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 1.135 2011/09/20 15:30:14 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 __arraycount 90 #define __arraycount(a) (sizeof(a) / sizeof(a[0])) 91 #endif 92 93 #ifndef __GNUC_PREREQ__ 94 #ifdef __GNUC__ 95 #define __GNUC_PREREQ__(x, y) \ 96 ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \ 97 (__GNUC__ > (x))) 98 #else 99 #define __GNUC_PREREQ__(x, y) 0 100 #endif 101 #endif 102 103 #ifndef __GNUC__ 104 #ifndef __attribute__ 105 #define __attribute__(a) 106 #endif 107 #endif 108 109 #ifndef MIN 110 #define MIN(a,b) (((a) < (b)) ? (a) : (b)) 111 #endif 112 113 #ifndef MAX 114 #define MAX(a,b) (((a) > (b)) ? (a) : (b)) 115 #endif 116 117 #ifndef HOWMANY 118 # define HOWMANY (256 * 1024) /* how much of the file to look at */ 119 #endif 120 #define MAXMAGIS 8192 /* max entries in any one magic file 121 or directory */ 122 #define MAXDESC 64 /* max leng of text description/MIME type */ 123 #define MAXstring 64 /* max leng of "string" types */ 124 125 #define MAGICNO 0xF11E041C 126 #define VERSIONNO 8 127 #define FILE_MAGICSIZE 232 128 129 #define FILE_LOAD 0 130 #define FILE_CHECK 1 131 #define FILE_COMPILE 2 132 #define FILE_LIST 3 133 134 union VALUETYPE { 135 uint8_t b; 136 uint16_t h; 137 uint32_t l; 138 uint64_t q; 139 uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */ 140 uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */ 141 uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */ 142 char s[MAXstring]; /* the search string or regex pattern */ 143 unsigned char us[MAXstring]; 144 float f; 145 double d; 146 }; 147 148 struct magic { 149 /* Word 1 */ 150 uint16_t cont_level; /* level of ">" */ 151 uint8_t flag; 152 #define INDIR 0x01 /* if '(...)' appears */ 153 #define OFFADD 0x02 /* if '>&' or '>...(&' appears */ 154 #define INDIROFFADD 0x04 /* if '>&(' appears */ 155 #define UNSIGNED 0x08 /* comparison is unsigned */ 156 #define NOSPACE 0x10 /* suppress space character before output */ 157 #define BINTEST 0x20 /* test is for a binary type (set only 158 for top-level tests) */ 159 #define TEXTTEST 0x40 /* for passing to file_softmagic */ 160 161 uint8_t factor; 162 163 /* Word 2 */ 164 uint8_t reln; /* relation (0=eq, '>'=gt, etc) */ 165 uint8_t vallen; /* length of string value, if any */ 166 uint8_t type; /* comparison type (FILE_*) */ 167 uint8_t in_type; /* type of indirection */ 168 #define FILE_INVALID 0 169 #define FILE_BYTE 1 170 #define FILE_SHORT 2 171 #define FILE_DEFAULT 3 172 #define FILE_LONG 4 173 #define FILE_STRING 5 174 #define FILE_DATE 6 175 #define FILE_BESHORT 7 176 #define FILE_BELONG 8 177 #define FILE_BEDATE 9 178 #define FILE_LESHORT 10 179 #define FILE_LELONG 11 180 #define FILE_LEDATE 12 181 #define FILE_PSTRING 13 182 #define FILE_LDATE 14 183 #define FILE_BELDATE 15 184 #define FILE_LELDATE 16 185 #define FILE_REGEX 17 186 #define FILE_BESTRING16 18 187 #define FILE_LESTRING16 19 188 #define FILE_SEARCH 20 189 #define FILE_MEDATE 21 190 #define FILE_MELDATE 22 191 #define FILE_MELONG 23 192 #define FILE_QUAD 24 193 #define FILE_LEQUAD 25 194 #define FILE_BEQUAD 26 195 #define FILE_QDATE 27 196 #define FILE_LEQDATE 28 197 #define FILE_BEQDATE 29 198 #define FILE_QLDATE 30 199 #define FILE_LEQLDATE 31 200 #define FILE_BEQLDATE 32 201 #define FILE_FLOAT 33 202 #define FILE_BEFLOAT 34 203 #define FILE_LEFLOAT 35 204 #define FILE_DOUBLE 36 