1 /* $NetBSD: file.h,v 1.7 2013/01/03 23:05:38 christos Exp $ */ 2 3 /* 4 * Copyright (c) Ian F. Darwin 1986-1995. 5 * Software written by Ian F. Darwin and others; 6 * maintained 1995-present by Christos Zoulas and others. 7 * 8 * Redistribution and use in source and binary forms, with or without 9 * modification, are permitted provided that the following conditions 10 * are met: 11 * 1. Redistributions of source code must retain the above copyright 12 * notice immediately at the beginning of the file, without modification, 13 * this list of conditions, and the following disclaimer. 14 * 2. Redistributions in binary form must reproduce the above copyright 15 * notice, this list of conditions and the following disclaimer in the 16 * documentation and/or other materials provided with the distribution. 17 * 18 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND 19 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 20 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 21 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR 22 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 23 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 24 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 25 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 26 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 27 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 28 * SUCH DAMAGE. 29 */ 30 /* 31 * file.h - definitions for file(1) program 32 * @(#)$File: file.h,v 1.140 2012/10/30 23:11:51 christos Exp $ 33 */ 34 35 #ifndef __file_h__ 36 #define __file_h__ 37 38 #ifdef HAVE_CONFIG_H 39 #include <config.h> 40 #endif 41 42 #ifdef WIN32 43 #ifdef _WIN64 44 #define SIZE_T_FORMAT "I64" 45 #else 46 #define SIZE_T_FORMAT "" 47 #endif 48 #define INT64_T_FORMAT "I64" 49 #else 50 #define SIZE_T_FORMAT "z" 51 #define INT64_T_FORMAT "ll" 52 #endif 53 54 #include <stdio.h> /* Include that here, to make sure __P gets defined */ 55 #include <errno.h> 56 #include <fcntl.h> /* For open and flags */ 57 #ifdef HAVE_STDINT_H 58 #ifndef __STDC_LIMIT_MACROS 59 #define __STDC_LIMIT_MACROS 60 #endif 61 #include <stdint.h> 62 #endif 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 #include <sys/param.h> 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 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 leng of text description/MIME type */ 134 #define MAXstring 64 /* max leng of "string" types */ 135 136 #define MAGICNO 0xF11E041C 137 #define VERSIONNO 9 138 #define FILE_MAGICSIZE 232 139 140 #define FILE_LOAD 0 141 #define FILE_CHECK 1 142 #define FILE_COMPILE 2 143 #define FILE_LIST 3 144 145 union VALUETYPE { 146 uint8_t b; 147 uint16_t h; 148 uint32_t l; 149 uint64_t q; 150 uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */ 151 uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */ 152 uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */ 153 char s[MAXstring]; /* the search string or regex pattern */ 154 unsigned char us[MAXstring]; 155 float f; 156 double d; 157 }; 158 159 struct magic { 160 /* Word 1 */ 161 uint16_t cont_level; /* level of ">" */ 162 uint8_t flag; 163 #define INDIR 0x01 /* if '(...)' appears */ 164 #define OFFADD 0x02 /* if '>&' or '>...