file.c 11 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497
  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. #include "file.h"
  32. #ifndef lint
  33. FILE_RCSID("@(#)$File: file.c,v 1.144 2011/05/10 17:08:14 christos Exp $")
  34. #endif /* lint */
  35. #include "magic.h"
  36. #include <stdlib.h>
  37. #include <unistd.h>
  38. #include <string.h>
  39. #ifdef RESTORE_TIME
  40. # if (__COHERENT__ >= 0x420)
  41. # include <sys/utime.h>
  42. # else
  43. # ifdef USE_UTIMES
  44. # include <sys/time.h>
  45. # else
  46. # include <utime.h>
  47. # endif
  48. # endif
  49. #endif
  50. #ifdef HAVE_UNISTD_H
  51. #include <unistd.h> /* for read() */
  52. #endif
  53. #ifdef HAVE_LOCALE_H
  54. #include <locale.h>
  55. #endif
  56. #ifdef HAVE_WCHAR_H
  57. #include <wchar.h>
  58. #endif
  59. #if defined(HAVE_GETOPT_H) && defined(HAVE_STRUCT_OPTION)
  60. #include <getopt.h>
  61. #ifndef HAVE_GETOPT_LONG
  62. int getopt_long(int argc, char * const *argv, const char *optstring, const struct option *longopts, int *longindex);
  63. #endif
  64. #else
  65. #include "mygetopt.h"
  66. #endif
  67. #ifdef S_IFLNK
  68. #define FILE_FLAGS "-bchikLlNnprsvz0"
  69. #else
  70. #define FILE_FLAGS "-bciklNnprsvz0"
  71. #endif
  72. # define USAGE \
  73. "Usage: %s [" FILE_FLAGS \
  74. "] [--apple] [--mime-encoding] [--mime-type]\n" \
  75. " [-e testname] [-F separator] [-f namefile] [-m magicfiles] " \
  76. "file ...\n" \
  77. " %s -C [-m magicfiles]\n" \
  78. " %s [--help]\n"
  79. private int /* Global command-line options */
  80. bflag = 0, /* brief output format */
  81. nopad = 0, /* Don't pad output */
  82. nobuffer = 0, /* Do not buffer stdout */
  83. nulsep = 0; /* Append '\0' to the separator */
  84. private const char *separator = ":"; /* Default field separator */
  85. private const struct option long_options[] = {
  86. #define OPT(shortname, longname, opt, doc) \
  87. {longname, opt, NULL, shortname},
  88. #define OPT_LONGONLY(longname, opt, doc) \
  89. {longname, opt, NULL, 0},
  90. #include "file_opts.h"
  91. #undef OPT
  92. #undef OPT_LONGONLY
  93. {0, 0, NULL, 0}
  94. };
  95. #define OPTSTRING "bcCde:f:F:hiklLm:nNprsvz0"
  96. private const struct {
  97. const char *name;
  98. int value;
  99. } nv[] = {
  100. { "apptype", MAGIC_NO_CHECK_APPTYPE },
  101. { "ascii", MAGIC_NO_CHECK_ASCII },
  102. { "cdf", MAGIC_NO_CHECK_CDF },
  103. { "compress", MAGIC_NO_CHECK_COMPRESS },
  104. { "elf", MAGIC_NO_CHECK_ELF },
  105. { "encoding", MAGIC_NO_CHECK_ENCODING },
  106. { "soft", MAGIC_NO_CHECK_SOFT },
  107. { "tar", MAGIC_NO_CHECK_TAR },
  108. { "text", MAGIC_NO_CHECK_TEXT }, /* synonym for ascii */
  109. { "tokens", MAGIC_NO_CHECK_TOKENS },
  110. };
  111. private char *progname; /* used throughout */
  112. private void usage(void);
  113. private void help(void);
  114. int main(int, char *[]);
  115. private int unwrap(struct magic_set *, const char *);
  116. private int process(struct magic_set *ms, const char *, int);
  117. private struct magic_set *load(const char *, int);
  118. /*
  119. * main - parse arguments and handle options
  120. */
  121. int
  122. main(int argc, char *argv[])
  123. {
  124. int c;
  125. size_t i;
  126. int action = 0, didsomefiles = 0, errflg = 0;
  127. int flags = 0, e = 0;
  128. struct magic_set *magic = NULL;
  129. int longindex;
  130. const char *magicfile = NULL; /* where the magic is */
  131. /* makes islower etc work for other langs */
  132. (void)setlocale(LC_CTYPE, "");
  133. #ifdef __EMX__
  134. /* sh-like wildcard expansion! Shouldn't hurt at least ... */
  135. _wildcard(&argc, &argv);
  136. #endif
  137. if ((progname = strrchr(argv[0], '/')) != NULL)
  138. progname++;
  139. else
  140. progname = argv[0];
  141. #ifdef S_IFLNK
  142. flags |= getenv("POSIXLY_CORRECT") ? MAGIC_SYMLINK : 0;
  143. #endif
  144. while ((c = getopt_long(argc, argv, OPTSTRING, long_options,
  145. &longindex)) != -1)
  146. switch (c) {
  147. case 0 :
  148. switch (longindex) {
  149. case 0:
  150. help();
  151. break;
  152. case 10:
  153. flags |= MAGIC_APPLE;
  154. break;
  155. case 11:
  156. flags |= MAGIC_MIME_TYPE;
  157. break;
  158. case 12:
  159. flags |= MAGIC_MIME_ENCODING;
