file.c 13 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594
  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 1.159 2014/11/28 02:46:39 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_WCHAR_H
  54. #include <wchar.h>
  55. #endif
  56. #if defined(HAVE_GETOPT_H) && defined(HAVE_STRUCT_OPTION)
  57. #include <getopt.h>
  58. #ifndef HAVE_GETOPT_LONG
  59. int getopt_long(int argc, char * const *argv, const char *optstring, const struct option *longopts, int *longindex);
  60. #endif
  61. #else
  62. #include "mygetopt.h"
  63. #endif
  64. #ifdef S_IFLNK
  65. #define FILE_FLAGS "-bcEhikLlNnprsvz0"
  66. #else
  67. #define FILE_FLAGS "-bcEiklNnprsvz0"
  68. #endif
  69. # define USAGE \
  70. "Usage: %s [" FILE_FLAGS \
  71. "] [--apple] [--mime-encoding] [--mime-type]\n" \
  72. " [-e testname] [-F separator] [-f namefile] [-m magicfiles] " \
  73. "file ...\n" \
  74. " %s -C [-m magicfiles]\n" \
  75. " %s [--help]\n"
  76. private int /* Global command-line options */
  77. bflag = 0, /* brief output format */
  78. nopad = 0, /* Don't pad output */
  79. nobuffer = 0, /* Do not buffer stdout */
  80. nulsep = 0; /* Append '\0' to the separator */
  81. private const char *separator = ":"; /* Default field separator */
  82. private const struct option long_options[] = {
  83. #define OPT(shortname, longname, opt, doc) \
  84. {longname, opt, NULL, shortname},
  85. #define OPT_LONGONLY(longname, opt, doc) \
  86. {longname, opt, NULL, 0},
  87. #include "file_opts.h"
  88. #undef OPT
  89. #undef OPT_LONGONLY
  90. {0, 0, NULL, 0}
  91. };
  92. #define OPTSTRING "bcCde:Ef:F:hiklLm:nNpP:rsvz0"
  93. private const struct {
  94. const char *name;
  95. int value;
  96. } nv[] = {
  97. { "apptype", MAGIC_NO_CHECK_APPTYPE },
  98. { "ascii", MAGIC_NO_CHECK_ASCII },
  99. { "cdf", MAGIC_NO_CHECK_CDF },
  100. { "compress", MAGIC_NO_CHECK_COMPRESS },
  101. { "elf", MAGIC_NO_CHECK_ELF },
  102. { "encoding", MAGIC_NO_CHECK_ENCODING },
  103. { "soft", MAGIC_NO_CHECK_SOFT },
  104. { "tar", MAGIC_NO_CHECK_TAR },
  105. { "text", MAGIC_NO_CHECK_TEXT }, /* synonym for ascii */
  106. { "tokens", MAGIC_NO_CHECK_TOKENS }, /* OBSOLETE: ignored for backwards compatibility */
  107. };
  108. private struct {
  109. const char *name;
  110. int tag;
  111. size_t value;
  112. } pm[] = {
  113. { "indir", MAGIC_PARAM_INDIR_MAX, 0 },
  114. { "name", MAGIC_PARAM_NAME_MAX, 0 },
  115. { "elf_phnum", MAGIC_PARAM_ELF_PHNUM_MAX, 0 },
  116. { "elf_shnum", MAGIC_PARAM_ELF_SHNUM_MAX, 0 },
  117. { "elf_notes", MAGIC_PARAM_ELF_NOTES_MAX, 0 },
  118. };
  119. private char *progname; /* used throughout */
  120. private void usage(void);
  121. private void docprint(const char *);
  122. private void help(void);
  123. private int unwrap(struct magic_set *, const char *);
  124. private int process(struct magic_set *ms, const char *, int);
  125. private struct magic_set *load(const char *, int);
