file.h revision 276415
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 1.161 2014/12/04 15:56:46 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_NAME || \ 238 (t) == FILE_USE) 239 240#define FILE_FMT_NONE 0 241#define FILE_FMT_NUM 1 /* "cduxXi" */ 242#define FILE_FMT_STR 2 /* "s" */ 243#define FILE_FMT_QUAD 3 /* "ll" */ 244#define FILE_FMT_FLOAT 4 /* "eEfFgG" */ 245#define FILE_FMT_DOUBLE 5 /* "eEfFgG" */ 246 247 /* Word 3 */ 248 uint8_t in_op; /* operator for indirection */ 249 uint8_t mask_op; /* operator for mask */ 250#ifdef ENABLE_CONDITIONALS 251 uint8_t cond; /* conditional type */ 252#else 253 uint8_t dummy; 254#endif 255 uint8_t factor_op; 256#define FILE_FACTOR_OP_PLUS '+' 257#define FILE_FACTOR_OP_MINUS '-' 258#define FILE_FACTOR_OP_TIMES '*' 259#define FILE_FACTOR_OP_DIV '/' 260#define FILE_FACTOR_OP_NONE '\0' 261 262#define FILE_OPS "&|^+-*/%" 263#define FILE_OPAND 0 264#define FILE_OPOR 1 265#define FILE_OPXOR 2 266#define FILE_OPADD 3 267#define FILE_OPMINUS 4 268#define FILE_OPMULTIPLY 5 269#define FILE_OPDIVIDE 6 270#define FILE_OPMODULO 7 271#define FILE_OPS_MASK 0x07 /* mask for above ops */ 272#define FILE_UNUSED_1 0x08 273#define FILE_UNUSED_2 0x10 274#define FILE_UNUSED_3 0x20 275#define FILE_OPINVERSE 0x40 276#define FILE_OPINDIRECT 0x80 277 278#ifdef ENABLE_CONDITIONALS 279#define COND_NONE 0 280#define COND_IF 1 281#define COND_ELIF 2 282#define COND_ELSE 3 283#endif /* ENABLE_CONDITIONALS */ 284 285 /* Word 4 */ 286 uint32_t offset; /* offset to magic number */ 287 /* Word 5 */ 288 int32_t in_offset; /* offset from indirection */ 289 /* Word 6 */ 290 uint32_t lineno; /* line number in magic file */ 291 /* Word 7,8 */ 292 union { 293 uint64_t _mask; /* for use with numeric and date types */ 294 struct { 295 uint32_t _count; /* repeat/line count */ 296 uint32_t _flags; /* modifier flags */ 297 } _s; /* for use with string types */ 298 } _u; 299#define num_mask _u._mask 300#define str_range _u._s._count 301#define str_flags _u._s._flags 302 /* Words 9-16 */ 303 union VALUETYPE value; /* either number or string */ 304 /* Words 17-32 */ 305 char desc[MAXDESC]; /* description */ 306 /* Words 33-52 */ 307 char mimetype[MAXMIME]; /* MIME type */ 308 /* Words 53-54 */ 309 char apple[8]; 310}; 311 312#define BIT(A) (1 << (A)) 313#define STRING_COMPACT_WHITESPACE BIT(0) 314#define STRING_COMPACT_OPTIONAL_WHITESPACE BIT(1) 315#define STRING_IGNORE_LOWERCASE BIT(2) 316#define STRING_IGNORE_UPPERCASE BIT(3) 317#define REGEX_OFFSET_START BIT(4) 318#define STRING_TEXTTEST BIT(5) 319#define STRING_BINTEST BIT(6) 320#define PSTRING_1_BE BIT(7) 321#define PSTRING_1_LE BIT(7) 322#define PSTRING_2_BE BIT(8) 323#define PSTRING_2_LE BIT(9) 324#define PSTRING_4_BE BIT(10) 325#define PSTRING_4_LE BIT(11) 326#define REGEX_LINE_COUNT BIT(11) 327#define PSTRING_LEN \ 328 (PSTRING_1_BE|PSTRING_2_LE|PSTRING_2_BE|PSTRING_4_LE|PSTRING_4_BE) 329#define PSTRING_LENGTH_INCLUDES_ITSELF BIT(12) 330#define STRING_TRIM BIT(13) 331#define CHAR_COMPACT_WHITESPACE 'W' 332#define CHAR_COMPACT_OPTIONAL_WHITESPACE 'w' 333#define CHAR_IGNORE_LOWERCASE 'c' 334#define CHAR_IGNORE_UPPERCASE 'C' 335#define CHAR_REGEX_OFFSET_START 's' 336#define CHAR_TEXTTEST 't' 337#define CHAR_TRIM 'T' 338#define CHAR_BINTEST 'b' 339#define CHAR_PSTRING_1_BE 'B' 340#define CHAR_PSTRING_1_LE 'B' 341#define CHAR_PSTRING_2_BE 'H' 342#define CHAR_PSTRING_2_LE 'h' 343#define CHAR_PSTRING_4_BE 'L' 344#define CHAR_PSTRING_4_LE 'l' 345#define CHAR_PSTRING_LENGTH_INCLUDES_ITSELF 'J' 346#define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE) 347#define STRING_DEFAULT_RANGE 100 348 349 350/* list of magic entries */ 351struct mlist { 352 struct magic *magic; /* array of magic entries */ 353 uint32_t nmagic; /* number of entries in array */ 354 void *map; /* internal resources used by entry */ 355 struct mlist *next, *prev; 356}; 357 358#ifdef __cplusplus 359#define