file.h 15 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503
  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.h - definitions for file(1) program
  30. * @(#)$File: file.h,v 1.135 2011/09/20 15:30:14 christos Exp $
  31. */
  32. #ifndef __file_h__
  33. #define __file_h__
  34. #ifdef HAVE_CONFIG_H
  35. #include <config.h>
  36. #endif
  37. #ifdef WIN32
  38. #ifdef _WIN64
  39. #define SIZE_T_FORMAT "I64"
  40. #else
  41. #define SIZE_T_FORMAT ""
  42. #endif
  43. #define INT64_T_FORMAT "I64"
  44. #else
  45. #define SIZE_T_FORMAT "z"
  46. #define INT64_T_FORMAT "ll"
  47. #endif
  48. #include <stdio.h> /* Include that here, to make sure __P gets defined */
  49. #include <errno.h>
  50. #include <fcntl.h> /* For open and flags */
  51. #ifdef HAVE_STDINT_H
  52. #ifndef __STDC_LIMIT_MACROS
  53. #define __STDC_LIMIT_MACROS
  54. #endif
  55. #include <stdint.h>
  56. #endif
  57. #ifdef HAVE_INTTYPES_H
  58. #include <inttypes.h>
  59. #endif
  60. #include <regex.h>
  61. #include <sys/types.h>
  62. #include <sys/param.h>
  63. /* Do this here and now, because struct stat gets re-defined on solaris */
  64. #include <sys/stat.h>
  65. #include <stdarg.h>
  66. #define ENABLE_CONDITIONALS
  67. #ifndef MAGIC
  68. #define MAGIC "/etc/magic"
  69. #endif
  70. #if defined(__EMX__) || defined (WIN32)
  71. #define PATHSEP ';'
  72. #else
  73. #define PATHSEP ':'
  74. #endif
  75. #define private static
  76. #ifndef protected
  77. #define protected
  78. #endif
  79. #define public
  80. #ifndef __arraycount
  81. #define __arraycount(a) (sizeof(a) / sizeof(a[0]))
  82. #endif
  83. #ifndef __GNUC_PREREQ__
  84. #ifdef __GNUC__
  85. #define __GNUC_PREREQ__(x, y) \
  86. ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \
  87. (__GNUC__ > (x)))
  88. #else
  89. #define __GNUC_PREREQ__(x, y) 0
  90. #endif
  91. #endif
  92. #ifndef __GNUC__
  93. #ifndef __attribute__
  94. #define __attribute__(a)
  95. #endif
  96. #endif
  97. #ifndef MIN
  98. #define MIN(a,b) (((a) < (b)) ? (a) : (b))
  99. #endif
  100. #ifndef MAX
  101. #define MAX(a,b) (((a) > (b)) ? (a) : (b))
  102. #endif
  103. #ifndef HOWMANY
  104. # define HOWMANY (256 * 1024) /* how much of the file to look at */
  105. #endif
  106. #define MAXMAGIS 8192 /* max entries in any one magic file
  107. or directory */
  108. #define MAXDESC 64 /* max leng of text description/MIME type */
  109. #define MAXstring 64 /* max leng of "string" types */
  110. #define MAGICNO 0xF11E041C
  111. #define VERSIONNO 8
  112. #define FILE_MAGICSIZE 232
  113. #define FILE_LOAD 0
  114. #define FILE_CHECK 1
  115. #define FILE_COMPILE 2
  116. #define FILE_LIST 3
  117. union VALUETYPE {
  118. uint8_t b;
  119. uint16_t h;
  120. uint32_t l;
  121. uint64_t q;
  122. uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */
  123. uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */
  124. uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */
  125. char s[MAXstring]; /* the search string or regex pattern */
  126. unsigned char us[MAXstring];
  127. float f;
  128. double d;
  129. };
  130. struct magic {
  131. /* Word 1 */
  132. uint16_t cont_level; /* level of ">" */
  133. uint8_t flag;
  134. #define INDIR 0x01 /* if '(...)' appears */
  135. #define OFFADD 0x02 /* if '>&' or '>...(&' appears */
  136. #define INDIROFFADD 0x04 /* if '>&(' appears */
  137. #define UNSIGNED 0x08 /* comparison is unsigned */
  138. #define NOSPACE 0x10 /* suppress space character before output */
  139. #define BINTEST 0x20 /* test is for a binary type (set only
  140. for top-level tests) */
  141. #define TEXTTEST 0x40 /* for passing to file_softmagic */
  142. uint8_t factor;
  143. /* Word 2 */
  144. uint8_t reln; /* relation (0=eq, '>'=gt, etc) */
  145. uint8_t vallen; /* length of string value, if any */
  146. uint8_t type; /* comparison type (FILE_*) */
  147. uint8_t in_type; /* type of indirection */
  148. #define FILE_INVALID 0
  149. #define FILE_BYTE 1
  150. #define FILE_SHORT 2
  151. #define FILE_DEFAULT 3
  152. #define FILE_LONG 4
  153. #define FILE_STRING 5
  154. #define FILE_DATE 6
  155. #define FILE_BESHORT 7
  156. #define FILE_BELONG 8
  157. #define FILE_BEDATE 9
  158. #define FILE_LESHORT 10
  159. #define FILE_LELONG 11
  160. #define FILE_LEDATE 12
  161. #define FILE_PSTRING 13
  162. #define FILE_LDATE 14
  163. #define FILE_BELDATE 15
  164. #define FILE_LELDATE 16
  165. #define FILE_REGEX 17
  166. #define FILE_BESTRING16 18
  167. #define FILE_LESTRING16 19
  168. #define FILE_SEARCH 20
  169. #define FILE_MEDATE 21
  170. #define FILE_MELDATE 22
  171. #define FILE_MELONG 23
  172. #define FILE_QUAD 24
  173. #define FILE_LEQUAD 25
  174. #define FILE_BEQUAD 26
  175. #define FILE_QDATE 27
  176. #define FILE_LEQDATE 28
  177. #define FILE_BEQDATE 29
  178. #define FILE_QLDATE 30
  179. #define FILE_LEQLDATE 31
  180. #define FILE_BEQLDATE 32
  181. #define FILE_FLOAT 33
  182. #define FILE_BEFLOAT 34
  183. #define FILE_LEFLOAT 35
  184. #define FILE_DOUBLE 36
  185. #define FILE_BEDOUBLE 37
  186. #define FILE_LEDOUBLE 38
  187. #define FILE_BEID3 39
  188. #define FILE_LEID3 40
  189. #define FILE_INDIRECT 41
  190. #define FILE_NAMES_SIZE 42/* size of array to contain all names */
  191. #define IS_STRING(t) \
  192. ((t) == FILE_STRING || \
  193. (t) == FILE_PSTRING || \
  194. (t) == FILE_BESTRING16 || \
  195. (t) == FILE_LESTRING16 || \
  196. (t) == FILE_REGEX || \
  197. (t) == FILE_SEARCH || \
  198. (t) == FILE_DEFAULT)
  199. #define FILE_FMT_NONE 0
  200. #define FILE_FMT_NUM 1 /* "cduxXi" */
  201. #define FILE_FMT_STR 2 /* "s" */
  202. #define FILE_FMT_QUAD 3 /* "ll" */
  203. #define FILE_FMT_FLOAT 4 /* "eEfFgG" */
  204. #define FILE_FMT_DOUBLE 5 /* "eEfFgG" */
  205. /* Word 3 */
  206. uint8_t in_op; /* operator for indirection */
