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.
6 * Redistribution and use in source and binary forms, with or without
7 * modification, are permitted provided that the following conditions
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.
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
29 * file.h - definitions for file(1) program
30 * @(#)$File: file.h,v 1.225 2021/02/05 22:29:07 christos Exp $
44 #ifdef HAVE_INTTYPES_H
48 #ifndef __STDC_LIMIT_MACROS
49 #define __STDC_LIMIT_MACROS
51 #ifndef __STDC_FORMAT_MACROS
52 #define __STDC_FORMAT_MACROS
58 # define SIZE_T_FORMAT PRIu64
60 # define SIZE_T_FORMAT PRIu32
62 # define INT64_T_FORMAT PRIi64
63 # define INTMAX_T_FORMAT PRIiMAX
66 # define SIZE_T_FORMAT "I64"
68 # define SIZE_T_FORMAT ""
70 # define INT64_T_FORMAT "I64"
71 # define INTMAX_T_FORMAT "I64"
74 # define SIZE_T_FORMAT "z"
75 # define INT64_T_FORMAT "ll"
76 # define INTMAX_T_FORMAT "j"
79 #include <stdio.h> /* Include that here, to make sure __P gets defined */
81 #include <fcntl.h> /* For open and flags */
84 #include <sys/types.h>
86 #include <sys/param.h>
88 /* Do this here and now, because struct stat gets re-defined on solaris */
93 #define ENABLE_CONDITIONALS
96 #define MAGIC "/etc/magic"
99 #if defined(__EMX__) || defined (WIN32)
105 #define private static
107 #if HAVE_VISIBILITY && !defined(WIN32)
108 #define public __attribute__ ((__visibility__("default")))
110 #define protected __attribute__ ((__visibility__("hidden")))
120 #define __arraycount(a) (sizeof(a) / sizeof(a[0]))
123 #ifndef __GNUC_PREREQ__
125 #define __GNUC_PREREQ__(x, y) \
126 ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \
129 #define __GNUC_PREREQ__(x, y) 0
134 #ifndef __attribute__
135 #define __attribute__(a)
140 #define MIN(a,b) (((a) < (b)) ? (a) : (b))
144 #define MAX(a,b) (((a) > (b)) ? (a) : (b))
152 # define FD_CLOEXEC 1
155 #define FILE_BADSIZE CAST(size_t, ~0ul)
156 #define MAXDESC 64 /* max len of text description/MIME type */
157 #define MAXMIME 80 /* max len of text MIME type */
158 #define MAXstring 128 /* max len of "string" types */
160 #define MAGICNO 0xF11E041C
162 #define FILE_MAGICSIZE 376
164 #define FILE_GUID_SIZE sizeof("XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX")
168 #define FILE_COMPILE 2
186 uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */
187 uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */
188 uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */
189 char s[MAXstring]; /* the search string or regex pattern */
190 unsigned char us[MAXstring];
198 uint16_t cont_level; /* level of ">" */
200 #define INDIR 0x01 /* if '(...)' appears */
201 #define OFFADD 0x02 /* if '>&' or '>...(&' appears */
202 #define INDIROFFADD 0x04 /* if '>&(' appears */
203 #define UNSIGNED 0x08 /* comparison is unsigned */
204 #define NOSPACE 0x10 /* suppress space character before output */
205 #define BINTEST 0x20 /* test is for a binary type (set only
206 for top-level tests) */
207 #define TEXTTEST 0x40 /* for passing to file_softmagic */
208 #define OFFNEGATIVE 0x80 /* relative to the end of file */
213 uint8_t reln; /* relation (0=eq, '>'=gt, etc) */
