1 /* $NetBSD: file.c,v 1.1.1.5 2013/03/23 15:49:16 christos 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 1.149 2013/01/07 18:20:19 christos Exp $") 39 #else 40 __RCSID("$NetBSD: file.c,v 1.1.1.5 2013/03/23 15:49:16 christos 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 private void usage(void); 131 private void docprint(const char *); 132 private void help(void); 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(stdout, "%s-%s\n", progname, VERSION); 259 (void)fprintf(stdout, "magic file from %s\n", 260 magicfile); 261 return 0; 262 case 'z': 263 flags |= MAGIC_COMPRESS; 264 break; 265 #ifdef S_IFLNK 266 case 'L': 267 flags |= MAGIC_SYMLINK; 268 break; 269 case 'h': 270 flags &= ~MAGIC_SYMLINK; 271 break; 272 #endif 273 case '?': 274 default: 275 errflg++; 276 break; 277 } 278 279 if (errflg) { 280 usage(); 281 } 282 if (e) 283 return e; 284 285 if (MAGIC_VERSION != magic_version()) 286 (void)fprintf(stderr, "%s: compiled magic version [%d] " 287 "does not match with shared library magic version [%d]\n", 288 progname, MAGIC_VERSION, magic_version()); 289 290 switch(action) { 291 case FILE_CHECK: 292 case FILE_COMPILE: 293 case FILE_LIST: 294 /* 295 * Don't try to check/compile ~/.magic unless we explicitly 296 * ask for it. 297 */ 298 magic = magic_open(flags|MAGIC_CHECK); 299 if (magic == NULL) { 300 (void)fprintf(stderr, "%s: %s\n", progname, 301 strerror(errno)); 302 return 1; 303 } 304 switch(action) { 305 case FILE_CHECK: 306 c = magic_check(magic, magicfile); 307 break; 308 case FILE_COMPILE: 309 c = magic_compile(magic, magicfile); 310 break; 311 case FILE_LIST: 312 c = magic_list(magic, magicfile); 313 break; 314 default: 315 abort(); 316 } 317 if (c == -1) { 318 (void)fprintf(stderr, "%s: %s\n", progname, 319 magic_error(magic)); 320 return 1; 321 } 322 return 0; 323 default: 324 if (magic == NULL) 325 if ((magic = load(magicfile, flags)) == NULL) 326 return 1; 327 break; 328 } 329 330 if (optind == argc) { 331 if (!didsomefiles) 332 usage(); 333 } 334 else { 335 size_t j, wid, nw; 336 for (wid = 0, j = (size_t)optind; j < (size_t)argc; j++) { 337 nw = file_mbswidth(argv[j]); 338 if (nw > wid) 339 wid = nw; 340 } 341 /* 342 * If bflag is only set twice, set it depending on 343 * number of files [this is undocumented, and subject to change] 344 */ 345 if (bflag == 2) { 346 bflag = optind >= argc - 1; 347 } 348 for (; optind < argc; optind++) 349 e |= process(magic, argv[optind], wid); 350 } 351 352 if (magic) 353 magic_close(magic); 354 return e; 355 } 356 357 358 private struct magic_set * 359 /*ARGSUSED*/ 360 load(const char *magicfile, int flags) 361 { 362 struct magic_set *magic = magic_open(flags); 363 if (magic == NULL) { 364 (void)fprintf(stderr, "%s: %s\n", progname, strerror(errno)); 365 return NULL; 366 } 367 if (magic_load(magic, magicfile) == -1) { 368 (void)fprintf(stderr, "%s: %s\n", 369 progname, magic_error(magic)); 370 magic_close(magic); 371 return NULL; 372 } 373 return magic; 374 } 375 376 /* 377 * unwrap -- read a file of filenames, do each one. 378 */ 379 private int 380 unwrap(struct magic_set *ms, const char *fn) 381 { 382 FILE *f; 383 ssize_t len; 384 char *line = NULL; 385 size_t llen = 0; 386 int wid = 0, cwid; 387 int e = 0; 388 389 if (strcmp("-", fn) == 0) { 390 f = stdin; 391 wid = 1; 392 } else { 393 if ((f = fopen(fn, "r")) == NULL) { 394 (void)fprintf(stderr, "%s: Cannot open `%s' (%s).