file.h revision 277592
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 1.164 2015/01/01 17:07:34 christos Exp $ 31 */ 32 33#ifndef __file_h__ 34#define __file_h__ 35 36#ifdef HAVE_CONFIG_H 37#include <config.h> 38#endif 39 40#ifdef WIN32 41 #ifdef _WIN64 42 #define SIZE_T_FORMAT "I64" 43 #else 44 #define SIZE_T_FORMAT "" 45 #endif 46 #define INT64_T_FORMAT "I64" 47#else 48 #define SIZE_T_FORMAT "z" 49 #define INT64_T_FORMAT "ll" 50#endif 51 52#include <stdio.h> /* Include that here, to make sure __P gets defined */ 53#include <errno.h> 54#include <fcntl.h> /* For open and flags */ 55#ifdef HAVE_STDINT_H 56#ifndef __STDC_LIMIT_MACROS 57#define __STDC_LIMIT_MACROS 58#endif 59#include <stdint.h> 60#endif 61#ifdef HAVE_INTTYPES_H 62#include <inttypes.h> 63#endif 64#include <regex.h> 65#include <time.h> 66#include <sys/types.h> 67#ifndef WIN32 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 74#define ENABLE_CONDITIONALS 75 76#ifndef MAGIC 77#define MAGIC "/etc/magic" 78#endif 79 80#if defined(__EMX__) || defined (WIN32) 81#define PATHSEP ';' 82#else 83#define PATHSEP ':' 84#endif 85 86#define private static 87 88#if HAVE_VISIBILITY && !defined(WIN32) 89#define public __attribute__ ((__visibility__("default"))) 90#ifndef protected 91#define protected __attribute__ ((__visibility__("hidden"))) 92#endif 93#else 94#define public 95#ifndef protected 96#define protected 97#endif 98#endif 99 100#ifndef __arraycount 101#define __arraycount(a) (sizeof(a) / sizeof(a[0])) 102#endif 103 104#ifndef __GNUC_PREREQ__ 105#ifdef __GNUC__ 106#define __GNUC_PREREQ__(x, y) \ 107 ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \ 108 (__GNUC__ > (x))) 109#else 110#define __GNUC_PREREQ__(x, y) 0 111#endif 112#endif 113 114#ifndef __GNUC__ 115#ifndef __attribute__ 116#define __attribute__(a) 117#endif 118#endif 119 120#ifndef MIN 121#define MIN(a,b) (((a) < (b)) ? (a) : (b)) 122#endif 123 124#ifndef MAX 125#define MAX(a,b) (((a) > (b)) ? (a) : (b)) 126#endif 127 128#ifndef HOWMANY 129# define HOWMANY (256 * 1024) /* how much of the file to look at */ 130#endif 131#define MAXMAGIS 8192 /* max entries in any one magic file 132 or directory */ 133#define MAXDESC 64 /* max len of text description/MIME type */ 134#define MAXMIME 80 /* max len of text MIME type */ 135#define MAXstring 64 /* max len of "string" types */ 136 137#define MAGICNO 0xF11E041C 138#define VERSIONNO 12 139#define FILE_MAGICSIZE 248 140 141#define FILE_LOAD 0 142#define FILE_CHECK 1 143#define FILE_COMPILE 2 144#define FILE_LIST 3 145 146union VALUETYPE { 147 uint8_t b; 148 uint16_t h; 149 uint32_t l; 150 uint64_t q; 151 uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */ 152 uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */ 153 uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */ 154 char s[MAXstring]; /* the search string or regex pattern */ 155 unsigned char us[MAXstring]; 156 float f; 157 double d; 158}; 159 160struct magic { 161 /* Word 1 */ 162 uint16_t cont_level; /* level of ">" */ 163 uint8_t flag; 164#define INDIR 0x01 /* if '(...)' appears */ 165#define OFFADD 0x02 /* if '>&' or '>...