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 1.135 2011/09/20 15:30:14 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 #ifndef protected 94 #define protected 95 #endif 96 #define public 97 98 #ifndef __arraycount 99 #define __arraycount(a) (sizeof(a) / sizeof(a[0])) 100 #endif 101 102 #ifndef __GNUC_PREREQ__ 103 #ifdef __GNUC__ 104 #define __GNUC_PREREQ__(x, y) \ 105 ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \ 106 (__GNUC__ > (x))) 107 #else 108 #define __GNUC_PREREQ__(x, y) 0 109 #endif 110 #endif 111 112 #ifndef MIN 113 #define MIN(a,b) (((a) < (b)) ? (a) : (b)) 114 #endif 115 116 #ifndef MAX 117 #define MAX(a,b) (((a) > (b)) ? (a) : (b)) 118 #endif 119 120 #ifndef HOWMANY 121 # define HOWMANY (256 * 1024) /* how much of the file to look at */ 122 #endif 123 #define MAXMAGIS 8192 /* max entries in any one magic file 124 or directory */ 125 #define MAXDESC 64 /* max leng of text description/MIME type */ 126 #define MAXstring 64 /* max leng of "string" types */ 127 128 #define MAGICNO 0xF11E041C 129 #define VERSIONNO 8 130 #define FILE_MAGICSIZE 232 131 132 #define FILE_LOAD 0 133 #define FILE_CHECK 1 134 #define FILE_COMPILE 2 135 #define FILE_LIST 3 136 137 union VALUETYPE { 138 uint8_t b; 139 uint16_t h; 140 uint32_t l; 141 uint64_t q; 142 uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */ 143 uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */ 144 uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */ 145 char s[MAXstring]; /* the search string or regex pattern */ 146 unsigned char us[MAXstring]; 147 float f; 148 double d; 149 }; 150 151 struct magic { 152 /* Word 1 */ 153 uint16_t cont_level; /* level of ">" */ 154 uint8_t flag; 155 #define INDIR 0x01 /* if '(...)' appears */ 156 #define OFFADD 0x02 /* if '>&' or '>...(&' appears */ 157 #define INDIROFFADD 0x04 /* if '>&(' appears */ 158 #define UNSIGNED 0x08 /* comparison is unsigned */ 159 #define NOSPACE 0x10 /* suppress space character before output */ 160 #define BINTEST 0x20 /* test is for a binary type (set only 161 for top-level tests) */ 162 #define TEXTTEST 0x40 /* for passing to file_softmagic */ 163 164 uint8_t factor; 165 166 /* Word 2 */ 167 uint8_t reln; /* relation (0=eq, '>'=gt, etc) */ 168 uint8_t vallen; /* length of string value, if any */ 169 uint8_t type; /* comparison type (FILE_*) */ 170 uint8_t in_type; /* type of indirection */ 171 #define FILE_INVALID 0 172 #define FILE_BYTE 1 173 #define FILE_SHORT 2 174 #define FILE_DEFAULT 3 175 #define FILE_LONG 4 176 #define FILE_STRING 5 177 #define FILE_DATE 6 178 #define FILE_BESHORT 7 179 #define FILE_BELONG 8 180 #define FILE_BEDATE 9 181 #define FILE_LESHORT 10 182 #define FILE_LELONG 11 183 #define FILE_LEDATE 12 184 #define FILE_PSTRING 13 185 #define FILE_LDATE 14 186 #define FILE_BELDATE 15 187 #define FILE_LELDATE 16 188 #define FILE_REGEX 17 189 #define FILE_BESTRING16 18 190 #define FILE_LESTRING16 19 191 #define FILE_SEARCH 20 192 #define FILE_MEDATE 21 193 #define FILE_MELDATE 22 194 #define FILE_MELONG 23 195 #define FILE_QUAD 24 196 #define FILE_LEQUAD 25 197 #define FILE_BEQUAD 26 198 #define FILE_QDATE 27 199 #define FILE_LEQDATE 28 200 #define FILE_BEQDATE 29 201 #define FILE_QLDATE 30 202 #define FILE_LEQLDATE 31 203 #define FILE_BEQLDATE 32 204 #define FILE_FLOAT 33 205 #define