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 1.124 2010/01/16 17:45:12 chl 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#include <stdio.h> /* Include that here, to make sure __P gets defined */ 41#include <errno.h> 42#include <fcntl.h> /* For open and flags */ 43#ifdef HAVE_STDINT_H 44#ifndef __STDC_LIMIT_MACROS 45#define __STDC_LIMIT_MACROS 46#endif 47#include <stdint.h> 48#endif 49#ifdef HAVE_INTTYPES_H 50#include <inttypes.h> 51#endif 52#include <regex.h> 53#include <sys/types.h> 54#include <sys/param.h> 55/* Do this here and now, because struct stat gets re-defined on solaris */ 56#include <sys/stat.h> 57#include <stdarg.h> 58 59#define ENABLE_CONDITIONALS 60 61#ifndef MAGIC 62#define MAGIC "/etc/magic" 63#endif 64 65#ifdef __EMX__ 66#define PATHSEP ';' 67#else 68#define PATHSEP ':' 69#endif 70 71#define private static 72#ifndef protected 73#define protected 74#endif 75#define public 76 77#ifndef __GNUC_PREREQ__ 78#ifdef __GNUC__ 79#define __GNUC_PREREQ__(x, y) \ 80 ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \ 81 (__GNUC__ > (x))) 82#else 83#define __GNUC_PREREQ__(x, y) 0 84#endif 85#endif 86 87#ifndef __GNUC__ 88#ifndef __attribute__ 89#define __attribute__(a) 90#endif 91#endif 92 93#ifndef MIN 94#define MIN(a,b) (((a) < (b)) ? (a) : (b)) 95#endif 96 97#ifndef MAX 98#define MAX(a,b) (((a) > (b)) ? (a) : (b)) 99#endif 100 101#ifndef HOWMANY 102# define HOWMANY (256 * 1024) /* how much of the file to look at */ 103#endif 104#define MAXMAGIS 8192 /* max entries in any one magic file 105 or directory */ 106#define MAXDESC 64 /* max leng of text description/MIME type */ 107#define MAXstring 32 /* max leng of "string" types */ 108 109#define MAGICNO 0xF11E041C 110#define VERSIONNO 7 111#define FILE_MAGICSIZE 200 112 113#define FILE_LOAD 0 114#define FILE_CHECK 1 115#define FILE_COMPILE 2 116 117union VALUETYPE { 118 uint8_t b; 119 uint16_t h; 120 uint32_t l; 121 uint64_t q; 122 uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */ 123 uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */ 124 uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */ 125 char s[MAXstring]; /* the search string or regex pattern */ 126 unsigned char us[MAXstring]; 127 float f; 128 double d; 129}; 130 131struct magic { 132 /* Word 1 */ 133 uint16_t cont_level; /* level of ">" */ 134 uint8_t flag; 135#define INDIR 0x01 /* if '(...)' appears */ 136#define OFFADD 0x02 /* if '>&' or '>...(&' appears */ 137#define INDIROFFADD 0x04 /* if '>&(' appears */ 138#define UNSIGNED 0x08 /* comparison is unsigned */ 139#define NOSPACE 0x10 /* suppress space character before output */ 140#define BINTEST 0x20 /* test is for a binary type (set only 141 for top-level tests) */ 142#define TEXTTEST 0x40 /* for passing to file_softmagic */ 143 144 uint8_t factor; 145 146 /* Word 2 */ 147 uint8_t reln; /* relation (0=eq, '>'=gt, etc) */ 148 uint8_t vallen; /* length of string value, if any */ 149 uint8_t type; /* comparison type (FILE_*) */ 150 uint8_t in_type; /* type of indirection */ 151#define FILE_INVALID 0 152#define FILE_BYTE 1 153#define FILE_SHORT 2 154#define FILE_DEFAULT 3 155#define FILE_LONG 4 156#define FILE_STRING 5 157#define FILE_DATE 6 158#define FILE_BESHORT 7 159#define