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 1.143 2011/03/20 20:36:52 christos 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 private int /* Global command-line options */ 90 bflag = 0, /* brief output format */ 91 nopad = 0, /* Don't pad output */ 92 nobuffer = 0, /* Do not buffer stdout */ 93 nulsep = 0; /* Append '\0' to the separator */ 94 95 private const char *separator = ":"; /* Default field separator */ 96 private const struct option long_options[] = { 97 #define OPT(shortname, longname, opt, doc) \ 98 {longname, opt, NULL, shortname}, 99 #define OPT_LONGONLY(longname, opt, doc) \ 100 {longname, opt, NULL, 0}, 101 #include "file_opts.h" 102 #undef OPT 103 #undef OPT_LONGONLY 104 {0, 0, NULL, 0} 105 }; 106 #define OPTSTRING "bcCde:f:F:hiklLm:nNprsvz0" 107 108 private const struct { 109 const char *name; 110 int value; 111 } nv[] = { 112 { "apptype", MAGIC_NO_CHECK_APPTYPE }, 113 { "ascii", MAGIC_NO_CHECK_ASCII }, 114 { "cdf", MAGIC_NO_CHECK_CDF }, 115 { "compress", MAGIC_NO_CHECK_COMPRESS }, 116 { "elf", MAGIC_NO_CHECK_ELF }, 117 { "encoding", MAGIC_NO_CHECK_ENCODING }, 118 { "soft", MAGIC_NO_CHECK_SOFT }, 119 { "tar", MAGIC_NO_CHECK_TAR }, 120 { "text", MAGIC_NO_CHECK_TEXT }, /* synonym for ascii */ 121 { "tokens", MAGIC_NO_CHECK_TOKENS }, 122 }; 123 124 private char *progname; /* used throughout */ 125 126 private void usage(void); 127 private void help(void); 128 int main(int, char *[]); 129 130 private int unwrap(struct magic_set *, const char *); 131 private int process(struct magic_set *ms, const char *, int); 132 private struct magic_set *load(const char *, int); 133 134 135 /* 136 * main - parse arguments and handle options 137 */ 138 int 139 main(int argc, char *argv[]) 140 { 141 int c; 142 size_t i; 143 int action = 0, didsomefiles = 0, errflg = 0; 144 int flags = 0, e = 0; 145 struct magic_set *magic = NULL; 146 int longindex; 147 const char *magicfile = NULL; /* where the magic is */ 148 149 /* makes islower etc work for other langs */ 150 (void)setlocale(LC_CTYPE, ""); 151 152 #ifdef __EMX__ 153 /* sh-like wildcard expansion! Shouldn't hurt at least ... */ 154 _wildcard(&argc, &argv); 155 #endif 156 157 if ((progname = strrchr(argv[0], '/')) != NULL) 158 progname++; 159 else 160 progname = argv[0]; 161 162 #ifdef S_IFLNK 163 flags |= getenv("POSIXLY_CORRECT") ? MAGIC_SYMLINK : 0; 164 #endif 165 while ((c = getopt_long(argc, argv, OPTSTRING, long_options, 166 &longindex)) != -1) 167 switch (c) { 168 case 0 : 169 switch (longindex) { 170 case 0: 171 help(); 172 break; 173 case 10: 174 flags |= MAGIC_APPLE; 175 break; 176 case 11: 177 flags |= MAGIC_MIME_TYPE; 178 break; 179 case 12: 180 flags |= MAGIC_MIME_ENCODING; 181 break; 182 } 183 break; 184 case '0': 185 nulsep = 1; 186 break; 187 case 'b': 188 bflag++; 189 break; 190 case 'c': 191 action = FILE_CHECK; 192 break; 193 case 'C': 194 action = FILE_COMPILE; 195 break; 196 case 'd': 197 flags |= MAGIC_DEBUG|MAGIC_CHECK; 198 break; 199 case 'e': 200 for (i = 0; i < sizeof(nv) / sizeof(nv[0]); i++) 201 if (strcmp(nv[i].name, optarg) == 0) 202 break; 203 204 if (i == sizeof(nv) / sizeof(nv[0])) 205 errflg++; 206 else 207 flags |= nv[i].value; 208 break; 209 210 case 'f': 211 if(action) 212 usage(); 213 if (magic == NULL) 214 if ((magic = load(magicfile, flags)) == NULL) 215 return 1; 216 e |= unwrap(magic, optarg); 217 ++didsomefiles; 218 break; 219 case 'F': 220 separator = optarg; 221 break; 222 case 'i': 223 flags |= MAGIC_MIME; 224 break; 225 case 'k': 226 flags |= MAGIC_CONTINUE; 227 break; 228 case 'l': 229 action = FILE_LIST; 230 break; 231 case 'm': 232 magicfile = optarg; 233 break; 234 case 'n': 235 ++nobuffer; 236 break; 237 case 'N': 238 ++nopad; 239 break; 240 #if defined(HAVE_UTIME) || defined(HAVE_UTIMES) 241 case 'p': 242 flags |= MAGIC_PRESERVE_ATIME; 243 break; 244 #endif 245 case 'r': 246 flags |= MAGIC_RAW; 247 break; 248 case 's': 249 flags |= MAGIC_DEVICES; 250 break; 251 case 'v': 252 if (magicfile == NULL) 253 magicfile = magic_getpath(magicfile, action); 254 (void)fprintf(stdout, "%s-%d.%.2d\n", progname, 255 FILE_VERSION_MAJOR, patchlevel); 256 (void)fprintf(stdout, "magic file from %s\n", 257 magicfile); 258 return 1; 259 case 'z': 260 flags |= MAGIC_COMPRESS; 261 break; 262 #ifdef S_IFLNK 263 case 'L': 264 flags |= MAGIC_SYMLINK; 265 break; 266 case 'h': 267 flags &= ~MAGIC_SYMLINK; 268 break; 269 #endif 270 case '?': 271 default: 272 errflg++; 273 break; 274 } 275 276 if (errflg) { 277 usage(); 278 } 279 if (e) 280 return e; 281 282 switch(action) { 283 case FILE_CHECK: 284 case FILE_COMPILE: 285 case FILE_LIST: 286 /* 287 * Don't try to check/compile ~/.magic unless we explicitly 288 * ask for it. 289 */ 290 magic = magic_open(flags|MAGIC_CHECK); 291 if (magic == NULL) { 292 (void)fprintf(stderr, "%s: %s\n", progname, 293 strerror(errno)); 294 return 1; 295 } 296 switch(action) { 297 case FILE_CHECK: 298 c = magic_check(magic, magicfile); 299 break; 300 case FILE_COMPILE: 301 c = magic_compile(magic, magicfile); 302 break; 303 case FILE_LIST: 304 c = magic_list(magic, magicfile); 305 break; 306 default: 307 abort(); 308 } 309 if (c == -1) { 310 (void)fprintf(stderr, "%s: %s\n", progname, 311 magic_error(magic)); 312 return 1; 313 } 314 return 0; 315 default: 316 if (magic == NULL) 317 if ((magic = load(magicfile, flags)) == NULL) 318 return 1; 319 break; 320 } 321 322 if (optind == argc) { 323 if (!didsomefiles) 324 usage(); 325 } 326 else { 327 size_t j, wid, nw; 328 for (wid = 0, j = (size_t)optind; j < (size_t)argc; j++) { 329 nw = file_mbswidth(argv[j]); 330 if (nw > wid) 331 wid = nw; 332 } 333 /* 334 * If bflag is only set twice, set it depending on 335 * number of files [this is undocumented, and subject to change] 336 */ 337 if (bflag == 2) { 338 bflag = optind >= argc - 1; 339 } 340 for (; optind < argc; optind++) 341 e |= process(magic, argv[optind], wid); 342 } 343 344 if (magic) 345 magic_close(magic); 346 return e; 347 } 348 349 350 private struct magic_set * 351 /*ARGSUSED*/ 352 load(const char *magicfile, int flags) 353 { 354 struct magic_set *magic = magic_open(flags); 355 if (magic == NULL) { 356 (void)fprintf(stderr, "%s: %s\n", progname, strerror(errno)); 357 return NULL; 358 } 359 if (magic_load(magic, magicfile) == -1) { 360 (void)fprintf(stderr, "%s: %s\n", 361 progname, magic_error(magic)); 