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 1.206 2019/05/07 02:27:11 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 #define INTMAX_T_FORMAT "I64" 46 #else 47 #define SIZE_T_FORMAT "z" 48 #define INT64_T_FORMAT "ll" 49 #define INTMAX_T_FORMAT "j" 50 #endif 51 52 #include <stdio.h> /* Include that here, to make sure __P gets defined */ 53 #include <errno.h> 54 #include <fcntl.h> /* For open and flags */ 55 #ifdef HAVE_STDINT_H 56 #ifndef __STDC_LIMIT_MACROS 57 #define __STDC_LIMIT_MACROS 58 #endif 59 #ifndef __STDC_FORMAT_MACROS 60 #define __STDC_FORMAT_MACROS 61 #endif 62 #include <stdint.h> 63 #endif 64 #ifdef HAVE_INTTYPES_H 65 #include <inttypes.h> 66 #endif 67 #ifdef PHP_WIN32 68 #include "win32/php_stdint.h" 69 #endif 70 71 #include "php.h" 72 #include "ext/standard/php_string.h" 73 #include "ext/pcre/php_pcre.h" 74 75 #include <sys/types.h> 76 #ifdef PHP_WIN32 77 #include "win32/param.h" 78 #else 79 #include <sys/param.h> 80 #endif 81 /* Do this here and now, because struct stat gets re-defined on solaris */ 82 #include <sys/stat.h> 83 #include <stdarg.h> 84 85 #define abort() zend_error_noreturn(E_ERROR, "fatal libmagic error") 86 87 #define ENABLE_CONDITIONALS 88 89 #ifndef MAGIC 90 #define MAGIC "/etc/magic" 91 #endif 92 93 #if defined(__EMX__) || defined(PHP_WIN32) 94 #define PATHSEP ';' 95 #else 96 #define PATHSEP ':' 97 #endif 98 99 #define private static 100 101 #if HAVE_VISIBILITY && !defined(WIN32) 102 #define public __attribute__ ((__visibility__("default"))) 103 #ifndef protected 104 #define protected __attribute__ ((__visibility__("hidden"))) 105 #endif 106 #else 107 #define public 108 #ifndef protected 109 #define protected 110 #endif 111 #endif 112 113 #ifndef __arraycount 114 #define __arraycount(a) (sizeof(a) / sizeof(a[0])) 115 #endif 116 117 #ifndef __GNUC_PREREQ__ 118 #ifdef __GNUC__ 119 #define __GNUC_PREREQ__(x, y) \ 120 ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \ 121 (__GNUC__ > (x))) 122 #else 123 #define __GNUC_PREREQ__(x, y) 0 124 #endif 125 #endif 126 127 #ifndef MIN 128 #define MIN(a,b) (((a) < (b)) ? (a) : (b)) 129 #endif 130 131 #ifndef MAX 132 #define MAX(a,b) (((a) > (b)) ? (a) : (b)) 133 #endif 134 135 #ifndef FILE_BYTES_MAX 136 # define FILE_BYTES_MAX (1024 * 1024) /* how much of the file to look at */ 137 #endif 138 #define MAXMAGIS 8192 /* max entries in any one magic file 139 or directory */ 140 #define MAXDESC 64 /* max len of text description/MIME type */ 141 #define MAXMIME 80 /* max len of text MIME type */ 142 #define MAXstring 96 /* max len of "string" types */ 143 144 #define MAGICNO 0xF11E041C 145 #define VERSIONNO 14 146 #define FILE_MAGICSIZE 344 147 148 #define FILE_LOAD 0 149 #define FILE_CHECK 1 150 #define FILE_COMPILE 2 151 #define FILE_LIST 3 152 153 struct buffer { 154 int fd; 155 zend_stat_t st; 156 const void *fbuf; 157 size_t flen; 158 zend_off_t eoff; 159 void *ebuf; 160 size_t elen; 161 }; 162 163 union VALUETYPE { 164 uint8_t b; 165 uint16_t h; 166 uint32_t l; 167 uint64_t q; 168 uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */ 169 uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */ 170 uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */ 171 char s[MAXstring]; /* the search string or regex pattern */ 172 unsigned char us[MAXstring]; 173 float f; 174 double d; 175 }; 176 177 struct magic { 178 /* Word 1 */ 179 uint16_t cont_level; /* level of ">" */ 180 uint8_t flag; 181 #define INDIR 0x01 /* if '(...)' appears */ 182 #define OFFADD 0x02 /* if '>&' or '>...