FILE_BELONG 8 160#define FILE_BEDATE 9 161#define FILE_LESHORT 10 162#define FILE_LELONG 11 163#define FILE_LEDATE 12 164#define FILE_PSTRING 13 165#define FILE_LDATE 14 166#define FILE_BELDATE 15 167#define FILE_LELDATE 16 168#define FILE_REGEX 17 169#define FILE_BESTRING16 18 170#define FILE_LESTRING16 19 171#define FILE_SEARCH 20 172#define FILE_MEDATE 21 173#define FILE_MELDATE 22 174#define FILE_MELONG 23 175#define FILE_QUAD 24 176#define FILE_LEQUAD 25 177#define FILE_BEQUAD 26 178#define FILE_QDATE 27 179#define FILE_LEQDATE 28 180#define FILE_BEQDATE 29 181#define FILE_QLDATE 30 182#define FILE_LEQLDATE 31 183#define FILE_BEQLDATE 32 184#define FILE_FLOAT 33 185#define FILE_BEFLOAT 34 186#define FILE_LEFLOAT 35 187#define FILE_DOUBLE 36 188#define FILE_BEDOUBLE 37 189#define FILE_LEDOUBLE 38 190#define FILE_BEID3 39 191#define FILE_LEID3 40 192#define FILE_INDIRECT 41 193#define FILE_NAMES_SIZE 42/* size of array to contain all names */ 194 195#define IS_STRING(t) \ 196 ((t) == FILE_STRING || \ 197 (t) == FILE_PSTRING || \ 198 (t) == FILE_BESTRING16 || \ 199 (t) == FILE_LESTRING16 || \ 200 (t) == FILE_REGEX || \ 201 (t) == FILE_SEARCH || \ 202 (t) == FILE_DEFAULT) 203 204#define FILE_FMT_NONE 0 205#define FILE_FMT_NUM 1 /* "cduxXi" */ 206#define FILE_FMT_STR 2 /* "s" */ 207#define FILE_FMT_QUAD 3 /* "ll" */ 208#define FILE_FMT_FLOAT 4 /* "eEfFgG" */ 209#define FILE_FMT_DOUBLE 5 /* "eEfFgG" */ 210 211 /* Word 3 */ 212 uint8_t in_op; /* operator for indirection */ 213 uint8_t mask_op; /* operator for mask */ 214#ifdef ENABLE_CONDITIONALS 215 uint8_t cond; /* conditional type */ 216#else 217 uint8_t dummy; 218#endif 219 uint8_t factor_op; 220#define FILE_FACTOR_OP_PLUS '+' 221#define FILE_FACTOR_OP_MINUS '-' 222#define FILE_FACTOR_OP_TIMES '*' 223#define FILE_FACTOR_OP_DIV '/' 224#define FILE_FACTOR_OP_NONE '\0' 225 226#define FILE_OPS "&|^+-*/%" 227#define FILE_OPAND 0 228#define FILE_OPOR 1 229#define FILE_OPXOR 2 230#define FILE_OPADD 3 231#define FILE_OPMINUS 4 232#define FILE_OPMULTIPLY 5 233#define FILE_OPDIVIDE 6 234#define FILE_OPMODULO 7 235#define FILE_OPS_MASK 0x07 /* mask for above ops */ 236#define FILE_UNUSED_1 0x08 237#define FILE_UNUSED_2 0x10 238#define FILE_UNUSED_3 0x20 239#define FILE_OPINVERSE 0x40 240#define FILE_OPINDIRECT 0x80 241 242#ifdef ENABLE_CONDITIONALS 243#define COND_NONE 0 244#define COND_IF 1 245#define COND_ELIF 2 246#define COND_ELSE 3 247#endif /* ENABLE_CONDITIONALS */ 248 249 /* Word 4 */ 250 uint32_t offset; /* offset to magic number */ 251 /* Word 5 */ 252 int32_t in_offset; /* offset from indirection */ 253 /* Word 6 */ 254 uint32_t lineno; /* line number in magic file */ 255 /* Word 7,8 */ 256 union { 257 uint64_t _mask; /* for use with numeric and date types */ 258 struct { 259 uint32_t _count; /* repeat/line count */ 260 uint32_t _flags; /* modifier flags */ 261 } _s; /* for use with string types */ 262 } _u; 263#define num_mask _u._mask 264#define str_range _u._s._count 265#define str_flags _u._s._flags 266 /* Words 9-16 */ 267 union VALUETYPE value; /* either number or string */ 268 /* Words 17-24 */ 269 char desc[MAXDESC]; /* description */ 270 /* Words 25-32 */ 271 char mimetype[MAXDESC]; /* MIME type */ 272 /* Words 33-34 */ 273 