  160. break;
  161. }
  162. break;
  163. case '0':
  164. nulsep = 1;
  165. break;
  166. case 'b':
  167. bflag++;
  168. break;
  169. case 'c':
  170. action = FILE_CHECK;
  171. break;
  172. case 'C':
  173. action = FILE_COMPILE;
  174. break;
  175. case 'd':
  176. flags |= MAGIC_DEBUG|MAGIC_CHECK;
  177. break;
  178. case 'e':
  179. for (i = 0; i < sizeof(nv) / sizeof(nv[0]); i++)
  180. if (strcmp(nv[i].name, optarg) == 0)
  181. break;
  182. if (i == sizeof(nv) / sizeof(nv[0]))
  183. errflg++;
  184. else
  185. flags |= nv[i].value;
  186. break;
  187. case 'f':
  188. if(action)
  189. usage();
  190. if (magic == NULL)
  191. if ((magic = load(magicfile, flags)) == NULL)
  192. return 1;
  193. e |= unwrap(magic, optarg);
  194. ++didsomefiles;
  195. break;
  196. case 'F':
  197. separator = optarg;
  198. break;
  199. case 'i':
  200. flags |= MAGIC_MIME;
  201. break;
  202. case 'k':
  203. flags |= MAGIC_CONTINUE;
  204. break;
  205. case 'l':
  206. action = FILE_LIST;
  207. break;
  208. case 'm':
  209. magicfile = optarg;
  210. break;
  211. case 'n':
  212. ++nobuffer;
  213. break;
  214. case 'N':
  215. ++nopad;
  216. break;
  217. #if defined(HAVE_UTIME) || defined(HAVE_UTIMES)
  218. case 'p':
  219. flags |= MAGIC_PRESERVE_ATIME;
  220. break;
  221. #endif
  222. case 'r':
  223. flags |= MAGIC_RAW;
  224. break;
  225. case 's':
  226. flags |= MAGIC_DEVICES;
  227. break;
  228. case 'v':
  229. if (magicfile == NULL)
  230. magicfile = magic_getpath(magicfile, action);
  231. (void)fprintf(stdout, "%s-%s\n", progname, VERSION);
  232. (void)fprintf(stdout, "magic file from %s\n",
  233. magicfile);
  234. return 1;
  235. case 'z':
  236. flags |= MAGIC_COMPRESS;
  237. break;
  238. #ifdef S_IFLNK
  239. case 'L':
  240. flags |= MAGIC_SYMLINK;
  241. break;
  242. case 'h':
  243. flags &= ~MAGIC_SYMLINK;
  244. break;
  245. #endif
  246. case '?':
  247. default:
  248. errflg++;
  249. break;
  250. }
  251. if (errflg) {
  252. usage();
  253. }
  254. if (e)
  255. return e;
  256. switch(action) {
  257. case FILE_CHECK:
  258. case FILE_COMPILE:
  259. case FILE_LIST:
  260. /*
  261. * Don't try to check/compile ~/.magic unless we explicitly
  262. * ask for it.
  263. */
  264. magic = magic_open(flags|MAGIC_CHECK);
  265. if (magic == NULL) {
  266. (void)fprintf(stderr, "%s: %s\n", progname,
  267. strerror(errno));
  268. return 1;
  269. }
  270. switch(action) {
  271. case FILE_CHECK:
  272. c = magic_check(magic, magicfile);
  273. break;
  274. case FILE_COMPILE:
  275. c = magic_compile(magic, magicfile);
  276. break;
  277. case FILE_LIST:
  278. c = magic_list(magic, magicfile);
  279. break;
  280. default:
  281. abort();
  282. }
  283. if (c == -1) {
  284. (void)fprintf(stderr, "%s: %s\n", progname,
  285. magic_error(magic));
  286. return 1;
  287. }
  288. return 0;
  289. default:
  290. if (magic == NULL)
  291. if ((magic = load(magicfile, flags)) == NULL)
  292. return 1;
  293. break;
  294. }
  295. if (optind == argc) {
  296. if (!didsomefiles)
  297. usage();
  298. }
  299. else {
  300. size_t j, wid, nw;
  301. for (wid = 0, j = (size_t)optind; j < (size_t)argc; j++) {
  302. nw = file_mbswidth(argv[j]);
  303. if (nw > wid)
  304. wid = nw;
  305. }
  306. /*
  307. * If bflag is only set twice, set it depending on
  308. * number of files [this is undocumented, and subject to change]
  309. */
  310. if (bflag == 2) {
  311. bflag = optind >= argc - 1;
  312. }
  313. for (; optind < argc; optind++)
  314. e |= process(magic, argv[optind], wid);
  315. }
  316. if (magic)
  317. magic_close(magic);
  318. return e;
  319. }
  320. private struct magic_set *
  321. /*ARGSUSED*/
  322. load(const char *magicfile, int flags)
  323. {
  324. struct magic_set *magic = magic_open(flags);
  325. if (magic == NULL) {
  326. (void)fprintf(stderr, "%s: %s\n", progname, strerror(errno));
  327. return NULL;
  328. }
  329. if (magic_load(magic, magicfile) == -1) {
  330. (void)fprintf(stderr, "%s: %s\n",
  331. progname, magic_error(magic));
  332. magic_close(magic);
  333. return NULL;
  334. }
  335. return magic;
  336. }
  337. /*
  338. * unwrap -- read a file of filenames, do each one.