  126. private void setparam(const char *);
  127. private void applyparam(magic_t);
  128. /*
  129. * main - parse arguments and handle options
  130. */
  131. int
  132. main(int argc, char *argv[])
  133. {
  134. int c;
  135. size_t i;
  136. int action = 0, didsomefiles = 0, errflg = 0;
  137. int flags = 0, e = 0;
  138. struct magic_set *magic = NULL;
  139. int longindex;
  140. const char *magicfile = NULL; /* where the magic is */
  141. /* makes islower etc work for other langs */
  142. #ifdef HAVE_SETLOCALE
  143. (void)setlocale(LC_CTYPE, "");
  144. #endif
  145. #ifdef __EMX__
  146. /* sh-like wildcard expansion! Shouldn't hurt at least ... */
  147. _wildcard(&argc, &argv);
  148. #endif
  149. if ((progname = strrchr(argv[0], '/')) != NULL)
  150. progname++;
  151. else
  152. progname = argv[0];
  153. #ifdef S_IFLNK
  154. flags |= getenv("POSIXLY_CORRECT") ? MAGIC_SYMLINK : 0;
  155. #endif
  156. while ((c = getopt_long(argc, argv, OPTSTRING, long_options,
  157. &longindex)) != -1)
  158. switch (c) {
  159. case 0 :
  160. switch (longindex) {
  161. case 0:
  162. help();
  163. break;
  164. case 10:
  165. flags |= MAGIC_APPLE;
  166. break;
  167. case 11:
  168. flags |= MAGIC_MIME_TYPE;
  169. break;
  170. case 12:
  171. flags |= MAGIC_MIME_ENCODING;
  172. break;
  173. }
  174. break;
  175. case '0':
  176. nulsep = 1;
  177. break;
  178. case 'b':
  179. bflag++;
  180. break;
  181. case 'c':
  182. action = FILE_CHECK;
  183. break;
  184. case 'C':
  185. action = FILE_COMPILE;
  186. break;
  187. case 'd':
  188. flags |= MAGIC_DEBUG|MAGIC_CHECK;
  189. break;
  190. case 'E':
  191. flags |= MAGIC_ERROR;
  192. break;
  193. case 'e':
  194. for (i = 0; i < sizeof(nv) / sizeof(nv[0]); i++)
  195. if (strcmp(nv[i].name, optarg) == 0)
  196. break;
  197. if (i == sizeof(nv) / sizeof(nv[0]))
  198. errflg++;
  199. else
  200. flags |= nv[i].value;
  201. break;
  202. case 'f':
  203. if(action)
  204. usage();
  205. if (magic == NULL)
  206. if ((magic = load(magicfile, flags)) == NULL)
  207. return 1;
  208. e |= unwrap(magic, optarg);
  209. ++didsomefiles;
  210. break;
  211. case 'F':
  212. separator = optarg;
  213. break;
  214. case 'i':
  215. flags |= MAGIC_MIME;
  216. break;
  217. case 'k':
  218. flags |= MAGIC_CONTINUE;
  219. break;
  220. case 'l':
  221. action = FILE_LIST;
  222. break;
  223. case 'm':
  224. magicfile = optarg;
  225. break;
  226. case 'n':
  227. ++nobuffer;
  228. break;
  229. case 'N':
  230. ++nopad;
  231. break;
  232. #if defined(HAVE_UTIME) || defined(HAVE_UTIMES)
  233. case 'p':
  234. flags |= MAGIC_PRESERVE_ATIME;
  235. break;
  236. #endif
  237. case 'P':
  238. setparam(optarg);
  239. break;
  240. case 'r':
  241. flags |= MAGIC_RAW;
  242. break;
  243. break;
  244. case 's':
  245. flags |= MAGIC_DEVICES;
  246. break;
  247. case 'v':
  248. if (magicfile == NULL)
