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 1.144 2013/02/18 15:40:59 christos Exp $ 31 */ 32 33 #ifndef __file_h__ 34 #define __file_h__ 35 36 #include "config.h" 37 38 #ifdef PHP_WIN32 39 #ifdef _WIN64 40 #define SIZE_T_FORMAT "I64" 41 #else 42 #define SIZE_T_FORMAT "" 43 #endif 44 #define INT64_T_FORMAT "I64" 45 #else 46 #define SIZE_T_FORMAT "z" 47 #define INT64_T_FORMAT "ll" 48 #endif 49 50 #include <stdio.h> /* Include that here, to make sure __P gets defined */ 51 #include <errno.h> 52 #include <fcntl.h> /* For open and flags */ 53 #ifdef HAVE_STDINT_H 54 #ifndef __STDC_LIMIT_MACROS 55 #define __STDC_LIMIT_MACROS 56 #endif 57 #include <stdint.h> 58 #endif 59 #ifdef HAVE_INTTYPES_H 60 #include <inttypes.h> 61 #endif 62 #ifdef PHP_WIN32 63 #include "win32/php_stdint.h" 64 #endif 65 66 #include "php.h" 67 #include "ext/standard/php_string.h" 68 #include "ext/pcre/php_pcre.h" 69 70 #include <sys/types.h> 71 #ifdef PHP_WIN32 72 #include "win32/param.h" 73 #else 74 #include <sys/param.h> 75 #endif 76 /* Do this here and now, because struct stat gets re-defined on solaris */ 77 #include <sys/stat.h> 78 #include <stdarg.h> 79 80 #define ENABLE_CONDITIONALS 81 82 #ifndef MAGIC 83 #define MAGIC "/etc/magic" 84 #endif 85 86 #if defined(__EMX__) || defined(PHP_WIN32) 87 #define PATHSEP ';' 88 #else 89 #define PATHSEP ':' 90 #endif 91 92 #define private static 93 94 #if HAVE_VISIBILITY 95 #define public __attribute__ ((__visibility__("default"))) 96 #ifndef protected 97 #define protected __attribute__ ((__visibility__("hidden"))) 98 #endif 99 #else 100 #define public 101 #ifndef protected 102 #define protected 103 #endif 104 #endif 105 106 #ifndef __arraycount 107 #define __arraycount(a) (sizeof(a) / sizeof(a[0])) 108 #endif 109 110 #ifndef __GNUC_PREREQ__ 111 #ifdef __GNUC__ 112 #define __GNUC_PREREQ__(x, y) \ 113 ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \ 114 (__GNUC__ > (x))) 115 #else 116 #define __GNUC_PREREQ__(x, y) 0 117 #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 len of text description/MIME type */ 134 #define MAXMIME 80 /* max len of text MIME type */ 135 #define MAXstring 64 /* max len of "string" types */ 136 137 #define MAGICNO 0xF11E041C 138 #define VERSIONNO 10 139 #define FILE_MAGICSIZE 248 140 141 #define FILE_LOAD 0 142 #define FILE_CHECK 1 143 #define FILE_COMPILE 2 144 #define FILE_LIST 3 145 146 union VALUETYPE { 147 uint8_t b; 148 uint16_t h; 149 uint32_t l; 150 uint64_t q; 151 uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */ 152 uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */ 153 uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */ 154 char s[MAXstring]; /* the search string or regex pattern */ 155 unsigned char us[MAXstring]; 156 float f; 157 double d; 158 }; 159 160 struct magic { 161 /* Word 1 */ 162 uint16_t cont_level; /* level of ">" */ 163 uint8_t flag; 164 #define INDIR 0x01 /* if '(...)' appears */ 165 #define OFFADD 0x02 /* if '>&' or '>...(&' appears */ 166 #define INDIROFFADD 0x04 /* if '>&(' appears */ 167 #define UNSIGNED 0x08 /* comparison is unsigned */ 168 #define NOSPACE 0x10 /* suppress space character before output */ 169 #define BINTEST 0x20 /* test is for a binary type (set only 170 for top-level tests) */ 171 #define TEXTTEST 0x40 /* for passing to file_softmagic */ 172 173 uint8_t factor; 174 175 /* Word 2 */ 176 uint8_t reln; /* relation (0=eq, '>'=gt, etc) */ 177 uint8_t vallen; /* length of string value, if any */ 178 uint8_t type; /* comparison type (FILE_*) */ 179 uint8_t in_type; /* type of indirection */ 180 #define FILE_INVALID 0 181 #define FILE_BYTE 1 182 #define FILE_SHORT 2 183 #define FILE_DEFAULT 3 184 #define FILE_LONG 4 185 #define