_format(); 188 diff_format = D_REVERSE; 189 break; 190 case 'H': 191 /* ignore but needed for compatibility with GNU diff */ 192 break; 193 case 'h': 194 /* silently ignore for backwards compatibility */ 195 break; 196 case 'B': 197 dflags |= D_SKIPBLANKLINES; 198 break; 199 case 'F': 200 if (dflags & D_PROTOTYPE) 201 conflicting_format(); 202 dflags |= D_MATCHLAST; 203 most_recent_pat = xstrdup(optarg); 204 break; 205 case 'I': 206 push_ignore_pats(optarg); 207 break; 208 case 'i': 209 dflags |= D_IGNORECASE; 210 break; 211 case 'L': 212 if (label[0] == NULL) 213 label[0] = optarg; 214 else if (label[1] == NULL) 215 label[1] = optarg; 216 else 217 usage(); 218 break; 219 case 'l': 220 lflag = true; 221 break; 222 case 'N': 223 Nflag = true; 224 break; 225 case 'n': 226 if (FORMAT_MISMATCHED(D_NREVERSE)) 227 conflicting_format(); 228 diff_format = D_NREVERSE; 229 break; 230 case 'p': 231 if (dflags & D_MATCHLAST) 232 conflicting_format(); 233 dflags |= D_PROTOTYPE; 234 break; 235 case 'P': 236 Pflag = true; 237 break; 238 case 'r': 239 rflag = true; 240 break; 241 case 'q': 242 if (FORMAT_MISMATCHED(D_BRIEF)) 243 conflicting_format(); 244 diff_format = D_BRIEF; 245 break; 246 case 'S': 247 start = optarg; 248 break; 249 case 's': 250 sflag = true; 251 break; 252 case 'T': 253 Tflag = true; 254 break; 255 case 't': 256 dflags |= D_EXPANDTABS; 257 break; 258 case 'U': 259 case 'u': 260 if (FORMAT_MISMATCHED(D_UNIFIED)) 261 conflicting_format(); 262 diff_format = D_UNIFIED; 263 if (optarg != NULL) { 264 l = strtol(optarg, &ep, 10); 265 if (*ep != '\0' || l < 0 || l >= INT_MAX) 266 usage(); 267 diff_context = (int)l; 268 } 269 break; 270 case 'w': 271 dflags |= D_IGNOREBLANKS; 272 break; 273 case 'W': 274 width = (int) strtonum(optarg, 1, INT_MAX, &errstr); 275 if (errstr) { 276 warnx("Invalid argument for width"); 277 usage(); 278 } 279 break; 280 case 'X': 281 read_excludes_file(optarg); 282 break; 283 case 'x': 284 push_excludes(optarg); 285 break; 286 case 'y': 287 if (FORMAT_MISMATCHED(D_SIDEBYSIDE)) 288 conflicting_format(); 289 diff_format = D_SIDEBYSIDE; 290 break; 291 case OPT_CHANGED_GROUP_FORMAT: 292 if (FORMAT_MISMATCHED(D_GFORMAT)) 293 conflicting_format(); 294 diff_format = D_GFORMAT; 295 group_format = optarg; 296 break; 297 case OPT_HORIZON_LINES: 298 break; /* XXX TODO for compatibility with GNU diff3 */ 299 case OPT_IGN_FN_CASE: 300 ignore_file_case = true; 301 break; 302 case OPT_NO_IGN_FN_CASE: 303 ignore_file_case = false; 304 break; 305 case OPT_NORMAL: 306 if (FORMAT_MISMATCHED(D_NORMAL)) 307 conflicting_format(); 308 diff_format = D_NORMAL; 309 break; 310 case OPT_TSIZE: 311 tabsize = (int) strtonum(optarg, 1, INT_MAX, &errstr); 312 if (errstr) { 313 warnx("Invalid argument for tabsize"); 314 usage(); 315 } 316 break; 317 case OPT_STRIPCR: 318 dflags |= D_STRIPCR; 