1 /* 2 * Copyright (c) Ian F. Darwin 1986-1995. 3 * Software written by Ian F. Darwin and others; 4 * maintained 1995-present by Christos Zoulas and others. 5 * 6 * Redistribution and use in source and binary forms, with or without 7 * modification, are permitted provided that the following conditions 8 * are met: 9 * 1. Redistributions of source code must retain the above copyright 10 * notice immediately at the beginning of the file, without modification, 11 * this list of conditions, and the following disclaimer. 12 * 2. Redistributions in binary form must reproduce the above copyright 13 * notice, this list of conditions and the following disclaimer in the 14 * documentation and/or other materials provided with the distribution. 15 * 16 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND 17 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 18 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 19 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR 20 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 21 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 22 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 23 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 24 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 25 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 26 * SUCH DAMAGE. 27 */ 28 /* 29 * file - find type of a file or files - main program. 30 */ 31 32 #include "file.h" 33 34 #ifndef lint 35 FILE_RCSID("@(#)$File: file.c,v 1.140 2010/11/30 14:58:53 rrt Exp $") 36 #endif /* lint */ 37 38 #include "magic.h" 39 40 #include <stdlib.h> 41 #include <unistd.h> 42 #include <string.h> 43 #ifdef RESTORE_TIME 44 # if (__COHERENT__ >= 0x420) 45 # include <sys/utime.h> 46 # else 47 # ifdef USE_UTIMES 48 # include <sys/time.h> 49 # else 50 # include <utime.h> 51 # endif 52 # endif 53 #endif 54 #ifdef HAVE_UNISTD_H 55 #include <unistd.h> /* for read() */ 56 #endif 57 #ifdef HAVE_LOCALE_H 58 #include <locale.h> 59 #endif 60 #ifdef HAVE_WCHAR_H 61 #include <wchar.h> 62 #endif 63 64 #if defined(HAVE_GETOPT_H) && defined(HAVE_STRUCT_OPTION) 65 #include <getopt.h> 66 #ifndef HAVE_GETOPT_LONG 67 int getopt_long(int argc, char * const *argv, const char *optstring, const struct option *longopts, int *longindex); 68 #endif 69 #else 70 #include "mygetopt.h" 71 #endif 72 73 #include "patchlevel.h" 74 75 #ifdef S_IFLNK 76 #define FILE_FLAGS "-bchikLlNnprsvz0" 77 #else 78 #define FILE_FLAGS "-bciklNnprsvz0" 79 #endif 80 81 # define USAGE \ 82 "Usage: %s [" FILE_FLAGS \ 83 "] [--apple] [--mime-encoding] [--mime-type]\n" \ 84 " [-e testname] [-F separator] [-f namefile] [-m magicfiles] " \ 85 "file ...\n" \ 86 " %s -C [-m magicfiles]\n" \ 87 " %s [--help]\n" 88 89 #ifndef MAXPATHLEN 90 #define MAXPATHLEN 1024 91 #endif 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 }, 126 }; 127 128 private char *progname; /* used throughout */ 129 130 private void usage(void); 131 private void help(void); 132 int main(int, char *[]); 133 134 private int unwrap(struct magic_set *, const char *); 135 private int process(struct magic_set *ms, const char *, int); 136 private struct magic_set *load(const char *, int); 137 138 139 /* 140 * main - parse arguments and handle options 141 */ 142 int 143 main(int argc, char *argv[]) 144 { 145 int c; 146 size_t i; 147 int action = 0, didsomefiles = 0, errflg = 0; 148 int flags = 0, e = 0; 149 struct magic_set *magic = NULL; 150 int longindex; 151 const char *magicfile = NULL; /* where the magic is */ 152 153 /* makes islower etc work for other langs */ 154 (void)setlocale(LC_CTYPE, ""); 155 156 #ifdef __EMX__ 157 /* sh-like