file.h revision 328875
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 1.183 2017/08/28 13:39:18 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#ifdef HAVE_STDINT_H 40#ifndef __STDC_LIMIT_MACROS 41#define __STDC_LIMIT_MACROS 42#endif 43 44#ifdef WIN32 45 #ifdef _WIN64 46 #define SIZE_T_FORMAT "I64" 47 #else 48 #define SIZE_T_FORMAT "" 49 #endif 50 #define INT64_T_FORMAT "I64" 51 #define INTMAX_T_FORMAT "I64" 52#else 53 #define SIZE_T_FORMAT "z" 54 #define INT64_T_FORMAT "ll" 55 #define INTMAX_T_FORMAT "j" 56#endif 57#include <stdint.h> 58#endif 59 60#include <stdio.h> /* Include that here, to make sure __P gets defined */ 61#include <errno.h> 62#include <fcntl.h> /* For open and flags */ 63#ifdef HAVE_INTTYPES_H 64#include <inttypes.h> 65#endif 66#include <regex.h> 67#include <time.h> 68#include <sys/types.h> 69#ifndef WIN32 70#include <sys/param.h> 71#endif 72/* Do this here and now, because struct stat gets re-defined on solaris */ 73#include <sys/stat.h> 74#include <stdarg.h> 75 76#define ENABLE_CONDITIONALS 77 78#ifndef MAGIC 79#define MAGIC "/etc/magic" 80#endif 81 82#if defined(__EMX__) || defined (WIN32) 83#define PATHSEP ';' 84#else 85#define PATHSEP ':' 86#endif 87 88#define private static 89 90#if HAVE_VISIBILITY && !defined(WIN32) 91#define public __attribute__ ((__visibility__("default"))) 92#ifndef protected 93#define protected __attribute__ ((__visibility__("hidden"))) 94#endif 95#else 96#define public 97#ifndef protected 98#define protected 99#endif 100#endif 101 102#ifndef __arraycount 103#define __arraycount(a) (sizeof(a) / sizeof(a[0])) 104#endif 105 106#ifndef __GNUC_PREREQ__ 107#ifdef __GNUC__ 108#define __GNUC_PREREQ__(x, y) \ 109 ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \ 110 (__GNUC__ > (x))) 111#else 112#define __GNUC_PREREQ__(x, y) 0 113#endif 114#endif 115 116#ifndef __GNUC__ 117#ifndef __attribute__ 118#define __attribute__(a) 119#endif 120#endif 121 122#ifndef MIN 123#define MIN(a,b) (((a) < (b)) ? (a) : (b)) 124#endif 125 126#ifndef MAX 127#define MAX(a,b) (((a) > (b)) ? (a) : (b)) 128#endif 129 130#ifndef FILE_BYTES_MAX 131# define FILE_BYTES_MAX (1024 * 1024) /* how much of the file to look at */ 132#endif 133#define MAXMAGIS 8192 /* max entries in any one magic file 134 or directory */ 135#define MAXDESC 64 /* max len of text description/MIME type */ 136#define MAXMIME 80 /* max len of text MIME type */ 137#define MAXstring 96 /* max len of "string" types */ 138 139#define MAGICNO 0xF11E041C 140#define VERSIONNO 14 141#define FILE_MAGICSIZE 344 142 143#define FILE_LOAD 0 144#define FILE_CHECK 1 145#define FILE_COMPILE 2 146#define FILE_LIST 3 147 148union VALUETYPE { 149 uint8_t b; 150 uint16_t h; 151 uint32_t l; 152 uint64_t q; 153 uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */ 154 uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */ 155 uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */ 156 char s[MAXstring]; /* the search string or regex pattern */ 157 unsigned char us[MAXstring]; 158 float f; 159 double d; 160}; 161 162struct magic { 163 /* Word 1 */ 164 uint16_t cont_level; /* level of ">" */ 165 uint8_t flag; 166#define INDIR 0x01 /* if '(...)' appears */ 167#define OFFADD 0x02 /* if '>&' or '>...