362 magic_close(magic); 363 return NULL; 364 } 365 return magic; 366 } 367 368 /* 369 * unwrap -- read a file of filenames, do each one. 370 */ 371 private int 372 unwrap(struct magic_set *ms, const char *fn) 373 { 374 FILE *f; 375 ssize_t len; 376 char *line = NULL; 377 size_t llen = 0; 378 int wid = 0, cwid; 379 int e = 0; 380 381 if (strcmp("-", fn) == 0) { 382 f = stdin; 383 wid = 1; 384 } else { 385 if ((f = fopen(fn, "r")) == NULL) { 386 (void)fprintf(stderr, "%s: Cannot open `%s' (%s).\n", 387 progname, fn, strerror(errno)); 388 return 1; 389 } 390 391 while ((len = getline(&line, &llen, f)) > 0) { 392 if (line[len - 1] == '\n') 393 line[len - 1] = '\0'; 394 cwid = file_mbswidth(line); 395 if (cwid > wid) 396 wid = cwid; 397 } 398 399 rewind(f); 400 } 401 402 while ((len = getline(&line, &llen, f)) > 0) { 403 if (line[len - 1] == '\n') 404 line[len - 1] = '\0'; 405 e |= process(ms, line, wid); 406 if(nobuffer) 407 (void)fflush(stdout); 408 } 409 410 free(line); 411 (void)fclose(f); 412 return e; 413 } 414 415 /* 416 * Called for each input file on the command line (or in a list of files) 417 */ 418 private int 419 process(struct magic_set *ms, const char *inname, int wid) 420 { 421 const char *type; 422 int std_in = strcmp(inname, "-") == 0; 423 424 if (wid > 0 && !bflag) { 425 (void)printf("%s", std_in ? "/dev/stdin" : inname); 426 if (nulsep) 427 (void)putc('\0', stdout); 428 (void)printf("%s", separator); 429 (void)printf("%*s ", 430 (int) (nopad ? 0 : (wid - file_mbswidth(inname))), ""); 431 } 432 433 type = magic_file(ms, std_in ? NULL : inname); 434 if (type == NULL) { 435 (void)printf("ERROR: %s\n", magic_error(ms)); 436 return 1; 437 } else { 438 (void)printf("%s\n", type); 439 return 0; 440 } 441 } 442 443 size_t 444 file_mbswidth(const char *s) 445 { 446 #if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH) 447 size_t bytesconsumed, old_n, n, width = 0; 448 mbstate_t state; 449 wchar_t nextchar; 450 (void)memset(&state, 0, sizeof(mbstate_t)); 451 old_n = n = strlen(s); 452 453 while (n > 0) { 454 bytesconsumed = mbrtowc(&nextchar, s, n, &state); 455 if (bytesconsumed == (size_t)(-1) || 456 bytesconsumed == (size_t)(-2)) { 457 /* Something went wrong, return something reasonable */ 458 return old_n; 459 } 460 if (s[0] == '\n') { 461 /* 462 * do what strlen() would do, so that caller 463 * is always right 464 */ 465 width++; 466 } else 467 width += wcwidth(nextchar); 468 469 s += bytesconsumed, n -= bytesconsumed; 470 } 471 return width; 472 #else 473 return strlen(s); 474 #endif 475 } 476 477 private void 478 usage(void) 479 { 480 (void)fprintf(stderr, USAGE, progname, progname, progname); 481 exit(1); 482 } 483 484 private void 485 help(void) 486 { 487 (void)fputs( 488 "Usage: file [OPTION...] [FILE...]\n" 489 "Determine type of FILEs.\n" 490 "\n", stdout); 491 #define OPT(shortname, longname, opt, doc) \ 492 fprintf(stdout, " -%c, --" longname doc, shortname); 493 #define OPT_LONGONLY(longname, opt, doc) \ 494 fprintf(stdout, " --" longname doc); 495 #include "file_opts.h" 496 #undef OPT 497 #undef OPT_LONGONLY 498 fprintf(stdout, "\nReport bugs to http://bugs.gw.com/\n"); 499 exit(0); 500 } 501