file.h 15 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529
  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 1.148 2014/02/12 23:20:53 christos Exp $
  31. */
  32. #ifndef __file_h__
  33. #define __file_h__
  34. #include "config.h"
  35. #ifdef PHP_WIN32
  36. #ifdef _WIN64
  37. #define SIZE_T_FORMAT "I64"
  38. #else
  39. #define SIZE_T_FORMAT ""
  40. #endif
  41. #define INT64_T_FORMAT "I64"
  42. #else
  43. #define SIZE_T_FORMAT "z"
  44. #define INT64_T_FORMAT "ll"
  45. #endif
  46. #include <stdio.h> /* Include that here, to make sure __P gets defined */
  47. #include <errno.h>
  48. #include <fcntl.h> /* For open and flags */
  49. #ifdef HAVE_STDINT_H
  50. #ifndef __STDC_LIMIT_MACROS
  51. #define __STDC_LIMIT_MACROS
  52. #endif
  53. #include <stdint.h>
  54. #endif
  55. #ifdef HAVE_INTTYPES_H
  56. #include <inttypes.h>
  57. #endif
  58. #ifdef PHP_WIN32
  59. #include "win32/php_stdint.h"
  60. #endif
  61. #include "php.h"
  62. #include "ext/standard/php_string.h"
  63. #include "ext/pcre/php_pcre.h"
  64. #include <sys/types.h>
  65. #ifdef PHP_WIN32
  66. #include "win32/param.h"
  67. #else
  68. #include <sys/param.h>
  69. #endif
  70. /* Do this here and now, because struct stat gets re-defined on solaris */
  71. #include <sys/stat.h>
  72. #include <stdarg.h>
  73. #define ENABLE_CONDITIONALS
  74. #ifndef MAGIC
  75. #define MAGIC "/etc/magic"
  76. #endif
  77. #if defined(__EMX__) || defined(PHP_WIN32)
  78. #define PATHSEP ';'
  79. #else
  80. #define PATHSEP ':'
  81. #endif
  82. #define private static
  83. #if HAVE_VISIBILITY
  84. #define public __attribute__ ((__visibility__("default")))
  85. #ifndef protected
  86. #define protected __attribute__ ((__visibility__("hidden")))
  87. #endif
  88. #else
  89. #define public
  90. #ifndef protected
  91. #define protected
  92. #endif
  93. #endif
  94. #ifndef __arraycount
  95. #define __arraycount(a) (sizeof(a) / sizeof(a[0]))
  96. #endif
  97. #ifndef __GNUC_PREREQ__
  98. #ifdef __GNUC__
  99. #define __GNUC_PREREQ__(x, y) \
  100. ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \
  101. (__GNUC__ > (x)))
  102. #else
  103. #define __GNUC_PREREQ__(x, y) 0
  104. #endif
  105. #endif
  106. #ifndef MIN
  107. #define MIN(a,b) (((a) < (b)) ? (a) : (b))
  108. #endif
  109. #ifndef MAX
  110. #define MAX(a,b) (((a) > (b)) ? (a) : (b))
  111. #endif
  112. #ifndef HOWMANY
  113. # define HOWMANY (256 * 1024) /* how much of the file to look at */
  114. #endif
  115. #define MAXMAGIS 8192 /* max entries in any one magic file
  116. or directory */
  117. #define MAXDESC 64 /* max len of text description/MIME type */
  118. #define MAXMIME 80 /* max len of text MIME type */
  119. #define MAXstring 64 /* max len of "string" types */
  120. #define MAGICNO 0xF11E041C
  121. #define VERSIONNO 11
  122. #define FILE_MAGICSIZE 248
  123. #define FILE_LOAD 0
  124. #define FILE_CHECK 1
  125. #define FILE_COMPILE 2
  126. #define FILE_LIST 3
  127. union VALUETYPE {
  128. uint8_t b;
  129. uint16_t h;
  130. uint32_t l;
  131. uint64_t q;
  132. uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */
  133. uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */
  134. uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */
  135. char s[MAXstring]; /* the search string or regex pattern */
  136. unsigned char us[MAXstring];
  137. float f;
  138. double d;
  139. };
  140. struct magic {
  141. /* Word 1 */
  142. uint16_t cont_level; /* level of ">" */
  143. uint8_t flag;
  144. #define INDIR 0x01 /* if '(...)' appears */
  145. #define OFFADD 0x02 /* if '>&' or '>...(&' appears */
  146. #define INDIROFFADD 0x04 /* if '>&(' appears */
  147. #define UNSIGNED 0x08 /* comparison is unsigned */
  148. #define NOSPACE 0x10 /* suppress space character before output */
  149. #define BINTEST 0x20 /* test is for a binary type (set only
  150. for top-level tests) */
  151. #define TEXTTEST 0x40 /* for passing to file_softmagic */
  152. uint8_t factor;
  153. /* Word 2 */
  154. uint8_t reln; /* relation (0=eq, '>'=gt, etc) */
  155. uint8_t vallen; /* length of string value, if any */
  156. uint8_t type; /* comparison type (FILE_*) */
  157. uint8_t in_type; /* type of indirection */
  158. #define FILE_INVALID 0
  159. #define FILE_BYTE 1
  160. #define FILE_SHORT 2
  161. #define FILE_DEFAULT 3
  162. #define FILE_LONG 4
  163. #define FILE_STRING 5
  164. #define FILE_DATE 6
  165. #define FILE_BESHORT 7
  166. #define FILE_BELONG 8
  167. #define FILE_BEDATE 9
  168. #define FILE_LESHORT 10
  169. #define FILE_LELONG 11
  170. #define FILE_LEDATE 12
  171. #define FILE_PSTRING 13
  172. #define FILE_LDATE 14
  173. #define FILE_BELDATE 15
  174. #define FILE_LELDATE 16
  175. #define FILE_REGEX 17
  176. #define FILE_BESTRING16 18
  177. #define FILE_LESTRING16 19
  178. #define FILE_SEARCH 20
  179. #define FILE_MEDATE 21
  180. #define FILE_MELDATE 22
  181. #define FILE_MELONG 23
  182. #define FILE_QUAD 24
  183. #define FILE_LEQUAD 25
  184. #define FILE_BEQUAD 26
  185. #define FILE_QDATE 27
  186. #define FILE_LEQDATE 28
  187. #define FILE_BEQDATE 29
  188. #define FILE_QLDATE 30
  189. #define FILE_LEQLDATE 31
  190. #define FILE_BEQLDATE 32
  191. #define FILE_FLOAT 33
  192. #define FILE_BEFLOAT 34
  193. #define FILE_LEFLOAT 35
  194. #define FILE_DOUBLE 36
  195. #define FILE_BEDOUBLE 37
  196. #define FILE_LEDOUBLE 38
  197. #define FILE_BEID3 39
  198. #define FILE_LEID3 40
  199. #define FILE_INDIRECT 41
  200. #define FILE_QWDATE 42
  201. #define FILE_LEQWDATE 43
  202. #define FILE_BEQWDATE 44
  203. #define FILE_NAME 45
  204. #define FILE_USE 46
  205. #define FILE_CLEAR 47
  206. #define FILE_NAMES_SIZE 48 /* size of array to contain all names */
  207. #define IS_LIBMAGIC_STRING(t) \
  208. ((t) == FILE_STRING || \
  209. (t) == FILE_PSTRING || \
  210. (t) == FILE_BESTRING16 || \
  211. (t) == FILE_LESTRING16 || \
  212. (t) == FILE_REGEX || \
  213. (t) == FILE_SEARCH || \
  214. (t) == FILE_NAME || \
  215. (t) == FILE_USE)
  216. #define FILE_FMT_NONE 0
  217. #define FILE_FMT_NUM 1 /* "cduxXi" */
  218. #define FILE_FMT_STR 2 /* "s" */
  219. #define FILE_FMT_QUAD 3 /* "ll" */
  220. #define FILE_FMT_FLOAT 4 /* "eEfFgG" */
  221. #define FILE_FMT_DOUBLE 5 /* "eEfFgG" */
  222. /* Word 3 */
  223. uint8_t in_op; /* operator for indirection */
  224. uint8_t mask_op; /* operator for mask */
  225. #ifdef ENABLE_CONDITIONALS
  226. uint8_t cond; /* conditional type */
  227. #else
  228. uint8_t dummy;
  229. #endif
  230. uint8_t factor_op;
  231. #define FILE_FACTOR_OP_PLUS '+'