(&' appears */ 168#define INDIROFFADD 0x04 /* if '>&(' appears */ 169#define UNSIGNED 0x08 /* comparison is unsigned */ 170#define NOSPACE 0x10 /* suppress space character before output */ 171#define BINTEST 0x20 /* test is for a binary type (set only 172 for top-level tests) */ 173#define TEXTTEST 0x40 /* for passing to file_softmagic */ 174 175 uint8_t factor; 176 177 /* Word 2 */ 178 uint8_t reln; /* relation (0=eq, '>'=gt, etc) */ 179 uint8_t vallen; /* length of string value, if any */ 180 uint8_t type; /* comparison type (FILE_*) */ 181 uint8_t in_type; /* type of indirection */ 182#define FILE_INVALID 0 183#define FILE_BYTE 1 184#define FILE_SHORT 2 185#define FILE_DEFAULT 3 186#define FILE_LONG 4 187#define FILE_STRING 5 188#define FILE_DATE 6 189#define FILE_BESHORT 7 190#define FILE_BELONG 8 191#define FILE_BEDATE 9 192#define FILE_LESHORT 10 193#define FILE_LELONG 11 194#define FILE_LEDATE 12 195#define FILE_PSTRING 13 196#define FILE_LDATE 14 197#define FILE_BELDATE 15 198#define FILE_LELDATE 16 199#define FILE_REGEX 17 200#define FILE_BESTRING16 18 201#define FILE_LESTRING16 19 202#define FILE_SEARCH 20 203#define FILE_MEDATE 21 204#define FILE_MELDATE 22 205#define FILE_MELONG 23 206#define FILE_QUAD 24 207#define FILE_LEQUAD 25 208#define FILE_BEQUAD 26 209#define FILE_QDATE 27 210#define FILE_LEQDATE 28 211#define FILE_BEQDATE 29 212#define FILE_QLDATE 30 213#define FILE_LEQLDATE 31 214#define FILE_BEQLDATE 32 215#define FILE_FLOAT 33 216#define FILE_BEFLOAT 34 217#define FILE_LEFLOAT 35 218#define FILE_DOUBLE 36 219#define FILE_BEDOUBLE 37 220#define FILE_LEDOUBLE 38 221#define FILE_BEID3 39 222#define FILE_LEID3 40 223#define FILE_INDIRECT 41 224#define FILE_QWDATE 42 225#define FILE_LEQWDATE 43 226#define FILE_BEQWDATE 44 227#define FILE_NAME 45 228#define FILE_USE 46 229#define FILE_CLEAR 47 230#define FILE_DER 48 231#define FILE_NAMES_SIZE 49 /* size of array to contain all names */ 232 233#define IS_STRING(t) \ 234 ((t) == FILE_STRING || \ 235 (t) == FILE_PSTRING || \ 236 (t) == FILE_BESTRING16 || \ 237 (t) == FILE_LESTRING16 || \ 238 (t) == FILE_REGEX || \ 239 (t) == FILE_SEARCH || \ 240 (t) == FILE_INDIRECT || \ 241 (t) == FILE_NAME || \ 242 (t) == FILE_USE) 243 244#define FILE_FMT_NONE 0 245#define FILE_FMT_NUM 1 /* "cduxXi" */ 246#define FILE_FMT_STR 2 /* "s" */ 247#define FILE_FMT_QUAD 3 /* "ll" */ 248#define FILE_FMT_FLOAT 4 /* "eEfFgG" */ 249#define FILE_FMT_DOUBLE 5 /* "eEfFgG" */ 250 251 /* Word 3 */ 252 uint8_t in_op; /* operator for indirection */ 253 uint8_t mask_op; /* operator for mask */ 254#ifdef ENABLE_CONDITIONALS 255 uint8_t cond; /* conditional type */ 256#else 257 uint8_t dummy; 258#endif 259 uint8_t factor_op; 260#define FILE_FACTOR_OP_PLUS '+' 261#define FILE_FACTOR_OP_MINUS '-' 262#define FILE_FACTOR_OP_TIMES '*' 263#define FILE_FACTOR_OP_DIV '/' 264#define FILE_FACTOR_OP_NONE '\0' 265 266#define FILE_OPS "&|^+-*/%" 267#define FILE_OPAND 0 268#define FILE_OPOR 1 269#define FILE_OPXOR 2 270#define FILE_OPADD 3 271#define FILE_OPMINUS 4 272#define FILE_OPMULTIPLY 5 273#define FILE_OPDIVIDE 6 274#define FILE_OPMODULO 7 275#define FILE_OPS_MASK 0x07 /* mask for above ops */ 276#define FILE_UNUSED_1 0x08 277#define FILE_UNUSED_2 0x10 278#define FILE_OPSIGNED 0x20 279#define FILE_OPINVERSE 0x40 280#define FILE_OPINDIRECT 0x80 281 282#ifdef ENABLE_CONDITIONALS 283#define COND_NONE 0 284#define COND_IF 1 285#define COND_ELIF 2 286#define COND_ELSE 3 287#endif /* ENABLE_CONDITIONALS */ 288 289 /* Word 4 */ 290 uint32_t offset; /* offset to magic number */ 291 /* Word 5 */ 292 int32_t in_offset; /* offset from indirection */ 293 /* Word 6 */ 294 uint32_t lineno; /* line number in magic file */ 295 /* Word 7,8 */ 296 union { 297 uint64_t _mask; /* for use with numeric and date types */ 298 struct { 299 uint32_t _count; /* repeat/line count */ 300 uint32_t _flags; /* modifier flags */ 301 } _s; /* for use with string types */ 302 } _u; 303#define num_mask _u._mask 304#define str_range _u._s._count 305#define str_flags _u._s._flags 306 /* Words 9-24 */ 307 union VALUETYPE value; /* either number or string */ 308 /* Words 25-40 */ 309 char desc[MAXDESC]; /* description */ 310 /* Words 41-60 */ 311 char mimetype[MAXMIME]; /* MIME type */ 312 /* Words 61-62 */ 313 char apple[8]; /* APPLE CREATOR/TYPE */ 314 /* Words 63-78 */ 315 char ext[64]; /* Popular extensions */ 316}; 317 318#define BIT(A) (1 << (A)) 319#define STRING_COMPACT_WHITESPACE BIT(0) 320#define STRING_COMPACT_OPTIONAL_WHITESPACE BIT(1) 321#define STRING_IGNORE_LOWERCASE BIT(2) 322#define STRING_IGNORE_UPPERCASE BIT(3) 323#define REGEX_OFFSET_START BIT(4) 324#define STRING_TEXTTEST BIT(5) 325#define STRING_BINTEST BIT(6) 326#define PSTRING_1_BE BIT(7) 327#define PSTRING_1_LE BIT(7) 328#define PSTRING_2_BE BIT(8) 329#define PSTRING_2_LE BIT(9) 330#define PSTRING_4_BE BIT(10) 331#define PSTRING_4_LE BIT(11) 332#define REGEX_LINE_COUNT BIT(11) 333#define PSTRING_LEN \ 334 (PSTRING_1_BE|PSTRING_2_LE|PSTRING_2_BE|PSTRING_4_LE|PSTRING_4_BE) 335#define PSTRING_LENGTH_INCLUDES_ITSELF BIT(12) 336#define STRING_TRIM BIT(13) 337#define CHAR_COMPACT_WHITESPACE 'W' 338#define CHAR_COMPACT_OPTIONAL_WHITESPACE 'w' 339#define CHAR_IGNORE_LOWERCASE 'c' 340#define CHAR_IGNORE_UPPERCASE 'C' 341#define CHAR_REGEX_OFFSET_START 's' 342#define CHAR_TEXTTEST 't' 343#define CHAR_TRIM 'T' 344#define CHAR_BINTEST 'b' 345#define CHAR_PSTRING_1_BE 'B' 346#define CHAR_PSTRING_1_LE 'B' 347#define CHAR_PSTRING_2_BE 'H' 348#define CHAR_PSTRING_2_LE 'h' 349#define CHAR_PSTRING_4_BE 'L' 350#define CHAR_PSTRING_4_LE 'l' 351#define CHAR_PSTRING_LENGTH_INCLUDES_ITSELF 'J' 352#define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE) 353#define STRING_DEFAULT_RANGE 100 354 355#define INDIRECT_RELATIVE BIT(0) 356#define CHAR_INDIRECT_RELATIVE 'r' 357 358/* list of magic entries */ 359struct mlist { 360 struct magic *magic; /* array of magic entries */ 361 uint32_t nmagic; /* number of entries in array */ 362 void *map; /* internal resources used by entry */ 363 struct mlist *next, *prev; 364}; 365 366#ifdef __cplusplus 367#define CAST(T, b) static_cast<T>(b) 368#define RCAST(T, b) reinterpret_cast<T>(b) 369#define CCAST(T, b) const_cast<T>(b) 370#else 371#define CAST(T, b) ((T)(b)) 372#define RCAST(T, b) ((T)(b)) 373#define CCAST(T, b) ((T)(uintptr_t)(b)) 374#endif 375 376struct level_info { 377 int32_t off; 378 int got_match; 379#ifdef ENABLE_CONDITIONALS 380 int last_match; 381 int last_cond; /* used for error checking by parse() */ 382#endif 383}; 384 385#define MAGIC_SETS 2 386 387struct magic_set { 388 struct mlist *mlist[MAGIC_SETS]; /* list of regular entries */ 389 struct cont { 390 size_t len; 391 struct level_info *li; 392 } c; 393 struct out { 394 char *buf; /* Accumulation buffer */ 395 char *pbuf; /* Printable buffer */ 396 } o; 397 uint32_t offset; 398 int error; 399 int flags; /* Control magic tests. */ 400 int event_flags; /* Note things that happened. */ 401#define EVENT_HAD_ERR 0x01 402 const char *file; 403 size_t line; /* current magic line number */ 404 405 /* data for searches */ 406 struct { 407 const char *s; /* start of search in original source */ 408 size_t s_len; /* length of search region */ 409 size_t offset; /* starting offset in source: XXX - should this be off_t? */ 410 size_t rm_len; /* match length */ 411 } search; 412 413 /* FIXME: Make the string dynamically allocated so that e.g. 414 strings matched in files can be longer