(&' appears */ 165 #define INDIROFFADD 0x04 /* if '>&(' appears */ 166 #define UNSIGNED 0x08 /* comparison is unsigned */ 167 #define NOSPACE 0x10 /* suppress space character before output */ 168 #define BINTEST 0x20 /* test is for a binary type (set only 169 for top-level tests) */ 170 #define TEXTTEST 0x40 /* for passing to file_softmagic */ 171 172 uint8_t factor; 173 174 /* Word 2 */ 175 uint8_t reln; /* relation (0=eq, '>'=gt, etc) */ 176 uint8_t vallen; /* length of string value, if any */ 177 uint8_t type; /* comparison type (FILE_*) */ 178 uint8_t in_type; /* type of indirection */ 179 #define FILE_INVALID 0 180 #define FILE_BYTE 1 181 #define FILE_SHORT 2 182 #define FILE_DEFAULT 3 183 #define FILE_LONG 4 184 #define FILE_STRING 5 185 #define FILE_DATE 6 186 #define FILE_BESHORT 7 187 #define FILE_BELONG 8 188 #define FILE_BEDATE 9 189 #define FILE_LESHORT 10 190 #define FILE_LELONG 11 191 #define FILE_LEDATE 12 192 #define FILE_PSTRING 13 193 #define FILE_LDATE 14 194 #define FILE_BELDATE 15 195 #define FILE_LELDATE 16 196 #define FILE_REGEX 17 197 #define FILE_BESTRING16 18 198 #define FILE_LESTRING16 19 199 #define FILE_SEARCH 20 200 #define FILE_MEDATE 21 201 #define FILE_MELDATE 22 202 #define FILE_MELONG 23 203 #define FILE_QUAD 24 204 #define FILE_LEQUAD 25 205 #define FILE_BEQUAD 26 206 #define FILE_QDATE 27 207 #define FILE_LEQDATE 28 208 #define FILE_BEQDATE 29 209 #define FILE_QLDATE 30 210 #define FILE_LEQLDATE 31 211 #define FILE_BEQLDATE 32 212 #define FILE_FLOAT 33 213 #define FILE_BEFLOAT 34 214 #define FILE_LEFLOAT 35 215 #define FILE_DOUBLE 36 216 #define FILE_BEDOUBLE 37 217 #define FILE_LEDOUBLE 38 218 #define FILE_BEID3 39 219 #define FILE_LEID3 40 220 #define FILE_INDIRECT 41 221 #define FILE_QWDATE 42 222 #define FILE_LEQWDATE 43 223 #define FILE_BEQWDATE 44 224 #define FILE_NAME 45 225 #define FILE_USE 46 226 #define FILE_NAMES_SIZE 47 /* size of array to contain all names */ 227 228 #define IS_STRING(t) \ 229 ((t) == FILE_STRING || \ 230 (t) == FILE_PSTRING || \ 231 (t) == FILE_BESTRING16 || \ 232 (t) == FILE_LESTRING16 || \ 233 (t) == FILE_REGEX || \ 234 (t) == FILE_SEARCH || \ 235 (t) == FILE_NAME || \ 236 (t) == FILE_USE || \ 237 (t) == FILE_DEFAULT) 238 239 #define FILE_FMT_NONE 0 240 #define FILE_FMT_NUM 1 /* "cduxXi" */ 241 #define FILE_FMT_STR 2 /* "s" */ 242 #define FILE_FMT_QUAD 3 /* "ll" */ 243 #define FILE_FMT_FLOAT 4 /* "eEfFgG" */ 244 #define FILE_FMT_DOUBLE 5 /* "eEfFgG" */ 245 246 /* Word 3 */ 247 uint8_t in_op; /* operator for indirection */ 248 uint8_t mask_op; /* operator for mask */ 249 #ifdef ENABLE_CONDITIONALS 250 uint8_t cond; /* conditional type */ 251 #else 252 uint8_t dummy; 253 #endif 254 uint8_t factor_op; 255 #define FILE_FACTOR_OP_PLUS '+' 256 #define FILE_FACTOR_OP_MINUS '-' 257 #define FILE_FACTOR_OP_TIMES '*' 258 #define FILE_FACTOR_OP_DIV '/' 259 #define FILE_FACTOR_OP_NONE '\0' 260 261 #define FILE_OPS "&|^+-*/%" 262 #define FILE_OPAND 0 263 #define FILE_OPOR 1 264 #define FILE_OPXOR 2 265 #define FILE_OPADD 3 266 #define FILE_OPMINUS 4 267 #define FILE_OPMULTIPLY 5 268 #define FILE_OPDIVIDE 6 269 #define FILE_OPMODULO 7 270 #define FILE_OPS_MASK 0x07 /* mask for above ops */ 271 #define FILE_UNUSED_1 0x08 272 #define FILE_UNUSED_2 0x10 273 #define FILE_UNUSED_3 0x20 274 #define FILE_OPINVERSE 0x40 275 #define FILE_OPINDIRECT 0x80 276 277 #ifdef ENABLE_CONDITIONALS 278 #define COND_NONE 0 279 #define COND_IF 1 280 #define COND_ELIF 2 281 #define COND_ELSE 3 282 #endif /* ENABLE_CONDITIONALS */ 283 284 /* Word 4 */ 285 uint32_t offset; /* offset to magic number */ 286 /* Word 5 */ 287 int32_t in_offset; /* offset from indirection */ 288 /* Word 6 */ 289 uint32_t lineno; /* line number in magic file */ 290 /* Word 7,8 */ 291 union { 292 uint64_t _mask; /* for use with numeric and date types */ 293 struct { 294 uint32_t _count; /* repeat/line count */ 295 uint32_t _flags; /* modifier flags */ 296 } _s; /* for use with string types */ 297 } _u; 298 #define num_mask _u._mask 299 #define str_range _u._s._count 300 #define