  339. */
  340. private int
  341. unwrap(struct magic_set *ms, const char *fn)
  342. {
  343. FILE *f;
  344. ssize_t len;
  345. char *line = NULL;
  346. size_t llen = 0;
  347. int wid = 0, cwid;
  348. int e = 0;
  349. if (strcmp("-", fn) == 0) {
  350. f = stdin;
  351. wid = 1;
  352. } else {
  353. if ((f = fopen(fn, "r")) == NULL) {
  354. (void)fprintf(stderr, "%s: Cannot open `%s' (%s).\n",
  355. progname, fn, strerror(errno));
  356. return 1;
  357. }
  358. while ((len = getline(&line, &llen, f)) > 0) {
  359. if (line[len - 1] == '\n')
  360. line[len - 1] = '\0';
  361. cwid = file_mbswidth(line);
  362. if (cwid > wid)
  363. wid = cwid;
  364. }
  365. rewind(f);
  366. }
  367. while ((len = getline(&line, &llen, f)) > 0) {
  368. if (line[len - 1] == '\n')
  369. line[len - 1] = '\0';
  370. e |= process(ms, line, wid);
  371. if(nobuffer)
  372. (void)fflush(stdout);
  373. }
  374. free(line);
  375. (void)fclose(f);
  376. return e;
  377. }
  378. /*
  379. * Called for each input file on the command line (or in a list of files)
  380. */
  381. private int
  382. process(struct magic_set *ms, const char *inname, int wid)
  383. {
  384. const char *type;
  385. int std_in = strcmp(inname, "-") == 0;
  386. if (wid > 0 && !bflag) {
  387. (void)printf("%s", std_in ? "/dev/stdin" : inname);
  388. if (nulsep)
  389. (void)putc('\0', stdout);
  390. (void)printf("%s", separator);
  391. (void)printf("%*s ",
  392. (int) (nopad ? 0 : (wid - file_mbswidth(inname))), "");
  393. }
  394. type = magic_file(ms, std_in ? NULL : inname);
  395. if (type == NULL) {
  396. (void)printf("ERROR: %s\n", magic_error(ms));
  397. return 1;
  398. } else {
  399. (void)printf("%s\n", type);
  400. return 0;
  401. }
  402. }
  403. size_t
  404. file_mbswidth(const char *s)
  405. {
  406. #if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH)
  407. size_t bytesconsumed, old_n, n, width = 0;
  408. mbstate_t state;
  409. wchar_t nextchar;
  410. (void)memset(&state, 0, sizeof(mbstate_t));
  411. old_n = n = strlen(s);
  412. while (n > 0) {
  413. bytesconsumed = mbrtowc(&nextchar, s, n, &state);
  414. if (bytesconsumed == (size_t)(-1) ||
  415. bytesconsumed == (size_t)(-2)) {
  416. /* Something went wrong, return something reasonable */
  417. return old_n;
  418. }
  419. if (s[0] == '\n') {
  420. /*
  421. * do what strlen() would do, so that caller
  422. * is always right
  423. */
  424. width++;
  425. } else
  426. width += wcwidth(nextchar);
  427. s += bytesconsumed, n -= bytesconsumed;
  428. }
  429. return width;
  430. #else
  431. return strlen(s);
  432. #endif
  433. }
  434. private void
  435. usage(void)
  436. {
  437. (void)fprintf(stderr, USAGE, progname, progname, progname);
  438. exit(1);
  439. }
  440. private void
  441. help(void)
  442. {
  443. (void)fputs(
  444. "Usage: file [OPTION...] [FILE...]\n"
  445. "Determine type of FILEs.\n"
  446. "\n", stdout);
  447. #define OPT(shortname, longname, opt, doc) \
  448. fprintf(stdout, " -%c, --" longname doc, shortname);
  449. #define OPT_LONGONLY(longname, opt, doc) \
  450. fprintf(stdout, " --" longname doc);
  451. #include "file_opts.h"
  452. #undef OPT
  453. #undef OPT_LONGONLY
  454. fprintf(stdout, "\nReport bugs to http://bugs.gw.com/\n");
  455. exit(0);
  456. }