FILE_BEFLOAT 34 206 #define FILE_LEFLOAT 35 207 #define FILE_DOUBLE 36 208 #define FILE_BEDOUBLE 37 209 #define FILE_LEDOUBLE 38 210 #define FILE_BEID3 39 211 #define FILE_LEID3 40 212 #define FILE_INDIRECT 41 213 #define FILE_NAMES_SIZE 42/* size of array to contain all names */ 214 215 #define IS_LIBMAGIC_STRING(t) \ 216 ((t) == FILE_STRING || \ 217 (t) == FILE_PSTRING || \ 218 (t) == FILE_BESTRING16 || \ 219 (t) == FILE_LESTRING16 || \ 220 (t) == FILE_REGEX || \ 221 (t) == FILE_SEARCH || \ 222 (t) == FILE_DEFAULT) 223 224 #define FILE_FMT_NONE 0 225 #define FILE_FMT_NUM 1 /* "cduxXi" */ 226 #define FILE_FMT_STR 2 /* "s" */ 227 #define FILE_FMT_QUAD 3 /* "ll" */ 228 #define FILE_FMT_FLOAT 4 /* "eEfFgG" */ 229 #define FILE_FMT_DOUBLE 5 /* "eEfFgG" */ 230 231 /* Word 3 */ 232 uint8_t in_op; /* operator for indirection */ 233 uint8_t mask_op; /* operator for mask */ 234 #ifdef ENABLE_CONDITIONALS 235 uint8_t cond; /* conditional type */ 236 #else 237 uint8_t dummy; 238 #endif 239 uint8_t factor_op; 240 #define FILE_FACTOR_OP_PLUS '+' 241 #define FILE_FACTOR_OP_MINUS '-' 242 #define FILE_FACTOR_OP_TIMES '*' 243 #define FILE_FACTOR_OP_DIV '/' 244 #define FILE_FACTOR_OP_NONE '\0' 245 246 #define FILE_OPS "&|^+-*/%" 247 #define FILE_OPAND 0 248 #define FILE_OPOR 1 249 #define FILE_OPXOR 2 250 #define FILE_OPADD 3 251 #define FILE_OPMINUS 4 252 #define FILE_OPMULTIPLY 5 253 #define FILE_OPDIVIDE 6 254 #define FILE_OPMODULO 7 255 #define FILE_OPS_MASK 0x07 /* mask for above ops */ 256 #define FILE_UNUSED_1 0x08 257 #define FILE_UNUSED_2 0x10 258 #define FILE_UNUSED_3 0x20 259 #define FILE_OPINVERSE 0x40 260 #define FILE_OPINDIRECT 0x80 261 262 #ifdef ENABLE_CONDITIONALS 263 #define COND_NONE 0 264 #define COND_IF 1 265 #define COND_ELIF 2 266 #define COND_ELSE 3 267 #endif /* ENABLE_CONDITIONALS */ 268 269 /* Word 4 */ 270 uint32_t offset; /* offset to magic number */ 271 /* Word 5 */ 272 int32_t in_offset; /* offset from indirection */ 273 /* Word 6 */ 274 uint32_t lineno; /* line number in magic file */ 275 /* Word 7,8 */ 276 union { 277 uint64_t _mask; /* for use with numeric and date types */ 278 struct { 279 uint32_t _count; /* repeat/line count */ 280 uint32_t _flags; /* modifier flags */ 281 } _s; /* for use with string types */ 282 } _u; 283 #define num_mask _u._mask 284 #define str_range _u._s._count 285 #define str_flags _u._s._flags 286 /* Words 9-16 */ 287 union VALUETYPE value; /* either number or string */ 288 /* Words 17-32 */ 289 char desc[MAXDESC]; /* description */ 290 /* Words 33-48 */ 291 char mimetype[MAXDESC]; /* MIME type */ 292 /* Words 49-50 */ 293 char apple[8]; 294 }; 295 296 #define BIT(A) (1 << (A)) 297 #define STRING_COMPACT_WHITESPACE BIT(0) 298 #define STRING_COMPACT_OPTIONAL_WHITESPACE BIT(1) 299 #define STRING_IGNORE_LOWERCASE BIT(2) 300 #define STRING_IGNORE_UPPERCASE BIT(3) 301 #define REGEX_OFFSET_START BIT(4) 302 #define STRING_TEXTTEST BIT(5) 303 #define STRING_BINTEST BIT(6) 304 #define PSTRING_1_BE BIT(7) 305 #define PSTRING_1_LE BIT(7) 306 #define PSTRING_2_BE BIT(8) 307 #define PSTRING_2_LE BIT(9) 308 #define PSTRING_4_BE BIT(10) 309 #define PSTRING_4_LE BIT(11) 310 #define PSTRING_LEN \ 311 (PSTRING_1_BE|PSTRING_2_LE|PSTRING_2_BE|PSTRING_4_LE|PSTRING_4_BE) 312 #define PSTRING_LENGTH_INCLUDES_ITSELF BIT(12) 313 #define