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 1.133 2011/05/13 22:15:40 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 <sys/types.h> 66 #include <sys/param.h> 67 /* Do this here and now, because struct stat gets re-defined on solaris */ 68 #include <sys/stat.h> 69 #include <stdarg.h> 70 71 #define ENABLE_CONDITIONALS 72 73 #ifndef MAGIC 74 #define MAGIC "/etc/magic" 75 #endif 76 77 #if defined(__EMX__) || defined (WIN32) 78 #define PATHSEP ';' 79 #else 80 #define PATHSEP ':' 81 #endif 82 83 #define private static 84 #ifndef protected 85 #define protected 86 #endif 87 #define public 88 89 #ifndef __GNUC_PREREQ__ 90 #ifdef __GNUC__ 91 #define __GNUC_PREREQ__(x, y) \ 92 ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \ 93 (__GNUC__ > (x))) 94 #else 95 #define __GNUC_PREREQ__(x, y) 0 96 #endif 97 #endif 98 99 #ifndef __GNUC__ 100 #ifndef __attribute__ 101 #define __attribute__(a) 102 #endif 103 #endif 104 105 #ifndef MIN 106 #define MIN(a,b) (((a) < (b)) ? (a) : (b)) 107 #endif 108 109 #ifndef MAX 110 #define MAX(a,b) (((a) > (b)) ? (a) : (b)) 111 #endif 112 113 #ifndef HOWMANY 114 # define HOWMANY (256 * 1024) /* how much of the file to look at */ 115 #endif 116 #define MAXMAGIS 8192 /* max entries in any one magic file 117 or directory */ 118 #define MAXDESC 64 /* max leng of text description/MIME type */ 119 #define MAXstring 64 /* max leng of "string" types */ 120 121 #define MAGICNO 0xF11E041C 122 #define VERSIONNO 8 123 #define FILE_MAGICSIZE 232 124 125 #define FILE_LOAD 0 126 #define FILE_CHECK 1 127 #define FILE_COMPILE 2 128 #define FILE_LIST 3 129 130 union VALUETYPE { 131 uint8_t b; 132 uint16_t h; 133 uint32_t l; 134 uint64_t q; 135 uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */ 136 uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */ 137 uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */ 138 char s[MAXstring]; /* the search string or regex pattern */ 139 unsigned char us[MAXstring]; 140 float f; 141 double d; 142 }; 143 144 struct magic { 145 /* Word 1 */ 146 uint16_t cont_level; /* level of ">" */ 147 uint8_t flag; 148 #define INDIR 0x01 /* if '(...)' appears */ 149 #define OFFADD 0x02 /* if '>&' or '>...(&' appears */ 150 #define INDIROFFADD 0x04 /* if '>&(' appears */ 151 #define UNSIGNED 0x08 /* comparison is unsigned */ 152 #define NOSPACE 0x10 /* suppress space character before output */ 153 #define BINTEST 0x20 /* test is for a binary type (set only 154 for top-level tests) */ 155 #define TEXTTEST 0x40 /* for passing to file_softmagic */ 156 157 uint8_t factor; 158 159 /* Word 2 */ 160 uint8_t reln; /* relation (0=eq, '>'=gt, etc) */ 161 uint8_t vallen; /* length of string value, if any */ 162 uint8_t type; /* comparison type (FILE_*) */ 163 uint8_t in_type; /* type of indirection */ 164 #define FILE_INVALID 0 165 #define FILE_BYTE 1 166 #define FILE_SHORT 2 167 #define FILE_DEFAULT 3 168 #define FILE_LONG 4 169 #define FILE_STRING 5 170 #define FILE_DATE 6 171 #define FILE_BESHORT 7 172 #define FILE_BELONG 8 173 #define FILE_BEDATE 9 174 #define FILE_LESHORT 10 175 #define FILE_LELONG 11 176 #define FILE_LEDATE 12 177 #define FILE_PSTRING 13 178 #define FILE_LDATE 14 179 #define FILE_BELDATE 15 180 #define FILE_LELDATE 16 181 #define FILE_REGEX 17 182 #define FILE_BESTRING16 18 183 #define FILE_LESTRING16 19 184 #define FILE_SEARCH 20 185 #define FILE_MEDATE 21 186 #define FILE_MELDATE 22 187 #define FILE_MELONG 23 188 #define FILE_QUAD 24 189 #define FILE_LEQUAD 25 190 #define FILE_BEQUAD 26 191 #define FILE_QDATE 27 192 #define FILE_LEQDATE 28 193 #define FILE_BEQDATE 29 194 #define FILE_QLDATE 30 195 #define FILE_LEQLDATE 31 196 #define FILE_BEQLDATE 32 197 #define FILE_FLOAT 33 198 #define FILE_BEFLOAT 34 199 #define FILE_LEFLOAT 35 200 #define FILE_DOUBLE 36 201 #define FILE_BEDOUBLE 37 202 #define FILE_LEDOUBLE 38 203 #define FILE_BEID3 39 204 #define FILE_LEID3 40 205 #define FILE_INDIRECT 41 206 #define FILE_NAMES_SIZE 42/* size of array to contain all names */ 207 208 #define IS_STRING(t) \ 209 ((t) == FILE_STRING || \ 210 (t) == FILE_PSTRING || \ 211 (t) == FILE_BESTRING16 || \ 212 (t) == FILE_LESTRING16 || \ 213 (t) == FILE_REGEX || \ 214 (t) == FILE_SEARCH || \ 215 (t) == FILE_DEFAULT) 216 217 #define FILE_FMT_NONE 0 218 #define FILE_FMT_NUM 1 /* "cduxXi" */ 219 #define FILE_FMT_STR 2 /* "s" */ 220 #define FILE_FMT_QUAD 3 /* "ll" */ 221 #define FILE_FMT_FLOAT 4 /* "eEfFgG" */ 222 #define FILE_FMT_DOUBLE 5 /* "eEfFgG" */ 223 224 /* Word 3 */ 225 uint8_t in_op; /* operator for indirection */ 226 uint8_t mask_op; /* operator for mask */ 227 #ifdef ENABLE_CONDITIONALS 228 uint8_t cond; /* conditional type */ 229 #else 230 uint8_t dummy; 231 #endif 232 uint8_t factor_op; 233 #define FILE_FACTOR_OP_PLUS '+' 234 #define FILE_FACTOR_OP_MINUS '-' 235 #define FILE_FACTOR_OP_TIMES '*' 236 #define FILE_FACTOR_OP_DIV '/' 237 #define FILE_FACTOR_OP_NONE '\0' 238 239 #define FILE_OPS "&|^+-*/%" 240 #define FILE_OPAND 0 241 #define FILE_OPOR 1 242 #define FILE_OPXOR 2 243 #define FILE_OPADD 3 244 #define FILE_OPMINUS 4 245 #define FILE_OPMULTIPLY 5 246 #define FILE_OPDIVIDE 6 247 #define FILE_OPMODULO 7 248 #define FILE_OPS_MASK 0x07 /* mask for above ops */ 249 #define FILE_UNUSED_1 0x08 250 #define FILE_UNUSED_2 0x10 251 #define FILE_UNUSED_3 0x20 252 #define FILE_OPINVERSE 0x40 253 #define FILE_OPINDIRECT 0x80 254 255 #ifdef ENABLE_CONDITIONALS 256 #define COND_NONE 0 257 #define COND_IF 1 258 #define COND_ELIF 2 259 #define COND_ELSE 3 260 #endif /* ENABLE_CONDITIONALS */ 261 262 /* Word 4 */ 263 uint32_t offset; /* offset to magic number */ 264 /* Word 5 */ 265 int32_t in_offset; /* offset from indirection */ 266 /* Word 6 */ 267 uint32_t lineno; /* line number in magic file */ 268 /* Word 7,8 */ 269 union { 270 uint64_t _mask; /* for use with numeric and date types */ 271 struct { 272 uint32_t _count; /* repeat/line count */ 273 uint32_t _flags; /* modifier flags */ 274 } _s; /* for use with string types */ 275 } _u; 276 #define num_mask _u._mask 277 #define str_range _u._s._count 278 #define str_flags _u._s._flags 279 /* Words 9-16 */ 280 union VALUETYPE value; /* either number or string */ 281 /* Words 17-32 */ 282 char desc[MAXDESC]; /* description */ 283 /* Words 33-48 */ 284 char mimetype[MAXDESC]; /* MIME type */ 285 /* Words 49-50 */ 286 char apple[8]; 287 }; 288 289 #define BIT(A) (1 << (A)) 290 #define STRING_COMPACT_WHITESPACE BIT(0) 291 #define STRING_COMPACT_OPTIONAL_WHITESPACE BIT(1) 292 #define STRING_IGNORE_LOWERCASE BIT(2) 293 #define STRING_IGNORE_UPPERCASE BIT(3) 294 #define REGEX_OFFSET_START BIT(4) 295 #define STRING_TEXTTEST BIT(5) 296 #define STRING_BINTEST BIT(6) 297 #define PSTRING_1_BE BIT(7) 298 #define PSTRING_1_LE BIT(7) 299 #define PSTRING_2_BE BIT(8) 300 #define PSTRING_2_LE BIT(9) 301 #define PSTRING_4_BE BIT(10) 302 #define PSTRING_4_LE BIT(11) 303 #define PSTRING_LEN \ 304 (PSTRING_1_BE|PSTRING_2_LE|PSTRING_2_BE|PSTRING_4_LE|PSTRING_4_BE) 305 #define