str_flags _u._s._flags 301 /* Words 9-16 */ 302 union VALUETYPE value; /* either number or string */ 303 /* Words 17-32 */ 304 char desc[MAXDESC]; /* description */ 305 /* Words 33-48 */ 306 char mimetype[MAXDESC]; /* MIME type */ 307 /* Words 49-50 */ 308 char apple[8]; 309 }; 310 311 #define BIT(A) (1 << (A)) 312 #define STRING_COMPACT_WHITESPACE BIT(0) 313 #define STRING_COMPACT_OPTIONAL_WHITESPACE BIT(1) 314 #define STRING_IGNORE_LOWERCASE BIT(2) 315 #define STRING_IGNORE_UPPERCASE BIT(3) 316 #define REGEX_OFFSET_START BIT(4) 317 #define STRING_TEXTTEST BIT(5) 318 #define STRING_BINTEST BIT(6) 319 #define PSTRING_1_BE BIT(7) 320 #define PSTRING_1_LE BIT(7) 321 #define PSTRING_2_BE BIT(8) 322 #define PSTRING_2_LE BIT(9) 323 #define PSTRING_4_BE BIT(10) 324 #define PSTRING_4_LE BIT(11) 325 #define PSTRING_LEN \ 326 (PSTRING_1_BE|PSTRING_2_LE|PSTRING_2_BE|PSTRING_4_LE|PSTRING_4_BE) 327 #define PSTRING_LENGTH_INCLUDES_ITSELF BIT(12) 328 #define STRING_TRIM BIT(13) 329 #define CHAR_COMPACT_WHITESPACE 'W' 330 #define CHAR_COMPACT_OPTIONAL_WHITESPACE 'w' 331 #define CHAR_IGNORE_LOWERCASE 'c' 332 #define CHAR_IGNORE_UPPERCASE 'C' 333 #define CHAR_REGEX_OFFSET_START 's' 334 #define CHAR_TEXTTEST 't' 335 #define CHAR_TRIM 'T' 336 #define CHAR_BINTEST 'b' 337 #define CHAR_PSTRING_1_BE 'B' 338 #define CHAR_PSTRING_1_LE 'B' 339 #define CHAR_PSTRING_2_BE 'H' 340 #define CHAR_PSTRING_2_LE 'h' 341 #define CHAR_PSTRING_4_BE 'L' 342 #define CHAR_PSTRING_4_LE 'l' 343 #define CHAR_PSTRING_LENGTH_INCLUDES_ITSELF 'J' 344 #define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE) 345 #define STRING_DEFAULT_RANGE 100 346 347 348 /* list of magic entries */ 349 struct mlist { 350 struct magic *magic; /* array of magic entries */ 351 uint32_t nmagic; /* number of entries in array */ 352 int mapped; /* allocation type: 0 => apprentice_file 353 * 1 => apprentice_map + malloc 354 * 2 => apprentice_map + mmap */ 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 365 366 struct 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 2 376 377 struct 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 }; 407 408 /* Type for Unicode characters */ 409 typedef unsigned long unichar; 410 411 struct stat; 412 #define FILE_T_LOCAL 1 413 #define FILE_T_WINDOWS 2 414 protected const char *file_fmttime(uint64_t, int, char *); 415 protected struct magic_set *file_ms_alloc(int); 416 protected void file_ms_free(struct magic_set *); 417 protected int file_buffer(struct magic_set *, int, const char *, const void *, 418 size_t); 419 protected int file_fsmagic(struct magic_set *, const char *, struct stat *); 420 protected int file_pipe2file(struct magic_set *, int, const void *, size_t); 421 protected int file_vprintf(struct magic_set *, const char *, va_list); 422 protected size_t file_printedlen(const struct magic_set *); 423 protected int file_replace(struct magic_set *, const char *, const char *); 424 protected int file_printf(struct magic_set *, const char *, ...) 425 __attribute__((__format__(__printf__, 2, 3))); 426 protected int file_reset(struct magic_set *); 427 protected int file_tryelf(struct magic_set *, int, const unsigned char *, 428 size_t); 429 protected int file_trycdf(struct magic_set *, int, const unsigned char *, 430 size_t); 431 #if HAVE_FORK 432 protected int file_zmagic(struct magic_set *, int, const char *, 433 const unsigned char *, size_t); 434 #endif 435 protected int file_ascmagic(struct magic_set *, const unsigned char *, size_t, 436 int); 437 protected int file_ascmagic_with_encoding(struct magic_set *, 438 const unsigned char *, size_t, unichar *, size_t, const char *, 439 