file.c revision 226048
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 35FILE_RCSID("@(#)$File: file.c,v 1.144 2011/05/10 17:08:14 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 67int 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#ifdef S_IFLNK 74#define FILE_FLAGS "-bchikLlNnprsvz0" 75#else 76#define FILE_FLAGS "-bciklNnprsvz0" 77#endif 78 79# define USAGE \ 80 "Usage: %s [" FILE_FLAGS \ 81 "] [--apple] [--mime-encoding] [--mime-type]\n" \ 82 " [-e testname] [-F separator] [-f namefile] [-m magicfiles] " \ 83 "file ...\n" \ 84 " %s -C [-m magicfiles]\n" \ 85 " %s [--help]\n" 86 87private int /* Global command-line options */ 88 bflag = 0, /* brief output format */ 89 nopad = 0, /* Don't pad output */ 90 nobuffer = 0, /* Do not buffer stdout */ 91 nulsep = 0; /* Append '\0' to the separator */ 92 93private const char *separator = ":"; /* Default field separator */ 94private const struct option long_options[] = { 95#define OPT(shortname, longname, opt, doc) \ 96 {longname, opt, NULL, shortname}, 97#define OPT_LONGONLY(longname, opt, doc) \ 98 {longname, opt, NULL, 0}, 99#include "file_opts.h" 100#undef OPT 101#undef OPT_LONGONLY 102 {0, 0, NULL, 0} 103}; 104#define OPTSTRING "bcCde:f:F:hiklLm:nNprsvz0" 105 106private const struct { 107 const char *name; 108 int value; 109} nv[] = { 110 { "apptype", MAGIC_NO_CHECK_APPTYPE }, 111 { "ascii", MAGIC_NO_CHECK_ASCII }, 112 { "cdf", MAGIC_NO_CHECK_CDF }, 113 { "compress", MAGIC_NO_CHECK_COMPRESS }, 114 { "elf", MAGIC_NO_CHECK_ELF }, 115 { "encoding", MAGIC_NO_CHECK_ENCODING }, 116 { "soft", MAGIC_NO_CHECK_SOFT }, 117 { "tar", MAGIC_NO_CHECK_TAR }, 118 { "text", MAGIC_NO_CHECK_TEXT }, /* synonym for ascii */ 119 { "tokens", MAGIC_NO_CHECK_TOKENS }, 120}; 121 122private char *progname; /* used throughout */ 123 124private void usage(void); 125private void help(void); 126int main(int, char *[]); 127 128private int unwrap(struct magic_set *, const char *); 129private int process(struct magic_set *ms, const char *, int); 130private struct magic_set *load(const char *, int); 131 132 133/* 134 * main - parse arguments and handle options 135 */ 136int 137main(int argc, char *argv[]) 138{ 139 int c; 140 size_t i; 141 int action = 0, didsomefiles = 0, errflg = 0; 142 int flags = 0, e = 0; 143 struct magic_set *magic = NULL; 144 int longindex; 145 const char *magicfile = NULL; /* where the magic is */ 146 147 /* makes islower etc work for other langs */ 148 (void)setlocale(LC_CTYPE, ""); 149 150#ifdef __EMX__ 151 /* sh-like wildcard expansion! Shouldn't hurt at least ... */ 152 _wildcard(&argc, &argv); 153#endif 154 155 if ((progname = strrchr(argv[0], '/')) != NULL) 156 progname++; 157 else 158 progname = argv[0]; 159 160#ifdef S_IFLNK 161 flags |= getenv("POSIXLY_CORRECT") ? MAGIC_SYMLINK : 0; 162#endif 163 while ((c = getopt_long(argc, argv, OPTSTRING, long_options, 164 &longindex)) != -1) 165 switch (c) { 166 case 0 : 167 switch (longindex) { 168 case 0: 169 help(); 170 break; 171 case 10: 172 flags |= MAGIC_APPLE; 173 break; 174 case 11: 175 flags |= MAGIC_MIME_TYPE; 176 break; 177 case 12: 178 flags |= MAGIC_MIME_ENCODING; 179 break; 180 } 181 break; 182 case '0': 183 nulsep = 1; 184 break; 185 case 'b': 186 bflag++; 187 break; 188 case 'c': 189 action = FILE_CHECK; 190 break; 