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 #include <wchar.h> 74 75 #define ENABLE_CONDITIONALS 76 77 #ifndef MAGIC 78 #define MAGIC "/etc/magic" 79 #endif 80 81 #if defined(__EMX__) || defined (WIN32) 82 #define PATHSEP ';' 83 #else 84 #define PATHSEP ':' 85 #endif 86 87 #define private static 88 89 #if HAVE_VISIBILITY && !defined(WIN32) 90 #define public __attribute__ ((__visibility__("default"))) 91 #ifndef protected 92 #define protected __attribute__ ((__visibility__("hidden"))) 93 #endif 94 #else 95 #define public 96 #ifndef protected 97 #define protected 98 #endif 99 #endif 100 101 #ifndef __arraycount 102 #define __arraycount(a) (sizeof(a) / sizeof(a[0])) 103 #endif 104 105 #ifndef __GNUC_PREREQ__ 106 #ifdef __GNUC__ 107 #define __GNUC_PREREQ__(x, y) \ 108 ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \ 109 (__GNUC__ > (x))) 110 #else 111 #define __GNUC_PREREQ__(x, y) 0 112 #endif 113 #endif 114 115 #ifndef __GNUC__ 116 #ifndef __attribute__ 117 #define __attribute__(a) 118 #endif 119 #endif 120 121 #ifndef MIN 122 #define MIN(a,b) (((a) < (b)) ? (a) : (b)) 123 #endif 124 125 #ifndef MAX 126 #define MAX(a,b) (((a) > (b)) ? (a) : (b)) 127 #endif 128 129 #ifndef HOWMANY 130 # define HOWMANY (256 * 1024) /* how much of the file to look at */ 131 #endif 132 #define MAXMAGIS 8192 /* max entries in any one magic file 133 or directory */ 134 #define MAXDESC 64 /* max len of text description/MIME type */ 135 #define MAXMIME 80 /* max len of text MIME type */ 136 #define MAXstring 64 /* max len of "string" types */ 137 138 #define MAGICNO 0xF11E041C 139 #define VERSIONNO 12 140 #define FILE_MAGICSIZE 248 141 142 #define FILE_LOAD 0 143 #define FILE_CHECK 1 144 #define FILE_COMPILE 2 145 #define FILE_LIST 3 146 147 union VALUETYPE { 148 uint8_t b; 149 uint16_t h; 150 uint32_t l; 151 uint64_t q; 152 uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */ 153 uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */ 154 uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */ 155 char s[MAXstring]; /* the search string or regex pattern */ 156 unsigned char us[MAXstring]; 157 float f; 158 double d; 159 }; 160 161 struct magic { 162 /* Word 1 */ 163 uint16_t cont_level; /* level of ">" */ 164 uint8_t flag; 165 #define INDIR 0x01 /* if '(...)' appears */ 166 #define OFFADD 0x02 /* if '>&' or '>...(&' appears */ 167 #define INDIROFFADD 0x04 /* if '>&(' appears */ 168 #define UNSIGNED 0x08 /* comparison is unsigned */ 169 #define NOSPACE 0x10 /* suppress space character before output */ 170 #define BINTEST 0x20 /* test is for a binary type (set only 171 for top-level tests) */ 172 #define TEXTTEST 0x40 /* for passing to file_softmagic */ 173 174 uint8_t factor; 175 176 /* Word 2 */ 177 uint8_t reln; /* relation (0=eq, '>'=gt, etc) */ 178 uint8_t vallen; /* length of string value, if any */ 179 uint8_t type; /* comparison type (FILE_*) */ 180 uint8_t in_type; /* type of indirection */ 181 #define FILE_INVALID 0 182 #define FILE_BYTE 1 183 #define FILE_SHORT 2 184 #define FILE_DEFAULT 3 185 #define FILE_LONG 4 186 #define FILE_STRING 5 187 #define FILE_DATE 6 188 #define FILE_BESHORT 7 189 #define FILE_BELONG 8 190 #define FILE_BEDATE 9 191 #define FILE_LESHORT 10 192 #define FILE_LELONG 11 193 #define FILE_LEDATE 12 194 #define FILE_PSTRING 13 