FILE_STRING 5 186 #define FILE_DATE 6 187 #define FILE_BESHORT 7 188 #define FILE_BELONG 8 189 #define FILE_BEDATE 9 190 #define FILE_LESHORT 10 191 #define FILE_LELONG 11 192 #define FILE_LEDATE 12 193 #define FILE_PSTRING 13 194 #define FILE_LDATE 14 195 #define FILE_BELDATE 15 196 #define FILE_LELDATE 16 197 #define FILE_REGEX 17 198 #define FILE_BESTRING16 18 199 #define FILE_LESTRING16 19 200 #define FILE_SEARCH 20 201 #define FILE_MEDATE 21 202 #define FILE_MELDATE 22 203 #define FILE_MELONG 23 204 #define FILE_QUAD 24 205 #define FILE_LEQUAD 25 206 #define FILE_BEQUAD 26 207 #define FILE_QDATE 27 208 #define FILE_LEQDATE 28 209 #define FILE_BEQDATE 29 210 #define FILE_QLDATE 30 211 #define FILE_LEQLDATE 31 212 #define FILE_BEQLDATE 32 213 #define FILE_FLOAT 33 214 #define FILE_BEFLOAT 34 215 #define FILE_LEFLOAT 35 216 #define FILE_DOUBLE 36 217 #define FILE_BEDOUBLE 37 218 #define FILE_LEDOUBLE 38 219 #define FILE_BEID3 39 220 #define FILE_LEID3 40 221 #define FILE_INDIRECT 41 222 #define FILE_QWDATE 42 223 #define FILE_LEQWDATE 43 224 #define FILE_BEQWDATE 44 225 #define FILE_NAME 45 226 #define FILE_USE 46 227 #define FILE_NAMES_SIZE 47 /* size of array to contain all names */ 228 229 #define IS_LIBMAGIC_STRING(t) \ 230 ((t) == FILE_STRING || \ 231 (t) == FILE_PSTRING || \ 232 (t) == FILE_BESTRING16 || \ 233 (t) == FILE_LESTRING16 || \ 234 (t) == FILE_REGEX || \ 235 (t) == FILE_SEARCH || \ 236 (t) == FILE_NAME || \ 237 (t) == FILE_USE || \ 238 (t) == FILE_DEFAULT) 239 240 #define FILE_FMT_NONE 0 241 #define FILE_FMT_NUM 1 /* "cduxXi" */ 242 #define FILE_FMT_STR 2 /* "s" */ 243 #define FILE_FMT_QUAD 3 /* "ll" */ 244 #define FILE_FMT_FLOAT 4 /* "eEfFgG" */ 245 #define FILE_FMT_DOUBLE 5 /* "eEfFgG" */ 246 247 /* Word 3 */ 248 uint8_t in_op; /* operator for indirection */ 249 uint8_t mask_op; /* operator for mask */ 250 #ifdef ENABLE_CONDITIONALS 251 uint8_t cond; /* conditional type */ 252 #else 253 uint8_t dummy; 254 #endif 255 uint8_t factor_op; 256 #define FILE_FACTOR_OP_PLUS '+' 257 #define FILE_FACTOR_OP_MINUS '-' 258 #define FILE_FACTOR_OP_TIMES '*' 259 #define FILE_FACTOR_OP_DIV '/' 260 #define FILE_FACTOR_OP_NONE '\0' 261 262 #define FILE_OPS "&|^+-*/%" 263 #define FILE_OPAND 0 264 #define FILE_OPOR 1 265 #define FILE_OPXOR 2 266 #define FILE_OPADD 3 267 #define FILE_OPMINUS 4 268 #define FILE_OPMULTIPLY 5 269 #define FILE_OPDIVIDE 6 270 #define FILE_OPMODULO 7 271 #define FILE_OPS_MASK 0x07 /* mask for above ops */ 272 #define FILE_UNUSED_1 0x08 273 #define FILE_UNUSED_2 0x10 274 #define FILE_UNUSED_3 0x20 275 #define FILE_OPINVERSE 0x40 276 #define FILE_OPINDIRECT 0x80 277 278 #ifdef ENABLE_CONDITIONALS 279 #define COND_NONE 0 280 #define COND_IF 1 281 #define COND_ELIF 2 282 #define COND_ELSE 3 283 #endif /* ENABLE_CONDITIONALS */ 284 285 /* Word 4 */ 286 uint32_t offset; /* offset to magic number */ 287 /* Word 5 */ 288 int32_t in_offset; /* offset from indirection */ 289 /* Word 6 */ 290 uint32_t lineno; /* line number in magic file */ 291 /* Word 7,8 */ 292 union { 293 uint64_t _mask; /* for use with numeric and date types */ 294 struct { 295 uint32_t _count; /* repeat/line count */ 296 uint32_t _flags; /* modifier flags */ 297 } _s; /* for use with string types */ 298 } _u; 299 #define num_mask _u._mask 300 #define str_range _u._s._count 301 #define str_flags _u._s._flags 302 /* Words 9-16 */ 303 union VALUETYPE value; /* either number or string */ 304 /* Words 17-32 */ 305 char desc[MAXDESC]; /* description */ 306 /* Words 33-52 */ 307 char mimetype[MAXMIME]; /* MIME type */ 308 /* Words 53-54 */ 309 char apple[8]; 310 }; 311 312 #define BIT(A) (1 << (A)) 313 #define STRING_COMPACT_WHITESPACE BIT(0) 314 #define