1 /* $NetBSD: file.c,v 1.3 2014/01/16 23:36:52 joerg Exp $ */ 2 3 /* 4 * Copyright (c) Ian F. Darwin 1986-1995. 5 * Software written by Ian F. Darwin and others; 6 * maintained 1995-present by Christos Zoulas and others. 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 immediately at the beginning of the file, without modification, 13 * this list of conditions, and the following disclaimer. 14 * 2. Redistributions in binary form must reproduce the above copyright 15 * notice, this list of conditions and the following disclaimer in the 16 * documentation and/or other materials provided with the distribution. 17 * 18 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND 19 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 20 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 21 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR 22 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 23 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 24 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 25 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 26 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 27 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 28 * SUCH DAMAGE. 29 */ 30 /* 31 * file - find type of a file or files - main program. 32 */ 33 34 #include "file.h" 35 36 #ifndef lint 37 #if 0 38 FILE_RCSID("@(#)$File: file.c,v 1.152 2013/06/26 14:46:54 christos Exp $") 39 #else 40 __RCSID("$NetBSD: file.c,v 1.3 2014/01/16 23:36:52 joerg Exp $"); 41 #endif 42 #endif /* lint */ 43 44 #include "magic.h" 45 46 #include <stdlib.h> 47 #include <unistd.h> 48 #include <string.h> 49 #ifdef RESTORE_TIME 50 # if (__COHERENT__ >= 0x420) 51 # include <sys/utime.h> 52 # else 53 # ifdef USE_UTIMES 54 # include <sys/time.h> 55 # else 56 # include <utime.h> 57 # endif 58 # endif 59 #endif 60 #ifdef HAVE_UNISTD_H 61 #include <unistd.h> /* for read() */ 62 #endif 63 #ifdef HAVE_LOCALE_H 64 #include <locale.h> 65 #endif 66 #ifdef HAVE_WCHAR_H 67 #include <wchar.h> 68 #endif 69 70 #if defined(HAVE_GETOPT_H) && defined(HAVE_STRUCT_OPTION) 71 #include <getopt.h> 72 #ifndef HAVE_GETOPT_LONG 73 int getopt_long(int argc, char * const *argv, const char *optstring, const struct option *longopts, int *longindex); 74 #endif 75 #else 76 #include "mygetopt.h" 77 #endif 78 79 #ifdef S_IFLNK 80 #define FILE_FLAGS "-bchikLlNnprsvz0" 81 #else 82 #define FILE_FLAGS "-bciklNnprsvz0" 83 #endif 84 85 # define USAGE \ 86 "Usage: %s [" FILE_FLAGS \ 87 "] [--apple] [--mime-encoding] [--mime-type]\n" \ 88 " [-e testname] [-F separator] [-f namefile] [-m magicfiles] " \ 89 "file ...\n" \ 90 " %s -C [-m magicfiles]\n" \ 91 " %s [--help]\n" 92 93 private int /* Global command-line options */ 94 bflag = 0, /* brief output format */ 95 nopad = 0, /* Don't pad output */ 96 nobuffer = 0, /* Do not buffer stdout */ 97 nulsep = 0; /* Append '\0' to the separator */ 98 99 private const char *separator = ":"; /* Default field separator */ 100 private const struct option long_options[] = { 101 #define OPT(shortname, longname, opt, doc) \ 102 {longname, opt, NULL, shortname}, 103 #define OPT_LONGONLY(longname, opt, doc) \ 104 {longname, opt, NULL, 0}, 105 #include "file_opts.h" 106 #undef OPT 107 #undef OPT_LONGONLY 108 {0, 0, NULL, 0} 109 }; 