205 #define FILE_BEDOUBLE 37 206 #define FILE_LEDOUBLE 38 207 #define FILE_BEID3 39 208 #define FILE_LEID3 40 209 #define FILE_INDIRECT 41 210 #define FILE_NAMES_SIZE 42/* size of array to contain all names */ 211 212 #define IS_STRING(t) \ 213 ((t) == FILE_STRING || \ 214 (t) == FILE_PSTRING || \ 215 (t) == FILE_BESTRING16 || \ 216 (t) == FILE_LESTRING16 || \ 217 (t) == FILE_REGEX || \ 218 (t) == FILE_SEARCH || \ 219 (t) == FILE_DEFAULT) 220 221 #define FILE_FMT_NONE 0 222 #define FILE_FMT_NUM 1 /* "cduxXi" */ 223 #define FILE_FMT_STR 2 /* "s" */ 224 #define FILE_FMT_QUAD 3 /* "ll" */ 225 #define FILE_FMT_FLOAT 4 /* "eEfFgG" */ 226 #define FILE_FMT_DOUBLE 5 /* "eEfFgG" */ 227 228 /* Word 3 */ 229 uint8_t in_op; /* operator for indirection */ 230 uint8_t mask_op; /* operator for mask */ 231 #ifdef ENABLE_CONDITIONALS 232 uint8_t cond; /* conditional type */ 233 #else 234 uint8_t dummy; 235 #endif 236 uint8_t factor_op; 237 #define FILE_FACTOR_OP_PLUS '+' 238 #define FILE_FACTOR_OP_MINUS '-' 239 #define FILE_FACTOR_OP_TIMES '*' 240 #define FILE_FACTOR_OP_DIV '/' 241 #define FILE_FACTOR_OP_NONE '\0' 242 243 #define FILE_OPS "&|^+-*/%" 244 #define FILE_OPAND 0 245 #define FILE_OPOR 1 246 #define FILE_OPXOR 2 247 #define FILE_OPADD 3 248 #define FILE_OPMINUS 4 249 #define FILE_OPMULTIPLY 5 250 #define FILE_OPDIVIDE 6 251 #define FILE_OPMODULO 7 252 #define FILE_OPS_MASK 0x07 /* mask for above ops */ 253 #define FILE_UNUSED_1 0x08 254 #define FILE_UNUSED_2 0x10 255 #define FILE_UNUSED_3 0x20 256 #define FILE_OPINVERSE 0x40 257 #define FILE_OPINDIRECT 0x80 258 259 #ifdef ENABLE_CONDITIONALS 260 #define COND_NONE 0 261 #define COND_IF 1 262 #define COND_ELIF 2 263 #define COND_ELSE 3 264 #endif /* ENABLE_CONDITIONALS */ 265 266 /* Word 4 */ 267 uint32_t offset; /* offset to magic number */ 268 /* Word 5 */ 269 int32_t in_offset; /* offset from indirection */ 270 /* Word 6 */ 271 uint32_t lineno; /* line number in magic file */ 272 /* Word 7,8 */ 273 union { 274 uint64_t _mask; /* for use with numeric and date types */ 275 struct { 276 uint32_t _count; /* repeat/line count */ 277 uint32_t _flags; /* modifier flags */ 278 } _s; /* for use with string types */ 279 } _u; 280 #define num_mask _u._mask 281 #define str_range _u._s._count 282 #define str_flags _u._s._flags 283 /* Words 9-16 */ 284 union VALUETYPE value; /* either number or string */ 285 /* Words 17-32 */ 286 char desc[MAXDESC]; /* description */ 287 /* Words 33-48 */ 288 char mimetype[MAXDESC]; /* MIME type */ 289 /* Words 49-50 */ 290 char apple[8]; 291 }; 292 293 #define BIT(A) (1 << (A)) 294 #define STRING_COMPACT_WHITESPACE BIT(0) 295 #define STRING_COMPACT_OPTIONAL_WHITESPACE BIT(1) 296 #define STRING_IGNORE_LOWERCASE BIT(2) 297 #define STRING_IGNORE_UPPERCASE BIT(3) 298 #define REGEX_OFFSET_START BIT(4) 299 #define STRING_TEXTTEST BIT(5) 300 #define STRING_BINTEST BIT(6) 301 #define PSTRING_1_BE BIT(7) 302 #define PSTRING_1_LE BIT(7) 303 #define PSTRING_2_BE BIT(8) 304 #define PSTRING_2_LE BIT(9) 305 #define PSTRING_4_BE BIT(10) 306 #define PSTRING_4_LE BIT(11) 307 #define PSTRING_LEN \ 308 (PSTRING_1_BE|PSTRING_2_LE|PSTRING_2_BE|PSTRING_4_LE|PSTRING_4_BE) 309 #define PSTRING_LENGTH_INCLUDES_ITSELF BIT(12) 310 #define CHAR_COMPACT_WHITESPACE 'W' 311 #define CHAR_COMPACT_OPTIONAL_WHITESPACE 'w' 312 #define CHAR_IGNORE_LOWERCASE 'c' 313 #define CHAR_IGNORE_UPPERCASE 'C' 314 #define