STRING_COMPACT_OPTIONAL_WHITESPACE BIT(1) 315 #define STRING_IGNORE_LOWERCASE BIT(2) 316 #define STRING_IGNORE_UPPERCASE BIT(3) 317 #define REGEX_OFFSET_START BIT(4) 318 #define STRING_TEXTTEST BIT(5) 319 #define STRING_BINTEST BIT(6) 320 #define PSTRING_1_BE BIT(7) 321 #define PSTRING_1_LE BIT(7) 322 #define PSTRING_2_BE BIT(8) 323 #define PSTRING_2_LE BIT(9) 324 #define PSTRING_4_BE BIT(10) 325 #define PSTRING_4_LE BIT(11) 326 #define PSTRING_LEN \ 327 (PSTRING_1_BE|PSTRING_2_LE|PSTRING_2_BE|PSTRING_4_LE|PSTRING_4_BE) 328 #define PSTRING_LENGTH_INCLUDES_ITSELF BIT(12) 329 #define STRING_TRIM BIT(13) 330 #define CHAR_COMPACT_WHITESPACE 'W' 331 #define CHAR_COMPACT_OPTIONAL_WHITESPACE 'w' 332 #define CHAR_IGNORE_LOWERCASE 'c' 333 #define CHAR_IGNORE_UPPERCASE 'C' 334 #define CHAR_REGEX_OFFSET_START 's' 335 #define CHAR_TEXTTEST 't' 336 #define CHAR_TRIM 'T' 337 #define CHAR_BINTEST 'b' 338 #define CHAR_PSTRING_1_BE 'B' 339 #define CHAR_PSTRING_1_LE 'B' 340 #define CHAR_PSTRING_2_BE 'H' 341 #define CHAR_PSTRING_2_LE 'h' 342 #define CHAR_PSTRING_4_BE 'L' 343 #define CHAR_PSTRING_4_LE 'l' 344 #define CHAR_PSTRING_LENGTH_INCLUDES_ITSELF 'J' 345 #define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE) 346 #define STRING_DEFAULT_RANGE 100 347 348 349 /* list of magic entries */ 350 struct mlist { 351 struct magic *magic; /* array of magic entries */ 352 uint32_t nmagic; /* number of entries in array */ 353 void *map; /* internal resources used by entry */ 354 struct mlist *next, *prev; 355 }; 356 357 #ifdef __cplusplus 358 #define CAST(T, b) static_cast<T>(b) 359 #define RCAST(T, b) reinterpret_cast<T>(b) 360 #else 361 #define CAST(T, b) (T)(b) 362 #define RCAST(T, b) (T)(b) 363 #endif 364 365 struct level_info { 366 int32_t off; 367 int got_match; 368 #ifdef ENABLE_CONDITIONALS 369 int last_match; 370 int last_cond; /* used for error checking by parse() */ 371 #endif 372 }; 373 374 #define MAGIC_SETS 2 375 376 struct magic_set { 377 struct mlist *mlist[MAGIC_SETS]; /* list of regular entries */ 378 struct cont { 379 size_t len; 380 struct level_info *li; 381 } c; 382 struct out { 383 char *buf; /* Accumulation buffer */ 384 char *pbuf; /* Printable buffer */ 385 } o; 386 uint32_t offset; 387 int error; 388 int flags; /* Control magic tests. */ 389 int event_flags; /* Note things that happened. */ 390 #define EVENT_HAD_ERR 0x01 391 const char *file; 392 size_t line; /* current magic line number */ 393 394 /* data for searches */ 395 struct { 396 const char *s; /* start of search in original source */ 397 size_t s_len; /* length of search region */ 398 size_t offset; /* starting offset in source: XXX - should this be off_t? */ 399 size_t rm_len; /* match length */ 400 } search; 401 402 /* FIXME: Make the string dynamically allocated so that e.g. 403 strings matched in files can be longer than MAXstring */ 404 union VALUETYPE ms_value; /* either number or string */ 405 }; 406 407 /* Type for Unicode characters */ 408 typedef unsigned long unichar; 409 410 struct stat; 411 #define FILE_T_LOCAL 1 412 #define FILE_T_WINDOWS 2 413 protected const char *file_fmttime(uint64_t, int, char *); 414 protected struct magic_set *file_ms_alloc(int); 415 protected void file_ms_free(struct magic_set *); 416 protected int file_buffer(struct magic_set *, php_stream *, const char *, const void *, 417 size_t); 418 protected int file_fsmagic(struct magic_set *, const char *, struct stat *, php_stream *); 419 protected int file_pipe2file(struct magic_set *, int, const void *, size_t); 420 protected int file_replace(struct magic_set *, const char *, const char *); 421 protected int file_printf(struct magic_set *, const char *, ...); 422 protected int file_reset(struct magic_set *); 