  232. #define FILE_FACTOR_OP_MINUS '-'
  233. #define FILE_FACTOR_OP_TIMES '*'
  234. #define FILE_FACTOR_OP_DIV '/'
  235. #define FILE_FACTOR_OP_NONE '\0'
  236. #define FILE_OPS "&|^+-*/%"
  237. #define FILE_OPAND 0
  238. #define FILE_OPOR 1
  239. #define FILE_OPXOR 2
  240. #define FILE_OPADD 3
  241. #define FILE_OPMINUS 4
  242. #define FILE_OPMULTIPLY 5
  243. #define FILE_OPDIVIDE 6
  244. #define FILE_OPMODULO 7
  245. #define FILE_OPS_MASK 0x07 /* mask for above ops */
  246. #define FILE_UNUSED_1 0x08
  247. #define FILE_UNUSED_2 0x10
  248. #define FILE_UNUSED_3 0x20
  249. #define FILE_OPINVERSE 0x40
  250. #define FILE_OPINDIRECT 0x80
  251. #ifdef ENABLE_CONDITIONALS
  252. #define COND_NONE 0
  253. #define COND_IF 1
  254. #define COND_ELIF 2
  255. #define COND_ELSE 3
  256. #endif /* ENABLE_CONDITIONALS */
  257. /* Word 4 */
  258. uint32_t offset; /* offset to magic number */
  259. /* Word 5 */
  260. int32_t in_offset; /* offset from indirection */
  261. /* Word 6 */
  262. uint32_t lineno; /* line number in magic file */
  263. /* Word 7,8 */
  264. union {
  265. uint64_t _mask; /* for use with numeric and date types */
  266. struct {
  267. uint32_t _count; /* repeat/line count */
  268. uint32_t _flags; /* modifier flags */
  269. } _s; /* for use with string types */
  270. } _u;
  271. #define num_mask _u._mask
  272. #define str_range _u._s._count
  273. #define str_flags _u._s._flags
  274. /* Words 9-16 */
  275. union VALUETYPE value; /* either number or string */
  276. /* Words 17-32 */
  277. char desc[MAXDESC]; /* description */
  278. /* Words 33-52 */
  279. char mimetype[MAXMIME]; /* MIME type */
  280. /* Words 53-54 */
  281. char apple[8];
  282. };
  283. #define BIT(A) (1 << (A))
  284. #define STRING_COMPACT_WHITESPACE BIT(0)
  285. #define STRING_COMPACT_OPTIONAL_WHITESPACE BIT(1)
  286. #define STRING_IGNORE_LOWERCASE BIT(2)
  287. #define STRING_IGNORE_UPPERCASE BIT(3)
  288. #define REGEX_OFFSET_START BIT(4)
  289. #define STRING_TEXTTEST BIT(5)
  290. #define STRING_BINTEST BIT(6)
  291. #define PSTRING_1_BE BIT(7)
  292. #define PSTRING_1_LE BIT(7)
  293. #define PSTRING_2_BE BIT(8)
  294. #define PSTRING_2_LE BIT(9)
  295. #define PSTRING_4_BE BIT(10)
  296. #define PSTRING_4_LE BIT(11)
  297. #define PSTRING_LEN \
  298. (PSTRING_1_BE|PSTRING_2_LE|PSTRING_2_BE|PSTRING_4_LE|PSTRING_4_BE)
  299. #define PSTRING_LENGTH_INCLUDES_ITSELF BIT(12)
  300. #define STRING_TRIM BIT(13)
  301. #define CHAR_COMPACT_WHITESPACE 'W'
  302. #define CHAR_COMPACT_OPTIONAL_WHITESPACE 'w'
  303. #define CHAR_IGNORE_LOWERCASE 'c'
  304. #define CHAR_IGNORE_UPPERCASE 'C'
  305. #define CHAR_REGEX_OFFSET_START 's'
  306. #define CHAR_TEXTTEST 't'
  307. #define CHAR_TRIM 'T'
  308. #define CHAR_BINTEST 'b'
  309. #define CHAR_PSTRING_1_BE 'B'
  310. #define CHAR_PSTRING_1_LE 'B'
  311. #define CHAR_PSTRING_2_BE 'H'
  312. #define CHAR_PSTRING_2_LE 'h'
  313. #define CHAR_PSTRING_4_BE 'L'
  314. #define CHAR_PSTRING_4_LE 'l'
  315. #define CHAR_PSTRING_LENGTH_INCLUDES_ITSELF 'J'
  316. #define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE)
  317. #define STRING_DEFAULT_RANGE 100
  318. /* list of magic entries */
  319. struct mlist {
  320. struct magic *magic; /* array of magic entries */
  321. uint32_t nmagic; /* number of entries in array */
