1 /* $NetBSD: file.h,v 1.3 2009/05/15 12:06:09 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 1.119 2009/02/04 18:24:32 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 #include <stdio.h> /* Include that here, to make sure __P gets defined */ 43 #include <errno.h> 44 #include <fcntl.h> /* For open and flags */ 45 #ifdef HAVE_STDINT_H 46 #include <stdint.h> 47 #endif 48 #ifdef HAVE_INTTYPES_H 49 #include <inttypes.h> 50 #endif 51 #include <regex.h> 52 #include <sys/types.h> 53 #include <sys/param.h> 54 /* Do this here and now, because struct stat gets re-defined on solaris */ 55 #include <sys/stat.h> 56 #include <stdarg.h> 57 58 #define ENABLE_CONDITIONALS 59 60 #ifndef MAGIC 61 #define MAGIC "/etc/magic" 62 #endif 63 64 #ifdef __EMX__ 65 #define PATHSEP ';' 66 #else 67 #define PATHSEP ':' 68 #endif 69 70 #define private static 71 #ifndef protected 72 #define protected 73 #endif 74 #define public 75 76 #ifndef __GNUC_PREREQ__ 77 #ifdef __GNUC__ 78 #define __GNUC_PREREQ__(x, y) \ 79 ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \ 80 (__GNUC__ > (x))) 81 #else 82 #define __GNUC_PREREQ__(x, y) 0 83 #endif 84 #endif 85 86 #ifndef __GNUC__ 87 #ifndef __attribute__ 88 #define __attribute__(a) 89 #endif 90 #endif 91 92 #ifndef MIN 93 #define MIN(a,b) (((a) < (b)) ? (a) : (b)) 94 #endif 95 96 #ifndef MAX 97 #define MAX(a,b) (((a) > (b)) ? (a) : (b)) 98 #endif 99 100 #ifndef HOWMANY 101 # define HOWMANY (256 * 1024) /* how much of the file to look at */ 102 #endif 103 #define MAXMAGIS 8192 /* max entries in any one magic file 104 or directory */ 105 #define MAXDESC 64 /* max leng of text description/MIME type */ 106 #define MAXstring 32 /* max leng of "string" types */ 107 108 #define MAGICNO 0xF11E041C 109 #define VERSIONNO 7 110 #define FILE_MAGICSIZE 200 111 112 #define FILE_LOAD 0 113 #define FILE_CHECK 1 114 #define FILE_COMPILE 2 115 116 union VALUETYPE { 117 uint8_t b; 118 uint16_t h; 119 uint32_t l; 120 uint64_t q; 121 uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */ 122 uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */ 123 uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */ 124 char s[MAXstring]; /* the search string or regex pattern */ 125 unsigned char us[MAXstring]; 126 float f; 127 double d; 128 }; 129 130 struct magic { 131 /* Word 1 */ 132 uint16_t cont_level; /* level of ">" */ 133 uint8_t flag; 134 #define INDIR 0x01 /* if '(...)' appears */ 135 #define OFFADD 0x02 /* if '>&' or '>...(&' appears */ 136 #define INDIROFFADD 0x04 /* if '>&(' appears */ 137 #define UNSIGNED 0x08 /* comparison is unsigned */ 138 #define NOSPACE 0x10 /* suppress space character before output */ 139 #define BINTEST 0x20 /* test is for a binary type (set only 140 for top-level tests) */ 141 #define TEXTTEST 0 /* for passing to file_softmagic */ 142 143 uint8_t factor; 144 145 /* Word 2 */ 146 uint8_t reln; /* relation (0=eq, '>'=gt, etc) */ 147 uint8_t vallen; /* length of string value, if any */ 148 uint8_t type; /* comparison type (FILE_*) */ 149 uint8_t in_type; /* type of indirection */ 150 #define FILE_INVALID 0 151 #define FILE_BYTE 1 152 #define FILE_SHORT 2 153 #define FILE_DEFAULT 3 154 #define FILE_LONG 4 155 #define FILE_STRING 5 156 #define FILE_DATE 