wildcard expansion! Shouldn't hurt at least ... */ 158 _wildcard(&argc, &argv); 159 #endif 160 161 if ((progname = strrchr(argv[0], '/')) != NULL) 162 progname++; 163 else 164 progname = argv[0]; 165 166 #ifdef S_IFLNK 167 flags |= getenv("POSIXLY_CORRECT") ? MAGIC_SYMLINK : 0; 168 #endif 169 while ((c = getopt_long(argc, argv, OPTSTRING, long_options, 170 &longindex)) != -1) 171 switch (c) { 172 case 0 : 173 switch (longindex) { 174 case 0: 175 help(); 176 break; 177 case 10: 178 flags |= MAGIC_APPLE; 179 break; 180 case 11: 181 flags |= MAGIC_MIME_TYPE; 182 break; 183 case 12: 184 flags |= MAGIC_MIME_ENCODING; 185 break; 186 } 187 break; 188 case '0': 189 nulsep = 1; 190 break; 191 case 'b': 192 bflag++; 193 break; 194 case 'c': 195 action = FILE_CHECK; 196 break; 197 case 'C': 198 action = FILE_COMPILE; 199 break; 200 case 'd': 201 flags |= MAGIC_DEBUG|MAGIC_CHECK; 202 break; 203 case 'e': 204 for (i = 0; i < sizeof(nv) / sizeof(nv[0]); i++) 205 if (strcmp(nv[i].name, optarg) == 0) 206 break; 207 208 if (i == sizeof(nv) / sizeof(nv[0])) 209 errflg++; 210 else 211 flags |= nv[i].value; 212 break; 213 214 case 'f': 215 if(action) 216 usage(); 217 if (magic == NULL) 218 if ((magic = load(magicfile, flags)) == NULL) 219 return 1; 220 e |= unwrap(magic, optarg); 221 ++didsomefiles; 222 break; 223 case 'F': 224 separator = optarg; 225 break; 226 case 'i': 227 flags |= MAGIC_MIME; 228 break; 229 case 'k': 230 flags |= MAGIC_CONTINUE; 231 break; 232 case 'l': 233 action = FILE_LIST; 234 break; 235 case 'm': 236 magicfile = optarg; 237 break; 238 case 'n': 239 ++nobuffer; 240 break; 241 case 'N': 242 ++nopad; 243 break; 244 #if defined(HAVE_UTIME) || defined(HAVE_UTIMES) 245 case 'p': 246 flags |= MAGIC_PRESERVE_ATIME; 247 break; 248 #endif 249 case 'r': 250 flags |= MAGIC_RAW; 251 break; 252 case 's': 253 flags |= MAGIC_DEVICES; 254 break; 255 case 'v': 256 if (magicfile == NULL) 257 magicfile = magic_getpath(magicfile, action); 258 (void)fprintf(stderr, "%s-%d.%.2d\n", progname, 259 FILE_VERSION_MAJOR, patchlevel); 260 (void)fprintf(stderr, "magic file from %s\n", 261 magicfile); 262 return 1; 263 case 'z': 264 flags |= MAGIC_COMPRESS; 265 break; 266 #ifdef S_IFLNK 267 case 'L': 268 flags |= MAGIC_SYMLINK; 269 break; 270 case 'h': 271 flags &= ~MAGIC_SYMLINK; 272 break; 273 #endif 274 case '?': 275 default: 276 errflg++; 277 break; 278 } 279 280 if (errflg) { 281 usage(); 282 } 283 if (e) 284 return e; 285 286 switch(action) { 287 case FILE_CHECK: 288 case FILE_COMPILE: 289 case FILE_LIST: 290 /* 291 * Don't try to check/compile ~/.magic unless we explicitly 292 * ask for it. 293 */ 294 magic = magic_open(flags|MAGIC_CHECK); 295 if (magic == NULL) { 296 (void)fprintf(stderr, "%s: %s\n", progname, 297 strerror(errno)); 298 return 1; 299 } 300 switch(action) { 301 case FILE_CHECK: 302 c = magic_check(magic, magicfile); 303 break; 304 case FILE_COMPILE: 305 c = magic_compile(magic, magicfile); 306 break; 307 case FILE_LIST: 308 c = magic_list(magic, magicfile); 309 break; 310 default: 311 abort(); 312 } 313 if (c == -1) { 314 (void)fprintf(stderr, "%s: %s\n", progname, 315 magic_error(magic)); 316 return 1; 317 } 318 return 0; 319 default: 320 if (magic == NULL) 321 if ((magic = load(magicfile, flags)) == NULL) 322 return 1; 323 break; 324 } 325 326 if (optind == argc) { 327 if (!didsomefiles) 328 usage(); 329 } 330 else { 331 size_t j, wid, nw; 332 for (wid = 0, j = (size_t)optind; j < (size_t)argc; j++) { 333 nw = file_mbswidth(argv[j]); 334 if (nw > wid) 335 wid = nw; 336 } 337 /* 338 * If bflag is only set twice, set it depending on 339 * number of files [this is undocumented, and