  322. void *map; /* internal resources used by entry */
  323. struct mlist *next, *prev;
  324. };
  325. #ifdef __cplusplus
  326. #define CAST(T, b) static_cast<T>(b)
  327. #define RCAST(T, b) reinterpret_cast<T>(b)
  328. #else
  329. #define CAST(T, b) (T)(b)
  330. #define RCAST(T, b) (T)(b)
  331. #endif
  332. struct level_info {
  333. int32_t off;
  334. int got_match;
  335. #ifdef ENABLE_CONDITIONALS
  336. int last_match;
  337. int last_cond; /* used for error checking by parse() */
  338. #endif
  339. };
  340. #define MAGIC_SETS 2
  341. struct magic_set {
  342. struct mlist *mlist[MAGIC_SETS]; /* list of regular entries */
  343. struct cont {
  344. size_t len;
  345. struct level_info *li;
  346. } c;
  347. struct out {
  348. char *buf; /* Accumulation buffer */
  349. char *pbuf; /* Printable buffer */
  350. } o;
  351. uint32_t offset;
  352. int error;
  353. int flags; /* Control magic tests. */
  354. int event_flags; /* Note things that happened. */
  355. #define EVENT_HAD_ERR 0x01
  356. const char *file;
  357. size_t line; /* current magic line number */
  358. /* data for searches */
  359. struct {
  360. const char *s; /* start of search in original source */
  361. size_t s_len; /* length of search region */
  362. size_t offset; /* starting offset in source: XXX - should this be off_t? */
  363. size_t rm_len; /* match length */
  364. } search;
  365. /* FIXME: Make the string dynamically allocated so that e.g.
  366. strings matched in files can be longer than MAXstring */
  367. union VALUETYPE ms_value; /* either number or string */
  368. };
  369. /* Type for Unicode characters */
  370. typedef unsigned long unichar;
  371. struct stat;
  372. #define FILE_T_LOCAL 1
  373. #define FILE_T_WINDOWS 2
  374. protected const char *file_fmttime(uint64_t, int, char *);
  375. protected struct magic_set *file_ms_alloc(int);
  376. protected void file_ms_free(struct magic_set *);
  377. protected int file_buffer(struct magic_set *, php_stream *, const char *, const void *,
  378. size_t);
  379. protected int file_fsmagic(struct magic_set *, const char *, struct stat *, php_stream *);
  380. protected int file_pipe2file(struct magic_set *, int, const void *, size_t);
  381. protected int file_replace(struct magic_set *, const char *, const char *);
  382. protected int file_printf(struct magic_set *, const char *, ...);
  383. protected int file_reset(struct magic_set *);
  384. protected int file_trycdf(struct magic_set *, int, const unsigned char *,
  385. size_t);
  386. #ifdef PHP_FILEINFO_UNCOMPRESS
  387. protected int file_zmagic(struct magic_set *, int, const char *,
  388. const unsigned char *, size_t);
  389. #endif
  390. protected int file_ascmagic(struct magic_set *, const unsigned char *, size_t,
  391. int);
  392. protected int file_ascmagic_with_encoding(struct magic_set *,
  393. const unsigned char *, size_t, unichar *, size_t, const char *,
  394. const char *, int);
  395. protected int file_encoding(struct magic_set *, const unsigned char *, size_t,
  396. unichar **, size_t *, const char **, const char **, const char **);
  397. protected int file_is_tar(struct magic_set *, const unsigned char *, size_t);
  398. protected int file_softmagic(struct magic_set *, const unsigned char *, size_t,