char apple[8]; 274}; 275 276#define BIT(A) (1 << (A)) 277#define STRING_COMPACT_WHITESPACE BIT(0) 278#define STRING_COMPACT_OPTIONAL_WHITESPACE BIT(1) 279#define STRING_IGNORE_LOWERCASE BIT(2) 280#define STRING_IGNORE_UPPERCASE BIT(3) 281#define REGEX_OFFSET_START BIT(4) 282#define STRING_TEXTTEST BIT(5) 283#define STRING_BINTEST BIT(6) 284#define CHAR_COMPACT_WHITESPACE 'W' 285#define CHAR_COMPACT_OPTIONAL_WHITESPACE 'w' 286#define CHAR_IGNORE_LOWERCASE 'c' 287#define CHAR_IGNORE_UPPERCASE 'C' 288#define CHAR_REGEX_OFFSET_START 's' 289#define CHAR_TEXTTEST 't' 290#define CHAR_BINTEST 'b' 291#define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE) 292#define STRING_DEFAULT_RANGE 100 293 294 295/* list of magic entries */ 296struct mlist { 297 struct magic *magic; /* array of magic entries */ 298 uint32_t nmagic; /* number of entries in array */ 299 int mapped; /* allocation type: 0 => apprentice_file 300 * 1 => apprentice_map + malloc 301 * 2 => apprentice_map + mmap */ 302 struct mlist *next, *prev; 303}; 304 305#ifdef __cplusplus 306#define CAST(T, b) static_cast<T>(b) 307#else 308#define CAST(T, b) (T)(b) 309#endif 310 311struct level_info { 312 int32_t off; 313 int got_match; 314#ifdef ENABLE_CONDITIONALS 315 int last_match; 316 int last_cond; /* used for error checking by parse() */ 317#endif 318}; 319struct magic_set { 320 struct mlist *mlist; 321 struct cont { 322 size_t len; 323 struct level_info *li; 324 } c; 325 struct out { 326 char *buf; /* Accumulation buffer */ 327 char *pbuf; /* Printable buffer */ 328 } o; 329 uint32_t offset; 330 int error; 331 int flags; /* Control magic tests. */ 332 int event_flags; /* Note things that happened. */ 333#define EVENT_HAD_ERR 0x01 334 const char *file; 335 size_t line; /* current magic line number */ 336 337 /* data for searches */ 338 struct { 339 const char *s; /* start of search in original source */ 340 size_t s_len; /* length of search region */ 341 size_t offset; /* starting offset in source: XXX - should this be off_t? */ 342 size_t rm_len; /* match length */ 343 } search; 344 345 /* FIXME: Make the string dynamically allocated so that e.g. 346 strings matched in files can be longer than MAXstring */ 347 union VALUETYPE ms_value; /* either number or string */ 348}; 349 350/* Type for Unicode characters */ 351typedef unsigned long unichar; 352 353struct stat; 354protected const char *file_fmttime(uint32_t, int); 355protected int file_buffer(struct magic_set *, int, const char *, const void *, 356 size_t); 357protected int file_fsmagic(struct magic_set *, const char *, struct stat *); 358protected int file_pipe2file(struct magic_set *, int, const void *, size_t); 359protected int file_vprintf(struct magic_set *, const char *, va_list); 360protected int file_printf(struct magic_set *, const char *, ...) 