than MAXstring */ 415 union VALUETYPE ms_value; /* either number or string */ 416 uint16_t indir_max; 417 uint16_t name_max; 418 uint16_t elf_shnum_max; 419 uint16_t elf_phnum_max; 420 uint16_t elf_notes_max; 421 uint16_t regex_max; 422 size_t bytes_max; /* number of bytes to read from file */ 423#define FILE_INDIR_MAX 50 424#define FILE_NAME_MAX 30 425#define FILE_ELF_SHNUM_MAX 32768 426#define FILE_ELF_PHNUM_MAX 2048 427#define FILE_ELF_NOTES_MAX 256 428#define FILE_REGEX_MAX 8192 429}; 430 431/* Type for Unicode characters */ 432typedef unsigned long unichar; 433 434struct stat; 435#define FILE_T_LOCAL 1 436#define FILE_T_WINDOWS 2 437protected const char *file_fmttime(uint64_t, int, char *); 438protected struct magic_set *file_ms_alloc(int); 439protected void file_ms_free(struct magic_set *); 440protected int file_buffer(struct magic_set *, int, const char *, const void *, 441 size_t); 442protected int file_fsmagic(struct magic_set *, const char *, struct stat *); 443protected int file_pipe2file(struct magic_set *, int, const void *, size_t); 444protected int file_vprintf(struct magic_set *, const char *, va_list) 445 __attribute__((__format__(__printf__, 2, 0))); 446protected size_t file_printedlen(const struct magic_set *); 447protected int file_replace(struct magic_set *, const char *, const char *); 448protected int file_printf(struct magic_set *, const char *, ...) 449 __attribute__((__format__(__printf__, 2, 3))); 450protected int file_reset(struct magic_set *, int); 451protected int file_tryelf(struct magic_set *, int, const unsigned char *, 452 size_t); 453protected int file_trycdf(struct magic_set *, int, const unsigned char *, 454 size_t); 455#if HAVE_FORK 456protected int file_zmagic(struct magic_set *, int, const char *, 457 const unsigned char *, size_t); 458#endif 459protected int file_ascmagic(struct magic_set *, const unsigned char *, size_t, 460 int); 461protected int file_ascmagic_with_encoding(struct magic_set *, 462 const unsigned char *, size_t, unichar *, size_t, const char *, 463 const char *, int); 464protected int file_encoding(struct magic_set *, const unsigned char *, size_t, 465 unichar **, size_t *, const char **, const char **, const char **); 466protected int file_is_tar(struct magic_set *, const unsigned char *, size_t); 467protected int file_softmagic(struct magic_set *, const unsigned char *, size_t, 468 uint16_t *, uint16_t *, int, int); 469protected int file_apprentice(struct magic_set *, const char *, int); 470protected int buffer_apprentice(struct magic_set *, struct magic **, 471 size_t *, size_t); 472protected int file_magicfind(struct magic_set *, const char *, struct mlist *); 473protected uint64_t file_signextend(struct magic_set *, struct magic *, 474 uint64_t); 475protected void file_badread(struct magic_set *); 476protected void file_badseek(struct magic_set *); 477protected void file_oomem(struct magic_set *, size_t); 478protected void file_error(struct magic_set *, int, const char *, ...) 479 __attribute__((__format__(__printf__, 3, 4))); 480protected void file_magerror(struct magic_set *, const char *, ...) 481 __attribute__((__format__(__printf__, 2, 3))); 482protected void file_magwarn(struct magic_set *, const char *, ...) 