const char *, int); 440 protected int file_encoding(struct magic_set *, const unsigned char *, size_t, 441 unichar **, size_t *, const char **, const char **, const char **); 442 protected int file_is_tar(struct magic_set *, const unsigned char *, size_t); 443 protected int file_softmagic(struct magic_set *, const unsigned char *, size_t, 444 int, int); 445 protected int file_apprentice(struct magic_set *, const char *, int); 446 protected int file_magicfind(struct magic_set *, const char *, struct mlist *); 447 protected uint64_t file_signextend(struct magic_set *, struct magic *, 448 uint64_t); 449 protected void file_delmagic(struct magic *, int type, size_t entries); 450 protected void file_badread(struct magic_set *); 451 protected void file_badseek(struct magic_set *); 452 protected void file_oomem(struct magic_set *, size_t); 453 protected void file_error(struct magic_set *, int, const char *, ...) 454 __attribute__((__format__(__printf__, 3, 4))); 455 protected void file_magerror(struct magic_set *, const char *, ...) 456 __attribute__((__format__(__printf__, 2, 3))); 457 protected void file_magwarn(struct magic_set *, const char *, ...) 458 __attribute__((__format__(__printf__, 2, 3))); 459 protected void file_mdump(struct magic *); 460 protected void file_showstr(FILE *, const char *, size_t); 461 protected size_t file_mbswidth(const char *); 462 protected const char *file_getbuffer(struct magic_set *); 463 protected ssize_t sread(int, void *, size_t, int); 464 protected int file_check_mem(struct magic_set *, unsigned int); 465 protected int file_looks_utf8(const unsigned char *, size_t, unichar *, 466 size_t *); 467 protected size_t file_pstring_length_size(const struct magic *); 468 protected size_t file_pstring_get_length(const struct magic *, const char *); 469 #ifdef __EMX__ 470 protected int file_os2_apptype(struct magic_set *, const char *, const void *, 471 size_t); 472 #endif /* __EMX__ */ 473 474 475 #ifndef COMPILE_ONLY 476 extern const char *file_names[]; 477 extern const size_t file_nnames; 478 #endif 479 480 #ifndef HAVE_STRERROR 481 extern int sys_nerr; 482 extern char *sys_errlist[]; 483 #define strerror(e) \ 484 (((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error") 485 #endif 486 487 #ifndef HAVE_STRTOUL 488 #define strtoul(a, b, c) strtol(a, b, c) 489 #endif 490 491 #ifndef HAVE_VASPRINTF 492 int vasprintf(char **, const char *, va_list); 493 #endif 494 #ifndef HAVE_ASPRINTF 495 int asprintf(char **ptr, const char *format_string, ...); 496 #endif 497 498 #ifndef HAVE_STRLCPY 499 size_t strlcpy(char *dst, const char *src, size_t siz); 500 #endif 501 #ifndef HAVE_STRLCAT 502 size_t strlcat(char *dst, const char *src, size_t siz); 503 #endif 504 #ifndef HAVE_GETLINE 505 ssize_t getline(char **dst, size_t *len, FILE *fp); 506 ssize_t getdelim(char **dst, size_t *len, int delimiter, FILE *fp); 507 #endif 508 #ifndef HAVE_CTIME_R 509 char *ctime_r(const time_t *, char *); 510 #endif 511 #ifndef HAVE_ASCTIME_R 512 char *asctime_r(const struct tm *, char *); 513 #endif 514 515 #if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK) 516 #define QUICK 517 #endif 518 519 #ifndef O_BINARY 520 #define O_BINARY 0 521 #endif 522 523 #ifndef __cplusplus 524 #if defined(__GNUC__) && (__GNUC__ >= 3) 525 #define FILE_RCSID(id) \ 526 static const char rcsid[] __attribute__((__used__)) = id; 527 #else 528 #define FILE_RCSID(id) \ 529 static const char *rcsid(const char *p) { \ 530 return rcsid(p = id); \ 531 } 532 #endif 533 #else 534 #define FILE_RCSID(id) 535 #endif 536 #ifndef __RCSID 537 #define __RCSID(a) 538 #endif 539 540 #endif /* __file_h__ */ 541