PSTRING_LENGTH_INCLUDES_ITSELF BIT(12) 306 #define CHAR_COMPACT_WHITESPACE 'W' 307 #define CHAR_COMPACT_OPTIONAL_WHITESPACE 'w' 308 #define CHAR_IGNORE_LOWERCASE 'c' 309 #define CHAR_IGNORE_UPPERCASE 'C' 310 #define CHAR_REGEX_OFFSET_START 's' 311 #define CHAR_TEXTTEST 't' 312 #define CHAR_BINTEST 'b' 313 #define CHAR_PSTRING_1_BE 'B' 314 #define CHAR_PSTRING_1_LE 'B' 315 #define CHAR_PSTRING_2_BE 'H' 316 #define CHAR_PSTRING_2_LE 'h' 317 #define CHAR_PSTRING_4_BE 'L' 318 #define CHAR_PSTRING_4_LE 'l' 319 #define CHAR_PSTRING_LENGTH_INCLUDES_ITSELF 'J' 320 #define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE) 321 #define STRING_DEFAULT_RANGE 100 322 323 324 /* list of magic entries */ 325 struct mlist { 326 struct magic *magic; /* array of magic entries */ 327 uint32_t nmagic; /* number of entries in array */ 328 int mapped; /* allocation type: 0 => apprentice_file 329 * 1 => apprentice_map + malloc 330 * 2 => apprentice_map + mmap */ 331 struct mlist *next, *prev; 332 }; 333 334 #ifdef __cplusplus 335 #define CAST(T, b) static_cast<T>(b) 336 #define RCAST(T, b) reinterpret_cast<T>(b) 337 #else 338 #define CAST(T, b) (T)(b) 339 #define RCAST(T, b) (T)(b) 340 #endif 341 342 struct level_info { 343 int32_t off; 344 int got_match; 345 #ifdef ENABLE_CONDITIONALS 346 int last_match; 347 int last_cond; /* used for error checking by parse() */ 348 #endif 349 }; 350 struct magic_set { 351 struct mlist *mlist; 352 struct cont { 353 size_t len; 354 struct level_info *li; 355 } c; 356 struct out { 357 char *buf; /* Accumulation buffer */ 358 char *pbuf; /* Printable buffer */ 359 } o; 360 uint32_t offset; 361 int error; 362 int flags; /* Control magic tests. */ 363 int event_flags; /* Note things that happened. */ 364 #define EVENT_HAD_ERR 0x01 365 const char *file; 366 size_t line; /* current magic line number */ 367 368 /* data for searches */ 369 struct { 370 const char *s; /* start of search in original source */ 371 size_t s_len; /* length of search region */ 372 size_t offset; /* starting offset in source: XXX - should this be off_t? */ 373 size_t rm_len; /* match length */ 374 } search; 375 376 /* FIXME: Make the string dynamically allocated so that e.g. 377 strings matched in files can be longer than MAXstring */ 378 union VALUETYPE ms_value; /* either number or string */ 379 }; 380 381 /* Type for Unicode characters */ 382 typedef unsigned long unichar; 383 384 struct stat; 385 protected const char *file_fmttime(uint32_t, int); 386 protected int file_buffer(struct magic_set *, int, const char *, const void *, 387 size_t); 388 protected int file_fsmagic(struct magic_set *, const char *, struct stat *); 389 protected int file_pipe2file(struct magic_set *, int, const void *, size_t); 390 protected int file_vprintf(struct magic_set *, const char *, va_list); 391 protected size_t file_printedlen(const struct magic_set *); 392 protected int file_replace(struct magic_set *, const char *, const char *); 393 protected int file_printf(struct magic_set *, const char *, ...) 