195 #define FILE_LDATE 14 196 #define FILE_BELDATE 15 197 #define FILE_LELDATE 16 198 #define FILE_REGEX 17 199 #define FILE_BESTRING16 18 200 #define FILE_LESTRING16 19 201 #define FILE_SEARCH 20 202 #define FILE_MEDATE 21 203 #define FILE_MELDATE 22 204 #define FILE_MELONG 23 205 #define FILE_QUAD 24 206 #define FILE_LEQUAD 25 207 #define FILE_BEQUAD 26 208 #define FILE_QDATE 27 209 #define FILE_LEQDATE 28 210 #define FILE_BEQDATE 29 211 #define FILE_QLDATE 30 212 #define FILE_LEQLDATE 31 213 #define FILE_BEQLDATE 32 214 #define FILE_FLOAT 33 215 #define FILE_BEFLOAT 34 216 #define FILE_LEFLOAT 35 217 #define FILE_DOUBLE 36 218 #define FILE_BEDOUBLE 37 219 #define FILE_LEDOUBLE 38 220 #define FILE_BEID3 39 221 #define FILE_LEID3 40 222 #define FILE_INDIRECT 41 223 #define FILE_QWDATE 42 224 #define FILE_LEQWDATE 43 225 #define FILE_BEQWDATE 44 226 #define FILE_NAME 45 227 #define FILE_USE 46 228 #define FILE_CLEAR 47 229 #define FILE_NAMES_SIZE 48 /* size of array to contain all names */ 230 231 #define IS_STRING(t) \ 232 ((t) == FILE_STRING || \ 233 (t) == FILE_PSTRING || \ 234 (t) == FILE_BESTRING16 || \ 235 (t) == FILE_LESTRING16 || \ 236 (t) == FILE_REGEX || \ 237 (t) == FILE_SEARCH || \ 238 (t) == FILE_INDIRECT || \ 239 (t) == FILE_NAME || \ 240 (t) == FILE_USE) 241 242 #define FILE_FMT_NONE 0 243 #define FILE_FMT_NUM 1 /* "cduxXi" */ 244 #define FILE_FMT_STR 2 /* "s" */ 245 #define FILE_FMT_QUAD 3 /* "ll" */ 246 #define FILE_FMT_FLOAT 4 /* "eEfFgG" */ 247 #define FILE_FMT_DOUBLE 5 /* "eEfFgG" */ 248 249 /* Word 3 */ 250 uint8_t in_op; /* operator for indirection */ 251 uint8_t mask_op; /* operator for mask */ 252 #ifdef ENABLE_CONDITIONALS 253 uint8_t cond; /* conditional type */ 254 #else 255 uint8_t dummy; 256 #endif 257 uint8_t factor_op; 258 #define FILE_FACTOR_OP_PLUS '+' 259 #define FILE_FACTOR_OP_MINUS '-' 260 #define FILE_FACTOR_OP_TIMES '*' 261 #define FILE_FACTOR_OP_DIV '/' 262 #define FILE_FACTOR_OP_NONE '\0' 263 264 #define FILE_OPS "&|^+-*/%" 265 #define FILE_OPAND 0 266 #define FILE_OPOR 1 267 #define FILE_OPXOR 2 268 #define FILE_OPADD 3 269 #define FILE_OPMINUS 4 270 #define FILE_OPMULTIPLY 5 271 #define FILE_OPDIVIDE 6 272 #define FILE_OPMODULO 7 273 #define FILE_OPS_MASK 0x07 /* mask for above ops */ 274 #define FILE_UNUSED_1 0x08 275 #define FILE_UNUSED_2 0x10 276 #define FILE_UNUSED_3 0x20 277 #define FILE_OPINVERSE 0x40 278 #define FILE_OPINDIRECT 0x80 279 280 #ifdef ENABLE_CONDITIONALS 281 #define COND_NONE 0 282 #define COND_IF 1 283 #define COND_ELIF 2 284 #define COND_ELSE 3 285 #endif /* ENABLE_CONDITIONALS */ 286 287 /* Word 4 */ 288 uint32_t offset; /* offset to magic number */ 289 /* Word 5 */ 290 int32_t in_offset; /* offset from indirection */ 291 /* Word 6 */ 292 uint32_t lineno; /* line number in magic file */ 293 /* Word 7,8 */ 294 union { 295 uint64_t _mask; /* for use with numeric and date types */ 296 struct { 297 uint32_t _count; /* repeat/line count */ 298 uint32_t _flags; /* modifier flags */ 299 } _s; /* for use with string types */ 300 } _u; 301 #define num_mask _u._mask 302 #define str_range _u._s._count 303 #define str_flags _u._s._flags 304 /* Words 9-16 */ 305 union VALUETYPE value; /* either number or string */ 306 /* Words 17-32 */ 307 char desc[MAXDESC]; /* description */ 308 /* Words 33-52 */ 309 