contrib/file-4/src/file.c

   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 #include "magic.h"
  34
  35 #include <stdio.h>
  36 #include <stdlib.h>
  37 #include <unistd.h>
  38 #include <string.h>
  39 #include <sys/types.h>
  40 #include <sys/param.h>  /* for MAXPATHLEN */
  41 #include <sys/stat.h>
  42 #include <fcntl.h>      /* for open() */
  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 #ifdef HAVE_GETOPT_H
  65 #include <getopt.h>     /* for long options (is this portable?)*/
  66 #else
  67 #undef HAVE_GETOPT_LONG
  68 #endif
  69
  70 #include <netinet/in.h>         /* for byte swapping */
  71
  72 #include "patchlevel.h"
  73
  74 #ifndef lint
  75 FILE_RCSID("@(#)$Id: file.c,v 1.95 2004/09/27 15:28:37 christos Exp $")
  76 #endif  /* lint */
  77
  78
  79 #ifdef S_IFLNK
  80 #define SYMLINKFLAG "L"
  81 #else
  82 #define SYMLINKFLAG ""
  83 #endif
  84
  85 # define USAGE  "Usage: %s [-bcik" SYMLINKFLAG "nNsvz] [-f namefile] [-F separator] [-m magicfiles] file...\n       %s -C -m magicfiles\n"
  86
  87 #ifndef MAXPATHLEN
  88 #define MAXPATHLEN      512
  89 #endif
  90
  91 private int             /* Global command-line options          */
  92         bflag = 0,      /* brief output format                  */
  93         nopad = 0,      /* Don't pad output                     */
  94         nobuffer = 0;   /* Do not buffer stdout                 */
  95
  96 private const char *magicfile = 0;      /* where the magic is   */
  97 private const char *default_magicfile = MAGIC;
  98 private char *separator = ":";  /* Default field separator      */
  99
 100 private char *progname;         /* used throughout              */
 101
 102 private struct magic_set *magic;
 103
 104 private void unwrap(char *);
 105 private void usage(void);
 106 #ifdef HAVE_GETOPT_LONG
 107 private void help(void);
 108 #endif
 109 #if 0
 110 private int byteconv4(int, int, int);
 111 private short byteconv2(int, int, int);
 112 #endif
 113
 114 int main(int, char *[]);
 115 private void process(const char *, int);
 116 private void load(const char *, int);
 117
 118
 119 /*
 120  * main - parse arguments and handle options
 121  */
 122 int
 123 main(int argc, char *argv[])
 124 {
 125         int c;
 126         int action = 0, didsomefiles = 0, errflg = 0;
 127         int flags = 0;
 128         char *home, *usermagic;
 129         struct stat sb;
 130 #define OPTSTRING       "bcCdf:F:ikLm:nNprsvz"
 131 #ifdef HAVE_GETOPT_LONG
 132         int longindex;
 133         private struct option long_options[] =
 134         {
 135                 {"version", 0, 0, 'v'},
 136                 {"help", 0, 0, 0},
 137                 {"brief", 0, 0, 'b'},
 138                 {"checking-printout", 0, 0, 'c'},
 139                 {"debug", 0, 0, 'd'},
 140                 {"files-from", 1, 0, 'f'},
 141                 {"separator", 1, 0, 'F'},
 142                 {"mime", 0, 0, 'i'},
 143                 {"keep-going", 0, 0, 'k'},
 144 #ifdef S_IFLNK
 145                 {"dereference", 0, 0, 'L'},
 146 #endif
 147                 {"magic-file", 1, 0, 'm'},
 148 #if defined(HAVE_UTIME) || defined(HAVE_UTIMES)
 149                 {"preserve-date", 0, 0, 'p'},
 150 #endif
 151                 {"uncompress", 0, 0, 'z'},
 152                 {"raw", 0, 0, 'r'},
 153                 {"no-buffer", 0, 0, 'n'},
 154                 {"no-pad", 0, 0, 'N'},
 155                 {"special-files", 0, 0, 's'},
 156                 {"compile", 0, 0, 'C'},
 157                 {0, 0, 0, 0},
 158         };
 159 #endif
 160
 161 #ifdef LC_CTYPE
 162         setlocale(LC_CTYPE, ""); /* makes islower etc work for other langs */
 163 #endif
 164
 165 #ifdef __EMX__
 166         /* sh-like wildcard expansion! Shouldn't hurt at least ... */
 167         _wildcard(&argc, &argv);
 168 #endif
 169
 170         if ((progname = strrchr(argv[0], '/')) != NULL)
 171                 progname++;
 172         else
 173                 progname = argv[0];