214 uint8_t vallen; /* length of string value, if any */
215 uint8_t type; /* comparison type (FILE_*) */
216 uint8_t in_type; /* type of indirection */
217 #define FILE_INVALID 0
220 #define FILE_DEFAULT 3
222 #define FILE_STRING 5
224 #define FILE_BESHORT 7
225 #define FILE_BELONG 8
226 #define FILE_BEDATE 9
227 #define FILE_LESHORT 10
228 #define FILE_LELONG 11
229 #define FILE_LEDATE 12
230 #define FILE_PSTRING 13
231 #define FILE_LDATE 14
232 #define FILE_BELDATE 15
233 #define FILE_LELDATE 16
234 #define FILE_REGEX 17
235 #define FILE_BESTRING16 18
236 #define FILE_LESTRING16 19
237 #define FILE_SEARCH 20
238 #define FILE_MEDATE 21
239 #define FILE_MELDATE 22
240 #define FILE_MELONG 23
242 #define FILE_LEQUAD 25
243 #define FILE_BEQUAD 26
244 #define FILE_QDATE 27
245 #define FILE_LEQDATE 28
246 #define FILE_BEQDATE 29
247 #define FILE_QLDATE 30
248 #define FILE_LEQLDATE 31
249 #define FILE_BEQLDATE 32
250 #define FILE_FLOAT 33
251 #define FILE_BEFLOAT 34
252 #define FILE_LEFLOAT 35
253 #define FILE_DOUBLE 36
254 #define FILE_BEDOUBLE 37
255 #define FILE_LEDOUBLE 38
256 #define FILE_BEID3 39
257 #define FILE_LEID3 40
258 #define FILE_INDIRECT 41
259 #define FILE_QWDATE 42
260 #define FILE_LEQWDATE 43
261 #define FILE_BEQWDATE 44
264 #define FILE_CLEAR 47
267 #define FILE_OFFSET 50
268 #define FILE_NAMES_SIZE 51 /* size of array to contain all names */
270 #define IS_STRING(t) \
271 ((t) == FILE_STRING || \
272 (t) == FILE_PSTRING || \
273 (t) == FILE_BESTRING16 || \
274 (t) == FILE_LESTRING16 || \
275 (t) == FILE_REGEX || \
276 (t) == FILE_SEARCH || \
277 (t) == FILE_INDIRECT || \
278 (t) == FILE_NAME || \
281 #define FILE_FMT_NONE 0
282 #define FILE_FMT_NUM 1 /* "cduxXi" */
283 #define FILE_FMT_STR 2 /* "s" */
284 #define FILE_FMT_QUAD 3 /* "ll" */
285 #define FILE_FMT_FLOAT 4 /* "eEfFgG" */
286 #define FILE_FMT_DOUBLE 5 /* "eEfFgG" */
289 uint8_t in_op; /* operator for indirection */
290 uint8_t mask_op; /* operator for mask */
291 #ifdef ENABLE_CONDITIONALS
292 uint8_t cond; /* conditional type */
297 #define FILE_FACTOR_OP_PLUS '+'
298 #define FILE_FACTOR_OP_MINUS '-'
299 #define FILE_FACTOR_OP_TIMES '*'
300 #define FILE_FACTOR_OP_DIV '/'
301 #define FILE_FACTOR_OP_NONE '\0'
303 #define FILE_OPS "&|^+-*/%"
308 #define FILE_OPMINUS 4
309 #define FILE_OPMULTIPLY 5
310 #define FILE_OPDIVIDE 6
311 #define FILE_OPMODULO 7
312 #define FILE_OPS_MASK 0x07 /* mask for above ops */
313 #define FILE_UNUSED_1 0x08
314 #define FILE_UNUSED_2 0x10
315 #define FILE_OPSIGNED 0x20
316 #define FILE_OPINVERSE 0x40
317 #define FILE_OPINDIRECT 0x80
319 #ifdef ENABLE_CONDITIONALS
324 #endif /* ENABLE_CONDITIONALS */
327 int32_t offset; /* offset to magic number */
329 int32_t in_offset; /* offset from indirection */
331 uint32_t lineno; /* line number in magic file */
334 uint64_t _mask; /* for use with numeric and date types */
336 uint32_t _count; /* repeat/line count */
337 uint32_t _flags; /* modifier flags */
338 } _s; /* for use with string types */
340 #define num_mask _u._mask
341 #define str_range _u._s._count
342 #define str_flags _u._s._flags