char mimetype[MAXMIME]; /* MIME type */ 310 /* Words 53-54 */ 311 char apple[8]; 312 }; 313 314 #define BIT(A) (1 << (A)) 315 #define STRING_COMPACT_WHITESPACE BIT(0) 316 #define STRING_COMPACT_OPTIONAL_WHITESPACE BIT(1) 317 #define STRING_IGNORE_LOWERCASE BIT(2) 318 #define STRING_IGNORE_UPPERCASE BIT(3) 319 #define REGEX_OFFSET_START BIT(4) 320 #define STRING_TEXTTEST BIT(5) 321 #define STRING_BINTEST BIT(6) 322 #define PSTRING_1_BE BIT(7) 323 #define PSTRING_1_LE BIT(7) 324 #define PSTRING_2_BE BIT(8) 325 #define PSTRING_2_LE BIT(9) 326 #define PSTRING_4_BE BIT(10) 327 #define PSTRING_4_LE BIT(11) 328 #define REGEX_LINE_COUNT BIT(11) 329 #define PSTRING_LEN \ 330 (PSTRING_1_BE|PSTRING_2_LE|PSTRING_2_BE|PSTRING_4_LE|PSTRING_4_BE) 331 #define PSTRING_LENGTH_INCLUDES_ITSELF BIT(12) 332 #define STRING_TRIM BIT(13) 333 #define CHAR_COMPACT_WHITESPACE 'W' 334 #define CHAR_COMPACT_OPTIONAL_WHITESPACE 'w' 335 #define CHAR_IGNORE_LOWERCASE 'c' 336 #define CHAR_IGNORE_UPPERCASE 'C' 337 #define CHAR_REGEX_OFFSET_START 's' 338 #define CHAR_TEXTTEST 't' 339 #define CHAR_TRIM 'T' 340 #define CHAR_BINTEST 'b' 341 #define CHAR_PSTRING_1_BE 'B' 342 #define CHAR_PSTRING_1_LE 'B' 343 #define CHAR_PSTRING_2_BE 'H' 344 #define CHAR_PSTRING_2_LE 'h' 345 #define CHAR_PSTRING_4_BE 'L' 346 #define CHAR_PSTRING_4_LE 'l' 347 #define CHAR_PSTRING_LENGTH_INCLUDES_ITSELF 'J' 348 #define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE) 349 #define STRING_DEFAULT_RANGE 100 350 351 #define INDIRECT_RELATIVE BIT(0) 352 #define CHAR_INDIRECT_RELATIVE 'r' 353 354 /* list of magic entries */ 355 struct mlist { 356 struct magic *magic; /* array of magic entries */ 357 uint32_t nmagic; /* number of entries in array */ 358 void *map; /* internal resources used by entry */ 359 struct mlist *next, *prev; 360 }; 361 362 #ifdef __cplusplus 363 #define CAST(T, b) static_cast<T>(b) 364 #define RCAST(T, b) reinterpret_cast<T>(b) 365 #else 366 #define CAST(T, b) (T)(b) 367 #define RCAST(T, b) (T)(b) 368 #endif 369 370 struct level_info { 371 int32_t off; 372 int got_match; 373 #ifdef ENABLE_CONDITIONALS 374 int last_match; 375 int last_cond; /* used for error checking by parse() */ 376 #endif 377 }; 378 379 #define MAGIC_SETS 2 380 381 struct magic_set { 382 struct mlist *mlist[MAGIC_SETS]; /* list of regular entries */ 383 struct cont { 384 size_t len; 385 struct level_info *li; 386 } c; 387 struct out { 388 char *buf; /* Accumulation buffer */ 389 char *pbuf; /* Printable buffer */ 390 } o; 391 uint32_t offset; 392 int error; 393 int flags; /* Control magic tests. */ 394 int event_flags; /* Note things that happened. */ 395 #define EVENT_HAD_ERR 0x01 396 const char *file; 397 size_t line; /* current magic line number */ 398 399 /* data for searches */ 400 struct { 401 const char *s; /* start of search in original source */ 402 size_t s_len; /* length of search region */ 403 size_t offset; /* starting offset in source: XXX - should this be off_t? */ 404 size_t rm_len; /* match length */ 405 } search; 406 407 /* FIXME: Make the string dynamically allocated so that e.g. 408 strings matched in files can be longer than MAXstring */ 409 union VALUETYPE ms_value; /* either number or string */ 410 uint16_t indir_max; 411 uint16_t name_max; 412 uint16_t elf_shnum_max; 413 uint16_t elf_phnum_max; 