  249. magicfile = magic_getpath(magicfile, action);
  250. (void)fprintf(stdout, "%s-%s\n", progname, VERSION);
  251. (void)fprintf(stdout, "magic file from %s\n",
  252. magicfile);
  253. return 0;
  254. case 'z':
  255. flags |= MAGIC_COMPRESS;
  256. break;
  257. #ifdef S_IFLNK
  258. case 'L':
  259. flags |= MAGIC_SYMLINK;
  260. break;
  261. case 'h':
  262. flags &= ~MAGIC_SYMLINK;
  263. break;
  264. #endif
  265. case '?':
  266. default:
  267. errflg++;
  268. break;
  269. }
  270. if (errflg) {
  271. usage();
  272. }
  273. if (e)
  274. return e;
  275. if (MAGIC_VERSION != magic_version())
  276. (void)fprintf(stderr, "%s: compiled magic version [%d] "
  277. "does not match with shared library magic version [%d]\n",
  278. progname, MAGIC_VERSION, magic_version());
  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. applyparam(magic);
  317. }
  318. if (optind == argc) {
  319. if (!didsomefiles)
  320. usage();
  321. }
  322. else {
  323. size_t j, wid, nw;
  324. for (wid = 0, j = (size_t)optind; j < (size_t)argc; j++) {
  325. nw = file_mbswidth(argv[j]);
  326. if (nw > wid)
  327. wid = nw;
  328. }
  329. /*
  330. * If bflag is only set twice, set it depending on
  331. * number of files [this is undocumented, and subject to change]
  332. */
  333. if (bflag == 2) {
  334. bflag = optind >= argc - 1;
  335. }
  336. for (; optind < argc; optind++)
  337. e |= process(magic, argv[optind], wid);
  338. }
  339. if (magic)
  340. magic_close(magic);
  341. return e;
  342. }
  343. private void
  344. applyparam(magic_t magic)
  345. {
  346. size_t i;
  347. for (i = 0; i < __arraycount(pm); i++) {
  348. if (pm[i].value == 0)
  349. continue;
  350. if (magic_setparam(magic, pm[i].tag, &pm[i].value) == -1) {
  351. (void)fprintf(stderr, "%s: Can't set %s %s\n", progname,
  352. pm[i].name, strerror(errno));
  353. exit(1);
  354. }
  355. }
  356. }
  357. private void
  358. setparam(const char *p)
  359. {
  360. size_t i;
  361. char *s;
  362. if ((s = strchr(p, '=')) == NULL)
  363. goto badparm;
  364. for (i = 0; i < __arraycount(pm); i++) {
  365. if (strncmp(p, pm[i].name, s - p) != 0)
  366. continue;
  367. pm[i].value = atoi(s + 1);
  368. return;
  369. }
  370. badparm:
  371. (void)fprintf(stderr, "%s: Unknown param %s\n", progname, p);
  372. exit(1);
  373. }
  374. private struct magic_set *
  375. /*ARGSUSED*/
  376. load(const char *magicfile, int flags)
  377. {
  378. struct magic_set *magic = magic_open(flags);
  379. if (magic == NULL) {
  380. (void)fprintf(stderr, "%s: %s\n", progname, strerror(errno));
  381. return NULL;
  382. }
  383. if (magic_load(magic, magicfile) == -1) {
  384. (void)fprintf(stderr, "%s: %s\n",
  385. progname, magic_error(magic));
  386. magic_close(magic);
  387. return NULL;
  388. }
  389. return magic;
  390. }
  391. /*
  392. * unwrap -- read a file of filenames, do each one.