 174
 175         magicfile = default_magicfile;
 176         if ((usermagic = getenv("MAGIC")) != NULL)
 177                 magicfile = usermagic;
 178         else
 179                 if ((home = getenv("HOME")) != NULL) {
 180                         if ((usermagic = malloc(strlen(home) + 8)) != NULL) {
 181                                 (void)strcpy(usermagic, home);
 182                                 (void)strcat(usermagic, "/.magic");
 183                                 if (stat(usermagic, &sb)<0)
 184                                         free(usermagic);
 185                                 else
 186                                         magicfile = usermagic;
 187                         }
 188                 }
 189
 190 #ifndef HAVE_GETOPT_LONG
 191         while ((c = getopt(argc, argv, OPTSTRING)) != -1)
 192 #else
 193         while ((c = getopt_long(argc, argv, OPTSTRING, long_options,
 194             &longindex)) != -1)
 195 #endif
 196                 switch (c) {
 197 #ifdef HAVE_GETOPT_LONG
 198                 case 0 :
 199                         if (longindex == 1)
 200                                 help();
 201                         break;
 202 #endif
 203                 case 'b':
 204                         ++bflag;
 205                         break;
 206                 case 'c':
 207                         action = FILE_CHECK;
 208                         break;
 209                 case 'C':
 210                         action = FILE_COMPILE;
 211                         break;
 212                 case 'd':
 213                         flags |= MAGIC_DEBUG|MAGIC_CHECK;
 214                         break;
 215                 case 'f':
 216                         if(action)
 217                                 usage();
 218                         load(magicfile, flags);
 219                         unwrap(optarg);
 220                         ++didsomefiles;
 221                         break;
 222                 case 'F':
 223                         separator = optarg;
 224                         break;
 225                 case 'i':
 226                         flags |= MAGIC_MIME;
 227                         break;
 228                 case 'k':
 229                         flags |= MAGIC_CONTINUE;
 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                         (void) fprintf(stdout, "%s-%d.%.2d\n", progname,
 253                                        FILE_VERSION_MAJOR, patchlevel);
 254                         (void) fprintf(stdout, "magic file from %s\n",
 255                                        magicfile);
 256                         return 1;
 257                 case 'z':
 258                         flags |= MAGIC_COMPRESS;
 259                         break;
 260 #ifdef S_IFLNK
 261                 case 'L':
 262                         flags |= MAGIC_SYMLINK;
 263                         break;
 264 #endif
 265                 case '?':
 266                 default:
 267                         errflg++;
 268                         break;
 269                 }
 270
 271         if (errflg) {
 272                 usage();
 273         }
 274
 275         switch(action) {
 276         case FILE_CHECK:
 277         case FILE_COMPILE:
 278                 magic = magic_open(flags|MAGIC_CHECK);
 279                 if (magic == NULL) {
 280                         (void)fprintf(stderr, "%s: %s\n", progname,
 281                             strerror(errno));
 282                         return 1;
 283                 }
 284                 c = action == FILE_CHECK ? magic_check(magic, magicfile) :
 285                     magic_compile(magic, magicfile);
 286                 if (c == -1) {
 287                         (void)fprintf(stderr, "%s: %s\n", progname,
 288                             magic_error(magic));
 289                         return -1;
 290                 }
 291                 return 0;
 292         default:
 293                 load(magicfile, flags);
 294                 break;
 295         }
 296
 297         if (optind == argc) {
 298                 if (!didsomefiles) {
 299                         usage();
 300                 }
 301         }
 302         else {
 303                 int i, wid, nw;
 304                 for (wid = 0, i = optind; i < argc; i++) {
 305                         nw = file_mbswidth(argv[i]);
 306                         if (nw > wid)
 307                                 wid = nw;
 308                 }
 309                 for (; optind < argc; optind++)
 310                         process(argv[optind], wid);
 311         }
 312
 313         magic_close(magic);
 314         return 0;
 315 }
 316
 317
 318 private void
 319 load(const char *m, int flags)
 320 {
 321         if (magic)
 322                 return;
 323         magic = magic_open(flags);
 324         if (magic == NULL) {
 325                 (void)fprintf(stderr, "%s: %s\n", progname, strerror(errno));
 326                 exit(1);
 327         }
 328         if (magic_load(magic, magicfile) == -1) {
 329                 (void)fprintf(stderr, "%s: %s\n",
 330                     progname, magic_error(magic));
 331                 exit(1);
 332         }
 333 }
 334
 335 /*
 336  * unwrap -- read a file of filenames, do each one.
 337  */
 338 private void
 339 unwrap(char *fn)
 340 {
 341         char buf[MAXPATHLEN];
 342         FILE *f;
 343         int wid = 0, cwid;
 344
 345         if (strcmp("-", fn) == 0) {
 346                 f = stdin;
 347                 wid = 1;
 348         } else {
 349                 if ((f = fopen(fn, "r")) == NULL) {
 350                         (void)fprintf(stderr, "%s: Cannot open `%s' (%s).\n",
 351                             progname, fn, strerror(errno));
 352                         exit(1);
 353                 }
 354
 355                 while (fgets(buf, MAXPATHLEN, f) != NULL) {
 356                         cwid = file_mbswidth(buf) - 1;
 357                         if (cwid > wid)
 358                                 wid = cwid;
 359                 }
 360
 361                 rewind(f);
 362         }
 363
 364         while (fgets(buf, MAXPATHLEN, f) != NULL) {
 365                 buf[file_mbswidth(buf)-1] = '\0';
 366                 process(buf, wid);
 367                 if(nobuffer)
 368                         (void) fflush(stdout);
 369         }
 370
 371         (void) fclose(f);
 372 }
 373
 374 private void
 375 process(const char *inname, int wid)
 376 {
 377         const char *type;
 378         int std_in = strcmp(inname, "-") == 0;
 379
 380         if (wid > 0 && !bflag)
 381                 (void) printf("%s%s%*s ", std_in ? "/dev/stdin" : inname,
 382                     separator, (int) (nopad ? 0 : (wid - file_mbswidth(inname))), "");
 383
 384         type = magic_file(magic, std_in ? NULL : inname);
 385         if (type == NULL)
 386                 printf("ERROR: %s\n", magic_error(magic));
 387         else
 388                 printf("%s\n", type);
 389 }
 390
 391
 392 #if 0
 393 /*
 394  * byteconv4
 395  * Input:
 396  *      from            4 byte quantity to convert
 397  *      same            whether to perform byte swapping
 398  *      big_endian      whether we are a big endian host
 399  */
 400 private int
 401 byteconv4(int from, int same, int big_endian)
 402 {
 403         if (same)
 404                 return from;
 405         else if (big_endian) {          /* lsb -> msb conversion on msb */
 406                 union {
 407                         int i;
 408                         char c[4];
 409                 } retval, tmpval;
 410
 411                 tmpval.i = from;
 412                 retval.c[0] = tmpval.c[3];
 413                 retval.c[1] = tmpval.c[2];
 414                 retval.c[2] = tmpval.c[1];
 415                 retval.c[3] = tmpval.c[0];
 416
 417                 return retval.i;
 418         }
 419         else