  207. uint8_t mask_op; /* operator for mask */
  208. #ifdef ENABLE_CONDITIONALS
  209. uint8_t cond; /* conditional type */
  210. #else
  211. uint8_t dummy;
  212. #endif
  213. uint8_t factor_op;
  214. #define FILE_FACTOR_OP_PLUS '+'
  215. #define FILE_FACTOR_OP_MINUS '-'
  216. #define FILE_FACTOR_OP_TIMES '*'
  217. #define FILE_FACTOR_OP_DIV '/'
  218. #define FILE_FACTOR_OP_NONE '\0'
  219. #define FILE_OPS "&|^+-*/%"
  220. #define FILE_OPAND 0
  221. #define FILE_OPOR 1
  222. #define FILE_OPXOR 2
  223. #define FILE_OPADD 3
  224. #define FILE_OPMINUS 4
  225. #define FILE_OPMULTIPLY 5
  226. #define FILE_OPDIVIDE 6
  227. #define FILE_OPMODULO 7
  228. #define FILE_OPS_MASK 0x07 /* mask for above ops */
  229. #define FILE_UNUSED_1 0x08
  230. #define FILE_UNUSED_2 0x10
  231. #define FILE_UNUSED_3 0x20
  232. #define FILE_OPINVERSE 0x40
  233. #define FILE_OPINDIRECT 0x80
  234. #ifdef ENABLE_CONDITIONALS
  235. #define COND_NONE 0
  236. #define COND_IF 1
  237. #define COND_ELIF 2
  238. #define COND_ELSE 3
  239. #endif /* ENABLE_CONDITIONALS */
  240. /* Word 4 */
  241. uint32_t offset; /* offset to magic number */
  242. /* Word 5 */
  243. int32_t in_offset; /* offset from indirection */
  244. /* Word 6 */
  245. uint32_t lineno; /* line number in magic file */
  246. /* Word 7,8 */
  247. union {
  248. uint64_t _mask; /* for use with numeric and date types */
  249. struct {
  250. uint32_t _count; /* repeat/line count */
  251. uint32_t _flags; /* modifier flags */
  252. } _s; /* for use with string types */
  253. } _u;
  254. #define num_mask _u._mask
  255. #define str_range _u._s._count
  256. #define str_flags _u._s._flags
  257. /* Words 9-16 */
  258. union VALUETYPE value; /* either number or string */
  259. /* Words 17-32 */
  260. char desc[MAXDESC]; /* description */
  261. /* Words 33-48 */
  262. char mimetype[MAXDESC]; /* MIME type */
  263. /* Words 49-50 */
  264. char apple[8];
  265. };
  266. #define BIT(A) (1 << (A))
  267. #define STRING_COMPACT_WHITESPACE BIT(0)
  268. #define STRING_COMPACT_OPTIONAL_WHITESPACE BIT(1)
  269. #define STRING_IGNORE_LOWERCASE BIT(2)
  270. #define STRING_IGNORE_UPPERCASE BIT(3)
  271. #define REGEX_OFFSET_START BIT(4)
  272. #define STRING_TEXTTEST BIT(5)
  273. #define STRING_BINTEST BIT(6)
  274. #define PSTRING_1_BE BIT(7)
  275. #define PSTRING_1_LE BIT(7)
  276. #define PSTRING_2_BE BIT(8)
  277. #define PSTRING_2_LE BIT(9)
  278. #define PSTRING_4_BE BIT(10)
  279. #define PSTRING_4_LE BIT(11)
  280. #define PSTRING_LEN \
  281. (PSTRING_1_BE|PSTRING_2_LE|PSTRING_2_BE|PSTRING_4_LE|PSTRING_4_BE)
  282. #define PSTRING_LENGTH_INCLUDES_ITSELF BIT(12)
  283. #define CHAR_COMPACT_WHITESPACE 'W'
  284. #define CHAR_COMPACT_OPTIONAL_WHITESPACE 'w'
  285. #define CHAR_IGNORE_LOWERCASE 'c'
  286. #define CHAR_IGNORE_UPPERCASE 'C'
  287. #define CHAR_REGEX_OFFSET_START 's'
  288. #define CHAR_TEXTTEST 't'