(&' appears */ 166#define INDIROFFADD 0x04 /* if '>&(' appears */ 167#define UNSIGNED 0x08 /* comparison is unsigned */ 168#define NOSPACE 0x10 /* suppress space character before output */ 169#define BINTEST 0x20 /* test is for a binary type (set only 170 for top-level tests) */ 171#define TEXTTEST 0x40 /* for passing to file_softmagic */ 172 173 uint8_t factor; 174 175 /* Word 2 */ 176 uint8_t reln; /* relation (0=eq, '>'=gt, etc) */ 177 uint8_t vallen; /* length of string value, if any */ 178 uint8_t type; /* comparison type (FILE_*) */ 179 uint8_t in_type; /* type of indirection */ 180#define FILE_INVALID 0 181#define FILE_BYTE 1 182#define FILE_SHORT 2 183#define FILE_DEFAULT 3 184#define FILE_LONG 4 185#define FILE_STRING 5 186#define FILE_DATE 6 187#define FILE_BESHORT 7 188#define FILE_BELONG 8 189#define FILE_BEDATE 9 190#define FILE_LESHORT 10 191#define FILE_LELONG 11 192#define FILE_LEDATE 12 193#define FILE_PSTRING 13 194#define FILE_LDATE 14 195#define FILE_BELDATE 15 196#define FILE_LELDATE 16 197#define FILE_REGEX 17 198#define FILE_BESTRING16 18 199#define FILE_LESTRING16 19 200#define FILE_SEARCH 20 201#define FILE_MEDATE 21 202#define FILE_MELDATE 22 203#define FILE_MELONG 23 204#define FILE_QUAD 24 205#define FILE_LEQUAD 25 206#define FILE_BEQUAD 26 207#define FILE_QDATE 27 208#define FILE_LEQDATE 28 209#define FILE_BEQDATE 29 210#define FILE_QLDATE 30 211#define FILE_LEQLDATE 31 212#define FILE_BEQLDATE 32 213#define FILE_FLOAT 33 214#define FILE_BEFLOAT 34 215#define FILE_LEFLOAT 35 216#define FILE_DOUBLE 36 217#define FILE_BEDOUBLE 37 218#define FILE_LEDOUBLE 38 219#define FILE_BEID3 39 220#define FILE_LEID3 40 221#define FILE_INDIRECT 41 222#define FILE_QWDATE 42 223#define FILE_LEQWDATE 43 224#define FILE_BEQWDATE 44 225#define FILE_NAME 45 226#define FILE_USE 46 227#define FILE_CLEAR 47 228#define FILE_NAMES_SIZE 48 /* size of array to contain all names */ 229 230#define IS_STRING(t) \ 231 ((t) == FILE_STRING || \ 232 (t) == FILE_PSTRING || \ 233 (t) == FILE_BESTRING16 || \ 234 (t) == FILE_LESTRING16 || \ 235 (t) == FILE_REGEX || \ 236 (t) == FILE_SEARCH || \ 237 (t) == FILE_INDIRECT || \ 238 (t) == FILE_NAME || \ 239 (t) == FILE_USE) 240 241#define FILE_FMT_NONE 0 242#define FILE_FMT_NUM 1 /* "cduxXi" */ 243#define FILE_FMT_STR 2 /* "s" */ 244#define FILE_FMT_QUAD 3 /* "ll" */ 245#define FILE_FMT_FLOAT 4 /* "eEfFgG" */ 246#define FILE_FMT_DOUBLE 5 /* "eEfFgG" */ 247 248 /* Word 3 */ 249 uint8_t in_op; /* operator for indirection */ 250 uint8_t mask_op; /* operator for mask */ 251#ifdef ENABLE_CONDITIONALS 252 uint8_t cond; /* conditional type */ 253#else 254 uint8_t dummy; 255#endif 256 uint8_t factor_op; 257#define FILE_FACTOR_OP_PLUS '+' 258#define FILE_FACTOR_OP_MINUS '-' 259#define FILE_FACTOR_OP_TIMES '*' 260#define FILE_FACTOR_OP_DIV '/' 261#define FILE_FACTOR_OP_NONE '\0' 262 263#define FILE_OPS "&|^+-*/%" 264#define FILE_OPAND 0 265#define FILE_OPOR 1 266#define FILE_OPXOR 2 267#define FILE_OPADD 3 268#define FILE_OPMINUS 4 269#define FILE_OPMULTIPLY 5 270#define FILE_OPDIVIDE 6 271#define FILE_OPMODULO 7 272#define FILE_OPS_MASK 0x07 /* mask for above ops */ 273#define FILE_UNUSED_1 0x08 274#define FILE_UNUSED_2 0x10 275#define FILE_UNUSED_3 0x20 276#define FILE_OPINVERSE 0x40 277#define FILE_OPINDIRECT 0x80 278 279#ifdef ENABLE_CONDITIONALS 280#define COND_NONE 0 281#define COND_IF 1 282#define COND_ELIF 2 283#define COND_ELSE 3 284#endif /* ENABLE_CONDITIONALS */ 285 286 /* Word 4 */ 287 uint32_t offset; /* offset to magic number */ 288 /* Word 5 */ 289 int32_t in_offset; /* offset from indirection */ 290 /* Word 6 */ 291 uint32_t lineno; /* line number in magic file */ 292 /* Word 7,8 */ 293 