110 #define OPTSTRING "bcCde:f:F:hiklLm:nNprsvz0" 111 112 private const struct { 113 const char *name; 114 int value; 115 } nv[] = { 116 { "apptype", MAGIC_NO_CHECK_APPTYPE }, 117 { "ascii", MAGIC_NO_CHECK_ASCII }, 118 { "cdf", MAGIC_NO_CHECK_CDF }, 119 { "compress", MAGIC_NO_CHECK_COMPRESS }, 120 { "elf", MAGIC_NO_CHECK_ELF }, 121 { "encoding", MAGIC_NO_CHECK_ENCODING }, 122 { "soft", MAGIC_NO_CHECK_SOFT }, 123 { "tar", MAGIC_NO_CHECK_TAR }, 124 { "text", MAGIC_NO_CHECK_TEXT }, /* synonym for ascii */ 125 { "tokens", MAGIC_NO_CHECK_TOKENS }, /* OBSOLETE: ignored for backwards compatibility */ 126 }; 127 128 private char *progname; /* used throughout */ 129 130 #ifdef __dead 131 __dead 132 #endif 133 private void usage(void); 134 private void docprint(const char *); 135 #ifdef __dead 136 __dead 137 #endif 138 private void help(void); 139 140 private int unwrap(struct magic_set *, const char *); 141 private int process(struct magic_set *ms, const char *, int); 142 private struct magic_set *load(const char *, int); 143 144 145 /* 146 * main - parse arguments and handle options 147 */ 148 int 149 main(int argc, char *argv[]) 150 { 151 int c; 152 size_t i; 153 int action = 0, didsomefiles = 0, errflg = 0; 154 int flags = 0, e = 0; 155 struct magic_set *magic = NULL; 156 int longindex; 157 const char *magicfile = NULL; /* where the magic is */ 158 159 /* makes islower etc work for other langs */ 160 (void)setlocale(LC_CTYPE, ""); 161 162 #ifdef __EMX__ 163 /* sh-like wildcard expansion! Shouldn't hurt at least ... */ 164 _wildcard(&argc, &argv); 165 #endif 166 167 if ((progname = strrchr(argv[0], '/')) != NULL) 168 progname++; 169 else 170 progname = argv[0]; 171 172 #ifdef S_IFLNK 173 flags |= getenv("POSIXLY_CORRECT") ? MAGIC_SYMLINK : 0; 174 #endif 175 while ((c = getopt_long(argc, argv, OPTSTRING, long_options, 176 &longindex)) != -1) 177 switch (c) { 178 case 0 : 179 switch (longindex) { 180 case 0: 181 help(); 182 break; 183 case 10: 184 flags |= MAGIC_APPLE; 185 break; 186 case 11: 187 flags |= MAGIC_MIME_TYPE; 188 break; 189 case 12: 190 flags |= MAGIC_MIME_ENCODING; 191 break; 192 } 193 break; 194 case '0': 195 nulsep = 1; 196 break; 197 case 'b': 198 bflag++; 199 break; 200 case 'c': 201 action = FILE_CHECK; 202 break; 203 case 'C': 204 action = FILE_COMPILE; 205 break; 206 case 'd': 207 flags |= MAGIC_DEBUG|MAGIC_CHECK; 208 break; 209 case 'e': 210 for (i = 0; i < sizeof(nv) / sizeof(nv[0]); i++) 211 if (strcmp(nv[i].name, optarg) == 0) 212 break; 213 214 if (i == sizeof(nv) / sizeof(nv[0])) 215 errflg++; 216 else 217 flags |= nv[i].value; 218 break; 219 220 case 'f': 221 if(action) 222 usage(); 223 if (magic == NULL) 224 if ((magic = load(magicfile, flags)) == NULL) 225 return 1; 226 e |= unwrap(magic, optarg); 227 ++didsomefiles; 228 break; 229 case 'F': 230 separator = optarg; 231 break; 232 case 'i': 233 flags |= MAGIC_MIME; 234 break; 235 case 'k': 236 flags |= MAGIC_CONTINUE; 237 break; 238 case 'l': 239 action = FILE_LIST; 240 break; 241 case 'm': 242 magicfile = optarg; 243 break; 244 case 'n': 245 ++nobuffer; 246 break; 247 case 'N': 248 ++nopad; 249 break; 250 #if defined(HAVE_UTIME) || defined(HAVE_UTIMES) 251 case 'p': 252 flags |= MAGIC_PRESERVE_ATIME; 253 break; 254 #endif 255 case 'r': 256 flags |= MAGIC_RAW; 257 break; 258 case 's': 259 flags |= MAGIC_DEVICES; 260 break; 261 case 'v': 262 if (magicfile == NULL) 263 magicfile = magic_getpath(magicfile, action); 264 (void)fprintf(stdout, "%s-%s\n", progname, VERSION); 265 (void)fprintf(stdout, "magic file from %s\n", 266 magicfile); 267 return 0; 268 case 'z': 269 flags |= MAGIC_COMPRESS; 270 break; 271 #ifdef S_IFLNK 272 case 'L': 273 flags |= MAGIC_SYMLINK; 274 break; 275 case 'h': 276 flags &= ~MAGIC_SYMLINK; 277 break; 278 #endif 279 case '?': 280 default: 281 errflg++; 282 break; 283 } 284 285 if (errflg) { 286 usage(); 287 } 288 if (e) 289 return e; 290 291 if (MAGIC_VERSION != magic_version()) 292 (void)fprintf(stderr, "%s: compiled magic version [%d] " 293 "does not match with shared library magic version [%d]\n", 294 progname, MAGIC_VERSION, magic_version()); 295 296 switch(action) { 297 case FILE_CHECK: 298 case FILE_COMPILE: 299 case FILE_LIST: 300 /* 301 * Don't try to check/compile ~/.magic unless we explicitly 302 * ask for it. 303 */ 304 magic = magic_open(flags|MAGIC_CHECK); 305 if (magic == NULL) { 306 (void)fprintf(stderr, "%s: %s\n", progname, 307 strerror(errno)); 308 return 1; 309 } 310 switch(action) { 311 case FILE_CHECK: 312 c = magic_check(magic, magicfile); 313 break; 314 case FILE_COMPILE: 315 c = magic_compile(magic, magicfile); 316 break; 317 case FILE_LIST: 318 c = magic_list(magic, magicfile); 319 break; 320 default: 321 abort(); 322 } 323 if (c == -1) { 324 (void)fprintf(stderr, "%s: %s\n", progname, 325 magic_error(magic)); 326 return 1; 327 } 328 return 0; 329 default: 330 if (magic == NULL) 331 if ((magic = load(magicfile, flags)) == NULL) 332 return 1; 333 break; 334 } 335 336 if (optind == argc) { 337 if (!didsomefiles) 338 usage(); 339 } 340 else { 341 size_t j, wid, nw; 342 for (wid = 0, j = (size_t)optind; j < (size_t)argc; j++) { 343 nw = file_mbswidth(argv[j]); 344 if (nw > wid) 345 wid = nw; 346 } 347 /* 348 * If bflag is only set twice, set it depending on 349 * number of files [this is undocumented, and subject to change] 350 */ 351 if (bflag == 2) { 352 bflag = optind >= argc - 1; 353 } 354 for (; optind < argc; optind++) 355 e |= process(magic, argv[optind], wid); 356 } 357 358 if (magic) 359 magic_close(magic); 360 return e; 361 } 362 363 364 private struct magic_set * 365 /*ARGSUSED*/ 366 load(const char *magicfile, int flags) 367 { 368 struct magic_set *magic = magic_open(flags); 369 if (magic == NULL) { 370 (void)fprintf(stderr, "%s: %s\n", progname, strerror(errno)); 371 return NULL; 372 } 373 if (magic_load(magic, magicfile) == -1) { 374 (void)fprintf(stderr, "%s: %s\n", 375 progname, magic_error(magic)); 376 magic_close(magic); 377 return NULL; 378 } 379 return magic; 380 } 381 382 /* 383 * unwrap -- read a file of filenames, do each one. 384 */ 385 private int 386 unwrap(struct magic_set *ms, const char *fn) 387 { 388 FILE *f; 389 ssize_t len; 390 char *line = NULL; 391 size_t llen = 0; 392 int wid = 0, cwid; 393 int e = 0; 394 395 if (strcmp("-", fn) == 0) { 396 f = stdin; 397 wid = 1; 398 } else { 399 if ((f = fopen(fn, "r")) == NULL) { 400 (void)fprintf(stderr, "%s: Cannot open `%s' (%s).