191 case 'C': 192 action = FILE_COMPILE; 193 break; 194 case 'd': 195 flags |= MAGIC_DEBUG|MAGIC_CHECK; 196 break; 197 case 'e': 198 for (i = 0; i < sizeof(nv) / sizeof(nv[0]); i++) 199 if (strcmp(nv[i].name, optarg) == 0) 200 break; 201 202 if (i == sizeof(nv) / sizeof(nv[0])) 203 errflg++; 204 else 205 flags |= nv[i].value; 206 break; 207 208 case 'f': 209 if(action) 210 usage(); 211 if (magic == NULL) 212 if ((magic = load(magicfile, flags)) == NULL) 213 return 1; 214 e |= unwrap(magic, optarg); 215 ++didsomefiles; 216 break; 217 case 'F': 218 separator = optarg; 219 break; 220 case 'i': 221 flags |= MAGIC_MIME; 222 break; 223 case 'k': 224 flags |= MAGIC_CONTINUE; 225 break; 226 case 'l': 227 action = FILE_LIST; 228 break; 229 case 'm': 230 magicfile = optarg; 231 break; 232 case 'n': 233 ++nobuffer; 234 break; 235 case 'N': 236 ++nopad; 237 break; 238#if defined(HAVE_UTIME) || defined(HAVE_UTIMES) 239 case 'p': 240 flags |= MAGIC_PRESERVE_ATIME; 241 break; 242#endif 243 case 'r': 244 flags |= MAGIC_RAW; 245 break; 246 case 's': 247 flags |= MAGIC_DEVICES; 248 break; 249 case 'v': 250 if (magicfile == NULL) 251 magicfile = magic_getpath(magicfile, action); 252 (void)fprintf(stdout, "%s-%s\n", progname, VERSION); 253 (void)fprintf(stdout, "magic file from %s\n", 254 magicfile); 255 return 1; 256 case 'z': 257 flags |= MAGIC_COMPRESS; 258 break; 259#ifdef S_IFLNK 260 case 'L': 261 flags |= MAGIC_SYMLINK; 262 break; 263 case 'h': 264 flags &= ~MAGIC_SYMLINK; 265 break; 266#endif 267 case '?': 268 default: 269 errflg++; 270 break; 271 } 272 273 if (errflg) { 274 usage(); 275 } 276 if (e) 277 return e; 278 279 switch(action) { 280 case FILE_CHECK: 281 case FILE_COMPILE: 282 case FILE_LIST: 283 /* 284 * Don't try to check/compile ~/.magic unless we explicitly 285 * ask for it. 286 */ 287 magic = magic_open(flags|MAGIC_CHECK); 288 if (magic == NULL) { 289 (void)fprintf(stderr, "%s: %s\n", progname, 290 strerror(errno)); 291 return 1; 292 } 293 switch(action) { 294 case FILE_CHECK: 295 c = magic_check(magic, magicfile); 296 break; 297 case FILE_COMPILE: 298 c = magic_compile(magic, magicfile); 299 break; 300 case FILE_LIST: 301 c = magic_list(magic, magicfile); 302 break; 303 default: 304 abort(); 305 } 306 if (c == -1) { 307 (void)fprintf(stderr, "%s: %s\n", progname, 308 magic_error(magic)); 309 return 1; 310 } 311 return 0; 312 default: 313 if (magic == NULL) 314 if ((magic = load(magicfile, flags)) == NULL) 315 return 1; 316 break; 317 } 318 319 if (optind == argc) { 320 if (!didsomefiles) 321 usage(); 322 } 323 else { 324 size_t j, wid, nw; 325 for (wid = 0, j = (size_t)optind; j < (size_t)argc; j++) { 326 nw = file_mbswidth(argv[j]); 327 if (nw > wid) 328 wid = nw; 329 } 330 /* 331 * If bflag is only set twice, set it depending on 332 * number of files [this is undocumented, and subject to change] 333 */ 334 if (bflag == 2) { 335 bflag = optind >= argc - 1; 336 } 337 for (; optind < argc; optind++) 338 e |= process(magic, argv[optind], wid); 339 } 340 341 if (magic) 342 magic_close(magic); 343 return e; 344} 345 346 347private struct magic_set * 348/*ARGSUSED*/ 349load(const char *magicfile, int flags) 350{ 351 struct magic_set *magic = magic_open(flags); 352 if (magic == NULL) { 353 (void)fprintf(stderr, "%s: %s\n", progname, strerror(errno)); 