  393. */
  394. private int
  395. unwrap(struct magic_set *ms, const char *fn)
  396. {
  397. FILE *f;
  398. ssize_t len;
  399. char *line = NULL;
  400. size_t llen = 0;
  401. int wid = 0, cwid;
  402. int e = 0;
  403. if (strcmp("-", fn) == 0) {
  404. f = stdin;
  405. wid = 1;
  406. } else {
  407. if ((f = fopen(fn, "r")) == NULL) {
  408. (void)fprintf(stderr, "%s: Cannot open `%s' (%s).\n",
  409. progname, fn, strerror(errno));
  410. return 1;
  411. }
  412. while ((len = getline(&line, &llen, f)) > 0) {
  413. if (line[len - 1] == '\n')
  414. line[len - 1] = '\0';
  415. cwid = file_mbswidth(line);
  416. if (cwid > wid)
  417. wid = cwid;
  418. }
  419. rewind(f);
  420. }
  421. while ((len = getline(&line, &llen, f)) > 0) {
  422. if (line[len - 1] == '\n')
  423. line[len - 1] = '\0';
  424. e |= process(ms, line, wid);
  425. if(nobuffer)
  426. (void)fflush(stdout);
  427. }
  428. free(line);
  429. (void)fclose(f);
  430. return e;
  431. }
  432. /*
  433. * Called for each input file on the command line (or in a list of files)
  434. */
  435. private int
  436. process(struct magic_set *ms, const char *inname, int wid)
  437. {
  438. const char *type;
  439. int std_in = strcmp(inname, "-") == 0;
  440. if (wid > 0 && !bflag) {
  441. (void)printf("%s", std_in ? "/dev/stdin" : inname);
  442. if (nulsep)
  443. (void)putc('\0', stdout);
  444. (void)printf("%s", separator);
  445. (void)printf("%*s ",
  446. (int) (nopad ? 0 : (wid - file_mbswidth(inname))), "");
  447. }
  448. type = magic_file(ms, std_in ? NULL : inname);
  449. if (type == NULL) {
  450. (void)printf("ERROR: %s\n", magic_error(ms));
  451. return 1;
  452. } else {
  453. (void)printf("%s\n", type);
  454. return 0;
  455. }
  456. }
  457. protected size_t
  458. file_mbswidth(const char *s)
  459. {
  460. #if defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH)
  461. size_t bytesconsumed, old_n, n, width = 0;
  462. mbstate_t state;
  463. wchar_t nextchar;
  464. (void)memset(&state, 0, sizeof(mbstate_t));
  465. old_n = n = strlen(s);
  466. while (n > 0) {
  467. bytesconsumed = mbrtowc(&nextchar, s, n, &state);
  468. if (bytesconsumed == (size_t)(-1) ||
  469. bytesconsumed == (size_t)(-2)) {
  470. /* Something went wrong, return something reasonable */
  471. return old_n;
  472. }
  473. if (s[0] == '\n') {
  474. /*
  475. * do what strlen() would do, so that caller
  476. * is always right
  477. */
  478. width++;
  479. } else {
  480. int w = wcwidth(nextchar);
  481. if (w > 0)
  482. width += w;
  483. }
  484. s += bytesconsumed, n -= bytesconsumed;
  485. }
  486. return width;
  487. #else
  488. return strlen(s);
  489. #endif
  490. }
  491. private void
  492. usage(void)
  493. {
  494. (void)fprintf(stderr, USAGE, progname, progname, progname);
  495. exit(1);
  496. }
  497. private void
  498. docprint(const char *opts)
  499. {
  500. size_t i;
  501. int comma;
  502. char *sp, *p;
  503. p = strstr(opts, "%o");
  504. if (p == NULL) {
  505. fprintf(stdout, "%s", opts);
  506. return;
  507. }
  508. for (sp = p - 1; sp > opts && *sp == ' '; sp--)
  509. continue;
  510. fprintf(stdout, "%.*s", (int)(p - opts), opts);
  511. comma = 0;
  512. for (i = 0; i < __arraycount(nv); i++) {
  513. fprintf(stdout, "%s%s", comma++ ? ", " : "", nv[i].name);
  514. if (i && i % 5 == 0) {
  515. fprintf(stdout, ",\n%*s", (int)(p - sp - 1), "");
  516. comma = 0;
  517. }
  518. }
  519. fprintf(stdout, "%s", opts + (p - opts) + 2);
  520. }
  521. private void
  522. help(void)
  523. {
  524. (void)fputs(
  525. "Usage: file [OPTION...] [FILE...]\n"
  526. "Determine type of FILEs.\n"
  527. "\n", stdout);
  528. #define OPT(shortname, longname, opt, doc) \
  529. fprintf(stdout, " -%c, --" longname, shortname), \
  530. docprint(doc);
  531. #define OPT_LONGONLY(longname, opt, doc) \
  532. fprintf(stdout, " --" longname), \
  533. docprint(doc);
  534. #include "file_opts.h"
  535. #undef OPT
  536. #undef OPT_LONGONLY
  537. fprintf(stdout, "\nReport bugs to http://bugs.gw.com/\n");
  538. exit(0);
  539. }