  289. #define CHAR_BINTEST 'b'
  290. #define CHAR_PSTRING_1_BE 'B'
  291. #define CHAR_PSTRING_1_LE 'B'
  292. #define CHAR_PSTRING_2_BE 'H'
  293. #define CHAR_PSTRING_2_LE 'h'
  294. #define CHAR_PSTRING_4_BE 'L'
  295. #define CHAR_PSTRING_4_LE 'l'
  296. #define CHAR_PSTRING_LENGTH_INCLUDES_ITSELF 'J'
  297. #define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE)
  298. #define STRING_DEFAULT_RANGE 100
  299. /* list of magic entries */
  300. struct mlist {
  301. struct magic *magic; /* array of magic entries */
  302. uint32_t nmagic; /* number of entries in array */
  303. int mapped; /* allocation type: 0 => apprentice_file
  304. * 1 => apprentice_map + malloc
  305. * 2 => apprentice_map + mmap */
  306. struct mlist *next, *prev;
  307. };
  308. #ifdef __cplusplus
  309. #define CAST(T, b) static_cast<T>(b)
  310. #define RCAST(T, b) reinterpret_cast<T>(b)
  311. #else
  312. #define CAST(T, b) (T)(b)
  313. #define RCAST(T, b) (T)(b)
  314. #endif
  315. struct level_info {
  316. int32_t off;
  317. int got_match;
  318. #ifdef ENABLE_CONDITIONALS
  319. int last_match;
  320. int last_cond; /* used for error checking by parse() */
  321. #endif
  322. };
  323. struct magic_set {
  324. struct mlist *mlist;
  325. struct cont {
  326. size_t len;
  327. struct level_info *li;
  328. } c;
  329. struct out {
  330. char *buf; /* Accumulation buffer */
  331. char *pbuf; /* Printable buffer */
  332. } o;
  333. uint32_t offset;
  334. int error;
  335. int flags; /* Control magic tests. */
  336. int event_flags; /* Note things that happened. */
  337. #define EVENT_HAD_ERR 0x01
  338. const char *file;
  339. size_t line; /* current magic line number */
  340. /* data for searches */
  341. struct {
  342. const char *s; /* start of search in original source */
  343. size_t s_len; /* length of search region */
  344. size_t offset; /* starting offset in source: XXX - should this be off_t? */
  345. size_t rm_len; /* match length */
  346. } search;
  347. /* FIXME: Make the string dynamically allocated so that e.g.
  348. strings matched in files can be longer than MAXstring */
  349. union VALUETYPE ms_value; /* either number or string */
  350. };
  351. /* Type for Unicode characters */
  352. typedef unsigned long unichar;
  353. struct stat;
  354. protected const char *file_fmttime(uint32_t, int);
  355. protected int file_buffer(struct magic_set *, int, const char *, const void *,
  356. size_t);
  357. protected int file_fsmagic(struct magic_set *, const char *, struct stat *);
  358. protected int file_pipe2file(struct magic_set *, int, const void *, size_t);
  359. protected int file_vprintf(struct magic_set *, const char *, va_list);
  360. protected size_t file_printedlen(const struct magic_set *);
  361. protected int file_replace(struct magic_set *, const char *, const char *);
  362. protected int file_printf(struct magic_set *, const char *, ...)