361 __attribute__((__format__(__printf__, 2, 3))); 362protected int file_reset(struct magic_set *); 363protected int file_tryelf(struct magic_set *, int, const unsigned char *, 364 size_t); 365protected int file_trymacho(struct magic_set *, int, const unsigned char *, size_t, const char *); 366protected int file_trycdf(struct magic_set *, int, const unsigned char *, 367 size_t); 368protected int file_zmagic(struct magic_set *, int, const char *, 369 const unsigned char *, size_t); 370protected int file_ascmagic(struct magic_set *, const unsigned char *, size_t); 371protected int file_ascmagic_with_encoding(struct magic_set *, 372 const unsigned char *, size_t, unichar *, size_t, const char *, 373 const char *); 374protected int file_encoding(struct magic_set *, const unsigned char *, size_t, 375 unichar **, size_t *, const char **, const char **, const char **); 376protected int file_is_tar(struct magic_set *, const unsigned char *, size_t); 377protected int file_softmagic(struct magic_set *, const unsigned char *, size_t, 378 int); 379protected struct mlist *file_apprentice(struct magic_set *, const char *, int); 380protected uint64_t file_signextend(struct magic_set *, struct magic *, 381 uint64_t); 382protected void file_delmagic(struct magic *, int type, size_t entries); 383protected void file_badread(struct magic_set *); 384protected void file_badseek(struct magic_set *); 385protected void file_oomem(struct magic_set *, size_t); 386protected void file_error(struct magic_set *, int, const char *, ...) 387 __attribute__((__format__(__printf__, 3, 4))); 388protected void file_magerror(struct magic_set *, const char *, ...) 389 __attribute__((__format__(__printf__, 2, 3))); 390protected void file_magwarn(struct magic_set *, const char *, ...) 391 __attribute__((__format__(__printf__, 2, 3))); 392protected void file_mdump(struct magic *); 393protected void file_showstr(FILE *, const char *, size_t); 394protected size_t file_mbswidth(const char *); 395protected const char *file_getbuffer(struct magic_set *); 396protected ssize_t sread(int, void *, size_t, int); 397protected int file_check_mem(struct magic_set *, unsigned int); 398protected int file_looks_utf8(const unsigned char *, size_t, unichar *, 399 size_t *); 400#ifdef __EMX__ 401protected int file_os2_apptype(struct magic_set *, const char *, const void *, 402 size_t); 403#endif /* __EMX__ */ 404 405 406#ifndef COMPILE_ONLY 407extern const char *file_names[]; 408extern const size_t file_nnames; 409#endif 410 411#ifndef HAVE_STRERROR 412extern int sys_nerr; 413extern char *sys_errlist[]; 414#define strerror(e) \ 415 (((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error") 416#endif 417 418#ifndef HAVE_STRTOUL 419#define strtoul(a, b, c) strtol(a, b, c) 420#endif 421 422#ifndef HAVE_VASPRINTF 423int vasprintf(char **, const char *, va_list); 424#endif 425#ifndef HAVE_ASPRINTF 426int asprintf(char **ptr, const char *format_string, ...); 427#endif 428 429#ifndef HAVE_STRLCPY 430size_t strlcpy(char *dst, const char *src, size_t siz); 431#endif 432#ifndef HAVE_STRLCAT 433size_t strlcat(char *dst, const char *src, size_t siz); 434#endif 435 436#if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK) 437#define QUICK 438#endif 439 440#ifndef O_BINARY 441#define O_BINARY 0 442#endif 443 444#ifndef __cplusplus 445#if defined(__GNUC__) && (__GNUC__ >= 3) 446#define FILE_RCSID(id) \ 447static const char rcsid[] __attribute__((__used__)) = id; 448#else 449#define FILE_RCSID(id) \ 450static const char *rcsid(const char *p) { \ 451 return rcsid(p = id); \ 452} 453#endif 454#else 455#define FILE_RCSID(id) 456#endif 457 458#endif /* __file_h__ */ 459