344 union VALUETYPE value; /* either number or string */
346 char desc[MAXDESC]; /* description */
348 char mimetype[MAXMIME]; /* MIME type */
350 char apple[8]; /* APPLE CREATOR/TYPE */
352 char ext[64]; /* Popular extensions */
355 #define BIT(A) (1 << (A))
356 #define STRING_COMPACT_WHITESPACE BIT(0)
357 #define STRING_COMPACT_OPTIONAL_WHITESPACE BIT(1)
358 #define STRING_IGNORE_LOWERCASE BIT(2)
359 #define STRING_IGNORE_UPPERCASE BIT(3)
360 #define REGEX_OFFSET_START BIT(4)
361 #define STRING_TEXTTEST BIT(5)
362 #define STRING_BINTEST BIT(6)
363 #define PSTRING_1_BE BIT(7)
364 #define PSTRING_1_LE BIT(7)
365 #define PSTRING_2_BE BIT(8)
366 #define PSTRING_2_LE BIT(9)
367 #define PSTRING_4_BE BIT(10)
368 #define PSTRING_4_LE BIT(11)
369 #define REGEX_LINE_COUNT BIT(11)
370 #define PSTRING_LEN \
371 (PSTRING_1_BE|PSTRING_2_LE|PSTRING_2_BE|PSTRING_4_LE|PSTRING_4_BE)
372 #define PSTRING_LENGTH_INCLUDES_ITSELF BIT(12)
373 #define STRING_TRIM BIT(13)
374 #define CHAR_COMPACT_WHITESPACE 'W'
375 #define CHAR_COMPACT_OPTIONAL_WHITESPACE 'w'
376 #define CHAR_IGNORE_LOWERCASE 'c'
377 #define CHAR_IGNORE_UPPERCASE 'C'
378 #define CHAR_REGEX_OFFSET_START 's'
379 #define CHAR_TEXTTEST 't'
380 #define CHAR_TRIM 'T'
381 #define CHAR_BINTEST 'b'
382 #define CHAR_PSTRING_1_BE 'B'
383 #define CHAR_PSTRING_1_LE 'B'
384 #define CHAR_PSTRING_2_BE 'H'
385 #define CHAR_PSTRING_2_LE 'h'
386 #define CHAR_PSTRING_4_BE 'L'
387 #define CHAR_PSTRING_4_LE 'l'
388 #define CHAR_PSTRING_LENGTH_INCLUDES_ITSELF 'J'
389 #define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE)
390 #define STRING_DEFAULT_RANGE 100
392 #define INDIRECT_RELATIVE BIT(0)
393 #define CHAR_INDIRECT_RELATIVE 'r'
395 /* list of magic entries */
397 struct magic *magic; /* array of magic entries */
398 uint32_t nmagic; /* number of entries in array */
399 void *map; /* internal resources used by entry */
400 struct mlist *next, *prev;
404 #define CAST(T, b) static_cast<T>(b)
405 #define RCAST(T, b) reinterpret_cast<T>(b)
406 #define CCAST(T, b) const_cast<T>(b)
408 #define CAST(T, b) ((T)(b))
409 #define RCAST(T, b) ((T)(uintptr_t)(b))
410 #define CCAST(T, b) ((T)(uintptr_t)(b))
416 #ifdef ENABLE_CONDITIONALS
418 int last_cond; /* used for error checking by parse() */
424 struct level_info *li;
430 struct mlist *mlist[MAGIC_SETS]; /* list of regular entries */
433 char *buf; /* Accumulation buffer */
434 size_t blen; /* Length of buffer */
435 char *pbuf; /* Printable buffer */
437 uint32_t offset; /* a copy of m->offset while we */
438 /* are working on the magic entry */
439 uint32_t eoffset; /* offset from end of file */
441 int flags; /* Control magic tests. */
442 int event_flags; /* Note things that happened. */
443 #define EVENT_HAD_ERR 0x01
445 size_t line; /* current magic line number */
446 mode_t mode; /* copy of current stat mode */
448 /* data for searches */
450 const char *s; /* start of search in original source */
451 size_t s_len; /* length of search region */
452 size_t offset; /* starting offset in source: XXX - should this be off_t? */
453 size_t rm_len; /* match length */
456 /* FIXME: Make the string dynamically allocated so that e.g.