union { 294 uint64_t _mask; /* for use with numeric and date types */ 295 struct { 296 uint32_t _count; /* repeat/line count */ 297 uint32_t _flags; /* modifier flags */ 298 } _s; /* for use with string types */ 299 } _u; 300#define num_mask _u._mask 301#define str_range _u._s._count 302#define str_flags _u._s._flags 303 /* Words 9-16 */ 304 union VALUETYPE value; /* either number or string */ 305 /* Words 17-32 */ 306 char desc[MAXDESC]; /* description */ 307 /* Words 33-52 */ 308 char mimetype[MAXMIME]; /* MIME type */ 309 /* Words 53-54 */ 310 char apple[8]; 311}; 312 313#define BIT(A) (1 << (A)) 314#define STRING_COMPACT_WHITESPACE BIT(0) 315#define STRING_COMPACT_OPTIONAL_WHITESPACE BIT(1) 316#define STRING_IGNORE_LOWERCASE BIT(2) 317#define STRING_IGNORE_UPPERCASE BIT(3) 318#define REGEX_OFFSET_START BIT(4) 319#define STRING_TEXTTEST BIT(5) 320#define STRING_BINTEST BIT(6) 321#define PSTRING_1_BE BIT(7) 322#define PSTRING_1_LE BIT(7) 323#define PSTRING_2_BE BIT(8) 324#define PSTRING_2_LE BIT(9) 325#define PSTRING_4_BE BIT(10) 326#define PSTRING_4_LE BIT(11) 327#define REGEX_LINE_COUNT BIT(11) 328#define PSTRING_LEN \ 329 (PSTRING_1_BE|PSTRING_2_LE|PSTRING_2_BE|PSTRING_4_LE|PSTRING_4_BE) 330#define PSTRING_LENGTH_INCLUDES_ITSELF BIT(12) 331#define STRING_TRIM BIT(13) 332#define CHAR_COMPACT_WHITESPACE 'W' 333#define CHAR_COMPACT_OPTIONAL_WHITESPACE 'w' 334#define CHAR_IGNORE_LOWERCASE 'c' 335#define CHAR_IGNORE_UPPERCASE 'C' 336#define CHAR_REGEX_OFFSET_START 's' 337#define CHAR_TEXTTEST 't' 338#define CHAR_TRIM 'T' 339#define CHAR_BINTEST 'b' 340#define CHAR_PSTRING_1_BE 'B' 341#define CHAR_PSTRING_1_LE 'B' 342#define CHAR_PSTRING_2_BE 'H' 343#define CHAR_PSTRING_2_LE 'h' 344#define CHAR_PSTRING_4_BE 'L' 345#define CHAR_PSTRING_4_LE 'l' 346#define CHAR_PSTRING_LENGTH_INCLUDES_ITSELF 'J' 347#define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE) 348#define STRING_DEFAULT_RANGE 100 349 350#define INDIRECT_RELATIVE BIT(0) 351#define CHAR_INDIRECT_RELATIVE 'r' 352 353/* list of magic entries */ 354struct mlist { 355 struct magic *magic; /* array of magic entries */ 356 uint32_t nmagic; /* number of entries in array */ 357 void *map; /* internal resources used by entry */ 358 struct mlist *next, *prev; 359}; 360 361#ifdef __cplusplus 362#define CAST(T, b) static_cast<T>(b) 363#define RCAST(T, b) reinterpret_cast<T>(b) 364#else 365#define CAST(T, b) (T)(b) 366#define RCAST(T, b) (T)(b) 367#endif 368 369struct level_info { 370 int32_t off; 371 int got_match; 372#ifdef ENABLE_CONDITIONALS 373 int last_match; 374 int last_cond; /* used for error checking by parse() */ 375#endif 376}; 377 378#define MAGIC_SETS 2 379 380struct magic_set { 381 struct mlist *mlist[MAGIC_SETS]; /* list of regular entries */ 382 struct cont { 383 size_t len; 384 struct level_info *li; 385 } c; 386 struct out { 387 char *buf; /* Accumulation buffer */ 388 char *pbuf; /* Printable buffer */ 389 } o; 390 uint32_t offset; 391 int error; 392 int flags; /* Control magic tests. */ 393 int event_flags; /* Note things that happened. */ 394#define EVENT_HAD_ERR 0x01 395 const char *file; 396 size_t line; /* current magic line number */ 397 398 /* data for searches */ 399 struct { 400 const char *s; /* start of search in original source */ 401 size_t s_len; /* length of search region */ 402 size_t offset; /* starting offset in source: XXX - should this be off_t? */ 403 size_t