1 /* $NetBSD: file.h,v 1.5 2011/09/16 21:06:26 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 1.133 2011/05/13 22:15:40 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 <sys/types.h> 68 #include <sys/param.h> 69 /* Do this here and now, because struct stat gets re-defined on solaris */ 70 #include <sys/stat.h> 71 #include <stdarg.h> 72 73 #define ENABLE_CONDITIONALS 74 75 #ifndef MAGIC 76 #define MAGIC "/etc/magic" 77 #endif 78 79 #if defined(__EMX__) || defined (WIN32) 80 #define PATHSEP ';' 81 #else 82 #define PATHSEP ':' 83 #endif 84 85 #define private static 86 #ifndef protected 87 #define protected 88 #endif 89 #define public 90 91 #ifndef __GNUC_PREREQ__ 92 #ifdef __GNUC__ 93 #define __GNUC_PREREQ__(x, y) \ 94 ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \ 95 (__GNUC__ > (x))) 96 #else 97 #define __GNUC_PREREQ__(x, y) 0 98 #endif 99 #endif 100 101 #ifndef __GNUC__ 102 #ifndef __attribute__ 103 #define __attribute__(a) 104 #endif 105 #endif 106 107 #ifndef MIN 108 #define MIN(a,b) (((a) < (b)) ? (a) : (b)) 109 #endif 110 111 #ifndef MAX 112 #define MAX(a,b) (((a) > (b)) ? (a) : (b)) 113 #endif 114 115 #ifndef HOWMANY 116 # define HOWMANY (256 * 1024) /* how much of the file to look at */ 117 #endif 118 #define MAXMAGIS 8192 /* max entries in any one magic file 119 or directory */ 120 #define MAXDESC 64 /* max leng of text description/MIME type */ 121 #define MAXstring 64 /* max leng of "string" types */ 122 123 #define MAGICNO 0xF11E041C 124 #define VERSIONNO 8 125 #define FILE_MAGICSIZE 232 126 127 #define FILE_LOAD 0 128 #define FILE_CHECK 1 129 #define FILE_COMPILE 2 130 #define FILE_LIST 3 131 132 union VALUETYPE { 133 uint8_t b; 134 uint16_t h; 135 uint32_t l; 136 uint64_t q; 137 uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */ 138 uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */ 139 uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */ 140 char s[MAXstring]; /* the search string or regex pattern */ 141 unsigned char us[MAXstring]; 142 float f; 143 double d; 144 }; 145 146 struct magic { 147 /* Word 1 */ 148 uint16_t cont_level; /* level of ">" */ 149 uint8_t flag; 150 #define INDIR 0x01 /* if '(...)' appears */ 151 #define OFFADD 0x02 /* if '>&' or '>...(&' appears */ 152 #define INDIROFFADD 0x04 /* if '>&(' appears */ 153 #define UNSIGNED 0x08 /* comparison is unsigned */ 154 #define NOSPACE 0x10 /* suppress space character before output */ 155 #define BINTEST 0x20 /* test is for a binary type (set only 156 for top-level tests) */ 157 #define TEXTTEST 0x40 /* for passing to file_softmagic */ 158 159 uint8_t factor; 160 161 /* Word 2 */ 162 uint8_t reln; /* relation (0=eq, '>'=gt, etc) */ 163 uint8_t vallen; /* length of string value, if any */ 164 uint8_t type; /* comparison type (FILE_*) */ 165 uint8_t in_type; /* type of indirection */ 166 #define FILE_INVALID 0 167 #define FILE_BYTE 1 168 #define FILE_SHORT 2 169 #define FILE_DEFAULT 3 170 #define FILE_LONG 4 171 #define FILE_STRING 5 172 #define FILE_DATE 6 173 #define FILE_BESHORT 7 174 #define FILE_BELONG 8 175 #define FILE_BEDATE 9 176 #define FILE_LESHORT 10 177 #define FILE_LELONG 11 178 #define FILE_LEDATE 12 179 #define FILE_PSTRING 13 180 #define FILE_LDATE 14 181 #define FILE_BELDATE 15 182 #define FILE_LELDATE 16 183 #define FILE_REGEX 17 184 #define FILE_BESTRING16 18 185 #define FILE_LESTRING16 19 186 #define FILE_SEARCH 20 187 #define FILE_MEDATE 21 188 #define FILE_MELDATE 22 189 #define FILE_MELONG 23 190 #define FILE_QUAD 24 191 #define