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 1.182 2017/04/07 19:46:44 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 #include <stdint.h> 60 #endif 61 #ifdef HAVE_INTTYPES_H 62 #include <inttypes.h> 63 #endif 64 #ifdef PHP_WIN32 65 #include "win32/php_stdint.h" 66 #endif 67 68 #include "php.h" 69 #include "ext/standard/php_string.h" 70 #include "ext/pcre/php_pcre.h" 71 72 #include <sys/types.h> 73 #ifdef PHP_WIN32 74 #include "win32/param.h" 75 #else 76 #include <sys/param.h> 77 #endif 78 /* Do this here and now, because struct stat gets re-defined on solaris */ 79 #include <sys/stat.h> 80 #include <stdarg.h> 81 82 #define ENABLE_CONDITIONALS 83 84 #ifndef MAGIC 85 #define MAGIC "/etc/magic" 86 #endif 87 88 #if defined(__EMX__) || defined(PHP_WIN32) 89 #define PATHSEP ';' 90 #else 91 #define PATHSEP ':' 92 #endif 93 94 #define private static 95 96 #if HAVE_VISIBILITY && !defined(WIN32) 97 #define public __attribute__ ((__visibility__("default"))) 98 #ifndef protected 99 #define protected __attribute__ ((__visibility__("hidden"))) 100 #endif 101 #else 102 #define public 103 #ifndef protected 104 #define protected 105 #endif 106 #endif 107 108 #ifndef __arraycount 109 #define __arraycount(a) (sizeof(a) / sizeof(a[0])) 110 #endif 111 112 #ifndef __GNUC_PREREQ__ 113 #ifdef __GNUC__ 114 #define __GNUC_PREREQ__(x, y) \ 115 ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \ 116 (__GNUC__ > (x))) 117 #else 118 #define __GNUC_PREREQ__(x, y) 0 119 #endif 120 #endif 121 122 #ifndef MIN 123 #define MIN(a,b) (((a) < (b)) ? (a) : (b)) 124 #endif 125 126 #ifndef MAX 127 #define MAX(a,b) (((a) > (b)) ? (a) : (b)) 128 #endif 129 130 #ifndef FILE_BYTES_MAX 131 # define FILE_BYTES_MAX (1024 * 1024) /* how much of the file to look at */ 132 #endif 133 #define MAXMAGIS 8192 /* max entries in any one magic file 134 or directory */ 135 #define MAXDESC 64 /* max len of text description/MIME type */ 136 #define MAXMIME 80 /* max len of text MIME type */ 137 #define MAXstring 96 /* max len of "string" types */ 138 139 #define MAGICNO 0xF11E041C 140 #define VERSIONNO 14 141 #define FILE_MAGICSIZE 344 142 143 #define FILE_LOAD 0 144 #define FILE_CHECK 1 145 #define FILE_COMPILE 2 146 #define FILE_LIST 3 147 148 struct buffer { 149 int fd; 150 zend_stat_t st; 151 const void *fbuf; 152 size_t flen; 153 zend_off_t eoff; 154 void *ebuf; 155 size_t elen; 156 }; 157 158 union VALUETYPE { 159 uint8_t b; 160 uint16_t h; 161 uint32_t l; 162 uint64_t q; 163 uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */ 164 uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */ 165 uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */ 166 char s[MAXstring]; /* the search string or regex pattern */ 167 unsigned char us[MAXstring]; 168 float f; 169 double d; 170 }; 171 172 struct magic { 173 /* Word 1 */ 174 uint16_t cont_level; /* level of ">" */ 175 uint8_t flag; 176 #define INDIR 0x01 /* if '(...)' appears */ 177 #define OFFADD 0x02 /* if '>&' or '>...