file.h 13 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450
  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 1.119 2009/02/04 18:24:32 christos Exp $
  31. */
  32. #ifndef __file_h__
  33. #define __file_h__
  34. #ifdef HAVE_CONFIG_H
  35. #include <config.h>
  36. #endif
  37. #include <stdio.h> /* Include that here, to make sure __P gets defined */
  38. #include <errno.h>
  39. #include <fcntl.h> /* For open and flags */
  40. #ifdef HAVE_STDINT_H
  41. #include <stdint.h>
  42. #endif
  43. #ifdef HAVE_INTTYPES_H
  44. #include <inttypes.h>
  45. #endif
  46. #include <regex.h>
  47. #include <sys/types.h>
  48. #include <sys/param.h>
  49. /* Do this here and now, because struct stat gets re-defined on solaris */
  50. #include <sys/stat.h>
  51. #include <stdarg.h>
  52. #define ENABLE_CONDITIONALS
  53. #ifndef MAGIC
  54. #define MAGIC "/etc/magic"
  55. #endif
  56. #ifdef __EMX__
  57. #define PATHSEP ';'
  58. #else
  59. #define PATHSEP ':'
  60. #endif
  61. #define private static
  62. #ifndef protected
  63. #define protected
  64. #endif
  65. #define public
  66. #ifndef __GNUC_PREREQ__
  67. #ifdef __GNUC__
  68. #define __GNUC_PREREQ__(x, y) \
  69. ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \
  70. (__GNUC__ > (x)))
  71. #else
  72. #define __GNUC_PREREQ__(x, y) 0
  73. #endif
  74. #endif
  75. #ifndef __GNUC__
  76. #ifndef __attribute__
  77. #define __attribute__(a)
  78. #endif
  79. #endif
  80. #ifndef MIN
  81. #define MIN(a,b) (((a) < (b)) ? (a) : (b))
  82. #endif
  83. #ifndef MAX
  84. #define MAX(a,b) (((a) > (b)) ? (a) : (b))
  85. #endif
  86. #ifndef HOWMANY
  87. # define HOWMANY (256 * 1024) /* how much of the file to look at */
  88. #endif
  89. #define MAXMAGIS 8192 /* max entries in any one magic file
  90. or directory */
  91. #define MAXDESC 64 /* max leng of text description/MIME type */
  92. #define MAXstring 32 /* max leng of "string" types */
  93. #define MAGICNO 0xF11E041C
  94. #define VERSIONNO 7
  95. #define FILE_MAGICSIZE 200
  96. #define FILE_LOAD 0
  97. #define FILE_CHECK 1
  98. #define FILE_COMPILE 2
  99. union VALUETYPE {
  100. uint8_t b;
  101. uint16_t h;
  102. uint32_t l;
  103. uint64_t q;
  104. uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */
  105. uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */
  106. uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */
  107. char s[MAXstring]; /* the search string or regex pattern */
  108. unsigned char us[MAXstring];
  109. float f;
  110. double d;
  111. };
  112. struct magic {
  113. /* Word 1 */
  114. uint16_t cont_level; /* level of ">" */
  115. uint8_t flag;
  116. #define INDIR 0x01 /* if '(...)' appears */
  117. #define OFFADD 0x02 /* if '>&' or '>...(&' appears */
  118. #define INDIROFFADD 0x04 /* if '>&(' appears */
  119. #define UNSIGNED 0x08 /* comparison is unsigned */
  120. #define NOSPACE 0x10 /* suppress space character before output */
  121. #define BINTEST 0x20 /* test is for a binary type (set only
  122. for top-level tests) */
  123. #define TEXTTEST 0 /* for passing to file_softmagic */
  124. uint8_t factor;
  125. /* Word 2 */
  126. uint8_t reln; /* relation (0=eq, '>'=gt, etc) */
  127. uint8_t vallen; /* length of string value, if any */
  128. uint8_t type; /* comparison type (FILE_*) */
  129. uint8_t in_type; /* type of indirection */