6 157 #define FILE_BESHORT 7 158 #define FILE_BELONG 8 159 #define FILE_BEDATE 9 160 #define FILE_LESHORT 10 161 #define FILE_LELONG 11 162 #define FILE_LEDATE 12 163 #define FILE_PSTRING 13 164 #define FILE_LDATE 14 165 #define FILE_BELDATE 15 166 #define FILE_LELDATE 16 167 #define FILE_REGEX 17 168 #define FILE_BESTRING16 18 169 #define FILE_LESTRING16 19 170 #define FILE_SEARCH 20 171 #define FILE_MEDATE 21 172 #define FILE_MELDATE 22 173 #define FILE_MELONG 23 174 #define FILE_QUAD 24 175 #define FILE_LEQUAD 25 176 #define FILE_BEQUAD 26 177 #define FILE_QDATE 27 178 #define FILE_LEQDATE 28 179 #define FILE_BEQDATE 29 180 #define FILE_QLDATE 30 181 #define FILE_LEQLDATE 31 182 #define FILE_BEQLDATE 32 183 #define FILE_FLOAT 33 184 #define FILE_BEFLOAT 34 185 #define FILE_LEFLOAT 35 186 #define FILE_DOUBLE 36 187 #define FILE_BEDOUBLE 37 188 #define FILE_LEDOUBLE 38 189 #define FILE_BEID3 39 190 #define FILE_LEID3 40 191 #define FILE_INDIRECT 41 192 #define FILE_NAMES_SIZE 42/* size of array to contain all names */ 193 194 #define IS_STRING(t) \ 195 ((t) == FILE_STRING || \ 196 (t) == FILE_PSTRING || \ 197 (t) == FILE_BESTRING16 || \ 198 (t) == FILE_LESTRING16 || \ 199 (t) == FILE_REGEX || \ 200 (t) == FILE_SEARCH || \ 201 (t) == FILE_DEFAULT) 202 203 #define FILE_FMT_NONE 0 204 #define FILE_FMT_NUM 1 /* "cduxXi" */ 205 #define FILE_FMT_STR 2 /* "s" */ 206 #define FILE_FMT_QUAD 3 /* "ll" */ 207 #define FILE_FMT_FLOAT 4 /* "eEfFgG" */ 208 #define FILE_FMT_DOUBLE 5 /* "eEfFgG" */ 209 210 /* Word 3 */ 211 uint8_t in_op; /* operator for indirection */ 212 uint8_t mask_op; /* operator for mask */ 213 #ifdef ENABLE_CONDITIONALS 214 uint8_t cond; /* conditional type */ 215 #else 216 uint8_t dummy; 217 #endif 218 uint8_t factor_op; 219 #define FILE_FACTOR_OP_PLUS '+' 220 #define FILE_FACTOR_OP_MINUS '-' 221 #define FILE_FACTOR_OP_TIMES '*' 222 #define FILE_FACTOR_OP_DIV '/' 223 #define FILE_FACTOR_OP_NONE '\0' 224 225 #define FILE_OPS "&|^+-*/%" 226 #define FILE_OPAND 0 227 #define FILE_OPOR 1 228 #define FILE_OPXOR 2 229 #define FILE_OPADD 3 230 #define FILE_OPMINUS 4 231 #define FILE_OPMULTIPLY 5 232 #define FILE_OPDIVIDE 6 233 #define FILE_OPMODULO 7 234 #define FILE_OPS_MASK 0x07 /* mask for above ops */ 235 #define FILE_UNUSED_1 0x08 236 #define FILE_UNUSED_2 0x10 237 #define FILE_UNUSED_3 0x20 238 #define FILE_OPINVERSE 0x40 239 #define FILE_OPINDIRECT 0x80 240 241 #ifdef ENABLE_CONDITIONALS 242 #define COND_NONE 0 243 #define COND_IF 1 244 #define COND_ELIF 2 245 #define COND_ELSE 3 246 #endif /* ENABLE_CONDITIONALS */ 247 248 /* Word 4 */ 249 uint32_t offset; /* offset to magic number */ 250 /* Word 5 */ 251 int32_t in_offset; /* offset from indirection */ 252 /* Word 6 */ 253 uint32_t lineno; /* line number in magic file */ 254 /* Word 7,8 */ 255 union { 256 uint64_t _mask; /* for use with numeric and date types */ 257 struct { 258 uint32_t _count; /* repeat/line count */ 259 uint32_t _flags; /* modifier flags */ 260 } _s; /* for use with string types */ 261 } _u; 262 #define num_mask _u._mask 263 #define str_range _u._s._count 264 #define str_flags _u._s._flags 265 /* Words 9-16 */ 266 union VALUETYPE value; /* either number or string */ 267 /* Words 17-24 */ 268 char