rm_len; /* match length */ 404 } search; 405 406 /* FIXME: Make the string dynamically allocated so that e.g. 407 strings matched in files can be longer than MAXstring */ 408 union VALUETYPE ms_value; /* either number or string */ 409 uint16_t indir_max; 410 uint16_t name_max; 411 uint16_t elf_shnum_max; 412 uint16_t elf_phnum_max; 413 uint16_t elf_notes_max; 414#define FILE_INDIR_MAX 15 415#define FILE_NAME_MAX 30 416#define FILE_ELF_SHNUM_MAX 32768 417#define FILE_ELF_PHNUM_MAX 128 418#define FILE_ELF_NOTES_MAX 256 419}; 420 421/* Type for Unicode characters */ 422typedef unsigned long unichar; 423 424struct stat; 425#define FILE_T_LOCAL 1 426#define FILE_T_WINDOWS 2 427protected const char *file_fmttime(uint64_t, int, char *); 428protected struct magic_set *file_ms_alloc(int); 429protected void file_ms_free(struct magic_set *); 430protected int file_buffer(struct magic_set *, int, const char *, const void *, 431 size_t); 432protected int file_fsmagic(struct magic_set *, const char *, struct stat *); 433protected int file_pipe2file(struct magic_set *, int, const void *, size_t); 434protected int file_vprintf(struct magic_set *, const char *, va_list) 435 __attribute__((__format__(__printf__, 2, 0))); 436protected size_t file_printedlen(const struct magic_set *); 437protected int file_replace(struct magic_set *, const char *, const char *); 438protected int file_printf(struct magic_set *, const char *, ...) 439 __attribute__((__format__(__printf__, 2, 3))); 440protected int file_reset(struct magic_set *); 441protected int file_tryelf(struct magic_set *, int, const unsigned char *, 442 size_t); 443protected int file_trycdf(struct magic_set *, int, const unsigned char *, 444 size_t); 445#if HAVE_FORK 446protected int file_zmagic(struct magic_set *, int, const char *, 447 const unsigned char *, size_t); 448#endif 449protected int file_ascmagic(struct magic_set *, const unsigned char *, size_t, 450 int); 451protected int file_ascmagic_with_encoding(struct magic_set *, 452 const unsigned char *, size_t, unichar *, size_t, const char *, 453 const char *, int); 454protected int file_encoding(struct magic_set *, const unsigned char *, size_t, 455 unichar **, size_t *, const char **, const char **, const char **); 456protected int file_is_tar(struct magic_set *, const unsigned char *, size_t); 457protected int file_softmagic(struct magic_set *, const unsigned char *, size_t, 458 uint16_t, uint16_t *, int, int); 459protected int file_apprentice(struct magic_set *, const char *, int); 460protected int buffer_apprentice(struct magic_set *, struct magic **, 461 size_t *, size_t); 462protected int file_magicfind(struct magic_set *, const char *, struct mlist *); 463protected uint64_t file_signextend(struct magic_set *, struct magic *, 464 uint64_t); 465protected void file_badread(struct magic_set *); 466protected void file_badseek(struct magic_set *); 467protected void file_oomem(struct magic_set *, size_t); 468protected void file_error(struct magic_set *, int, const char *, ...) 469 __attribute__((__format__(__printf__, 3, 4))); 470protected void file_magerror(struct magic_set *, const char *, ...) 471 __attribute__((__format__(__printf__, 2, 3))); 472protected void file_magwarn(struct magic_set *, const char *, ...) 