  399. size_t, int, int);
  400. protected int file_apprentice(struct magic_set *, const char *, int);
  401. protected int file_magicfind(struct magic_set *, const char *, struct mlist *);
  402. protected uint64_t file_signextend(struct magic_set *, struct magic *,
  403. uint64_t);
  404. protected void file_delmagic(struct magic *, int type, size_t entries);
  405. protected void file_badread(struct magic_set *);
  406. protected void file_badseek(struct magic_set *);
  407. protected void file_oomem(struct magic_set *, size_t);
  408. protected void file_error(struct magic_set *, int, const char *, ...);
  409. protected void file_magerror(struct magic_set *, const char *, ...);
  410. protected void file_magwarn(struct magic_set *, const char *, ...);
  411. protected void file_showstr(FILE *, const char *, size_t);
  412. protected size_t file_mbswidth(const char *);
  413. protected const char *file_getbuffer(struct magic_set *);
  414. protected ssize_t sread(int, void *, size_t, int);
  415. protected int file_check_mem(struct magic_set *, unsigned int);
  416. protected int file_looks_utf8(const unsigned char *, size_t, unichar *,
  417. size_t *);
  418. protected size_t file_pstring_length_size(const struct magic *);
  419. protected size_t file_pstring_get_length(const struct magic *, const char *);
  420. protected size_t file_printedlen(const struct magic_set *ms);
  421. #ifdef __EMX__
  422. protected int file_os2_apptype(struct magic_set *, const char *, const void *,
  423. size_t);
  424. #endif /* __EMX__ */
  425. extern const char *file_names[];
  426. extern const size_t file_nnames;
  427. #ifndef HAVE_STRERROR
  428. extern int sys_nerr;
  429. extern char *sys_errlist[];
  430. #define strerror(e) \
  431. (((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error")
  432. #endif
  433. #ifndef HAVE_STRTOUL
  434. #define strtoul(a, b, c) strtol(a, b, c)
  435. #endif
  436. #ifndef strlcpy
  437. size_t strlcpy(char *, const char *, size_t);
  438. #endif
  439. #ifndef strlcat
  440. size_t strlcat(char *, const char *, size_t);
  441. #endif
  442. #ifndef HAVE_STRCASESTR
  443. char *strcasestr(const char *, const char *);
  444. #endif
  445. #ifndef HAVE_GETLINE
  446. ssize_t getline(char **, size_t *, FILE *);
  447. ssize_t getdelim(char **, size_t *, int, FILE *);
  448. #endif
  449. #ifndef HAVE_CTIME_R
  450. char *ctime_r(const time_t *, char *);
  451. #endif
  452. #ifndef HAVE_ASCTIME_R
  453. char *asctime_r(const struct tm *, char *);
  454. #endif
  455. #if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK)
  456. #define QUICK
  457. #endif
  458. #ifndef O_BINARY
  459. #define O_BINARY 0
  460. #endif
  461. #ifndef __cplusplus
  462. #if defined(__GNUC__) && (__GNUC__ >= 3)
  463. #define FILE_RCSID(id) \
  464. static const char rcsid[] __attribute__((__used__)) = id;
  465. #else
  466. #define FILE_RCSID(id) \
  467. static const char *rcsid(const char *p) { \
  468. return rcsid(p = id); \
  469. }
  470. #endif
  471. #else
  472. #define FILE_RCSID(id)
  473. #endif
  474. #ifdef PHP_WIN32
  475. #define FINFO_LSEEK_FUNC _lseek
  476. #define FINFO_READ_FUNC _read
  477. #else
  478. #define FINFO_LSEEK_FUNC lseek
  479. #define FINFO_READ_FUNC read
  480. #endif
  481. #endif /* __file_h__ */