  363. __attribute__((__format__(__printf__, 2, 3)));
  364. protected int file_reset(struct magic_set *);
  365. protected int file_tryelf(struct magic_set *, int, const unsigned char *,
  366. size_t);
  367. protected int file_trycdf(struct magic_set *, int, const unsigned char *,
  368. size_t);
  369. #if HAVE_FORK
  370. protected int file_zmagic(struct magic_set *, int, const char *,
  371. const unsigned char *, size_t);
  372. #endif
  373. protected int file_ascmagic(struct magic_set *, const unsigned char *, size_t,
  374. int);
  375. protected int file_ascmagic_with_encoding(struct magic_set *,
  376. const unsigned char *, size_t, unichar *, size_t, const char *,
  377. const char *, int);
  378. protected int file_encoding(struct magic_set *, const unsigned char *, size_t,
  379. unichar **, size_t *, const char **, const char **, const char **);
  380. protected int file_is_tar(struct magic_set *, const unsigned char *, size_t);
  381. protected int file_softmagic(struct magic_set *, const unsigned char *, size_t,
  382. int, int);
  383. protected struct mlist *file_apprentice(struct magic_set *, const char *, int);
  384. protected uint64_t file_signextend(struct magic_set *, struct magic *,
  385. uint64_t);
  386. protected void file_delmagic(struct magic *, int type, size_t entries);
  387. protected void file_badread(struct magic_set *);
  388. protected void file_badseek(struct magic_set *);
  389. protected void file_oomem(struct magic_set *, size_t);
  390. protected void file_error(struct magic_set *, int, const char *, ...)
  391. __attribute__((__format__(__printf__, 3, 4)));
  392. protected void file_magerror(struct magic_set *, const char *, ...)
  393. __attribute__((__format__(__printf__, 2, 3)));
  394. protected void file_magwarn(struct magic_set *, const char *, ...)
  395. __attribute__((__format__(__printf__, 2, 3)));
  396. protected void file_mdump(struct magic *);
  397. protected void file_showstr(FILE *, const char *, size_t);
  398. protected size_t file_mbswidth(const char *);
  399. protected const char *file_getbuffer(struct magic_set *);
  400. protected ssize_t sread(int, void *, size_t, int);
  401. protected int file_check_mem(struct magic_set *, unsigned int);
  402. protected int file_looks_utf8(const unsigned char *, size_t, unichar *,
  403. size_t *);
  404. protected size_t file_pstring_length_size(const struct magic *);
  405. protected size_t file_pstring_get_length(const struct magic *, const char *);
  406. #ifdef __EMX__
  407. protected int file_os2_apptype(struct magic_set *, const char *, const void *,
  408. size_t);
  409. #endif /* __EMX__ */
  410. #ifndef COMPILE_ONLY
  411. extern const char *file_names[];
  412. extern const size_t file_nnames;
  413. #endif
  414. #ifndef HAVE_STRERROR
  415. extern int sys_nerr;
  416. extern char *sys_errlist[];
  417. #define strerror(e) \
  418. (((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error")
  419. #endif
  420. #ifndef HAVE_STRTOUL
  421. #define strtoul(a, b, c) strtol(a, b, c)
  422. #endif
  423. #ifndef HAVE_VASPRINTF
  424. int vasprintf(char **, const char *, va_list);
  425. #endif
  426. #ifndef HAVE_ASPRINTF
  427. int asprintf(char **ptr, const char *format_string, ...);
  428. #endif
  429. #ifndef HAVE_STRLCPY
  430. size_t strlcpy(char *dst, const char *src, size_t siz);
  431. #endif
  432. #ifndef HAVE_STRLCAT
  433. size_t strlcat(char *dst, const char *src, size_t siz);
  434. #endif
  435. #ifndef HAVE_GETLINE
  436. ssize_t getline(char **dst, size_t *len, FILE *fp);
  437. ssize_t getdelim(char **dst, size_t *len, int delimiter, FILE *fp);
  438. #endif
  439. #if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK)
  440. #define QUICK
  441. #endif
  442. #ifndef O_BINARY
  443. #define O_BINARY 0
  444. #endif
  445. #ifndef __cplusplus
  446. #if defined(__GNUC__) && (__GNUC__ >= 3)
  447. #define FILE_RCSID(id) \
  448. static const char rcsid[] __attribute__((__used__)) = id;
  449. #else
  450. #define FILE_RCSID(id) \
  451. static const char *rcsid(const char *p) { \
  452. return rcsid(p = id); \
  453. }
  454. #endif
  455. #else
  456. #define FILE_RCSID(id)
  457. #endif
  458. #endif /* __file_h__ */