CAST(T, b) static_cast<T>(b) 360#define RCAST(T, b) reinterpret_cast<T>(b) 361#else 362#define CAST(T, b) (T)(b) 363#define RCAST(T, b) (T)(b) 364#endif 365 366struct level_info { 367 int32_t off; 368 int got_match; 369#ifdef ENABLE_CONDITIONALS 370 int last_match; 371 int last_cond; /* used for error checking by parse() */ 372#endif 373}; 374 375#define MAGIC_SETS 2 376 377struct magic_set { 378 struct mlist *mlist[MAGIC_SETS]; /* list of regular entries */ 379 struct cont { 380 size_t len; 381 struct level_info *li; 382 } c; 383 struct out { 384 char *buf; /* Accumulation buffer */ 385 char *pbuf; /* Printable buffer */ 386 } o; 387 uint32_t offset; 388 int error; 389 int flags; /* Control magic tests. */ 390 int event_flags; /* Note things that happened. */ 391#define EVENT_HAD_ERR 0x01 392 const char *file; 393 size_t line; /* current magic line number */ 394 395 /* data for searches */ 396 struct { 397 const char *s; /* start of search in original source */ 398 size_t s_len; /* length of search region */ 399 size_t offset; /* starting offset in source: XXX - should this be off_t? */ 400 size_t rm_len; /* match length */ 401 } search; 402 403 /* FIXME: Make the string dynamically allocated so that e.g. 404 strings matched in files can be longer than MAXstring */ 405 union VALUETYPE ms_value; /* either number or string */ 406 uint16_t indir_max; 407 uint16_t name_max; 408 uint16_t elf_shnum_max; 409 uint16_t elf_phnum_max; 410#define FILE_INDIR_MAX 15 411#define FILE_NAME_MAX 30 412#define FILE_ELF_SHNUM_MAX 32768 413#define FILE_ELF_PHNUM_MAX 128 414}; 415 416/* Type for Unicode characters */ 417typedef unsigned long unichar; 418 419struct stat; 420#define FILE_T_LOCAL 1 421#define FILE_T_WINDOWS 2 422protected const char *file_fmttime(uint64_t, int, char *); 423protected struct magic_set *file_ms_alloc(int); 424protected void file_ms_free(struct magic_set *); 425protected int file_buffer(struct magic_set *, int, const char *, const void *, 426 size_t); 427protected int file_fsmagic(struct magic_set *, const char *, struct stat *); 428protected int file_pipe2file(struct magic_set *, int, const void *, size_t); 429protected int file_vprintf(struct magic_set *, const char *, va_list) 430 __attribute__((__format__(__printf__, 2, 0))); 431protected size_t file_printedlen(const struct magic_set *); 432protected int file_replace(struct magic_set *, const char *, const char *); 433protected int file_printf(struct magic_set *, const char *, ...) 434 __attribute__((__format__(__printf__, 2, 3))); 435protected int file_reset(struct magic_set *); 436protected int file_tryelf(struct magic_set *, int, const unsigned char *, 437 size_t); 438protected int file_trycdf(struct magic_set *, int, const unsigned char *, 439 size_t); 440#if HAVE_FORK 441protected int file_zmagic(struct magic_set *, int, const char *, 442 const unsigned char *, size_t); 443#endif 444protected int file_ascmagic(struct magic_set *, const unsigned char *, size_t, 445 int); 446protected int file_ascmagic_with_encoding(struct magic_set *, 447 const unsigned char *, size_t, unichar *, size_t, const char *, 448 const char *, int); 449protected int file_encoding(struct magic_set *, const unsigned char *, size_t, 450 unichar **, size_t *, const char **, const char **, const char **); 451protected int file_is_tar(struct magic_set *, const unsigned char *, size_t); 452protected int file_softmagic(struct magic_set *, const unsigned char *, size_t, 453 uint16_t, uint16_t *, int, int); 454protected int file_apprentice(struct magic_set *, const char *, int); 455protected int buffer_apprentice(struct magic_set *, struct magic **, 456 size_t *, size_t); 457protected int file_magicfind(struct magic_set *, const char *, struct mlist *); 458protected uint64_t file_signextend(struct magic_set *, struct magic *, 459 uint64_t); 460protected void file_badread(struct magic_set *); 461protected void file_badseek(struct magic_set *); 462protected void file_oomem(struct magic_set *, size_t); 463protected void file_error(struct magic_set *, int, const char *, ...) 