 420                 return ntohl(from);     /* msb -> lsb conversion on lsb */
 421 }
 422
 423 /*
 424  * byteconv2
 425  * Same as byteconv4, but for shorts
 426  */
 427 private short
 428 byteconv2(int from, int same, int big_endian)
 429 {
 430         if (same)
 431                 return from;
 432         else if (big_endian) {          /* lsb -> msb conversion on msb */
 433                 union {
 434                         short s;
 435                         char c[2];
 436                 } retval, tmpval;
 437
 438                 tmpval.s = (short) from;
 439                 retval.c[0] = tmpval.c[1];
 440                 retval.c[1] = tmpval.c[0];
 441
 442                 return retval.s;
 443         }
 444         else
 445                 return ntohs(from);     /* msb -> lsb conversion on lsb */
 446 }
 447 #endif
 448
 449 size_t
 450 file_mbswidth(const char *s)
 451 {
 452 #if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH)
 453         size_t bytesconsumed, old_n, n, width = 0;
 454         mbstate_t state;
 455         wchar_t nextchar;
 456         (void)memset(&state, 0, sizeof(mbstate_t));
 457         old_n = n = strlen(s);
 458
 459         while (n > 0) {
 460                 bytesconsumed = mbrtowc(&nextchar, s, n, &state);
 461                 if (bytesconsumed == (size_t)(-1) ||
 462                     bytesconsumed == (size_t)(-2)) {
 463                         /* Something went wrong, return something reasonable */
 464                         return old_n;
 465                 }
 466                 if (s[0] == '\n') {
 467                         /*
 468                          * do what strlen() would do, so that caller
 469                          * is always right
 470                          */
 471                         width++;
 472                 } else
 473                         width += wcwidth(nextchar);
 474
 475                 s += bytesconsumed, n -= bytesconsumed;
 476         }
 477         return width;
 478 #else
 479         return strlen(s);
 480 #endif
 481 }
 482
 483 private void
 484 usage(void)
 485 {
 486         (void)fprintf(stderr, USAGE, progname, progname);
 487 #ifdef HAVE_GETOPT_LONG
 488         (void)fputs("Try `file --help' for more information.\n", stderr);
 489 #endif
 490         exit(1);
 491 }
 492
 493 #ifdef HAVE_GETOPT_LONG
 494 private void
 495 help(void)
 496 {
 497         puts(
 498 "Usage: file [OPTION]... [FILE]...\n"
 499 "Determine file type of FILEs.\n"
 500 "\n"
 501 "  -m, --magic-file LIST      use LIST as a colon-separated list of magic\n"
 502 "                               number files\n"
 503 "  -z, --uncompress           try to look inside compressed files\n"
 504 "  -b, --brief                do not prepend filenames to output lines\n"
 505 "  -c, --checking-printout    print the parsed form of the magic file, use in\n"
 506 "                               conjunction with -m to debug a new magic file\n"
 507 "                               before installing it\n"
 508 "  -f, --files-from FILE      read the filenames to be examined from FILE\n"
 509 "  -F, --separator string     use string as separator instead of `:'\n"
 510 "  -i, --mime                 output mime type strings\n"
 511 "  -k, --keep-going           don't stop at the first match\n"
 512 "  -L, --dereference          causes symlinks to be followed\n"
 513 "  -n, --no-buffer            do not buffer output\n"
 514 "  -N, --no-pad               do not pad output\n"
 515 "  -p, --preserve-date        preserve access times on files\n"
 516 "  -r, --raw                  don't translate unprintable chars to \\ooo\n"
 517 "  -s, --special-files        treat special (block/char devices) files as\n"
 518 "                             ordinary ones\n"
 519 "      --help                 display this help and exit\n"
 520 "      --version              output version information and exit\n"
 521 );
 522         exit(0);
 523 }
 524 #endif