(&' appears */ 178 #define INDIROFFADD 0x04 /* if '>&(' appears */ 179 #define UNSIGNED 0x08 /* comparison is unsigned */ 180 #define NOSPACE 0x10 /* suppress space character before output */ 181 #define BINTEST 0x20 /* test is for a binary type (set only 182 for top-level tests) */ 183 #define TEXTTEST 0x40 /* for passing to file_softmagic */ 184 185 uint8_t factor; 186 187 /* Word 2 */ 188 uint8_t reln; /* relation (0=eq, '>'=gt, etc) */ 189 uint8_t vallen; /* length of string value, if any */ 190 uint8_t type; /* comparison type (FILE_*) */ 191 uint8_t in_type; /* type of indirection */ 192 #define FILE_INVALID 0 193 #define FILE_BYTE 1 194 #define FILE_SHORT 2 195 #define FILE_DEFAULT 3 196 #define FILE_LONG 4 197 #define FILE_STRING 5 198 #define FILE_DATE 6 199 #define FILE_BESHORT 7 200 #define FILE_BELONG 8 201 #define FILE_BEDATE 9 202 #define FILE_LESHORT 10 203 #define FILE_LELONG 11 204 #define FILE_LEDATE 12 205 #define FILE_PSTRING 13 206 #define FILE_LDATE 14 207 #define FILE_BELDATE 15 208 #define FILE_LELDATE 16 209 #define FILE_REGEX 17 210 #define FILE_BESTRING16 18 211 #define FILE_LESTRING16 19 212 #define FILE_SEARCH 20 213 #define FILE_MEDATE 21 214 #define FILE_MELDATE 22 215 #define FILE_MELONG 23 216 #define FILE_QUAD 24 217 #define FILE_LEQUAD 25 218 #define FILE_BEQUAD 26 219 #define FILE_QDATE 27 220 #define FILE_LEQDATE 28 221 #define FILE_BEQDATE 29 222 #define FILE_QLDATE 30 223 #define FILE_LEQLDATE 31 224 #define FILE_BEQLDATE 32 225 #define FILE_FLOAT 33 226 #define FILE_BEFLOAT 34 227 #define FILE_LEFLOAT 35 228 #define FILE_DOUBLE 36 229 #define FILE_BEDOUBLE 37 230 #define FILE_LEDOUBLE 38 231 #define FILE_BEID3 39 232 #define FILE_LEID3 40 233 #define FILE_INDIRECT 41 234 #define FILE_QWDATE 42 235 #define FILE_LEQWDATE 43 236 #define FILE_BEQWDATE 44 237 #define FILE_NAME 45 238 #define FILE_USE 46 239 #define FILE_CLEAR 47 240 #define FILE_DER 48 241 #define FILE_NAMES_SIZE 49 /* size of array to contain all names */ 242 243 #define IS_LIBMAGIC_STRING(t) \ 244 ((t) == FILE_STRING || \ 245 (t) == FILE_PSTRING || \ 246 (t) == FILE_BESTRING16 || \ 247 (t) == FILE_LESTRING16 || \ 248 (t) == FILE_REGEX || \ 249 (t) == FILE_SEARCH || \ 250 (t) == FILE_INDIRECT || \ 251 (t) == FILE_NAME || \ 252 (t) == FILE_USE) 253 254 #define FILE_FMT_NONE 0 255 #define FILE_FMT_NUM 1 /* "cduxXi" */ 256 #define FILE_FMT_STR 2 /* "s" */ 257 #define FILE_FMT_QUAD 3 /* "ll" */ 258 #define FILE_FMT_FLOAT 4 /* "eEfFgG" */ 259 #define FILE_FMT_DOUBLE 5 /* "eEfFgG" */ 260 261 /* Word 3 */ 262 uint8_t in_op; /* operator for indirection */ 263 uint8_t mask_op; /* operator for mask */ 264 #ifdef ENABLE_CONDITIONALS 265 uint8_t cond; /* conditional type */ 266 #else 267 uint8_t dummy; 268 #endif 269 uint8_t factor_op; 270 #define FILE_FACTOR_OP_PLUS '+' 271 #define FILE_FACTOR_OP_MINUS '-' 272 #define FILE_FACTOR_OP_TIMES '*' 273 #define FILE_FACTOR_OP_DIV '/' 274 #define FILE_FACTOR_OP_NONE '\0' 275 276 #define FILE_OPS "&|^+-*/%" 277 #define FILE_OPAND 0 278 #define FILE_OPOR 1 279 #define FILE_OPXOR 2 280 #define FILE_OPADD 3 281 #define FILE_OPMINUS 4 282 #define FILE_OPMULTIPLY 5 283 #define FILE_OPDIVIDE 6 284 #define FILE_OPMODULO 7 285 #define FILE_OPS_MASK 0x07 /* mask for above ops */ 286 #define FILE_UNUSED_1 0x08 287 #define FILE_UNUSED_2 0x10 288 #define FILE_OPSIGNED 0x20 289 #define FILE_OPINVERSE 0x40 290 #define FILE_OPINDIRECT 0x80 291 292 #ifdef ENABLE_CONDITIONALS 293 #define COND_NONE 0 294 #define COND_IF 1 295 #define COND_ELIF 2 296 #define COND_ELSE 3 297 #endif /* ENABLE_CONDITIONALS */ 298 299 /* Word 4 */ 300 int32_t offset; /* offset to magic number */ 301 /* Word 5 */ 302 int32_t