319 break; 320 case OPT_SUPPRESS_COMMON: 321 suppress_common = 1; 322 break; 323 case OPT_COLOR: 324 if (optarg == NULL || strncmp(optarg, "auto", 4) == 0) 325 colorflag = COLORFLAG_AUTO; 326 else if (strncmp(optarg, "always", 6) == 0) 327 colorflag = COLORFLAG_ALWAYS; 328 else if (strncmp(optarg, "never", 5) == 0) 329 colorflag = COLORFLAG_NEVER; 330 else 331 errx(2, "unsupported --color value '%s' (must be always, auto, or never)", 332 optarg); 333 break; 334 case OPT_NO_DEREFERENCE: 335 rflag = true; 336 noderef = true; 337 break; 338 default: 339 usage(); 340 break; 341 } 342 lastch = ch; 343 newarg = optind != prevoptind; 344 prevoptind = optind; 345 } 346 if (diff_format == D_UNSET && (dflags & D_PROTOTYPE) != 0) 347 diff_format = D_CONTEXT; 348 if (diff_format == D_UNSET) 349 diff_format = D_NORMAL; 350 argc -= optind; 351 argv += optind; 352 353 if (do_color()) { 354 char *p; 355 const char *env; 356 357 color = true; 358 add_code = "32"; 359 del_code = "31"; 360 env = getenv("DIFFCOLORS"); 361 if (env != NULL && *env != '\0' && (p = strdup(env))) { 362 add_code = p; 363 strsep(&p, ":"); 364 if (p != NULL) 365 del_code = p; 366 } 367 } 368 369 #ifdef __OpenBSD__ 370 if (pledge("stdio rpath tmppath", NULL) == -1) 371 err(2, "pledge"); 372 #endif 373 374 /* 375 * Do sanity checks, fill in stb1 and stb2 and call the appropriate 376 * driver routine. Both drivers use the contents of stb1 and stb2. 377 */ 378 if (argc != 2) 379 usage(); 380 checked_regcomp(ignore_pats, &ignore_re); 381 checked_regcomp(most_recent_pat, &most_recent_re); 382 if (strcmp(argv[0], "-") == 0) { 383 fstat(STDIN_FILENO, &stb1); 384 gotstdin = 1; 385 } else if (stat(argv[0], &stb1) != 0) { 386 if (!Nflag || errno != ENOENT) 387 err(2, "%s", argv[0]); 388 dflags |= D_EMPTY1; 389 memset(&stb1, 0, sizeof(struct stat)); 390 } 391 392 if (strcmp(argv[1], "-") == 0) { 393 fstat(STDIN_FILENO, &stb2); 394 gotstdin = 1; 395 } else if (stat(argv[1], &stb2) != 0) { 396 if (!Nflag || errno != ENOENT) 397 err(2, "%s", argv[1]); 398 dflags |= D_EMPTY2; 399 memset(&stb2, 0, sizeof(stb2)); 400 stb2.st_mode = stb1.st_mode; 401 } 402 403 if (dflags & D_EMPTY1 && dflags & D_EMPTY2){ 404 warn("%s", argv[0]); 405 warn("%s", argv[1]); 406 exit(2); 407 } 408 409 if (stb1.st_mode == 0) 410 stb1.st_mode = stb2.st_mode; 411 412 if (gotstdin && (S_ISDIR(stb1.st_mode) || S_ISDIR(stb2.st_mode))) 413 errx(2, "can't compare - to a directory"); 414 set_argstr(oargv, argv); 415 if (S_ISDIR(stb1.st_mode) && S_ISDIR(stb2.st_mode)) { 416 if (diff_format == D_IFDEF) 417 errx(2, "-D option not supported with directories"); 418 diffdir(argv[0], argv[1], dflags); 419 } else { 420 if (S_ISDIR(stb1.st_mode)) { 421 argv[0] = splice(argv[0], argv[1]); 422 if (stat(argv[0], &stb1) == -1) 423 err(2, "%s", argv[0]); 424 } 425 if (S_ISDIR(stb2.st_mode)) { 426 argv[1] = splice(argv[1], argv[0]); 427 if (stat(argv[1], &stb2) == -1) 428 err(2, "%s", argv[1]); 429 } 430 print_status(diffreg(argv[0], argv[1], dflags, 1), argv[0], 431 argv[1], ""); 432 } 433 exit(status); 434 } 435 436 static void 437 checked_regcomp(char const *pattern, regex_t *comp) 438 { 439 char buf[BUFSIZ]; 440 int error; 441 442 if (pattern == NULL) 443 return; 444 445 error = regcomp(comp, pattern, REG_NEWLINE | REG_EXTENDED); 446 if (error != 0) { 447 regerror(error, comp, buf, sizeof(buf)); 448 if (*pattern != '\0') 449 errx(2, "%s: %s", pattern, buf); 450 else 451 errx(2, "%s", buf); 452 } 453 } 454 455 static void 456 set_argstr(char **av, char **ave) 457 { 458 size_t argsize; 459 char **ap; 460 461 argsize = 4 + *ave - *av + 1; 462 diffargs = xmalloc(argsize); 463 strlcpy(diffargs, "diff", argsize); 464 for (ap = av + 1; ap < ave; ap++) { 465 if (strcmp(*ap, "--") != 0) { 466 strlcat(diffargs, " ", argsize); 467 strlcat(diffargs, *ap, argsize); 468 } 469 } 470 } 471 472 /* 473 * Read in an excludes file and push each line. 474 */ 475 static void 476 read_excludes_file(char *file) 477 { 478 FILE *fp; 479 char *buf, *pattern; 480 size_t len; 481 482 if (strcmp(file, "-") == 0) 483 fp = stdin; 484 else if ((fp = fopen(file, "r")) == NULL) 485 err(2, "%s", file); 486 while ((buf = fgetln(fp, &len)) != NULL) { 487 if (buf[len - 1] == '\n') 488 len--; 489 if ((pattern = strndup(buf, len)) == NULL) 490 err(2, "xstrndup"); 491 push_excludes(pattern); 492 } 493 if (strcmp(file, "-") != 0) 494 fclose(fp); 495 } 496 497 /* 498 * Push a pattern onto the excludes list. 499 */ 500 static void 501 push_excludes(char *pattern) 502 { 503 struct excludes *entry; 504 505 entry = xmalloc(sizeof(*entry)); 506 entry->pattern = pattern; 507 entry->next = excludes_list; 508 excludes_list = entry; 509 } 510 511 static void 512 push_ignore_pats(char *pattern) 513 { 514 size_t len; 515 516 if (ignore_pats == NULL) 517 ignore_pats = xstrdup(pattern); 518 else { 519 /* old + "|" + new + NUL */ 520 len = strlen(ignore_pats) + strlen(pattern) + 2; 521 ignore_pats = xreallocarray(ignore_pats, 1, len); 522 strlcat(ignore_pats, "|", len); 523 strlcat(ignore_pats, pattern, len); 524 } 525 } 526 527 void 528 print_status(int val, char *path1, char *path2, const char *entry) 529 { 530 if (label[0] != NULL) 531 path1 = label[0]; 532 if (label[1] != NULL) 533 path2 = label[1]; 534 535 switch (val) { 536 case D_BINARY: 537 printf("Binary files %s%s and %s%s differ\n", 538 path1, entry, path2, entry); 539 break; 540 case D_DIFFER: 541 if (diff_format == D_BRIEF) 542 printf("Files %s%s and %s%s differ\n", 543 path1, entry, path2, entry); 544 break; 545 case D_SAME: 546 if (sflag) 547 printf("Files %s%s and %s%s are identical\n", 548 path1, entry, path2, entry); 549 break; 550 case