394 __attribute__((__format__(__printf__, 2, 3))); 395 protected int file_reset(struct magic_set *); 396 protected int file_tryelf(struct magic_set *, int, const unsigned char *, 397 size_t); 398 protected int file_trycdf(struct magic_set *, int, const unsigned char *, 399 size_t); 400 #if HAVE_FORK 401 protected int file_zmagic(struct magic_set *, int, const char *, 402 const unsigned char *, size_t); 403 #endif 404 protected int file_ascmagic(struct magic_set *, const unsigned char *, size_t); 405 protected int file_ascmagic_with_encoding(struct magic_set *, 406 const unsigned char *, size_t, unichar *, size_t, const char *, 407 const char *); 408 protected int file_encoding(struct magic_set *, const unsigned char *, size_t, 409 unichar **, size_t *, const char **, const char **, const char **); 410 protected int file_is_tar(struct magic_set *, const unsigned char *, size_t); 411 protected int file_softmagic(struct magic_set *, const unsigned char *, size_t, 412 int); 413 protected struct mlist *file_apprentice(struct magic_set *, const char *, int); 414 protected uint64_t file_signextend(struct magic_set *, struct magic *, 415 uint64_t); 416 protected void file_delmagic(struct magic *, int type, size_t entries); 417 protected void file_badread(struct magic_set *); 418 protected void file_badseek(struct magic_set *); 419 protected void file_oomem(struct magic_set *, size_t); 420 protected void file_error(struct magic_set *, int, const char *, ...) 421 __attribute__((__format__(__printf__, 3, 4))); 422 protected void file_magerror(struct magic_set *, const char *, ...) 423 __attribute__((__format__(__printf__, 2, 3))); 424 protected void file_magwarn(struct magic_set *, const char *, ...) 425 __attribute__((__format__(__printf__, 2, 3))); 426 protected void file_mdump(struct magic *); 427 protected void file_showstr(FILE *, const char *, size_t); 428 protected size_t file_mbswidth(const char *); 429 protected const char *file_getbuffer(struct magic_set *); 430 protected ssize_t sread(int, void *, size_t, int); 431 protected int file_check_mem(struct magic_set *, unsigned int); 432 protected int file_looks_utf8(const unsigned char *, size_t, unichar *, 433 size_t *); 434 protected size_t file_pstring_length_size(const struct magic *); 435 protected size_t file_pstring_get_length(const struct magic *, const char *); 436 #ifdef __EMX__ 437 protected int file_os2_apptype(struct magic_set *, const char *, const void *, 438 size_t); 439 #endif /* __EMX__ */ 440 441 442 #ifndef COMPILE_ONLY 443 extern const char *file_names[]; 444 extern const size_t file_nnames; 445 #endif 446 447 #ifndef HAVE_STRERROR 448 extern int sys_nerr; 449 extern char *sys_errlist[]; 450 #define strerror(e) \ 451 (((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error") 452 #endif 453 454 #ifndef HAVE_STRTOUL 455 #define strtoul(a, b, c) strtol(a, b, c) 456 #endif 457 458 #ifndef HAVE_VASPRINTF 459 int vasprintf(char **, const char *, va_list); 460 #endif 461 #ifndef HAVE_ASPRINTF 462 int asprintf(char **ptr, const char *format_string, ...); 463 #endif 464 465 #ifndef HAVE_STRLCPY 466 size_t strlcpy(char *dst, const char *src, size_t siz); 467 #endif 468 #ifndef HAVE_STRLCAT 469 size_t strlcat(char *dst, const char *src, size_t siz); 470 #endif 471 #ifndef HAVE_GETLINE 472 ssize_t getline(char **dst, size_t *len, FILE *fp); 473 ssize_t getdelim(char **dst, size_t *len, int delimiter, FILE *fp); 474 #endif 475 476 #if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK) 477 #define QUICK 478 #endif 479 480 #ifndef O_BINARY 481 #define O_BINARY 0 482 #endif 483 484 #ifndef __cplusplus 485 #if defined(__GNUC__) && (__GNUC__ >= 3) 486 #define FILE_RCSID(id) \ 487 static const char rcsid[] __attribute__((__used__)) = id; 488 #else 489 #define FILE_RCSID(id) \ 490 static const char *rcsid(const char *p) { \ 491 return rcsid(p = id); \ 492 } 493 #endif 494 #else 495 #define FILE_RCSID(id) 496 #endif 497 498 #endif /* __file_h__ */ 499