  130. #define FILE_INVALID 0
  131. #define FILE_BYTE 1
  132. #define FILE_SHORT 2
  133. #define FILE_DEFAULT 3
  134. #define FILE_LONG 4
  135. #define FILE_STRING 5
  136. #define FILE_DATE 6
  137. #define FILE_BESHORT 7
  138. #define FILE_BELONG 8
  139. #define FILE_BEDATE 9
  140. #define FILE_LESHORT 10
  141. #define FILE_LELONG 11
  142. #define FILE_LEDATE 12
  143. #define FILE_PSTRING 13
  144. #define FILE_LDATE 14
  145. #define FILE_BELDATE 15
  146. #define FILE_LELDATE 16
  147. #define FILE_REGEX 17
  148. #define FILE_BESTRING16 18
  149. #define FILE_LESTRING16 19
  150. #define FILE_SEARCH 20
  151. #define FILE_MEDATE 21
  152. #define FILE_MELDATE 22
  153. #define FILE_MELONG 23
  154. #define FILE_QUAD 24
  155. #define FILE_LEQUAD 25
  156. #define FILE_BEQUAD 26
  157. #define FILE_QDATE 27
  158. #define FILE_LEQDATE 28
  159. #define FILE_BEQDATE 29
  160. #define FILE_QLDATE 30
  161. #define FILE_LEQLDATE 31
  162. #define FILE_BEQLDATE 32
  163. #define FILE_FLOAT 33
  164. #define FILE_BEFLOAT 34
  165. #define FILE_LEFLOAT 35
  166. #define FILE_DOUBLE 36
  167. #define FILE_BEDOUBLE 37
  168. #define FILE_LEDOUBLE 38
  169. #define FILE_BEID3 39
  170. #define FILE_LEID3 40
  171. #define FILE_INDIRECT 41
  172. #define FILE_NAMES_SIZE 42/* size of array to contain all names */
  173. #define IS_STRING(t) \
  174. ((t) == FILE_STRING || \
  175. (t) == FILE_PSTRING || \
  176. (t) == FILE_BESTRING16 || \
  177. (t) == FILE_LESTRING16 || \
  178. (t) == FILE_REGEX || \
  179. (t) == FILE_SEARCH || \
  180. (t) == FILE_DEFAULT)
  181. #define FILE_FMT_NONE 0
  182. #define FILE_FMT_NUM 1 /* "cduxXi" */
  183. #define FILE_FMT_STR 2 /* "s" */
  184. #define FILE_FMT_QUAD 3 /* "ll" */
  185. #define FILE_FMT_FLOAT 4 /* "eEfFgG" */
  186. #define FILE_FMT_DOUBLE 5 /* "eEfFgG" */
  187. /* Word 3 */
  188. uint8_t in_op; /* operator for indirection */
  189. uint8_t mask_op; /* operator for mask */
  190. #ifdef ENABLE_CONDITIONALS
  191. uint8_t cond; /* conditional type */
  192. #else
  193. uint8_t dummy;
  194. #endif
  195. uint8_t factor_op;
  196. #define FILE_FACTOR_OP_PLUS '+'
  197. #define FILE_FACTOR_OP_MINUS '-'
  198. #define FILE_FACTOR_OP_TIMES '*'
  199. #define FILE_FACTOR_OP_DIV '/'
  200. #define FILE_FACTOR_OP_NONE '\0'
  201. #define FILE_OPS "&|^+-*/%"
  202. #define FILE_OPAND 0
  203. #define FILE_OPOR 1
  204. #define FILE_OPXOR 2
  205. #define FILE_OPADD 3
  206. #define FILE_OPMINUS 4
  207. #define FILE_OPMULTIPLY 5
  208. #define FILE_OPDIVIDE 6
  209. #define FILE_OPMODULO 7
  210. #define FILE_OPS_MASK 0x07 /* mask for above ops */
  211. #define FILE_UNUSED_1 0x08
  212. #define FILE_UNUSED_2 0x10
  213. #define FILE_UNUSED_3 0x20
  214. #define FILE_OPINVERSE 0x40
  215. #define FILE_OPINDIRECT 0x80
  216. #ifdef ENABLE_CONDITIONALS
  217. #define COND_NONE 0
  218. #define COND_IF 1
  219. #define COND_ELIF 2
  220. #define COND_ELSE 3
  221. #endif /* ENABLE_CONDITIONALS */
  222. /* Word 4 */
  223. uint32_t offset; /* offset to magic number */
  224. /* Word 5 */
  225. int32_t in_offset; /* offset from indirection */
  226. /* Word 6 */
  227. uint32_t lineno; /* line number in magic file */
  228. /* Word 7,8 */
  229. union {
  230. uint64_t _mask; /* for use with numeric and date types */
  231. struct {
  232. uint32_t _count; /* repeat/line count */