D_MISMATCH1: 551 printf("File %s%s is a directory while file %s%s is a regular file\n", 552 path1, entry, path2, entry); 553 break; 554 case D_MISMATCH2: 555 printf("File %s%s is a regular file while file %s%s is a directory\n", 556 path1, entry, path2, entry); 557 break; 558 case D_SKIPPED1: 559 printf("File %s%s is not a regular file or directory and was skipped\n", 560 path1, entry); 561 break; 562 case D_SKIPPED2: 563 printf("File %s%s is not a regular file or directory and was skipped\n", 564 path2, entry); 565 break; 566 case D_ERROR: 567 break; 568 } 569 } 570 571 static void 572 usage(void) 573 { 574 (void)fprintf(stderr, 575 "usage: diff [-aBbdilpTtw] [-c | -e | -f | -n | -q | -u] [--ignore-case]\n" 576 " [--no-ignore-case] [--normal] [--strip-trailing-cr] [--tabsize]\n" 577 " [-I pattern] [-F pattern] [-L label] file1 file2\n" 578 " diff [-aBbdilpTtw] [-I pattern] [-L label] [--ignore-case]\n" 579 " [--no-ignore-case] [--normal] [--strip-trailing-cr] [--tabsize]\n" 580 " [-F pattern] -C number file1 file2\n" 581 " diff [-aBbdiltw] [-I pattern] [--ignore-case] [--no-ignore-case]\n" 582 " [--normal] [--strip-trailing-cr] [--tabsize] -D string file1 file2\n" 583 " diff [-aBbdilpTtw] [-I pattern] [-L label] [--ignore-case]\n" 584 " [--no-ignore-case] [--normal] [--tabsize] [--strip-trailing-cr]\n" 585 " [-F pattern] -U number file1 file2\n" 586 " diff [-aBbdilNPprsTtw] [-c | -e | -f | -n | -q | -u] [--ignore-case]\n" 587 " [--no-ignore-case] [--normal] [--tabsize] [-I pattern] [-L label]\n" 588 " [-F pattern] [-S name] [-X file] [-x pattern] dir1 dir2\n" 589 " diff [-aBbditwW] [--expand-tabs] [--ignore-all-blanks]\n" 590 " [--ignore-blank-lines] [--ignore-case] [--minimal]\n" 591 " [--no-ignore-file-name-case] [--strip-trailing-cr]\n" 592 " [--suppress-common-lines] [--tabsize] [--text] [--width]\n" 593 " -y | --side-by-side file1 file2\n"); 594 595 exit(2); 596 } 597 598 static void 599 conflicting_format(void) 600 { 601 602 fprintf(stderr, "error: conflicting output format options.\n"); 603 usage(); 604 } 605 606 static bool 607 do_color(void) 608 { 609 const char *p, *p2; 610 611 switch (colorflag) { 612 case COLORFLAG_AUTO: 613 p = getenv("CLICOLOR"); 614 p2 = getenv("COLORTERM"); 615 if ((p != NULL && *p != '\0') || (p2 != NULL && *p2 != '\0')) 616 return isatty(STDOUT_FILENO); 617 break; 618 case COLORFLAG_ALWAYS: 619 return (true); 620 case COLORFLAG_NEVER: 621 return (false); 622 } 623 624 return (false); 625 } 626 627 static char * 628 splice(char *dir, char *path) 629 { 630 char *tail, *buf; 631 size_t dirlen; 632 633 dirlen = strlen(dir); 634 while (dirlen != 0 && dir[dirlen - 1] == '/') 635 dirlen--; 636 if ((tail = strrchr(path, '/')) == NULL) 637 tail = path; 638 else 639 tail++; 640 xasprintf(&buf, "%.*s/%s", (int)dirlen, dir, tail); 641 return (buf); 642 } 643