file.h revision 192348
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 1.119 2009/02/04 18:24:32 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#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#include <stdint.h> 45#endif 46#ifdef HAVE_INTTYPES_H 47#include <inttypes.h> 48#endif 49#include <regex.h> 50#include <sys/types.h> 51#include <sys/param.h> 52/* Do this here and now, because struct stat gets re-defined on solaris */ 53#include <sys/stat.h> 54#include <stdarg.h> 55 56#define ENABLE_CONDITIONALS 57 58#ifndef MAGIC 59#define MAGIC "/etc/magic" 60#endif 61 62#ifdef __EMX__ 63#define PATHSEP ';' 64#else 65#define PATHSEP ':' 66#endif 67 68#define private static 69#ifndef protected 70#define protected 71#endif 72#define public 73 74#ifndef __GNUC_PREREQ__ 75#ifdef __GNUC__ 76#define __GNUC_PREREQ__(x, y) \ 77 ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \ 78 (__GNUC__ > (x))) 79#else 80#define __GNUC_PREREQ__(x, y) 0 81#endif 82#endif 83 84#ifndef __GNUC__ 85#ifndef __attribute__ 86#define __attribute__(a) 87#endif 88#endif 89 90#ifndef MIN 91#define MIN(a,b) (((a) < (b)) ? (a) : (b)) 92#endif 93 94#ifndef MAX 95#define MAX(a,b) (((a) > (b)) ? (a) : (b)) 96#endif 97 98#ifndef HOWMANY 99# define HOWMANY (256 * 1024) /* how much of the file to look at */ 100#endif 101#define MAXMAGIS 8192 /* max entries in any one magic file 102 or directory */ 103#define MAXDESC 64 /* max leng of text description/MIME type */ 104#define MAXstring 32 /* max leng of "string" types */ 105 106#define MAGICNO 0xF11E041C 107#define VERSIONNO 7 108#define FILE_MAGICSIZE 200 109 110#define FILE_LOAD 0 111#define FILE_CHECK 1 112#define FILE_COMPILE 2 113 114union VALUETYPE { 115 uint8_t b; 116 uint16_t h; 117 uint32_t l; 118 uint64_t q; 119 uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */ 120 uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */ 121 uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */ 122 char s[MAXstring]; /* the search string or regex pattern */ 123 unsigned char us[MAXstring]; 124 float f; 125 double d; 126}; 127 128struct magic { 129 /* Word 1 */ 130 uint16_t cont_level; /* level of ">" */ 131 uint8_t flag; 132#define INDIR 0x01 /* if '(...)' appears */ 133#define OFFADD 0x02 /* if '>&' or '>...(&' appears */ 134#define INDIROFFADD 0x04 /* if '>&(' appears */ 135#define UNSIGNED 0x08 /* comparison is unsigned */ 136#define NOSPACE 0x10 /* suppress space character before output */ 137#define BINTEST 0x20 /* test is for a binary type (set only 138 for top-level tests) */ 139#define TEXTTEST 0 /* for passing to file_softmagic */ 140 141 uint8_t factor; 142 143 /* Word 2 */ 144 uint8_t reln; /* relation (0=eq, '>'=gt, etc) */ 145 uint8_t vallen; /* length of string value, if any */ 146 uint8_t type; /* comparison type (FILE_*) */ 147 uint8_t in_type; /* type of indirection */ 148#define FILE_INVALID 0 149#define FILE_BYTE 1 150#define FILE_SHORT 2 151#define FILE_DEFAULT 3 152#define FILE_LONG 4 153#define FILE_STRING 5 154#define FILE_DATE 6 155#define FILE_BESHORT 7 156#define FILE_BELONG 8 157#define FILE_BEDATE 9 158#define FILE_LESHORT 10 159#define FILE_LELONG 11 160#define FILE_LEDATE 12 161#define FILE_PSTRING 13 162#define FILE_LDATE 14 163#define FILE_BELDATE 15 164#define FILE_LELDATE 16 165#define FILE_REGEX 17 166#define FILE_BESTRING16 18 167#define FILE_LESTRING16 19 168#define FILE_SEARCH 20 169#define FILE_MEDATE 21 170#define FILE_MELDATE 22 171#define FILE_MELONG 23 172#define FILE_QUAD 24 173#define FILE_LEQUAD 25 174#define FILE_BEQUAD 26 175#define FILE_QDATE 27 176#define FILE_LEQDATE 28 177#define FILE_BEQDATE 29 178#define FILE_QLDATE 30 179#define FILE_LEQLDATE 31 180#define FILE_BEQLDATE 32 181#define FILE_FLOAT 33 182#define FILE_BEFLOAT 34 183#define FILE_LEFLOAT 35 184#define FILE_DOUBLE 36 185#define FILE_BEDOUBLE 37 186#define FILE_LEDOUBLE 38 187#define FILE_BEID3 39 188#define FILE_LEID3 40 189#define FILE_INDIRECT 41 190#define FILE_NAMES_SIZE 42/* size of array to contain all names */ 191 192#define IS_STRING(t) \ 193 ((t) == FILE_STRING || \ 194 (t) == FILE_PSTRING || \ 195 (t) == FILE_BESTRING16 || \ 196 (t) == FILE_LESTRING16 || \ 197 (t) == FILE_REGEX || \ 198 (t) == FILE_SEARCH || \ 199 (t) == FILE_DEFAULT) 200 201#define FILE_FMT_NONE 0 202#define FILE_FMT_NUM 1 /* "cduxXi" */ 203#define FILE_FMT_STR 2 /* "s" */ 204#define FILE_FMT_QUAD 3 /* "ll" */ 205#define FILE_FMT_FLOAT 4 /* "eEfFgG" */ 206#define FILE_FMT_DOUBLE 5 /* "eEfFgG" */ 207 208 /* Word 3 */ 209 uint8_t in_op; /* operator for indirection */ 210 uint8_t mask_op; /* operator for mask */ 211#ifdef ENABLE_CONDITIONALS 212 uint8_t cond; /* conditional type */ 213#else 214 uint8_t dummy; 215#endif 216 uint8_t factor_op; 217#define FILE_FACTOR_OP_PLUS '+' 218#define FILE_FACTOR_OP_MINUS '-' 219#define FILE_FACTOR_OP_TIMES '*' 220#define FILE_FACTOR_OP_DIV '/' 221#define FILE_FACTOR_OP_NONE '\0' 222 223#define FILE_OPS "&|^+-*/%" 224#define FILE_OPAND 0 225#define FILE_OPOR 1 226#define FILE_OPXOR 2 227#define FILE_OPADD 3 228#define FILE_OPMINUS 4 229#define FILE_OPMULTIPLY 5 230#define FILE_OPDIVIDE 6 231#define FILE_OPMODULO 7 232#define FILE_OPS_MASK 0x07 /* mask for above ops */ 233#define FILE_UNUSED_1 0x08 234#define FILE_UNUSED_2 0x10 235#define FILE_UNUSED_3 0x20 236#define FILE_OPINVERSE 0x40 237#define FILE_OPINDIRECT 0x80 238 239#ifdef ENABLE_CONDITIONALS 240#define COND_NONE 0 241#define COND_IF 1 242#define COND_ELIF 2 243#define COND_ELSE 3 244#endif /* ENABLE_CONDITIONALS */ 245 246 /* Word 4 */ 247 uint32_t offset; /* offset to magic number */ 248 /* Word 5 */ 249 int32_t in_offset; /* offset from indirection */ 250 /* Word 6 */ 251 uint32_t lineno; /* line number in magic file */ 252 /* Word 7,8 */ 253 union { 254 uint64_t _mask; /* for use with numeric and date types */ 255 struct { 256 uint32_t _count; /* repeat/line count */ 257 uint32_t _flags; /* modifier flags */ 258 } _s; /* for use with string types */ 259 } _u; 260#define num_mask _u._mask 261#define str_range _u._s._count 262#define str_flags _u._s._flags 263 /* Words 9-16 */ 264 union VALUETYPE value; /* either number or string */ 265 /* Words 17-24 */ 266 char desc[MAXDESC]; /* description */ 267 /* Words 25-32 */ 268 char mimetype[MAXDESC]; /* MIME type */ 269 /* Words 33-34 */ 270 char apple[8]; 271}; 272 273#define BIT(A) (1 << (A)) 274#define STRING_COMPACT_BLANK BIT(0) 275#define STRING_COMPACT_OPTIONAL_BLANK BIT(1) 276#define STRING_IGNORE_LOWERCASE BIT(2) 277#define STRING_IGNORE_UPPERCASE BIT(3) 278#define REGEX_OFFSET_START BIT(4) 279#define CHAR_COMPACT_BLANK 'B' 280#define CHAR_COMPACT_OPTIONAL_BLANK 'b' 281#define CHAR_IGNORE_LOWERCASE 'c' 282#define CHAR_IGNORE_UPPERCASE 'C' 283#define CHAR_REGEX_OFFSET_START 's' 284#define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE) 285#define STRING_DEFAULT_RANGE 100 286 287 288/* list of magic entries */ 289struct mlist { 290 struct magic *magic; /* array of magic entries */ 291 uint32_t nmagic; /* number of entries in array */ 292 int mapped; /* allocation type: 0 => apprentice_file 293 * 1 => apprentice_map + malloc 294 * 2 => apprentice_map + mmap */ 295 struct mlist *next, *prev; 296}; 297 298#ifdef __cplusplus 299#define CAST(T, b) static_cast<T>(b) 300#else 301#define CAST(T, b) (b) 302#endif 303 304struct level_info { 305 int32_t off; 306 int got_match; 307#ifdef ENABLE_CONDITIONALS 308 int last_match; 309 int last_cond; /* used for error checking by parse() */ 310#endif 311}; 312struct magic_set { 313 struct mlist *mlist; 314 struct cont { 315 size_t len; 316 struct level_info *li; 317 } c; 318 struct out { 319 char *buf; /* Accumulation buffer */ 320 char *pbuf; /* Printable buffer */ 321 } o; 322 uint32_t offset; 323 int error; 324 int flags; /* Control magic tests. */ 325 int event_flags; /* Note things that happened. */ 326#define EVENT_HAD_ERR 0x01 327 const char *file; 328 size_t line; /* current magic line number */ 329 330 /* data for searches */ 331 struct { 332 const char *s; /* start of search in original source */ 333 size_t s_len; /* length of search region */ 334 size_t offset; /* starting offset in source: XXX - should this be off_t? */ 335 size_t rm_len; /* match length */ 336 } search; 337 338 /* FIXME: Make the string dynamically allocated so that e.g. 339 strings matched in files can be