FILE_LEQUAD 25 192 #define FILE_BEQUAD 26 193 #define FILE_QDATE 27 194 #define FILE_LEQDATE 28 195 #define FILE_BEQDATE 29 196 #define FILE_QLDATE 30 197 #define FILE_LEQLDATE 31 198 #define FILE_BEQLDATE 32 199 #define FILE_FLOAT 33 200 #define FILE_BEFLOAT 34 201 #define FILE_LEFLOAT 35 202 #define FILE_DOUBLE 36 203 #define FILE_BEDOUBLE 37 204 #define FILE_LEDOUBLE 38 205 #define FILE_BEID3 39 206 #define FILE_LEID3 40 207 #define FILE_INDIRECT 41 208 #define FILE_NAMES_SIZE 42/* size of array to contain all names */ 209 210 #define IS_STRING(t) \ 211 ((t) == FILE_STRING || \ 212 (t) == FILE_PSTRING || \ 213 (t) == FILE_BESTRING16 || \ 214 (t) == FILE_LESTRING16 || \ 215 (t) == FILE_REGEX || \ 216 (t) == FILE_SEARCH || \ 217 (t) == FILE_DEFAULT) 218 219 #define FILE_FMT_NONE 0 220 #define FILE_FMT_NUM 1 /* "cduxXi" */ 221 #define FILE_FMT_STR 2 /* "s" */ 222 #define FILE_FMT_QUAD 3 /* "ll" */ 223 #define FILE_FMT_FLOAT 4 /* "eEfFgG" */ 224 #define FILE_FMT_DOUBLE 5 /* "eEfFgG" */ 225 226 /* Word 3 */ 227 uint8_t in_op; /* operator for indirection */ 228 uint8_t mask_op; /* operator for mask */ 229 #ifdef ENABLE_CONDITIONALS 230 uint8_t cond; /* conditional type */ 231 #else 232 uint8_t dummy; 233 #endif 234 uint8_t factor_op; 235 #define FILE_FACTOR_OP_PLUS '+' 236 #define FILE_FACTOR_OP_MINUS '-' 237 #define FILE_FACTOR_OP_TIMES '*' 238 #define FILE_FACTOR_OP_DIV '/' 239 #define FILE_FACTOR_OP_NONE '\0' 240 241 #define FILE_OPS "&|^+-*/%" 242 #define FILE_OPAND 0 243 #define FILE_OPOR 1 244 #define FILE_OPXOR 2 245 #define FILE_OPADD 3 246 #define FILE_OPMINUS 4 247 #define FILE_OPMULTIPLY 5 248 #define FILE_OPDIVIDE 6 249 #define FILE_OPMODULO 7 250 #define FILE_OPS_MASK 0x07 /* mask for above ops */ 251 #define FILE_UNUSED_1 0x08 252 #define FILE_UNUSED_2 0x10 253 #define FILE_UNUSED_3 0x20 254 #define FILE_OPINVERSE 0x40 255 #define FILE_OPINDIRECT 0x80 256 257 #ifdef ENABLE_CONDITIONALS 258 #define COND_NONE 0 259 #define COND_IF 1 260 #define COND_ELIF 2 261 #define COND_ELSE 3 262 #endif /* ENABLE_CONDITIONALS */ 263 264 /* Word 4 */ 265 uint32_t offset; /* offset to magic number */ 266 /* Word 5 */ 267 int32_t in_offset; /* offset from indirection */ 268 /* Word 6 */ 269 uint32_t lineno; /* line number in magic file */ 270 /* Word 7,8 */ 271 union { 272 uint64_t _mask; /* for use with numeric and date types */ 273 struct { 274 uint32_t _count; /* repeat/line count */ 275 uint32_t _flags; /* modifier flags */ 276 } _s; /* for use with string types */ 277 } _u; 278 #define num_mask _u._mask 279 #define str_range _u._s._count 280 #define str_flags _u._s._flags 281 /* Words 9-16 */ 282 union VALUETYPE value; /* either number or string */ 283 /* Words 17-32 */ 284 char desc[MAXDESC]; /* description */ 285 /* Words 33-48 */ 286 char mimetype[MAXDESC]; /* MIME type */ 287 /* Words 49-50 */ 288 char apple[8]; 289 }; 290 291 #define BIT(A) (1 << (A)) 292 #define STRING_COMPACT_WHITESPACE BIT(0) 293 #define STRING_COMPACT_OPTIONAL_WHITESPACE BIT(1) 294 #define STRING_IGNORE_LOWERCASE BIT(2) 295 #define STRING_IGNORE_UPPERCASE BIT(3) 296 #define REGEX_OFFSET_START BIT(4) 297 #define STRING_TEXTTEST BIT(5) 298 #define STRING_BINTEST BIT(6) 299 #define PSTRING_1_BE BIT(7) 300 #define PSTRING_1_LE BIT(7) 301 #define PSTRING_2_BE BIT(8) 302 #define PSTRING_2_LE BIT(9) 303 #define PSTRING_4_BE BIT(10) 304 #define PSTRING_4_LE BIT(11) 305 #define PSTRING_LEN \ 306 (PSTRING_1_BE|PSTRING_2_LE|PSTRING_2_BE|PSTRING_4_LE|PSTRING_4_BE) 307 #define PSTRING_LENGTH_INCLUDES_ITSELF