CHAR_REGEX_OFFSET_START 's' 315 #define CHAR_TEXTTEST 't' 316 #define CHAR_BINTEST 'b' 317 #define CHAR_PSTRING_1_BE 'B' 318 #define CHAR_PSTRING_1_LE 'B' 319 #define CHAR_PSTRING_2_BE 'H' 320 #define CHAR_PSTRING_2_LE 'h' 321 #define CHAR_PSTRING_4_BE 'L' 322 #define CHAR_PSTRING_4_LE 'l' 323 #define CHAR_PSTRING_LENGTH_INCLUDES_ITSELF 'J' 324 #define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE) 325 #define STRING_DEFAULT_RANGE 100 326 327 328 /* list of magic entries */ 329 struct mlist { 330 struct magic *magic; /* array of magic entries */ 331 uint32_t nmagic; /* number of entries in array */ 332 int mapped; /* allocation type: 0 => apprentice_file 333 * 1 => apprentice_map + malloc 334 * 2 => apprentice_map + mmap */ 335 struct mlist *next, *prev; 336 }; 337 338 #ifdef __cplusplus 339 #define CAST(T, b) static_cast<T>(b) 340 #define RCAST(T, b) reinterpret_cast<T>(b) 341 #else 342 #define CAST(T, b) (T)(b) 343 #define RCAST(T, b) (T)(b) 344 #endif 345 346 struct level_info { 347 int32_t off; 348 int got_match; 349 #ifdef ENABLE_CONDITIONALS 350 int last_match; 351 int last_cond; /* used for error checking by parse() */ 352 #endif 353 }; 354 struct magic_set { 355 struct mlist *mlist; 356 struct cont { 357 size_t len; 358 struct level_info *li; 359 } c; 360 struct out { 361 char *buf; /* Accumulation buffer */ 362 char *pbuf; /* Printable buffer */ 363 } o; 364 uint32_t offset; 365 int error; 366 int flags; /* Control magic tests. */ 367 int event_flags; /* Note things that happened. */ 368 #define EVENT_HAD_ERR 0x01 369 const char *file; 370 size_t line; /* current magic line number */ 371 372 /* data for searches */ 373 struct { 374 const char *s; /* start of search in original source */ 375 size_t s_len; /* length of search region */ 376 size_t offset; /* starting offset in source: XXX - should this be off_t? */ 377 size_t rm_len; /* match length */ 378 } search; 379 380 /* FIXME: Make the string dynamically allocated so that e.g. 381 strings matched in files can be longer than MAXstring */ 382 union VALUETYPE ms_value; /* either number or string */ 383 }; 384 385 /* Type for Unicode characters */ 386 typedef unsigned long unichar; 387 388 struct stat; 389 protected const char *file_fmttime(uint32_t, int); 390 protected int file_buffer(struct magic_set *, int, const char *, const void *, 391 size_t); 392 protected int file_fsmagic(struct magic_set *, const char *, struct stat *); 393 protected int file_pipe2file(struct magic_set *, int, const void *, size_t); 394 protected int file_vprintf(struct magic_set *, const char *, va_list); 395 protected size_t file_printedlen(const struct magic_set *); 396 protected int file_replace(struct magic_set *, const char *, const char *); 397 protected int file_printf(struct magic_set *, const char *, ...) 398 __attribute__((__format__(__printf__, 2, 3))); 399 protected int file_reset(struct magic_set *); 400 protected int file_tryelf(struct magic_set *, int, const unsigned char *, 401 size_t); 402 protected int file_trycdf(struct magic_set *, int, const unsigned char *, 403 size_t); 404 #if HAVE_FORK 405 protected int file_zmagic(struct magic_set *, int, const char *, 406 const unsigned char *, size_t); 407 #endif 408 protected int file_ascmagic(struct magic_set *, const unsigned char *, size_t, 409 int); 410 protected int file_ascmagic_with_encoding(struct magic_set *, 411 const unsigned