(&' appears */ 183 #define INDIROFFADD 0x04 /* if '>&(' appears */ 184 #define UNSIGNED 0x08 /* comparison is unsigned */ 185 #define NOSPACE 0x10 /* suppress space character before output */ 186 #define BINTEST 0x20 /* test is for a binary type (set only 187 for top-level tests) */ 188 #define TEXTTEST 0x40 /* for passing to file_softmagic */ 189 190 uint8_t factor; 191 192 /* Word 2 */ 193 uint8_t reln; /* relation (0=eq, '>'=gt, etc) */ 194 uint8_t vallen; /* length of string value, if any */ 195 uint8_t type; /* comparison type (FILE_*) */ 196 uint8_t in_type; /* type of indirection */ 197 #define FILE_INVALID 0 198 #define FILE_BYTE 1 199 #define FILE_SHORT 2 200 #define FILE_DEFAULT 3 201 #define FILE_LONG 4 202 #define FILE_STRING 5 203 #define FILE_DATE 6 204 #define FILE_BESHORT 7 205 #define FILE_BELONG 8 206 #define FILE_BEDATE 9 207 #define FILE_LESHORT 10 208 #define FILE_LELONG 11 209 #define FILE_LEDATE 12 210 #define FILE_PSTRING 13 211 #define FILE_LDATE 14 212 #define FILE_BELDATE 15 213 #define FILE_LELDATE 16 214 #define FILE_REGEX 17 215 #define FILE_BESTRING16 18 216 #define FILE_LESTRING16 19 217 #define FILE_SEARCH 20 218 #define FILE_MEDATE 21 219 #define FILE_MELDATE 22 220 #define FILE_MELONG 23 221 #define FILE_QUAD 24 222 #define FILE_LEQUAD 25 223 #define FILE_BEQUAD 26 224 #define FILE_QDATE 27 225 #define FILE_LEQDATE 28 226 #define FILE_BEQDATE 29 227 #define FILE_QLDATE 30 228 #define FILE_LEQLDATE 31 229 #define FILE_BEQLDATE 32 230 #define FILE_FLOAT 33 231 #define FILE_BEFLOAT 34 232 #define FILE_LEFLOAT 35 233 #define FILE_DOUBLE 36 234 #define FILE_BEDOUBLE 37 235 #define FILE_LEDOUBLE 38 236 #define FILE_BEID3 39 237 #define FILE_LEID3 40 238 #define FILE_INDIRECT 41 239 #define FILE_QWDATE 42 240 #define FILE_LEQWDATE 43 241 #define FILE_BEQWDATE 44 242 #define FILE_NAME 45 243 #define FILE_USE 46 244 #define FILE_CLEAR 47 245 #define FILE_DER 48 246 #define FILE_NAMES_SIZE 49 /* size of array to contain all names */ 247 248 #define IS_LIBMAGIC_STRING(t) \ 249 ((t) == FILE_STRING || \ 250 (t) == FILE_PSTRING || \ 251 (t) == FILE_BESTRING16 || \ 252 (t) == FILE_LESTRING16 || \ 253 (t) == FILE_REGEX || \ 254 (t) == FILE_SEARCH || \ 255 (t) == FILE_INDIRECT || \ 256 (t) == FILE_NAME || \ 257 (t) == FILE_USE) 258 259 #define FILE_FMT_NONE 0 260 #define FILE_FMT_NUM 1 /* "cduxXi" */ 261 #define FILE_FMT_STR 2 /* "s" */ 262 #define FILE_FMT_QUAD 3 /* "ll" */ 263 #define FILE_FMT_FLOAT 4 /* "eEfFgG" */ 264 #define FILE_FMT_DOUBLE 5 /* "eEfFgG" */ 265 266 /* Word 3 */ 267 uint8_t in_op; /* operator for indirection */ 268 uint8_t mask_op; /* operator for mask */ 269 #ifdef ENABLE_CONDITIONALS 270 uint8_t cond; /* conditional type */ 271 #else 272 uint8_t dummy; 273 #endif 274 uint8_t factor_op; 275 #define FILE_FACTOR_OP_PLUS '+' 276 #define FILE_FACTOR_OP_MINUS '-' 277 #define FILE_FACTOR_OP_TIMES '*' 278 #define FILE_FACTOR_OP_DIV '/' 279 #define FILE_FACTOR_OP_NONE '\0' 280 281 #define FILE_OPS "&|^+-*/%" 282 #define FILE_OPAND 0 283 #define FILE_OPOR 1 284 #define FILE_OPXOR 2 285 #define FILE_OPADD 3 286 #define FILE_OPMINUS 4 287 #define FILE_OPMULTIPLY 5 288 #define FILE_OPDIVIDE 6 289 #define FILE_OPMODULO 7 290 #define FILE_OPS_MASK 0x07 /* mask for above ops */ 291 #define FILE_UNUSED_1 0x08 292 #define FILE_UNUSED_2 0x10 293 #define FILE_OPSIGNED 0x20 294 #define