in_offset; /* offset from indirection */ 303 /* Word 6 */ 304 uint32_t lineno; /* line number in magic file */ 305 /* Word 7,8 */ 306 union { 307 uint64_t _mask; /* for use with numeric and date types */ 308 struct { 309 uint32_t _count; /* repeat/line count */ 310 uint32_t _flags; /* modifier flags */ 311 } _s; /* for use with string types */ 312 } _u; 313 #define num_mask _u._mask 314 #define str_range _u._s._count 315 #define str_flags _u._s._flags 316 /* Words 9-24 */ 317 union VALUETYPE value; /* either number or string */ 318 /* Words 25-40 */ 319 char desc[MAXDESC]; /* description */ 320 /* Words 41-60 */ 321 char mimetype[MAXMIME]; /* MIME type */ 322 /* Words 61-62 */ 323 char apple[8]; /* APPLE CREATOR/TYPE */ 324 /* Words 63-78 */ 325 char ext[64]; /* Popular extensions */ 326 }; 327 328 #define BIT(A) (1 << (A)) 329 #define STRING_COMPACT_WHITESPACE BIT(0) 330 #define STRING_COMPACT_OPTIONAL_WHITESPACE BIT(1) 331 #define STRING_IGNORE_LOWERCASE BIT(2) 332 #define STRING_IGNORE_UPPERCASE BIT(3) 333 #define REGEX_OFFSET_START BIT(4) 334 #define STRING_TEXTTEST BIT(5) 335 #define STRING_BINTEST BIT(6) 336 #define PSTRING_1_BE BIT(7) 337 #define PSTRING_1_LE BIT(7) 338 #define PSTRING_2_BE BIT(8) 339 #define PSTRING_2_LE BIT(9) 340 #define PSTRING_4_BE BIT(10) 341 #define PSTRING_4_LE BIT(11) 342 #define REGEX_LINE_COUNT BIT(11) 343 #define PSTRING_LEN \ 344 (PSTRING_1_BE|PSTRING_2_LE|PSTRING_2_BE|PSTRING_4_LE|PSTRING_4_BE) 345 #define PSTRING_LENGTH_INCLUDES_ITSELF BIT(12) 346 #define STRING_TRIM BIT(13) 347 #define CHAR_COMPACT_WHITESPACE 'W' 348 #define CHAR_COMPACT_OPTIONAL_WHITESPACE 'w' 349 #define CHAR_IGNORE_LOWERCASE 'c' 350 #define CHAR_IGNORE_UPPERCASE 'C' 351 #define CHAR_REGEX_OFFSET_START 's' 352 #define CHAR_TEXTTEST 't' 353 #define CHAR_TRIM 'T' 354 #define CHAR_BINTEST 'b' 355 #define CHAR_PSTRING_1_BE 'B' 356 #define CHAR_PSTRING_1_LE 'B' 357 #define CHAR_PSTRING_2_BE 'H' 358 #define CHAR_PSTRING_2_LE 'h' 359 #define CHAR_PSTRING_4_BE 'L' 360 #define CHAR_PSTRING_4_LE 'l' 361 #define CHAR_PSTRING_LENGTH_INCLUDES_ITSELF 'J' 362 #define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE) 363 #define STRING_DEFAULT_RANGE 100 364 365 #define INDIRECT_RELATIVE BIT(0) 366 #define CHAR_INDIRECT_RELATIVE 'r' 367 368 /* list of magic entries */ 369 struct mlist { 370 struct magic *magic; /* array of magic entries */ 371 uint32_t nmagic; /* number of entries in array */ 372 void *map; /* internal resources used by entry */ 373 struct mlist *next, *prev; 374 }; 375 376 #ifdef __cplusplus 377 #define CAST(T, b) static_cast<T>(b) 378 #define RCAST(T, b) reinterpret_cast<T>(b) 379 #define CCAST(T, b) const_cast<T>(b) 380 #else 381 #define CAST(T, b) ((T)(b)) 382 #define RCAST(T, b) ((T)(b)) 383 #define CCAST(T, b) ((T)(uintptr_t)(b)) 384 #endif 385 386 struct level_info { 387 int32_t off; 388 int got_match; 389 #ifdef ENABLE_CONDITIONALS 390 int last_match; 391 int last_cond; /* used for error checking by parse() */ 392 #endif 393 }; 394 395 #define MAGIC_SETS 2 396 397 struct magic_set { 398 struct mlist *mlist[MAGIC_SETS]; /* list of regular entries */ 399 struct cont { 400 size_t len; 401 struct level_info *li; 402 } c; 403 struct out { 404 char *buf; /* Accumulation buffer */ 405 char *pbuf; /* Printable buffer */ 406 } o; 407 uint32_t offset; /* a copy of m->offset