414 uint16_t elf_notes_max; 415 #define FILE_INDIR_MAX 15 416 #define FILE_NAME_MAX 30 417 #define FILE_ELF_SHNUM_MAX 32768 418 #define FILE_ELF_PHNUM_MAX 128 419 #define FILE_ELF_NOTES_MAX 256 420 }; 421 422 /* Type for Unicode characters */ 423 typedef unsigned long unichar; 424 425 struct stat; 426 #define FILE_T_LOCAL 1 427 #define FILE_T_WINDOWS 2 428 protected const char *file_fmttime(uint64_t, int, char *); 429 protected struct magic_set *file_ms_alloc(int); 430 protected void file_ms_free(struct magic_set *); 431 protected int file_buffer(struct magic_set *, int, const char *, const void *, 432 size_t); 433 protected int file_fsmagic(struct magic_set *, const char *, struct stat *); 434 protected int file_pipe2file(struct magic_set *, int, const void *, size_t); 435 protected int file_vprintf(struct magic_set *, const char *, va_list) 436 __attribute__((__format__(__printf__, 2, 0))); 437 protected size_t file_printedlen(const struct magic_set *); 438 protected int file_replace(struct magic_set *, const char *, const char *); 439 protected int file_printf(struct magic_set *, const char *, ...) 440 __attribute__((__format__(__printf__, 2, 3))); 441 protected int file_reset(struct magic_set *); 442 protected int file_tryelf(struct magic_set *, int, const unsigned char *, 443 size_t); 444 protected int file_trycdf(struct magic_set *, int, const unsigned char *, 445 size_t); 446 #if HAVE_FORK 447 protected int file_zmagic(struct magic_set *, int, const char *, 448 const unsigned char *, size_t); 449 #endif 450 protected int file_ascmagic(struct magic_set *, const unsigned char *, size_t, 451 int); 452 protected int file_ascmagic_with_encoding(struct magic_set *, 453 const unsigned char *, size_t, unichar *, size_t, const char *, 454 const char *, int); 455 protected int file_encoding(struct magic_set *, const unsigned char *, size_t, 456 unichar **, size_t *, const char **, const char **, const char **); 457 protected int file_is_tar(struct magic_set *, const unsigned char *, size_t); 458 protected int file_softmagic(struct magic_set *, const unsigned char *, size_t, 459 uint16_t, uint16_t *, int, int); 460 protected int file_apprentice(struct magic_set *, const char *, int); 461 protected int buffer_apprentice(struct magic_set *, struct magic **, 462 size_t *, size_t); 463 protected int file_magicfind(struct magic_set *, const char *, struct mlist *); 464 protected uint64_t file_signextend(struct magic_set *, struct magic *, 465 uint64_t); 466 protected void file_badread(struct magic_set *); 467 protected void file_badseek(struct magic_set *); 468 protected void file_oomem(struct magic_set *, size_t); 469 protected void file_error(struct magic_set *, int, const char *, ...) 470 __attribute__((__format__(__printf__, 3, 4))); 471 protected void file_magerror(struct magic_set *, const char *, ...) 472 __attribute__((__format__(__printf__, 2, 3))); 473 protected void file_magwarn(struct magic_set *, const char *, ...) 