464 __attribute__((__format__(__printf__, 3, 4))); 465protected void file_magerror(struct magic_set *, const char *, ...) 466 __attribute__((__format__(__printf__, 2, 3))); 467protected void file_magwarn(struct magic_set *, const char *, ...) 468 __attribute__((__format__(__printf__, 2, 3))); 469protected void file_mdump(struct magic *); 470protected void file_showstr(FILE *, const char *, size_t); 471protected size_t file_mbswidth(const char *); 472protected const char *file_getbuffer(struct magic_set *); 473protected ssize_t sread(int, void *, size_t, int); 474protected int file_check_mem(struct magic_set *, unsigned int); 475protected int file_looks_utf8(const unsigned char *, size_t, unichar *, 476 size_t *); 477protected size_t file_pstring_length_size(const struct magic *); 478protected size_t file_pstring_get_length(const struct magic *, const char *); 479#ifdef __EMX__ 480protected int file_os2_apptype(struct magic_set *, const char *, const void *, 481 size_t); 482#endif /* __EMX__ */ 483 484#if defined(HAVE_LOCALE_H) 485#include <locale.h> 486#endif 487#if defined(HAVE_XLOCALE_H) 488#include <xlocale.h> 489#endif 490 491typedef struct { 492 const char *pat; 493#if defined(HAVE_NEWLOCALE) && defined(HAVE_USELOCALE) && defined(HAVE_FREELOCALE) 494#define USE_C_LOCALE 495 locale_t old_lc_ctype; 496 locale_t c_lc_ctype; 497#endif 498 int rc; 499 regex_t rx; 500} file_regex_t; 501 502protected int file_regcomp(file_regex_t *, const char *, int); 503protected int file_regexec(file_regex_t *, const char *, size_t, regmatch_t *, 504 int); 505protected void file_regfree(file_regex_t *); 506protected void file_regerror(file_regex_t *, int, struct magic_set *); 507 508typedef struct { 509 char *buf; 510 uint32_t offset; 511} file_pushbuf_t; 512 513protected file_pushbuf_t *file_push_buffer(struct magic_set *); 514protected char *file_pop_buffer(struct magic_set *, file_pushbuf_t *); 515 516#ifndef COMPILE_ONLY 517extern const char *file_names[]; 518extern const size_t file_nnames; 519#endif 520 521#ifndef HAVE_STRERROR 522extern int sys_nerr; 523extern char *sys_errlist[]; 524#define strerror(e) \ 525 (((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error") 526#endif 527 528#ifndef HAVE_STRTOUL 529#define strtoul(a, b, c) strtol(a, b, c) 530#endif 531 532#ifndef HAVE_PREAD 533ssize_t pread(int, void *, size_t, off_t); 534#endif 535#ifndef HAVE_VASPRINTF 536int vasprintf(char **, const char *, va_list); 537#endif 538#ifndef HAVE_ASPRINTF 539int asprintf(char **, const char *, ...); 540#endif 541 542#ifndef HAVE_STRLCPY 543size_t strlcpy(char *, const char *, size_t); 544#endif 545#ifndef HAVE_STRLCAT 546size_t strlcat(char *, const char *, size_t); 547#endif 548#ifndef HAVE_STRCASESTR 549char *strcasestr(const char *, const char *); 550#endif 551#ifndef HAVE_GETLINE 552ssize_t getline(char **, size_t *, FILE *); 553ssize_t getdelim(char **, size_t *, int, FILE *); 554#endif 555#ifndef HAVE_CTIME_R 556char *ctime_r(const time_t *, char *); 557#endif 558#ifndef HAVE_ASCTIME_R 559char *asctime_r(const struct tm *, char *); 560#endif 561#ifndef HAVE_FMTCHECK 562const char *fmtcheck(const char *, const char *) 563 __attribute__((__format_arg__(2))); 564#endif 565 566#if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK) 567#define QUICK 568#endif 569 570#ifndef O_BINARY 571#define O_BINARY 0 572#endif 573 574#ifndef __cplusplus 575#if defined(__GNUC__) && (__GNUC__ >= 3) 576#define FILE_RCSID(id) \ 577static const char rcsid[] __attribute__((__used__)) = id; 578#else 579#define FILE_RCSID(id) \ 580static const char *rcsid(const char *p) { \ 581 return rcsid(p = id); \ 582} 583#endif 584#else 585#define FILE_RCSID(id) 586#endif 587 588#endif /* __file_h__ */ 589