while we */ 408 /* are working on the magic entry */ 409 uint32_t eoffset; /* offset from end of file */ 410 int error; 411 int flags; /* Control magic tests. */ 412 int event_flags; /* Note things that happened. */ 413 #define EVENT_HAD_ERR 0x01 414 const char *file; 415 size_t line; /* current magic line number */ 416 417 /* data for searches */ 418 struct { 419 const char *s; /* start of search in original source */ 420 size_t s_len; /* length of search region */ 421 size_t offset; /* starting offset in source: XXX - should this be off_t? */ 422 size_t rm_len; /* match length */ 423 } search; 424 425 /* FIXME: Make the string dynamically allocated so that e.g. 426 strings matched in files can be longer than MAXstring */ 427 union VALUETYPE ms_value; /* either number or string */ 428 uint16_t indir_max; 429 uint16_t name_max; 430 uint16_t elf_shnum_max; 431 uint16_t elf_phnum_max; 432 uint16_t elf_notes_max; 433 uint16_t regex_max; 434 size_t bytes_max; /* number of bytes to read from file */ 435 #define FILE_INDIR_MAX 50 436 #define FILE_NAME_MAX 30 437 #define FILE_ELF_SHNUM_MAX 32768 438 #define FILE_ELF_PHNUM_MAX 2048 439 #define FILE_ELF_NOTES_MAX 256 440 #define FILE_REGEX_MAX 8192 441 }; 442 443 /* Type for Unicode characters */ 444 typedef unsigned long unichar; 445 446 #define FILE_T_LOCAL 1 447 #define FILE_T_WINDOWS 2 448 protected const char *file_fmttime(uint64_t, int, char *); 449 protected struct magic_set *file_ms_alloc(int); 450 protected void file_ms_free(struct magic_set *); 451 protected int file_buffer(struct magic_set *, php_stream *, const char *, const void *, 452 size_t); 453 protected int file_fsmagic(struct magic_set *, const char *, zend_stat_t *, php_stream *); 454 protected int file_pipe2file(struct magic_set *, int, const void *, size_t); 455 protected size_t file_printedlen(const struct magic_set *); 456 protected int file_replace(struct magic_set *, const char *, const char *); 457 protected int file_printf(struct magic_set *, const char *, ...); 458 protected int file_reset(struct magic_set *, int); 459 protected int file_tryelf(struct magic_set *, const struct buffer *); 460 protected int file_trycdf(struct magic_set *, const struct buffer *); 461 #ifdef PHP_FILEINFO_UNCOMPRESS 462 protected int file_zmagic(struct magic_set *, const struct buffer *, 463 const char *); 464 #endif 465 protected int file_ascmagic(struct magic_set *, const struct buffer *, 466 int); 467 protected int file_ascmagic_with_encoding(struct magic_set *, 468 const struct buffer *, unichar *, size_t, const char *, const char *, int); 469 protected int file_encoding(struct magic_set *, const struct buffer *, 470 unichar **, size_t *, const char **, const char **, const char **); 471 protected int file_is_tar(struct magic_set *, const struct buffer *); 472 protected int file_softmagic(struct magic_set *, const struct buffer *, 473 uint16_t *, uint16_t *, int, int); 474 protected int file_apprentice(struct magic_set *, const char *, int); 475 protected int buffer_apprentice(struct magic_set *, struct magic **, 476 size_t *, size_t); 477 protected int file_magicfind(struct magic_set *, const char *, struct mlist *); 478 protected uint64_t file_signextend(struct magic_set *, struct magic *, 479 uint64_t); 480 protected void file_badread(struct magic_set *); 481 protected void