subject to change] 340 */ 341 if (bflag == 2) { 342 bflag = optind >= argc - 1; 343 } 344 for (; optind < argc; optind++) 345 e |= process(magic, argv[optind], wid); 346 } 347 348 if (magic) 349 magic_close(magic); 350 return e; 351 } 352 353 354 private struct magic_set * 355 /*ARGSUSED*/ 356 load(const char *magicfile, int flags) 357 { 358 struct magic_set *magic = magic_open(flags); 359 if (magic == NULL) { 360 (void)fprintf(stderr, "%s: %s\n", progname, strerror(errno)); 361 return NULL; 362 } 363 if (magic_load(magic, magicfile) == -1) { 364 (void)fprintf(stderr, "%s: %s\n", 365 progname, magic_error(magic)); 366 magic_close(magic); 367 return NULL; 368 } 369 return magic; 370 } 371 372 /* 373 * unwrap -- read a file of filenames, do each one. 374 */ 375 private int 376 unwrap(struct magic_set *ms, const char *fn) 377 { 378 char buf[MAXPATHLEN]; 379 FILE *f; 380 int wid = 0, cwid; 381 int e = 0; 382 383 if (strcmp("-", fn) == 0) { 384 f = stdin; 385 wid = 1; 386 } else { 387 if ((f = fopen(fn, "r")) == NULL) { 388 (void)fprintf(stderr, "%s: Cannot open `%s' (%s).\n", 389 progname, fn, strerror(errno)); 390 return 1; 391 } 392 393 while (fgets(buf, sizeof(buf), f) != NULL) { 394 buf[strcspn(buf, "\n")] = '\0'; 395 cwid = file_mbswidth(buf); 396 if (cwid > wid) 397 wid = cwid; 398 } 399 400 rewind(f); 401 } 402 403 while (fgets(buf, sizeof(buf), f) != NULL) { 404 buf[strcspn(buf, "\n")] = '\0'; 405 e |= process(ms, buf, wid); 406 if(nobuffer) 407 (void)fflush(stdout); 408 } 409 410 (void)fclose(f); 411 return e; 412 } 413 414 /* 415 * Called for each input file on the command line (or in a list of files) 416 */ 417 private int 418 process(struct magic_set *ms, const char *inname, int wid) 419 { 420 const char *type; 421 int std_in = strcmp(inname, "-") == 0; 422 423 if (wid > 0 && !bflag) { 424 (void)printf("%s", std_in ? "/dev/stdin" : inname); 425 if (nulsep) 426 (void)putc('\0', stdout); 427 (void)printf("%s", separator); 428 (void)printf("%*s ", 429 (int) (nopad ? 0 : (wid - file_mbswidth(inname))), ""); 430 } 431 432 type = magic_file(ms, std_in ? NULL : inname); 433 if (type == NULL) { 434 (void)printf("ERROR: %s\n", magic_error(ms)); 435 return 1; 436 } else { 437 (void)printf("%s\n", type); 438 return 0; 439 } 440 } 441 442 size_t 443 file_mbswidth(const char *s) 444 { 445 #if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH) 446 size_t bytesconsumed, old_n, n, width = 0; 447 mbstate_t state; 448 wchar_t nextchar; 449 (void)memset(&state, 0, sizeof(mbstate_t)); 450 old_n = n = strlen(s); 451 452 while (n > 0) { 453 bytesconsumed = mbrtowc(&nextchar, s, n, &state); 454 if (bytesconsumed == (size_t)(-1) || 455 bytesconsumed == (size_t)(-2)) { 456 /* Something went wrong, return something reasonable */ 457 return old_n; 458 } 459 if (s[0] == '\n') { 460 /* 461 * do what strlen() would do, so that caller 462 * is always right 463 */ 464 width++; 465 } else 466 width += wcwidth(nextchar); 467 468 s += bytesconsumed, n -= bytesconsumed; 469 } 470 return width; 471 #else 472 return strlen(s); 473 #endif 474 } 475 476 private void 477 usage(void) 478 { 479 (void)fprintf(stderr, USAGE, progname, progname, progname); 480 exit(1); 481 } 482 483 private void 484 help(void) 485 { 486 (void)fputs( 487 "Usage: file [OPTION...] [FILE...]\n" 488 "Determine type of FILEs.\n" 489 "\n", stderr); 490 #define OPT(shortname, longname, opt, doc) \ 491 fprintf(stderr, " -%c, --" longname doc, shortname); 492 #define OPT_LONGONLY(longname, opt, doc) \ 493 fprintf(stderr, " --" longname doc); 494 #include "file_opts.h" 495 #undef OPT 496 #undef OPT_LONGONLY 497 exit(0); 498 } 499