473 __attribute__((__format__(__printf__, 2, 3))); 474protected void file_mdump(struct magic *); 475protected void file_showstr(FILE *, const char *, size_t); 476protected size_t file_mbswidth(const char *); 477protected const char *file_getbuffer(struct magic_set *); 478protected ssize_t sread(int, void *, size_t, int); 479protected int file_check_mem(struct magic_set *, unsigned int); 480protected int file_looks_utf8(const unsigned char *, size_t, unichar *, 481 size_t *); 482protected size_t file_pstring_length_size(const struct magic *); 483protected size_t file_pstring_get_length(const struct magic *, const char *); 484protected char * file_printable(char *, size_t, const char *); 485#ifdef __EMX__ 486protected int file_os2_apptype(struct magic_set *, const char *, const void *, 487 size_t); 488#endif /* __EMX__ */ 489 490#if defined(HAVE_LOCALE_H) 491#include <locale.h> 492#endif 493#if defined(HAVE_XLOCALE_H) 494#include <xlocale.h> 495#endif 496 497typedef struct { 498 const char *pat; 499#if defined(HAVE_NEWLOCALE) && defined(HAVE_USELOCALE) && defined(HAVE_FREELOCALE) 500#define USE_C_LOCALE 501 locale_t old_lc_ctype; 502 locale_t c_lc_ctype; 503#endif 504 int rc; 505 regex_t rx; 506} file_regex_t; 507 508protected int file_regcomp(file_regex_t *, const char *, int); 509protected int file_regexec(file_regex_t *, const char *, size_t, regmatch_t *, 510 int); 511protected void file_regfree(file_regex_t *); 512protected void file_regerror(file_regex_t *, int, struct magic_set *); 513 514typedef struct { 515 char *buf; 516 uint32_t offset; 517} file_pushbuf_t; 518 519protected file_pushbuf_t *file_push_buffer(struct magic_set *); 520protected char *file_pop_buffer(struct magic_set *, file_pushbuf_t *); 521 522#ifndef COMPILE_ONLY 523extern const char *file_names[]; 524extern const size_t file_nnames; 525#endif 526 527#ifndef HAVE_STRERROR 528extern int sys_nerr; 529extern char *sys_errlist[]; 530#define strerror(e) \ 531 (((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error") 532#endif 533 534#ifndef HAVE_STRTOUL 535#define strtoul(a, b, c) strtol(a, b, c) 536#endif 537 538#ifndef HAVE_PREAD 539ssize_t pread(int, void *, size_t, off_t); 540#endif 541#ifndef HAVE_VASPRINTF 542int vasprintf(char **, const char *, va_list); 543#endif 544#ifndef HAVE_ASPRINTF 545int asprintf(char **, const char *, ...); 546#endif 547 548#ifndef HAVE_STRLCPY 549size_t strlcpy(char *, const char *, size_t); 550#endif 551#ifndef HAVE_STRLCAT 552size_t strlcat(char *, const char *, size_t); 553#endif 554#ifndef HAVE_STRCASESTR 555char *strcasestr(const char *, const char *); 556#endif 557#ifndef HAVE_GETLINE 558ssize_t getline(char **, size_t *, FILE *); 559ssize_t getdelim(char **, size_t *, int, FILE *); 560#endif 561#ifndef HAVE_CTIME_R 562char *ctime_r(const time_t *, char *); 563#endif 564#ifndef HAVE_ASCTIME_R 565char *asctime_r(const struct tm *, char *); 566#endif 567#ifndef HAVE_FMTCHECK 568const char *fmtcheck(const char *, const char *) 569 __attribute__((__format_arg__(2))); 570#endif 571 572#if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK) 573#define QUICK 574#endif 575 576#ifndef O_BINARY 577#define O_BINARY 0 578#endif 579 580#ifndef __cplusplus 581#if defined(__GNUC__) && (__GNUC__ >= 3) 582#define FILE_RCSID(id) \ 583static const char rcsid[] __attribute__((__used__)) = id; 584#else 585#define FILE_RCSID(id) \ 586static const char *rcsid(const char *p) { \ 587 return rcsid(p = id); \ 588} 589#endif 590#else 591#define FILE_RCSID(id) 592#endif 593 594#endif /* __file_h__ */ 595