char *, size_t, unichar *, size_t, const char *, 412 const char *, int); 413 protected int file_encoding(struct magic_set *, const unsigned char *, size_t, 414 unichar **, size_t *, const char **, const char **, const char **); 415 protected int file_is_tar(struct magic_set *, const unsigned char *, size_t); 416 protected int file_softmagic(struct magic_set *, const unsigned char *, size_t, 417 int, int); 418 protected struct mlist *file_apprentice(struct magic_set *, const char *, int); 419 protected uint64_t file_signextend(struct magic_set *, struct magic *, 420 uint64_t); 421 protected void file_delmagic(struct magic *, int type, size_t entries); 422 protected void file_badread(struct magic_set *); 423 protected void file_badseek(struct magic_set *); 424 protected void file_oomem(struct magic_set *, size_t); 425 protected void file_error(struct magic_set *, int, const char *, ...) 426 __attribute__((__format__(__printf__, 3, 4))); 427 protected void file_magerror(struct magic_set *, const char *, ...) 428 __attribute__((__format__(__printf__, 2, 3))); 429 protected void file_magwarn(struct magic_set *, const char *, ...) 430 __attribute__((__format__(__printf__, 2, 3))); 431 protected void file_mdump(struct magic *); 432 protected void file_showstr(FILE *, const char *, size_t); 433 protected size_t file_mbswidth(const char *); 434 protected const char *file_getbuffer(struct magic_set *); 435 protected ssize_t sread(int, void *, size_t, int); 436 protected int file_check_mem(struct magic_set *, unsigned int); 437 protected int file_looks_utf8(const unsigned char *, size_t, unichar *, 438 size_t *); 439 protected size_t file_pstring_length_size(const struct magic *); 440 protected size_t file_pstring_get_length(const struct magic *, const char *); 441 #ifdef __EMX__ 442 protected int file_os2_apptype(struct magic_set *, const char *, const void *, 443 size_t); 444 #endif /* __EMX__ */ 445 446 447 #ifndef COMPILE_ONLY 448 extern const char *file_names[]; 449 extern const size_t file_nnames; 450 #endif 451 452 #ifndef HAVE_STRERROR 453 extern int sys_nerr; 454 extern char *sys_errlist[]; 455 #define strerror(e) \ 456 (((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error") 457 #endif 458 459 #ifndef HAVE_STRTOUL 460 #define strtoul(a, b, c) strtol(a, b, c) 461 #endif 462 463 #ifndef HAVE_VASPRINTF 464 int vasprintf(char **, const char *, va_list); 465 #endif 466 #ifndef HAVE_ASPRINTF 467 int asprintf(char **ptr, const char *format_string, ...); 468 #endif 469 470 #ifndef HAVE_STRLCPY 471 size_t strlcpy(char *dst, const char *src, size_t siz); 472 #endif 473 #ifndef HAVE_STRLCAT 474 size_t strlcat(char *dst, const char *src, size_t siz); 475 #endif 476 #ifndef HAVE_GETLINE 477 ssize_t getline(char **dst, size_t *len, FILE *fp); 478 ssize_t getdelim(char **dst, size_t *len, int delimiter, FILE *fp); 479 #endif 480 481 #if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK) 482 #define QUICK 483 #endif 484 485 #ifndef O_BINARY 486 #define O_BINARY 0 487 #endif 488 489 #ifndef __cplusplus 490 #if defined(__GNUC__) && (__GNUC__ >= 3) 491 #define FILE_RCSID(id) \ 492 static const char rcsid[] __attribute__((__used__)) = id; 493 #else 494 #define FILE_RCSID(id) \ 495 static const char *rcsid(const char *p) { \ 496 return rcsid(p = id); \ 497 } 498 #endif 499 #else 500 #define FILE_RCSID(id) 501 #endif 502 503 #endif /* __file_h__ */ 504