  233. uint32_t _flags; /* modifier flags */
  234. } _s; /* for use with string types */
  235. } _u;
  236. #define num_mask _u._mask
  237. #define str_range _u._s._count
  238. #define str_flags _u._s._flags
  239. /* Words 9-16 */
  240. union VALUETYPE value; /* either number or string */
  241. /* Words 17-24 */
  242. char desc[MAXDESC]; /* description */
  243. /* Words 25-32 */
  244. char mimetype[MAXDESC]; /* MIME type */
  245. /* Words 33-34 */
  246. char apple[8];
  247. };
  248. #define BIT(A) (1 << (A))
  249. #define STRING_COMPACT_BLANK BIT(0)
  250. #define STRING_COMPACT_OPTIONAL_BLANK BIT(1)
  251. #define STRING_IGNORE_LOWERCASE BIT(2)
  252. #define STRING_IGNORE_UPPERCASE BIT(3)
  253. #define REGEX_OFFSET_START BIT(4)
  254. #define CHAR_COMPACT_BLANK 'B'
  255. #define CHAR_COMPACT_OPTIONAL_BLANK 'b'
  256. #define CHAR_IGNORE_LOWERCASE 'c'
  257. #define CHAR_IGNORE_UPPERCASE 'C'
  258. #define CHAR_REGEX_OFFSET_START 's'
  259. #define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE)
  260. #define STRING_DEFAULT_RANGE 100
  261. /* list of magic entries */
  262. struct mlist {
  263. struct magic *magic; /* array of magic entries */
  264. uint32_t nmagic; /* number of entries in array */
  265. int mapped; /* allocation type: 0 => apprentice_file
  266. * 1 => apprentice_map + malloc
  267. * 2 => apprentice_map + mmap */
  268. struct mlist *next, *prev;
  269. };
  270. #ifdef __cplusplus
  271. #define CAST(T, b) static_cast<T>(b)
  272. #else
  273. #define CAST(T, b) (b)
  274. #endif
  275. struct level_info {
  276. int32_t off;
  277. int got_match;
  278. #ifdef ENABLE_CONDITIONALS
  279. int last_match;
  280. int last_cond; /* used for error checking by parse() */
  281. #endif
  282. };
  283. struct magic_set {
  284. struct mlist *mlist;
  285. struct cont {
  286. size_t len;
  287. struct level_info *li;
  288. } c;
  289. struct out {
  290. char *buf; /* Accumulation buffer */
  291. char *pbuf; /* Printable buffer */
  292. } o;
  293. uint32_t offset;
  294. int error;
  295. int flags; /* Control magic tests. */
  296. int event_flags; /* Note things that happened. */
  297. #define EVENT_HAD_ERR 0x01
  298. const char *file;
  299. size_t line; /* current magic line number */
  300. /* data for searches */
  301. struct {
  302. const char *s; /* start of search in original source */
  303. size_t s_len; /* length of search region */
  304. size_t offset; /* starting offset in source: XXX - should this be off_t? */
  305. size_t rm_len; /* match length */
  306. } search;
  307. /* FIXME: Make the string dynamically allocated so that e.g.
  308. strings matched in files can be longer than MAXstring */
  309. union VALUETYPE ms_value; /* either number or string */
  310. };
  311. /* Type for Unicode characters */
  312. typedef unsigned long unichar;
  313. struct stat;
  314. protected const char *file_fmttime(uint32_t, int);
  315. protected int file_buffer(struct magic_set *, int, const char *, const void *,
  316. size_t);
  317. protected int file_fsmagic(struct magic_set *, const char *, struct stat *);
  318. protected int file_pipe2file(struct magic_set *, int, const void *, size_t);
  319. protected int file_vprintf(struct magic_set *, const char *, va_list);
  320. protected int file_printf(struct magic_set *, const char *, ...)