423 protected int file_trycdf(struct magic_set *, int, const unsigned char *, 424 size_t); 425 #ifdef PHP_FILEINFO_UNCOMPRESS 426 protected int file_zmagic(struct magic_set *, int, const char *, 427 const unsigned char *, size_t); 428 #endif 429 protected int file_ascmagic(struct magic_set *, const unsigned char *, size_t, 430 int); 431 protected int file_ascmagic_with_encoding(struct magic_set *, 432 const unsigned char *, size_t, unichar *, size_t, const char *, 433 const char *, int); 434 protected int file_encoding(struct magic_set *, const unsigned char *, size_t, 435 unichar **, size_t *, const char **, const char **, const char **); 436 protected int file_is_tar(struct magic_set *, const unsigned char *, size_t); 437 protected int file_softmagic(struct magic_set *, const unsigned char *, size_t, 438 size_t, int, int); 439 protected int file_apprentice(struct magic_set *, const char *, int); 440 protected int file_magicfind(struct magic_set *, const char *, struct mlist *); 441 protected uint64_t file_signextend(struct magic_set *, struct magic *, 442 uint64_t); 443 protected void file_delmagic(struct magic *, int type, size_t entries); 444 protected void file_badread(struct magic_set *); 445 protected void file_badseek(struct magic_set *); 446 protected void file_oomem(struct magic_set *, size_t); 447 protected void file_error(struct magic_set *, int, const char *, ...); 448 protected void file_magerror(struct magic_set *, const char *, ...); 449 protected void file_magwarn(struct magic_set *, const char *, ...); 450 protected void file_showstr(FILE *, const char *, size_t); 451 protected size_t file_mbswidth(const char *); 452 protected const char *file_getbuffer(struct magic_set *); 453 protected ssize_t sread(int, void *, size_t, int); 454 protected int file_check_mem(struct magic_set *, unsigned int); 455 protected int file_looks_utf8(const unsigned char *, size_t, unichar *, 456 size_t *); 457 protected size_t file_pstring_length_size(const struct magic *); 458 protected size_t file_pstring_get_length(const struct magic *, const char *); 459 protected size_t file_printedlen(const struct magic_set *ms); 460 #ifdef __EMX__ 461 protected int file_os2_apptype(struct magic_set *, const char *, const void *, 462 size_t); 463 #endif /* __EMX__ */ 464 465 extern const char *file_names[]; 466 extern const size_t file_nnames; 467 468 #ifndef HAVE_STRERROR 469 extern int sys_nerr; 470 extern char *sys_errlist[]; 471 #define strerror(e) \ 472 (((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error") 473 #endif 474 475 #ifndef HAVE_STRTOUL 476 #define strtoul(a, b, c) strtol(a, b, c) 477 #endif 478 479 #ifndef strlcpy 480 size_t strlcpy(char *dst, const char *src, size_t siz); 481 #endif 482 #ifndef strlcat 483 size_t strlcat(char *dst, const char *src, size_t siz); 484 #endif 485 #ifndef HAVE_GETLINE 486 ssize_t getline(char **dst, size_t *len, FILE *fp); 487 ssize_t getdelim(char **dst, size_t *len, int delimiter, FILE *fp); 488 #endif 489 490 #if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK) 491 #define QUICK 492 #endif 493 494 #ifndef O_BINARY 495 #define O_BINARY 0 496 #endif 497 498 #ifndef __cplusplus 499 #if defined(__GNUC__) && (__GNUC__ >= 3) 500 #define FILE_RCSID(id) \ 501 static const char rcsid[] __attribute__((__used__)) = id; 502 #else 503 #define FILE_RCSID(id) \ 504 static const char *rcsid(const char *p) { \ 505 return rcsid(p = id); \ 506 } 507 #endif 508 #else 509 #define FILE_RCSID(id) 510 #endif 511 512 #ifdef PHP_WIN32 513 #define FINFO_LSEEK_FUNC _lseek 514 #define FINFO_READ_FUNC _read 515 #else 516 #define FINFO_LSEEK_FUNC lseek 517 #define FINFO_READ_FUNC read 518 #endif 519 520 #endif /* __file_h__ */ 521