BIT(12) 308 #define CHAR_COMPACT_WHITESPACE 'W' 309 #define CHAR_COMPACT_OPTIONAL_WHITESPACE 'w' 310 #define CHAR_IGNORE_LOWERCASE 'c' 311 #define CHAR_IGNORE_UPPERCASE 'C' 312 #define CHAR_REGEX_OFFSET_START 's' 313 #define CHAR_TEXTTEST 't' 314 #define CHAR_BINTEST 'b' 315 #define CHAR_PSTRING_1_BE 'B' 316 #define CHAR_PSTRING_1_LE 'B' 317 #define CHAR_PSTRING_2_BE 'H' 318 #define CHAR_PSTRING_2_LE 'h' 319 #define CHAR_PSTRING_4_BE 'L' 320 #define CHAR_PSTRING_4_LE 'l' 321 #define CHAR_PSTRING_LENGTH_INCLUDES_ITSELF 'J' 322 #define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE) 323 #define STRING_DEFAULT_RANGE 100 324 325 326 /* list of magic entries */ 327 struct mlist { 328 struct magic *magic; /* array of magic entries */ 329 uint32_t nmagic; /* number of entries in array */ 330 int mapped; /* allocation type: 0 => apprentice_file 331 * 1 => apprentice_map + malloc 332 * 2 => apprentice_map + mmap */ 333 struct mlist *next, *prev; 334 }; 335 336 #ifdef __cplusplus 337 #define CAST(T, b) static_cast<T>(b) 338 #define RCAST(T, b) reinterpret_cast<T>(b) 339 #else 340 #define CAST(T, b) (T)(b) 341 #define RCAST(T, b) (T)(b) 342 #endif 343 344 struct level_info { 345 int32_t off; 346 int got_match; 347 #ifdef ENABLE_CONDITIONALS 348 int last_match; 349 int last_cond; /* used for error checking by parse() */ 350 #endif 351 }; 352 struct magic_set { 353 struct mlist *mlist; 354 struct cont { 355 size_t len; 356 struct level_info *li; 357 } c; 358 struct out { 359 char *buf; /* Accumulation buffer */ 360 char *pbuf; /* Printable buffer */ 361 } o; 362 uint32_t offset; 363 int error; 364 int flags; /* Control magic tests. */ 365 int event_flags; /* Note things that happened. */ 366 #define EVENT_HAD_ERR 0x01 367 const char *file; 368 size_t line; /* current magic line number */ 369 370 /* data for searches */ 371 struct { 372 const char *s; /* start of search in original source */ 373 size_t s_len; /* length of search region */ 374 size_t offset; /* starting offset in source: XXX - should this be off_t? */ 375 size_t rm_len; /* match length */ 376 } search; 377 378 /* FIXME: Make the string dynamically allocated so that e.g. 379 strings matched in files can be longer than MAXstring */ 380 union VALUETYPE ms_value; /* either number or string */ 381 }; 382 383 /* Type for Unicode characters */ 384 typedef unsigned long unichar; 385 386 struct stat; 387 protected const char *file_fmttime(uint32_t, int); 388 protected int file_buffer(struct magic_set *, int, const char *, const void *, 389 size_t); 390 protected int file_fsmagic(struct magic_set *, const char *, struct stat *); 391 protected int file_pipe2file(struct magic_set *, int, const void *, size_t); 392 protected int file_vprintf(struct magic_set *, const char *, va_list); 393 protected size_t file_printedlen(const struct magic_set *); 394 protected int file_replace(struct magic_set *, const char *, const char *); 395 protected int file_printf(struct magic_set *, const char *, ...) 