desc[MAXDESC]; /* description */ 269 /* Words 25-32 */ 270 char mimetype[MAXDESC]; /* MIME type */ 271 /* Words 33-34 */ 272 char apple[8]; 273 }; 274 275 #define BIT(A) (1 << (A)) 276 #define STRING_COMPACT_BLANK BIT(0) 277 #define STRING_COMPACT_OPTIONAL_BLANK BIT(1) 278 #define STRING_IGNORE_LOWERCASE BIT(2) 279 #define STRING_IGNORE_UPPERCASE BIT(3) 280 #define REGEX_OFFSET_START BIT(4) 281 #define CHAR_COMPACT_BLANK 'B' 282 #define CHAR_COMPACT_OPTIONAL_BLANK 'b' 283 #define CHAR_IGNORE_LOWERCASE 'c' 284 #define CHAR_IGNORE_UPPERCASE 'C' 285 #define CHAR_REGEX_OFFSET_START 's' 286 #define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE) 287 #define STRING_DEFAULT_RANGE 100 288 289 290 /* list of magic entries */ 291 struct mlist { 292 struct magic *magic; /* array of magic entries */ 293 uint32_t nmagic; /* number of entries in array */ 294 int mapped; /* allocation type: 0 => apprentice_file 295 * 1 => apprentice_map + malloc 296 * 2 => apprentice_map + mmap */ 297 struct mlist *next, *prev; 298 }; 299 300 #ifdef __cplusplus 301 #define CAST(T, b) static_cast<T>(b) 302 #else 303 #define CAST(T, b) (T)(b) 304 #endif 305 306 struct level_info { 307 int32_t off; 308 int got_match; 309 #ifdef ENABLE_CONDITIONALS 310 int last_match; 311 int last_cond; /* used for error checking by parse() */ 312 #endif 313 }; 314 struct magic_set { 315 struct mlist *mlist; 316 struct cont { 317 size_t len; 318 struct level_info *li; 319 } c; 320 struct out { 321 char *buf; /* Accumulation buffer */ 322 char *pbuf; /* Printable buffer */ 323 } o; 324 uint32_t offset; 325 int error; 326 int flags; /* Control magic tests. */ 327 int event_flags; /* Note things that happened. */ 328 #define EVENT_HAD_ERR 0x01 329 const char *file; 330 size_t line; /* current magic line number */ 331 332 /* data for searches */ 333 struct { 334 const char *s; /* start of search in original source */ 335 size_t s_len; /* length of search region */ 336 size_t offset; /* starting offset in source: XXX - should this be off_t? */ 337 size_t rm_len; /* match length */ 338 } search; 339 340 /* FIXME: Make the string dynamically allocated so that e.g. 341 strings matched in files can be longer than MAXstring */ 342 union VALUETYPE ms_value; /* either number or string */ 343 }; 344 345 /* Type for Unicode characters */ 346 typedef unsigned long unichar; 347 348 struct stat; 349 protected const char *file_fmttime(uint32_t, int); 350 protected int file_buffer(struct magic_set *, int, const char *, const void *, 351 size_t); 352 protected int file_fsmagic(struct magic_set *, const char *, struct stat *); 353 protected int file_pipe2file(struct magic_set *, int, const void *, size_t); 354 protected int file_vprintf(struct magic_set *, const char *, va_list); 355 protected int file_printf(struct magic_set *, const char *, ...) 