CHAR_COMPACT_WHITESPACE 'W' 314 #define CHAR_COMPACT_OPTIONAL_WHITESPACE 'w' 315 #define CHAR_IGNORE_LOWERCASE 'c' 316 #define CHAR_IGNORE_UPPERCASE 'C' 317 #define CHAR_REGEX_OFFSET_START 's' 318 #define CHAR_TEXTTEST 't' 319 #define CHAR_BINTEST 'b' 320 #define CHAR_PSTRING_1_BE 'B' 321 #define CHAR_PSTRING_1_LE 'B' 322 #define CHAR_PSTRING_2_BE 'H' 323 #define CHAR_PSTRING_2_LE 'h' 324 #define CHAR_PSTRING_4_BE 'L' 325 #define CHAR_PSTRING_4_LE 'l' 326 #define CHAR_PSTRING_LENGTH_INCLUDES_ITSELF 'J' 327 #define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE) 328 #define STRING_DEFAULT_RANGE 100 329 330 331 /* list of magic entries */ 332 struct mlist { 333 struct magic *magic; /* array of magic entries */ 334 uint32_t nmagic; /* number of entries in array */ 335 int mapped; /* allocation type: 0 => apprentice_file 336 * 1 => apprentice_map + malloc 337 * 2 => apprentice_map + mmap */ 338 struct mlist *next, *prev; 339 }; 340 341 #ifdef __cplusplus 342 #define CAST(T, b) static_cast<T>(b) 343 #define RCAST(T, b) reinterpret_cast<T>(b) 344 #else 345 #define CAST(T, b) (T)(b) 346 #define RCAST(T, b) (T)(b) 347 #endif 348 349 struct level_info { 350 int32_t off; 351 int got_match; 352 #ifdef ENABLE_CONDITIONALS 353 int last_match; 354 int last_cond; /* used for error checking by parse() */ 355 #endif 356 }; 357 struct magic_set { 358 struct mlist *mlist; 359 struct cont { 360 size_t len; 361 struct level_info *li; 362 } c; 363 struct out { 364 char *buf; /* Accumulation buffer */ 365 char *pbuf; /* Printable buffer */ 366 } o; 367 uint32_t offset; 368 int error; 369 int flags; /* Control magic tests. */ 370 int event_flags; /* Note things that happened. */ 371 #define EVENT_HAD_ERR 0x01 372 const char *file; 373 size_t line; /* current magic line number */ 374 375 /* data for searches */ 376 struct { 377 const char *s; /* start of search in original source */ 378 size_t s_len; /* length of search region */ 379 size_t offset; /* starting offset in source: XXX - should this be off_t? */ 380 size_t rm_len; /* match length */ 381 } search; 382 383 /* FIXME: Make the string dynamically allocated so that e.g. 384 strings matched in files can be longer than MAXstring */ 385 union VALUETYPE ms_value; /* either number or string */ 386 }; 387 388 /* Type for Unicode characters */ 389 typedef unsigned long unichar; 390 391 struct stat; 392 protected const char *file_fmttime(uint32_t, int); 393 protected int file_buffer(struct magic_set *, php_stream *, const char *, const void *, 394 size_t); 395 protected int file_fsmagic(struct magic_set *ms, const char *fn, struct stat *sb, php_stream *stream); 396 protected int file_pipe2file(struct magic_set *, int, const void *, size_t); 397 protected int file_replace(struct magic_set *, const char *, const char *); 398 protected int file_printf(struct magic_set *, const char *, ...); 399 protected int file_reset(struct magic_set *); 400 protected int file_tryelf(struct magic_set *, int, const unsigned char *, 401 size_t); 402 protected int file_trycdf(struct magic_set *, int, const unsigned char *, 403 size_t); 404 #ifdef PHP_FILEINFO_UNCOMPRESS 405 protected int file_zmagic(struct magic_set *, int, const char *, 406 const unsigned char *, size_t); 407 #endif 408 protected int file_ascmagic(struct magic_set *, const unsigned char *, size_t, 409 int); 410 protected int file_ascmagic_with_encoding(struct magic_set *, 411 const unsigned char *, size_t, unichar *, size_t, const char *, 412 const char *, int); 413 protected int file_encoding(struct magic_set *, const unsigned char *, size_t, 414 unichar **, size_t *, const char **, const char **, const char **); 415 protected int file_is_tar(struct magic_set *, const unsigned char *, size_t); 416 protected int file_softmagic(struct magic_set *, const unsigned char *, size_t, 417 int, int); 418 protected struct mlist *file_apprentice(struct magic_set *, const char *, int); 419 protected uint64_t file_signextend(struct magic_set *, struct magic *, 420 uint64_t); 421 protected void file_delmagic(struct magic *, int type, size_t entries); 422 protected void file_badread(struct magic_set *); 423 protected void file_badseek(struct magic_set *); 424 protected void file_oomem(struct magic_set *, size_t); 425 protected void file_error(struct magic_set *, int, const char *, ...); 426 protected void file_magerror(struct magic_set *, const char *, ...); 427 protected void file_magwarn(struct magic_set *, const char *, ...); 428 protected void file_showstr(FILE *, const char *, size_t); 429 protected size_t file_mbswidth(const char *); 430 protected const char *file_getbuffer(struct magic_set *); 431 protected ssize_t sread(int, void *, size_t, int); 432 protected int file_check_mem(struct magic_set *, unsigned int); 433 protected int file_looks_utf8(const unsigned char *, size_t, unichar *, 434 size_t *); 435 protected size_t file_pstring_length_size(const struct magic *); 436 protected size_t file_pstring_get_length(const struct magic *, const char *); 437 protected size_t file_printedlen(const struct magic_set *ms); 438 #ifdef __EMX__ 439 protected int file_os2_apptype(struct magic_set *, const char *, const void *, 440 size_t); 441 #endif /* __EMX__ */ 442 443 extern const char *file_names[]; 444 extern const size_t file_nnames; 445 446 #ifndef HAVE_STRERROR 447 extern int sys_nerr; 448 extern char *sys_errlist[]; 449 #define strerror(e) \ 450 (((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error") 451 #endif 452 453 #ifndef HAVE_STRTOUL 454 #define strtoul(a, b, c) strtol(a, b, c) 455 #endif 456 457 #ifndef strlcpy 458 size_t strlcpy(char *dst, const char *src, size_t siz); 459 #endif 460 #ifndef strlcat 461 size_t strlcat(char *dst, const char *src, size_t siz); 462 #endif 463 #ifndef HAVE_GETLINE 464 ssize_t getline(char **dst, size_t *len, FILE *fp); 465 ssize_t getdelim(char **dst, size_t *len, int delimiter, FILE *fp); 466 #endif 467 468 #if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK) 469 #define QUICK 470 #endif 471 472 #ifndef O_BINARY 473 #define O_BINARY 0 474 #endif 475 476 #ifndef __cplusplus 477 #if defined(__GNUC__) && (__GNUC__ >= 3) 478 #define FILE_RCSID(id) \ 479 static const char rcsid[] __attribute__((__used__)) = id; 480 #else 481 #define FILE_RCSID(id) \ 482 static const char *rcsid(const char *p) { \ 483 return rcsid(p = id); \ 484 } 485 #endif 486 #else 487 #define FILE_RCSID(id) 488 #endif 489 490 #ifdef PHP_WIN32 491 #define FINFO_LSEEK_FUNC _lseek 492 #define FINFO_READ_FUNC _read 493 #else 494 #define FINFO_LSEEK_FUNC lseek 495 #define FINFO_READ_FUNC read 496 #endif 497 498 #endif /* __file_h__ */ 499