354 return NULL; 355 } 356 if (magic_load(magic, magicfile) == -1) { 357 (void)fprintf(stderr, "%s: %s\n", 358 progname, magic_error(magic)); 359 magic_close(magic); 360 return NULL; 361 } 362 return magic; 363} 364 365/* 366 * unwrap -- read a file of filenames, do each one. 367 */ 368private int 369unwrap(struct magic_set *ms, const char *fn) 370{ 371 FILE *f; 372 ssize_t len; 373 char *line = NULL; 374 size_t llen = 0; 375 int wid = 0, cwid; 376 int e = 0; 377 378 if (strcmp("-", fn) == 0) { 379 f = stdin; 380 wid = 1; 381 } else { 382 if ((f = fopen(fn, "r")) == NULL) { 383 (void)fprintf(stderr, "%s: Cannot open `%s' (%s).\n", 384 progname, fn, strerror(errno)); 385 return 1; 386 } 387 388 while ((len = getline(&line, &llen, f)) > 0) { 389 if (line[len - 1] == '\n') 390 line[len - 1] = '\0'; 391 cwid = file_mbswidth(line); 392 if (cwid > wid) 393 wid = cwid; 394 } 395 396 rewind(f); 397 } 398 399 while ((len = getline(&line, &llen, f)) > 0) { 400 if (line[len - 1] == '\n') 401 line[len - 1] = '\0'; 402 e |= process(ms, line, wid); 403 if(nobuffer) 404 (void)fflush(stdout); 405 } 406 407 free(line); 408 (void)fclose(f); 409 return e; 410} 411 412/* 413 * Called for each input file on the command line (or in a list of files) 414 */ 415private int 416process(struct magic_set *ms, const char *inname, int wid) 417{ 418 const char *type; 419 int std_in = strcmp(inname, "-") == 0; 420 421 if (wid > 0 && !bflag) { 422 (void)printf("%s", std_in ? "/dev/stdin" : inname); 423 if (nulsep) 424 (void)putc('\0', stdout); 425 (void)printf("%s", separator); 426 (void)printf("%*s ", 427 (int) (nopad ? 0 : (wid - file_mbswidth(inname))), ""); 428 } 429 430 type = magic_file(ms, std_in ? NULL : inname); 431 if (type == NULL) { 432 (void)printf("ERROR: %s\n", magic_error(ms)); 433 return 1; 434 } else { 435 (void)printf("%s\n", type); 436 return 0; 437 } 438} 439 440size_t 441file_mbswidth(const char *s) 442{ 443#if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH) 444 size_t bytesconsumed, old_n, n, width = 0; 445 mbstate_t state; 446 wchar_t nextchar; 447 (void)memset(&state, 0, sizeof(mbstate_t)); 448 old_n = n = strlen(s); 449 450 while (n > 0) { 451 bytesconsumed = mbrtowc(&nextchar, s, n, &state); 452 if (bytesconsumed == (size_t)(-1) || 453 bytesconsumed == (size_t)(-2)) { 454 /* Something went wrong, return something reasonable */ 455 return old_n; 456 } 457 if (s[0] == '\n') { 458 /* 459 * do what strlen() would do, so that caller 460 * is always right 461 */ 462 width++; 463 } else 464 width += wcwidth(nextchar); 465 466 s += bytesconsumed, n -= bytesconsumed; 467 } 468 return width; 469#else 470 return strlen(s); 471#endif 472} 473 474private void 475usage(void) 476{ 477 (void)fprintf(stderr, USAGE, progname, progname, progname); 478 exit(1); 479} 480 481private void 482help(void) 483{ 484 (void)fputs( 485"Usage: file [OPTION...] [FILE...]\n" 486"Determine type of FILEs.\n" 487"\n", stdout); 488#define OPT(shortname, longname, opt, doc) \ 489 fprintf(stdout, " -%c, --" longname doc, shortname); 490#define OPT_LONGONLY(longname, opt, doc) \ 491 fprintf(stdout, " --" longname doc); 492#include "file_opts.h" 493#undef OPT 494#undef OPT_LONGONLY 495 fprintf(stdout, "\nReport bugs to http://bugs.gw.com/\n"); 496 exit(0); 497} 498