  321. __attribute__((__format__(__printf__, 2, 3)));
  322. protected int file_reset(struct magic_set *);
  323. protected int file_tryelf(struct magic_set *, int, const unsigned char *,
  324. size_t);
  325. protected int file_trycdf(struct magic_set *, int, const unsigned char *,
  326. size_t);
  327. protected int file_zmagic(struct magic_set *, int, const char *,
  328. const unsigned char *, size_t);
  329. protected int file_ascmagic(struct magic_set *, const unsigned char *, size_t);
  330. protected int file_ascmagic_with_encoding(struct magic_set *,
  331. const unsigned char *, size_t, unichar *, size_t, const char *,
  332. const char *);
  333. protected int file_encoding(struct magic_set *, const unsigned char *, size_t,
  334. unichar **, size_t *, const char **, const char **, const char **);
  335. protected int file_is_tar(struct magic_set *, const unsigned char *, size_t);
  336. protected int file_softmagic(struct magic_set *, const unsigned char *, size_t,
  337. int);
  338. protected struct mlist *file_apprentice(struct magic_set *, const char *, int);
  339. protected uint64_t file_signextend(struct magic_set *, struct magic *,
  340. uint64_t);
  341. protected void file_delmagic(struct magic *, int type, size_t entries);
  342. protected void file_badread(struct magic_set *);
  343. protected void file_badseek(struct magic_set *);
  344. protected void file_oomem(struct magic_set *, size_t);
  345. protected void file_error(struct magic_set *, int, const char *, ...)
  346. __attribute__((__format__(__printf__, 3, 4)));
  347. protected void file_magerror(struct magic_set *, const char *, ...)
  348. __attribute__((__format__(__printf__, 2, 3)));
  349. protected void file_magwarn(struct magic_set *, const char *, ...)
  350. __attribute__((__format__(__printf__, 2, 3)));
  351. protected void file_mdump(struct magic *);
  352. protected void file_showstr(FILE *, const char *, size_t);
  353. protected size_t file_mbswidth(const char *);
  354. protected const char *file_getbuffer(struct magic_set *);
  355. protected ssize_t sread(int, void *, size_t, int);
  356. protected int file_check_mem(struct magic_set *, unsigned int);
  357. protected int file_looks_utf8(const unsigned char *, size_t, unichar *,
  358. size_t *);
  359. #ifdef __EMX__
  360. protected int file_os2_apptype(struct magic_set *, const char *, const void *,
  361. size_t);
  362. #endif /* __EMX__ */
  363. #ifndef COMPILE_ONLY
  364. extern const char *file_names[];
  365. extern const size_t file_nnames;
  366. #endif
  367. #ifndef HAVE_STRERROR
  368. extern int sys_nerr;
  369. extern char *sys_errlist[];
  370. #define strerror(e) \
  371. (((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error")
  372. #endif
  373. #ifndef HAVE_STRTOUL
  374. #define strtoul(a, b, c) strtol(a, b, c)
  375. #endif
  376. #ifndef HAVE_VASPRINTF
  377. int vasprintf(char **, const char *, va_list);
  378. #endif
  379. #ifndef HAVE_ASPRINTF
  380. int asprintf(char **ptr, const char *format_string, ...);
  381. #endif
  382. #ifndef HAVE_STRLCPY
  383. size_t strlcpy(char *dst, const char *src, size_t siz);
  384. #endif
  385. #ifndef HAVE_STRLCAT
  386. size_t strlcat(char *dst, const char *src, size_t siz);
  387. #endif
  388. #if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK)
  389. #define QUICK
  390. #endif
  391. #ifndef O_BINARY
  392. #define O_BINARY 0
  393. #endif
  394. #ifndef __cplusplus
  395. #ifdef __GNUC__
  396. #define FILE_RCSID(id) \
  397. static const char rcsid[] __attribute__((__used__)) = id;
  398. #else
  399. #define FILE_RCSID(id) \
  400. static const char *rcsid(const char *p) { \
  401. return rcsid(p = id); \
  402. }
  403. #endif
  404. #else
  405. #define FILE_RCSID(id)
  406. #endif
  407. #endif /* __file_h__ */