396 __attribute__((__format__(__printf__, 2, 3))); 397 protected int file_reset(struct magic_set *); 398 protected int file_tryelf(struct magic_set *, int, const unsigned char *, 399 size_t); 400 protected int file_trycdf(struct magic_set *, int, const unsigned char *, 401 size_t); 402 #if HAVE_FORK 403 protected int file_zmagic(struct magic_set *, int, const char *, 404 const unsigned char *, size_t); 405 #endif 406 protected int file_ascmagic(struct magic_set *, const unsigned char *, size_t); 407 protected int file_ascmagic_with_encoding(struct magic_set *, 408 const unsigned char *, size_t, unichar *, size_t, const char *, 409 const char *); 410 protected int file_encoding(struct magic_set *, const unsigned char *, size_t, 411 unichar **, size_t *, const char **, const char **, const char **); 412 protected int file_is_tar(struct magic_set *, const unsigned char *, size_t); 413 protected int file_softmagic(struct magic_set *, const unsigned char *, size_t, 414 int); 415 protected struct mlist *file_apprentice(struct magic_set *, const char *, int); 416 protected uint64_t file_signextend(struct magic_set *, struct magic *, 417 uint64_t); 418 protected void file_delmagic(struct magic *, int type, size_t entries); 419 protected void file_badread(struct magic_set *); 420 protected void file_badseek(struct magic_set *); 421 protected void file_oomem(struct magic_set *, size_t); 422 protected void file_error(struct magic_set *, int, const char *, ...) 423 __attribute__((__format__(__printf__, 3, 4))); 424 protected void file_magerror(struct magic_set *, const char *, ...) 425 __attribute__((__format__(__printf__, 2, 3))); 426 protected void file_magwarn(struct magic_set *, const char *, ...) 427 __attribute__((__format__(__printf__, 2, 3))); 428 protected void file_mdump(struct magic *); 429 protected void file_showstr(FILE *, const char *, size_t); 430 protected size_t file_mbswidth(const char *); 431 protected const char *file_getbuffer(struct magic_set *); 432 protected ssize_t sread(int, void *, size_t, int); 433 protected int file_check_mem(struct magic_set *, unsigned int); 434 protected int file_looks_utf8(const unsigned char *, size_t, unichar *, 435 size_t *); 436 protected size_t file_pstring_length_size(const struct magic *); 437 protected size_t file_pstring_get_length(const struct magic *, const char *); 438 #ifdef __EMX__ 439 protected int file_os2_apptype(struct magic_set *, const char *, const void *, 440 size_t); 441 #endif /* __EMX__ */ 442 443 444 #ifndef COMPILE_ONLY 445 extern const char *file_names[]; 446 extern const size_t file_nnames; 447 #endif 448 449 #ifndef HAVE_STRERROR 450 extern int sys_nerr; 451 extern char *sys_errlist[]; 452 #define strerror(e) \ 453 (((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error") 454 #endif 455 456 #ifndef HAVE_STRTOUL 457 #define strtoul(a, b, c) strtol(a, b, c) 458 #endif 459 460 #ifndef HAVE_VASPRINTF 461 int vasprintf(char **, const char *, va_list); 462 #endif 463 #ifndef HAVE_ASPRINTF 464 int asprintf(char **ptr, const char *format_string, ...); 465 #endif 466 467 #ifndef HAVE_STRLCPY 468 size_t strlcpy(char *dst, const char *src, size_t siz); 469 #endif 470 #ifndef HAVE_STRLCAT 471 size_t strlcat(char *dst, const char *src, size_t siz); 472 #endif 473 #ifndef HAVE_GETLINE 474 ssize_t getline(char **dst, size_t *len, FILE *fp); 475 ssize_t getdelim(char **dst, size_t *len, int delimiter, FILE *fp); 476 #endif 477 478 #if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK) 479 #define QUICK 480 #endif 481 482 #ifndef O_BINARY 483 #define O_BINARY 0 484 #endif 485 486 #ifndef __cplusplus 487 #if defined(__GNUC__) && (__GNUC__ >= 3) 488 #define FILE_RCSID(id) \ 489 static const char rcsid[] __attribute__((__used__)) = id; 490 #else 491 #define FILE_RCSID(id) \ 492 static const char *rcsid(const char *p) { \ 493 return rcsid(p = id); \ 494 } 495 #endif 496 #else 497 #define FILE_RCSID(id) 498 #endif 499 #ifndef __RCSID 500 #define __RCSID(a) 501 #endif 502 503 #endif /* __file_h__ */ 504