FILE_OPINVERSE 0x40 295 #define FILE_OPINDIRECT 0x80 296 297 #ifdef ENABLE_CONDITIONALS 298 #define COND_NONE 0 299 #define COND_IF 1 300 #define COND_ELIF 2 301 #define COND_ELSE 3 302 #endif /* ENABLE_CONDITIONALS */ 303 304 /* Word 4 */ 305 int32_t offset; /* offset to magic number */ 306 /* Word 5 */ 307 int32_t in_offset; /* offset from indirection */ 308 /* Word 6 */ 309 uint32_t lineno; /* line number in magic file */ 310 /* Word 7,8 */ 311 union { 312 uint64_t _mask; /* for use with numeric and date types */ 313 struct { 314 uint32_t _count; /* repeat/line count */ 315 uint32_t _flags; /* modifier flags */ 316 } _s; /* for use with string types */ 317 } _u; 318 #define num_mask _u._mask 319 #define str_range _u._s._count 320 #define str_flags _u._s._flags 321 /* Words 9-24 */ 322 union VALUETYPE value; /* either number or string */ 323 /* Words 25-40 */ 324 char desc[MAXDESC]; /* description */ 325 /* Words 41-60 */ 326 char mimetype[MAXMIME]; /* MIME type */ 327 /* Words 61-62 */ 328 char apple[8]; /* APPLE CREATOR/TYPE */ 329 /* Words 63-78 */ 330 char ext[64]; /* Popular extensions */ 331 }; 332 333 #define BIT(A) (1 << (A)) 334 #define STRING_COMPACT_WHITESPACE BIT(0) 335 #define STRING_COMPACT_OPTIONAL_WHITESPACE BIT(1) 336 #define STRING_IGNORE_LOWERCASE BIT(2) 337 #define STRING_IGNORE_UPPERCASE BIT(3) 338 #define REGEX_OFFSET_START BIT(4) 339 #define STRING_TEXTTEST BIT(5) 340 #define STRING_BINTEST BIT(6) 341 #define PSTRING_1_BE BIT(7) 342 #define PSTRING_1_LE BIT(7) 343 #define PSTRING_2_BE BIT(8) 344 #define PSTRING_2_LE BIT(9) 345 #define PSTRING_4_BE BIT(10) 346 #define PSTRING_4_LE BIT(11) 347 #define REGEX_LINE_COUNT BIT(11) 348 #define PSTRING_LEN \ 349 (PSTRING_1_BE|PSTRING_2_LE|PSTRING_2_BE|PSTRING_4_LE|PSTRING_4_BE) 350 #define PSTRING_LENGTH_INCLUDES_ITSELF BIT(12) 351 #define STRING_TRIM BIT(13) 352 #define CHAR_COMPACT_WHITESPACE 'W' 353 #define CHAR_COMPACT_OPTIONAL_WHITESPACE 'w' 354 #define CHAR_IGNORE_LOWERCASE 'c' 355 #define CHAR_IGNORE_UPPERCASE 'C' 356 #define CHAR_REGEX_OFFSET_START 's' 357 #define CHAR_TEXTTEST 't' 358 #define CHAR_TRIM 'T' 359 #define CHAR_BINTEST 'b' 360 #define CHAR_PSTRING_1_BE 'B' 361 #define CHAR_PSTRING_1_LE 'B' 362 #define CHAR_PSTRING_2_BE 'H' 363 #define CHAR_PSTRING_2_LE 'h' 364 #define CHAR_PSTRING_4_BE 'L' 365 #define CHAR_PSTRING_4_LE 'l' 366 #define CHAR_PSTRING_LENGTH_INCLUDES_ITSELF 'J' 367 #define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE) 368 #define STRING_DEFAULT_RANGE 100 369 370 #define INDIRECT_RELATIVE BIT(0) 371 #define CHAR_INDIRECT_RELATIVE 'r' 372 373 /* list of magic entries */ 374 struct mlist { 375 struct magic *magic; /* array of magic entries */ 376 uint32_t nmagic; /* number of entries in array */ 377 void *map; /* internal resources used by entry */ 378 struct mlist *next, *prev; 379 }; 380 381 #ifdef __cplusplus 382 #define CAST(T, b) static_cast<T>(b) 383 #define RCAST(T, b) reinterpret_cast<T>(b) 384 #define CCAST(T, b) const_cast<T>(b) 385 #else 386 #define CAST(T, b) ((T)(b)) 387 #define RCAST(T, b) ((T)(uintptr_t)(b)) 388 #define CCAST(T, b) ((T)(uintptr_t)(b)) 389 #endif 390 391 struct level_info { 392 int32_t off; 393 int got_match; 394 #ifdef ENABLE_CONDITIONALS 395 int last_match; 396 int last_cond; /* used for error checking by parse() */ 397 #endif 398 }; 399 400 #define MAGIC_SETS 2 401 402 struct magic_set { 403 struct mlist *mlist[MAGIC_SETS]; /* list of regular entries */ 404 struct cont { 405 size_t len; 406 struct level_info *li; 407 } c; 408 struct out { 409 char *buf; /* Accumulation buffer */ 410 char *pbuf; /* Printable buffer */ 411 } o; 412 uint32_t offset; /* a copy of m->offset while we */ 413 /* are working on the magic entry */ 414 uint32_t eoffset; /* offset from end of file */ 415 int error; 416 int flags; /* Control magic tests. */ 417 int event_flags; /* Note things that happened. */ 418 #define EVENT_HAD_ERR 0x01 419 const char *file; 420 size_t line; /* current magic line number */ 421 mode_t mode; /* copy of current stat mode */ 422 423 /* data for searches */ 424 struct { 425 const char *s; /* start of search in original source */ 426 size_t s_len; /* length of search region */ 427 size_t offset; /* starting offset in source: XXX - should this be off_t? */ 428 size_t rm_len; /* match length */ 429 } search; 430 431 /* FIXME: Make the string dynamically allocated so that e.g. 432 strings matched in files can be longer than MAXstring */ 433 union VALUETYPE ms_value; /* either number or string */ 434 uint16_t indir_max; 435 uint16_t name_max; 436 uint16_t elf_shnum_max; 437 uint16_t elf_phnum_max; 438 uint16_t elf_notes_max; 439 uint16_t regex_max; 440 size_t bytes_max; /* number of bytes to read from file */ 441 #define FILE_INDIR_MAX 50 442 #define FILE_NAME_MAX 30 443 #define FILE_ELF_SHNUM_MAX 32768 444 #define FILE_ELF_PHNUM_MAX 2048 445 #define FILE_ELF_NOTES_MAX 256 446 #define FILE_REGEX_MAX 8192 447 }; 448 449 /* Type for Unicode characters */ 450 typedef unsigned long unichar; 451 452 #define FILE_T_LOCAL 1 453 #define FILE_T_WINDOWS 2 454 protected const char *file_fmttime(uint64_t, int, char *); 455 protected struct magic_set *file_ms_alloc(int); 456 protected void file_ms_free(struct magic_set *); 457 protected int file_buffer(struct magic_set *, php_stream *, zend_stat_t *, const char *, const void *, 458 size_t); 459 protected int file_fsmagic(struct magic_set *, const char *, zend_stat_t *); 460 protected int file_pipe2file(struct magic_set *, int, const void *, size_t); 461 protected int file_separator(struct magic_set *); 462 protected size_t file_printedlen(const struct magic_set *); 463 protected int file_replace(struct magic_set *, const char *, const char *); 464 protected int file_printf(struct magic_set *, const char *, ...); 465 protected int file_reset(struct magic_set *, int); 466 protected int file_tryelf(struct magic_set *, const struct buffer *); 467 protected int file_trycdf(struct magic_set *, const struct buffer *); 468 #ifdef PHP_FILEINFO_UNCOMPRESS 469 protected int file_zmagic(struct magic_set *, const struct buffer *, 470 const char *); 471 #endif 472 protected int file_ascmagic(struct magic_set *, const struct buffer *, 473 int); 474 protected int file_ascmagic_with_encoding(struct magic_set *, 475 const struct buffer *, unichar *, size_t, const char *, const char *, int); 476 protected int file_encoding(struct magic_set *, const struct buffer *, 477 unichar **, size_t *, const char **, const char **, const char **); 478 protected int file_is_json(struct magic_set *, const struct buffer *); 479 protected int file_is_tar(struct magic_set *, const struct buffer *); 480 protected int file_softmagic(struct magic_set *, const struct buffer *, 481 uint16_t *, uint16_t *, int, int); 482 protected int file_apprentice(struct magic_set *, const char *, int); 483 protected int buffer_apprentice(struct magic_set *, struct magic **, 484 size_t *, size_t); 485 protected int file_magicfind(struct magic_set *, const char *, struct mlist *); 486 protected uint64_t file_signextend(struct magic_set *, struct magic *, 487 uint64_t); 488 protected void file_badread(struct magic_set *); 489 protected void file_badseek(struct magic_set *); 490 protected void file_oomem(struct magic_set *, size_t); 491 protected void file_error(struct magic_set *, int, const char *, ...); 492 protected void file_magerror(struct magic_set *, const char *, ...); 493 protected void file_magwarn(struct magic_set *, const char *, ...); 494 protected void file_showstr(FILE *, const char *, size_t); 495 protected size_t file_mbswidth(const char *); 496 protected const char *file_getbuffer(struct magic_set *); 497 protected ssize_t sread(int, void *, size_t, int); 498 protected int file_check_mem(struct magic_set *, unsigned int); 499 protected int file_looks_utf8(const unsigned char *, size_t, unichar *, 500 size_t *); 501 protected size_t file_pstring_length_size(const struct magic *); 502 protected size_t file_pstring_get_length(const struct magic *, const char *); 503 protected char * file_printable(char *, size_t, const char *, size_t); 504 #ifdef __EMX__ 505 protected int file_os2_apptype(struct magic_set *, const char *, const void *, 506 size_t); 507 #endif /* __EMX__ */ 508 509 protected void buffer_init(struct buffer *, int, const zend_stat_t *, 510 const void *, size_t); 511 protected void buffer_fini(struct buffer *); 512 protected int buffer_fill(const struct buffer *); 513 514 public void 515 convert_libmagic_pattern(zval *pattern, char *val, size_t len, uint32_t options); 516 517 typedef struct { 518 char *buf; 519 uint32_t offset; 520 } file_pushbuf_t; 521 522 protected file_pushbuf_t *file_push_buffer(struct magic_set *); 523 protected char *file_pop_buffer(struct magic_set *, file_pushbuf_t *); 524 525 extern const char *file_names[]; 526 extern const size_t file_nnames; 527 528 #ifndef strlcpy 529 size_t strlcpy(char *, const char *, size_t); 530 #endif 531 #ifndef strlcat 532 size_t strlcat(char *, const char *, size_t); 533 #endif 534 #ifndef HAVE_STRCASESTR 535 char *strcasestr(const char *, const char *); 536 #endif 537 #ifndef HAVE_GETLINE 538 ssize_t getline(char **, size_t *, FILE *); 539 ssize_t getdelim(char **, size_t *, int, FILE *); 540 #endif 541 #ifndef HAVE_CTIME_R 542 char *ctime_r(const time_t *, char *); 543 #endif 544 #ifndef HAVE_ASCTIME_R 545 char *asctime_r(const struct tm *, char *); 546 #endif 547 548 #if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK) 549 #define QUICK 550 #endif 551 552 #ifndef O_BINARY 553 #define O_BINARY 0 554 #endif 555 #ifndef O_NONBLOCK 556 #define O_NONBLOCK 0 557 #endif 558 559 #ifndef __cplusplus 560 #if defined(__GNUC__) && (__GNUC__ >= 3) 561 #define FILE_RCSID(id) \ 562 static const char rcsid[] __attribute__((__used__)) = id; 563 #else 564 #define FILE_RCSID(id) \ 565 static const char *rcsid(const char *p) { \ 566 return rcsid(p = id); \ 567 } 568 #endif 569 #else 570 #define FILE_RCSID(id) 571 #endif 572 573 #ifdef PHP_WIN32 574 #ifdef _WIN64 575 #define FINFO_LSEEK_FUNC _lseeki64 576 #else 577 #define FINFO_LSEEK_FUNC _lseek 578 #endif 579 #define FINFO_READ_FUNC _read 580 #else 581 #define FINFO_LSEEK_FUNC lseek 582 #define FINFO_READ_FUNC read 583 #endif 584 #ifndef __RCSID 585 #define __RCSID(a) 586 #endif 587 588 #endif /* __file_h__ */ 589