474 __attribute__((__format__(__printf__, 2, 3))); 475 protected void file_mdump(struct magic *); 476 protected void file_showstr(FILE *, const char *, size_t); 477 protected size_t file_mbswidth(const char *); 478 protected const char *file_getbuffer(struct magic_set *); 479 protected ssize_t sread(int, void *, size_t, int); 480 protected int file_check_mem(struct magic_set *, unsigned int); 481 protected int file_looks_utf8(const unsigned char *, size_t, unichar *, 482 size_t *); 483 protected size_t file_pstring_length_size(const struct magic *); 484 protected size_t file_pstring_get_length(const struct magic *, const char *); 485 protected char * file_printable(char *, size_t, const char *); 486 #ifdef __EMX__ 487 protected int file_os2_apptype(struct magic_set *, const char *, const void *, 488 size_t); 489 #endif /* __EMX__ */ 490 491 #if defined(HAVE_LOCALE_H) 492 #include <locale.h> 493 #endif 494 #if defined(HAVE_XLOCALE_H) 495 #include <xlocale.h> 496 #endif 497 498 typedef struct { 499 const char *pat; 500 #if defined(HAVE_NEWLOCALE) && defined(HAVE_USELOCALE) && defined(HAVE_FREELOCALE) 501 #define USE_C_LOCALE 502 locale_t old_lc_ctype; 503 locale_t c_lc_ctype; 504 #endif 505 int rc; 506 regex_t rx; 507 } file_regex_t; 508 509 protected int file_regcomp(file_regex_t *, const char *, int); 510 protected int file_regexec(file_regex_t *, const char *, size_t, regmatch_t *, 511 int); 512 protected void file_regfree(file_regex_t *); 513 protected void file_regerror(file_regex_t *, int, struct magic_set *); 514 515 typedef struct { 516 char *buf; 517 uint32_t offset; 518 } file_pushbuf_t; 519 520 protected file_pushbuf_t *file_push_buffer(struct magic_set *); 521 protected char *file_pop_buffer(struct magic_set *, file_pushbuf_t *); 522 523 #ifndef COMPILE_ONLY 524 extern const char *file_names[]; 525 extern const size_t file_nnames; 526 #endif 527 528 #ifndef HAVE_STRERROR 529 extern int sys_nerr; 530 extern char *sys_errlist[]; 531 #define strerror(e) \ 532 (((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error") 533 #endif 534 535 #ifndef HAVE_STRTOUL 536 #define strtoul(a, b, c) strtol(a, b, c) 537 #endif 538 539 #ifndef HAVE_PREAD 540 ssize_t pread(int, void *, size_t, off_t); 541 #endif 542 #ifndef HAVE_VASPRINTF 543 int vasprintf(char **, const char *, va_list); 544 #endif 545 #ifndef HAVE_ASPRINTF 546 int asprintf(char **, const char *, ...); 547 #endif 548 549 #ifndef HAVE_STRLCPY 550 size_t strlcpy(char *, const char *, size_t); 551 #endif 552 #ifndef HAVE_STRLCAT 553 size_t strlcat(char *, const char *, size_t); 554 #endif 555 #ifndef HAVE_STRCASESTR 556 char *strcasestr(const char *, const char *); 557 #endif 558 #ifndef HAVE_GETLINE 559 ssize_t getline(char **, size_t *, FILE *); 560 ssize_t getdelim(char **, size_t *, int, FILE *); 561 #endif 562 #ifndef HAVE_CTIME_R 563 char *ctime_r(const time_t *, char *); 564 #endif 565 #ifndef HAVE_ASCTIME_R 566 char *asctime_r(const struct tm *, char *); 567 #endif 568 #ifndef HAVE_FMTCHECK 569 const char *fmtcheck(const char *, const char *) 570 __attribute__((__format_arg__(2))); 571 #endif 572 573 #if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK) 574 #define QUICK 575 #endif 576 577 #ifndef O_BINARY 578 #define O_BINARY 0 579 #endif 580 581 #ifndef __cplusplus 582 #if defined(__GNUC__) && (__GNUC__ >= 3) 583 #define FILE_RCSID(id) \ 584 static const char rcsid[] __attribute__((__used__)) = id; 585 #else 586 #define FILE_RCSID(id) \ 587 static const char *rcsid(const char *p) { \ 588 return rcsid(p = id); \ 589 } 590 #endif 591 #else 592 #define FILE_RCSID(id) 593 #endif 594 595 #endif /* __file_h__ */ 596