\n", 395 progname, fn, strerror(errno)); 396 return 1; 397 } 398 399 while ((len = getline(&line, &llen, f)) > 0) { 400 if (line[len - 1] == '\n') 401 line[len - 1] = '\0'; 402 cwid = file_mbswidth(line); 403 if (cwid > wid) 404 wid = cwid; 405 } 406 407 rewind(f); 408 } 409 410 while ((len = getline(&line, &llen, f)) > 0) { 411 if (line[len - 1] == '\n') 412 line[len - 1] = '\0'; 413 e |= process(ms, line, wid); 414 if(nobuffer) 415 (void)fflush(stdout); 416 } 417 418 free(line); 419 (void)fclose(f); 420 return e; 421 } 422 423 /* 424 * Called for each input file on the command line (or in a list of files) 425 */ 426 private int 427 process(struct magic_set *ms, const char *inname, int wid) 428 { 429 const char *type; 430 int std_in = strcmp(inname, "-") == 0; 431 432 if (wid > 0 && !bflag) { 433 (void)printf("%s", std_in ? "/dev/stdin" : inname); 434 if (nulsep) 435 (void)putc('\0', stdout); 436 (void)printf("%s", separator); 437 (void)printf("%*s ", 438 (int) (nopad ? 0 : (wid - file_mbswidth(inname))), ""); 439 } 440 441 type = magic_file(ms, std_in ? NULL : inname); 442 if (type == NULL) { 443 (void)printf("ERROR: %s\n", magic_error(ms)); 444 return 1; 445 } else { 446 (void)printf("%s\n", type); 447 return 0; 448 } 449 } 450 451 protected size_t 452 file_mbswidth(const char *s) 453 { 454 #if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH) 455 size_t bytesconsumed, old_n, n, width = 0; 456 mbstate_t state; 457 wchar_t nextchar; 458 (void)memset(&state, 0, sizeof(mbstate_t)); 459 old_n = n = strlen(s); 460 461 while (n > 0) { 462 bytesconsumed = mbrtowc(&nextchar, s, n, &state); 463 if (bytesconsumed == (size_t)(-1) || 464 bytesconsumed == (size_t)(-2)) { 465 /* Something went wrong, return something reasonable */ 466 return old_n; 467 } 468 if (s[0] == '\n') { 469 /* 470 * do what strlen() would do, so that caller 471 * is always right 472 */ 473 width++; 474 } else 475 width += wcwidth(nextchar); 476 477 s += bytesconsumed, n -= bytesconsumed; 478 } 479 return width; 480 #else 481 return strlen(s); 482 #endif 483 } 484 485 private void 486 usage(void) 487 { 488 (void)fprintf(stderr, USAGE, progname, progname, progname); 489 exit(1); 490 } 491 492 private void 493 docprint(const char *opts) 494 { 495 size_t i; 496 int comma; 497 char *sp, *p; 498 499 p = strstr(opts, "%o"); 500 if (p == NULL) { 501 fprintf(stdout, "%s", opts); 502 return; 503 } 504 505 for (sp = p - 1; sp > opts && *sp == ' '; sp--) 506 continue; 507 508 fprintf(stdout, "%.*s", (int)(p - opts), opts); 509 510 comma = 0; 511 for (i = 0; i < __arraycount(nv); i++) { 512 fprintf(stdout, "%s%s", comma++ ? ", " : "", nv[i].name); 513 if (i && i % 5 == 0) { 514 fprintf(stdout, ",\n%*s", (int)(p - sp - 1), ""); 515 comma = 0; 516 } 517 } 518 519 fprintf(stdout, "%s", opts + (p - opts) + 2); 520 } 521 522 private void 523 help(void) 524 { 525 (void)fputs( 526 "Usage: file [OPTION...] [FILE...]\n" 527 "Determine type of FILEs.\n" 528 "\n", stdout); 529 #define OPT(shortname, longname, opt, doc) \ 530 fprintf(stdout, " -%c, --" longname, shortname), \ 531 docprint(doc); 532 #define OPT_LONGONLY(longname, opt, doc) \ 533 fprintf(stdout, " --" longname), \ 534 docprint(doc); 535 #include "file_opts.h" 536 #undef OPT 537 #undef OPT_LONGONLY 538 fprintf(stdout, "\nReport bugs to http://bugs.gw.com/\n"); 539 exit(0); 540 } 541