file_badseek(struct magic_set *); 482 protected void file_oomem(struct magic_set *, size_t); 483 protected void file_error(struct magic_set *, int, const char *, ...); 484 protected void file_magerror(struct magic_set *, const char *, ...); 485 protected void file_magwarn(struct magic_set *, const char *, ...); 486 protected void file_showstr(FILE *, const char *, size_t); 487 protected size_t file_mbswidth(const char *); 488 protected const char *file_getbuffer(struct magic_set *); 489 protected ssize_t sread(int, void *, size_t, int); 490 protected int file_check_mem(struct magic_set *, unsigned int); 491 protected int file_looks_utf8(const unsigned char *, size_t, unichar *, 492 size_t *); 493 protected size_t file_pstring_length_size(const struct magic *); 494 protected size_t file_pstring_get_length(const struct magic *, const char *); 495 protected char * file_printable(char *, size_t, const char *); 496 #ifdef __EMX__ 497 protected int file_os2_apptype(struct magic_set *, const char *, const void *, 498 size_t); 499 #endif /* __EMX__ */ 500 501 protected void buffer_init(struct buffer *, int, const void *, size_t); 502 protected void buffer_fini(struct buffer *); 503 protected int buffer_fill(const struct buffer *); 504 505 public void 506 convert_libmagic_pattern(zval *pattern, char *val, size_t len, uint32_t options); 507 508 typedef struct { 509 char *buf; 510 uint32_t offset; 511 } file_pushbuf_t; 512 513 protected file_pushbuf_t *file_push_buffer(struct magic_set *); 514 protected char *file_pop_buffer(struct magic_set *, file_pushbuf_t *); 515 516 extern const char *file_names[]; 517 extern const size_t file_nnames; 518 519 #ifndef HAVE_STRERROR 520 extern int sys_nerr; 521 extern char *sys_errlist[]; 522 #define strerror(e) \ 523 (((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error") 524 #endif 525 526 #ifndef HAVE_STRTOUL 527 #define strtoul(a, b, c) strtol(a, b, c) 528 #endif 529 530 #ifndef strlcpy 531 size_t strlcpy(char *, const char *, size_t); 532 #endif 533 #ifndef strlcat 534 size_t strlcat(char *, const char *, size_t); 535 #endif 536 #ifndef HAVE_STRCASESTR 537 char *strcasestr(const char *, const char *); 538 #endif 539 #ifndef HAVE_GETLINE 540 ssize_t getline(char **, size_t *, FILE *); 541 ssize_t getdelim(char **, size_t *, int, FILE *); 542 #endif 543 #ifndef HAVE_CTIME_R 544 char *ctime_r(const time_t *, char *); 545 #endif 546 #ifndef HAVE_ASCTIME_R 547 char *asctime_r(const struct tm *, char *); 548 #endif 549 550 #if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK) 551 #define QUICK 552 #endif 553 554 #ifndef O_BINARY 555 #define O_BINARY 0 556 #endif 557 558 #ifndef __cplusplus 559 #if defined(__GNUC__) && (__GNUC__ >= 3) 560 #define FILE_RCSID(id) \ 561 static const char rcsid[] __attribute__((__used__)) = id; 562 #else 563 #define FILE_RCSID(id) \ 564 static const char *rcsid(const char *p) { \ 565 return rcsid(p = id); \ 566 } 567 #endif 568 #else 569 #define FILE_RCSID(id) 570 #endif 571 572 #ifdef PHP_WIN32 573 #ifdef _WIN64 574 #define FINFO_LSEEK_FUNC _lseeki64 575 #else 576 #define FINFO_LSEEK_FUNC _lseek 577 #endif 578 #define FINFO_READ_FUNC _read 579 #else 580 #define FINFO_LSEEK_FUNC lseek 581 #define FINFO_READ_FUNC read 582 #endif 583 #ifndef __RCSID 584 #define __RCSID(a) 585 #endif 586 587 #endif /* __file_h__ */ 588