\n", 401 progname, fn, strerror(errno)); 402 return 1; 403 } 404 405 while ((len = getline(&line, &llen, f)) > 0) { 406 if (line[len - 1] == '\n') 407 line[len - 1] = '\0'; 408 cwid = file_mbswidth(line); 409 if (cwid > wid) 410 wid = cwid; 411 } 412 413 rewind(f); 414 } 415 416 while ((len = getline(&line, &llen, f)) > 0) { 417 if (line[len - 1] == '\n') 418 line[len - 1] = '\0'; 419 e |= process(ms, line, wid); 420 if(nobuffer) 421 (void)fflush(stdout); 422 } 423 424 free(line); 425 (void)fclose(f); 426 return e; 427 } 428 429 /* 430 * Called for each input file on the command line (or in a list of files) 431 */ 432 private int 433 process(struct magic_set *ms, const char *inname, int wid) 434 { 435 const char *type; 436 int std_in = strcmp(inname, "-") == 0; 437 438 if (wid > 0 && !bflag) { 439 (void)printf("%s", std_in ? "/dev/stdin" : inname); 440 if (nulsep) 441 (void)putc('\0', stdout); 442 (void)printf("%s", separator); 443 (void)printf("%*s ", 444 (int) (nopad ? 0 : (wid - file_mbswidth(inname))), ""); 445 } 446 447 type = magic_file(ms, std_in ? NULL : inname); 448 if (type == NULL) { 449 (void)printf("ERROR: %s\n", magic_error(ms)); 450 return 1; 451 } else { 452 (void)printf("%s\n", type); 453 return 0; 454 } 455 } 456 457 protected size_t 458 file_mbswidth(const char *s) 459 { 460 #if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH) 461 size_t bytesconsumed, old_n, n, width = 0; 462 mbstate_t state; 463 wchar_t nextchar; 464 (void)memset(&state, 0, sizeof(mbstate_t)); 465 old_n = n = strlen(s); 466 467 while (n > 0) { 468 bytesconsumed = mbrtowc(&nextchar, s, n, &state); 469 if (bytesconsumed == (size_t)(-1) || 470 bytesconsumed == (size_t)(-2)) { 471 /* Something went wrong, return something reasonable */ 472 return old_n; 473 } 474 if (s[0] == '\n') { 475 /* 476 * do what strlen() would do, so that caller 477 * is always right 478 */ 479 width++; 480 } else { 481 int w = wcwidth(nextchar); 482 if (w > 0) 483 width += w; 484 } 485 486 s += bytesconsumed, n -= bytesconsumed; 487 } 488 return width; 489 #else 490 return strlen(s); 491 #endif 492 } 493 494 private void 495 usage(void) 496 { 497 (void)fprintf(stderr, USAGE, progname, progname, progname); 498 exit(1); 499 } 500 501 private void 502 docprint(const char *opts) 503 { 504 size_t i; 505 int comma; 506 char *sp, *p; 507 508 p = strstr(opts, "%o"); 509 if (p == NULL) { 510 fprintf(stdout, "%s", opts); 511 return; 512 } 513 514 for (sp = p - 1; sp > opts && *sp == ' '; sp--) 515 continue; 516 517 fprintf(stdout, "%.*s", (int)(p - opts), opts); 518 519 comma = 0; 520 for (i = 0; i < __arraycount(nv); i++) { 521 fprintf(stdout, "%s%s", comma++ ? ", " : "", nv[i].name); 522 if (i && i % 5 == 0) { 523 fprintf(stdout, ",\n%*s", (int)(p - sp - 1), ""); 524 comma = 0; 525 } 526 } 527 528 fprintf(stdout, "%s", opts + (p - opts) + 2); 529 } 530 531 private void 532 help(void) 533 { 534 (void)fputs( 535 "Usage: file [OPTION...] [FILE...]\n" 536 "Determine type of FILEs.\n" 537 "\n", stdout); 538 #define OPT(shortname, longname, opt, doc) \ 539 fprintf(stdout, " -%c, --" longname, shortname), \ 540 docprint(doc); 541 #define OPT_LONGONLY(longname, opt, doc) \ 542 fprintf(stdout, " --" longname), \ 543 docprint(doc); 544 #include "file_opts.h" 545 #undef OPT 546 #undef OPT_LONGONLY 547 fprintf(stdout, "\nReport bugs to http://bugs.gw.com/\n"); 548 exit(0); 549 } 550