longer than MAXstring */ 340 union VALUETYPE ms_value; /* either number or string */ 341}; 342 343/* Type for Unicode characters */ 344typedef unsigned long unichar; 345 346struct stat; 347protected const char *file_fmttime(uint32_t, int); 348protected int file_buffer(struct magic_set *, int, const char *, const void *, 349 size_t); 350protected int file_fsmagic(struct magic_set *, const char *, struct stat *); 351protected int file_pipe2file(struct magic_set *, int, const void *, size_t); 352protected int file_vprintf(struct magic_set *, const char *, va_list); 353protected int file_printf(struct magic_set *, const char *, ...) 354 __attribute__((__format__(__printf__, 2, 3))); 355protected int file_reset(struct magic_set *); 356protected int file_tryelf(struct magic_set *, int, const unsigned char *, 357 size_t); 358protected int file_trycdf(struct magic_set *, int, const unsigned char *, 359 size_t); 360protected int file_zmagic(struct magic_set *, int, const char *, 361 const unsigned char *, size_t); 362protected int file_ascmagic(struct magic_set *, const unsigned char *, size_t); 363protected int file_ascmagic_with_encoding(struct magic_set *, 364 const unsigned char *, size_t, unichar *, size_t, const char *, 365 const char *); 366protected int file_encoding(struct magic_set *, const unsigned char *, size_t, 367 unichar **, size_t *, const char **, const char **, const char **); 368protected int file_is_tar(struct magic_set *, const unsigned char *, size_t); 369protected int file_softmagic(struct magic_set *, const unsigned char *, size_t, 370 int); 371protected struct mlist *file_apprentice(struct magic_set *, const char *, int); 372protected uint64_t file_signextend(struct magic_set *, struct magic *, 373 uint64_t); 374protected void file_delmagic(struct magic *, int type, size_t entries); 375protected void file_badread(struct magic_set *); 376protected void file_badseek(struct magic_set *); 377protected void file_oomem(struct magic_set *, size_t); 378protected void file_error(struct magic_set *, int, const char *, ...) 379 __attribute__((__format__(__printf__, 3, 4))); 380protected void file_magerror(struct magic_set *, const char *, ...) 381 __attribute__((__format__(__printf__, 2, 3))); 382protected void file_magwarn(struct magic_set *, const char *, ...) 383 __attribute__((__format__(__printf__, 2, 3))); 384protected void file_mdump(struct magic *); 385protected void file_showstr(FILE *, const char *, size_t); 386protected size_t file_mbswidth(const char *); 387protected const char *file_getbuffer(struct magic_set *); 388protected ssize_t sread(int, void *, size_t, int); 389protected int file_check_mem(struct magic_set *, unsigned int); 390protected int file_looks_utf8(const unsigned char *, size_t, unichar *, 391 size_t *); 392#ifdef __EMX__ 393protected int file_os2_apptype(struct magic_set *, const char *, const void *, 394 size_t); 395#endif /* __EMX__ */ 396 397 398#ifndef COMPILE_ONLY 399extern const char *file_names[]; 400extern const size_t file_nnames; 401#endif 402 403#ifndef HAVE_STRERROR 404extern int sys_nerr; 405extern char *sys_errlist[]; 406#define strerror(e) \ 407 (((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error") 408#endif 409 410#ifndef HAVE_STRTOUL 411#define strtoul(a, b, c) strtol(a, b, c) 412#endif 413 414#ifndef HAVE_VASPRINTF 415int vasprintf(char **, const char *, va_list); 416#endif 417#ifndef HAVE_ASPRINTF 418int asprintf(char **ptr, const char *format_string, ...); 419#endif 420 421#ifndef HAVE_STRLCPY 422size_t strlcpy(char *dst, const char *src, size_t siz); 423#endif 424#ifndef HAVE_STRLCAT 425size_t strlcat(char *dst, const char *src, size_t siz); 426#endif 427 428#if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK) 429#define QUICK 430#endif 431 432#ifndef O_BINARY 433#define O_BINARY 0 434#endif 435 436#ifndef __cplusplus 437#ifdef __GNUC__ 438#define FILE_RCSID(id) \ 439static const char rcsid[] __attribute__((__used__)) = id; 440#else 441#define FILE_RCSID(id) \ 442static const char *rcsid(const char *p) { \ 443 return rcsid(p = id); \ 444} 445#endif 446#else 447#define FILE_RCSID(id) 448#endif 449 450#endif /* __file_h__ */ 451