483 __attribute__((__format__(__printf__, 2, 3))); 484protected void file_mdump(struct magic *); 485protected void file_showstr(FILE *, const char *, size_t); 486protected size_t file_mbswidth(const char *); 487protected const char *file_getbuffer(struct magic_set *); 488protected ssize_t sread(int, void *, size_t, int); 489protected int file_check_mem(struct magic_set *, unsigned int); 490protected int file_looks_utf8(const unsigned char *, size_t, unichar *, 491 size_t *); 492protected size_t file_pstring_length_size(const struct magic *); 493protected size_t file_pstring_get_length(const struct magic *, const char *); 494protected char * file_printable(char *, size_t, const char *); 495#ifdef __EMX__ 496protected int file_os2_apptype(struct magic_set *, const char *, const void *, 497 size_t); 498#endif /* __EMX__ */ 499 500#if defined(HAVE_LOCALE_H) 501#include <locale.h> 502#endif 503#if defined(HAVE_XLOCALE_H) 504#include <xlocale.h> 505#endif 506 507typedef struct { 508 const char *pat; 509#if defined(HAVE_NEWLOCALE) && defined(HAVE_USELOCALE) && defined(HAVE_FREELOCALE) 510#define USE_C_LOCALE 511 locale_t old_lc_ctype; 512 locale_t c_lc_ctype; 513#else 514 char *old_lc_ctype; 515#endif 516 int rc; 517 regex_t rx; 518} file_regex_t; 519 520protected int file_regcomp(file_regex_t *, const char *, int); 521protected int file_regexec(file_regex_t *, const char *, size_t, regmatch_t *, 522 int); 523protected void file_regfree(file_regex_t *); 524protected void file_regerror(file_regex_t *, int, struct magic_set *); 525 526typedef struct { 527 char *buf; 528 uint32_t offset; 529} file_pushbuf_t; 530 531protected file_pushbuf_t *file_push_buffer(struct magic_set *); 532protected char *file_pop_buffer(struct magic_set *, file_pushbuf_t *); 533 534#ifndef COMPILE_ONLY 535extern const char *file_names[]; 536extern const size_t file_nnames; 537#endif 538 539#ifndef HAVE_STRERROR 540extern int sys_nerr; 541extern char *sys_errlist[]; 542#define strerror(e) \ 543 (((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error") 544#endif 545 546#ifndef HAVE_STRTOUL 547#define strtoul(a, b, c) strtol(a, b, c) 548#endif 549 550#ifndef HAVE_PREAD 551ssize_t pread(int, void *, size_t, off_t); 552#endif 553#ifndef HAVE_VASPRINTF 554int vasprintf(char **, const char *, va_list); 555#endif 556#ifndef HAVE_ASPRINTF 557int asprintf(char **, const char *, ...); 558#endif 559#ifndef HAVE_DPRINTF 560int dprintf(int, const char *, ...); 561#endif 562 563#ifndef HAVE_STRLCPY 564size_t strlcpy(char *, const char *, size_t); 565#endif 566#ifndef HAVE_STRLCAT 567size_t strlcat(char *, const char *, size_t); 568#endif 569#ifndef HAVE_STRCASESTR 570char *strcasestr(const char *, const char *); 571#endif 572#ifndef HAVE_GETLINE 573ssize_t getline(char **, size_t *, FILE *); 574ssize_t getdelim(char **, size_t *, int, FILE *); 575#endif 576#ifndef HAVE_CTIME_R 577char *ctime_r(const time_t *, char *); 578#endif 579#ifndef HAVE_ASCTIME_R 580char *asctime_r(const struct tm *, char *); 581#endif 582#ifndef HAVE_GMTIME_R 583struct tm *gmtime_r(const time_t *, struct tm *); 584#endif 585#ifndef HAVE_LOCALTIME_R 586struct tm *localtime_r(const time_t *, struct tm *); 587#endif 588#ifndef HAVE_FMTCHECK 589const char *fmtcheck(const char *, const char *) 590 __attribute__((__format_arg__(2))); 591#endif 592 593#if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK) 594#define QUICK 595#endif 596 597#ifndef O_BINARY 598#define O_BINARY 0 599#endif 600 601#ifndef __cplusplus 602#if defined(__GNUC__) && (__GNUC__ >= 3) 603#define FILE_RCSID(id) \ 604static const char rcsid[] __attribute__((__used__)) = id; 605#else 606#define FILE_RCSID(id) \ 607static const char *rcsid(const char *p) { \ 608 return rcsid(p = id); \ 609} 610#endif 611#else 612#define FILE_RCSID(id) 613#endif 614#ifndef __RCSID 615#define __RCSID(a) 616#endif 617 618#endif /* __file_h__ */ 619