356 __attribute__((__format__(__printf__, 2, 3))); 357 protected int file_reset(struct magic_set *); 358 protected int file_tryelf(struct magic_set *, int, const unsigned char *, 359 size_t); 360 protected int file_trycdf(struct magic_set *, int, const unsigned char *, 361 size_t); 362 protected int file_zmagic(struct magic_set *, int, const char *, 363 const unsigned char *, size_t); 364 protected int file_ascmagic(struct magic_set *, const unsigned char *, size_t); 365 protected int file_ascmagic_with_encoding(struct magic_set *, 366 const unsigned char *, size_t, unichar *, size_t, const char *, 367 const char *); 368 protected int file_encoding(struct magic_set *, const unsigned char *, size_t, 369 unichar **, size_t *, const char **, const char **, const char **); 370 protected int file_is_tar(struct magic_set *, const unsigned char *, size_t); 371 protected int file_softmagic(struct magic_set *, const unsigned char *, size_t, 372 int); 373 protected struct mlist *file_apprentice(struct magic_set *, const char *, int); 374 protected uint64_t file_signextend(struct magic_set *, struct magic *, 375 uint64_t); 376 protected void file_delmagic(struct magic *, int type, size_t entries); 377 protected void file_badread(struct magic_set *); 378 protected void file_badseek(struct magic_set *); 379 protected void file_oomem(struct magic_set *, size_t); 380 protected void file_error(struct magic_set *, int, const char *, ...) 381 __attribute__((__format__(__printf__, 3, 4))); 382 protected void file_magerror(struct magic_set *, const char *, ...) 383 __attribute__((__format__(__printf__, 2, 3))); 384 protected void file_magwarn(struct magic_set *, const char *, ...) 385 __attribute__((__format__(__printf__, 2, 3))); 386 protected void file_mdump(struct magic *); 387 protected void file_showstr(FILE *, const char *, size_t); 388 protected size_t file_mbswidth(const char *); 389 protected const char *file_getbuffer(struct magic_set *); 390 protected ssize_t sread(int, void *, size_t, int); 391 protected int file_check_mem(struct magic_set *, unsigned int); 392 protected int file_looks_utf8(const unsigned char *, size_t, unichar *, 393 size_t *); 394 #ifdef __EMX__ 395 protected int file_os2_apptype(struct magic_set *, const char *, const void *, 396 size_t); 397 #endif /* __EMX__ */ 398 399 400 #ifndef COMPILE_ONLY 401 extern const char *file_names[]; 402 extern const size_t file_nnames; 403 #endif 404 405 #ifndef HAVE_STRERROR 406 extern int sys_nerr; 407 extern char *sys_errlist[]; 408 #define strerror(e) \ 409 (((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error") 410 #endif 411 412 #ifndef HAVE_STRTOUL 413 #define strtoul(a, b, c) strtol(a, b, c) 414 #endif 415 416 #ifndef HAVE_VASPRINTF 417 int vasprintf(char **, const char *, va_list); 418 #endif 419 #ifndef HAVE_ASPRINTF 420 int asprintf(char **ptr, const char *format_string, ...); 421 #endif 422 423 #ifndef HAVE_STRLCPY 424 size_t strlcpy(char *dst, const char *src, size_t siz); 425 #endif 426 #ifndef HAVE_STRLCAT 427 size_t strlcat(char *dst, const char *src, size_t siz); 428 #endif 429 430 #if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK) 431 #define QUICK 432 #endif 433 434 #ifndef O_BINARY 435 #define O_BINARY 0 436 #endif 437 438 #ifndef __cplusplus 439 #ifdef __GNUC__ 440 #define FILE_RCSID(id) \ 441 static const char rcsid[] __attribute__((__used__)) = id; 442 #else 443 #define FILE_RCSID(id) \ 444 static const char *rcsid(const char *p) { \ 445 return rcsid(p = id); \ 446 } 447 #endif 448 #else 449 #define FILE_RCSID(id) 450 #endif 451 #ifndef __RCSID 452 #define __RCSID(a) 453 #endif 454 455 #endif /* __file_h__ */ 456