457 strings matched in files can be longer than MAXstring */
458 union VALUETYPE ms_value; /* either number or string */
461 uint16_t elf_shnum_max;
462 uint16_t elf_phnum_max;
463 uint16_t elf_notes_max;
465 size_t bytes_max; /* number of bytes to read from file */
466 size_t encoding_max; /* bytes to look for encoding */
467 #ifndef FILE_BYTES_MAX
468 # define FILE_BYTES_MAX (1024 * 1024) /* how much of the file to look at */
470 #define FILE_ELF_NOTES_MAX 256
471 #define FILE_ELF_PHNUM_MAX 2048
472 #define FILE_ELF_SHNUM_MAX 32768
473 #define FILE_INDIR_MAX 50
474 #define FILE_NAME_MAX 50
475 #define FILE_REGEX_MAX 8192
476 #define FILE_ENCODING_MAX (64 * 1024)
479 /* Type for Unicode characters */
480 typedef unsigned long file_unichar_t;
483 #define FILE_T_LOCAL 1
484 #define FILE_T_WINDOWS 2
485 protected const char *file_fmttime(char *, size_t, uint64_t, int);
486 protected struct magic_set *file_ms_alloc(int);
487 protected void file_ms_free(struct magic_set *);
488 protected int file_default(struct magic_set *, size_t);
489 protected int file_buffer(struct magic_set *, int, struct stat *, const char *,
490 const void *, size_t);
491 protected int file_fsmagic(struct magic_set *, const char *, struct stat *);
492 protected int file_pipe2file(struct magic_set *, int, const void *, size_t);
493 protected int file_vprintf(struct magic_set *, const char *, va_list)
494 __attribute__((__format__(__printf__, 2, 0)));
495 protected int file_separator(struct magic_set *);
496 protected char *file_copystr(char *, size_t, size_t, const char *);
497 protected int file_checkfmt(char *, size_t, const char *);
498 protected size_t file_printedlen(const struct magic_set *);
499 protected int file_print_guid(char *, size_t, const uint64_t *);
500 protected int file_parse_guid(const char *, uint64_t *);
501 protected int file_replace(struct magic_set *, const char *, const char *);
502 protected int file_printf(struct magic_set *, const char *, ...)
503 __attribute__((__format__(__printf__, 2, 3)));
504 protected int file_reset(struct magic_set *, int);
505 protected int file_tryelf(struct magic_set *, const struct buffer *);
506 protected int file_trycdf(struct magic_set *, const struct buffer *);
508 protected int file_zmagic(struct magic_set *, const struct buffer *,
511 protected int file_ascmagic(struct magic_set *, const struct buffer *,
513 protected int file_ascmagic_with_encoding(struct magic_set *,
514 const struct buffer *, file_unichar_t *, size_t, const char *, const char *, int);
515 protected int file_encoding(struct magic_set *, const struct buffer *,
516 file_unichar_t **, size_t *, const char **, const char **, const char **);
517 protected int file_is_json(struct magic_set *, const struct buffer *);
518 protected int file_is_csv(struct magic_set *, const struct buffer *, int);
519 protected int file_is_tar(struct magic_set *, const struct buffer *);
520 protected int file_softmagic(struct magic_set *, const struct buffer *,
521 uint16_t *, uint16_t *, int, int);
522 protected int file_apprentice(struct magic_set *, const char *, int);
523 protected int buffer_apprentice(struct magic_set *, struct magic **,
525 protected int file_magicfind(struct magic_set *, const char *, struct mlist *);
526 protected uint64_t file_signextend(struct magic_set *, struct magic *,
528 protected void file_badread(struct magic_set *);
529 protected void file_badseek(struct magic_set *);
530 protected void file_oomem(struct magic_set *, size_t);
531 protected void file_error(struct magic_set *, int, const char *, ...)
532 __attribute__((__format__(__printf__, 3, 4)));
533 protected void file_magerror(struct magic_set *, const char *, ...)
534 __attribute__((__format__(__printf__, 2, 3)));
535 protected void file_magwarn(struct magic_set *, const char *, ...)
536 __attribute__((__format__(__printf__, 2, 3)));
537 protected void file_mdump(struct magic *);
538 protected void file_showstr(FILE *, const char *, size_t);
539 protected size_t file_mbswidth(const char *);
540 protected const char *file_getbuffer(struct magic_set *);
541 protected ssize_t sread(int, void *, size_t, int);
542 protected int file_check_mem(struct magic_set *, unsigned int);
543 protected int file_looks_utf8(const unsigned char *, size_t, file_unichar_t *,
545 protected size_t file_pstring_length_size(struct magic_set *,
546 const struct magic *);
547 protected size_t file_pstring_get_length(struct magic_set *,
548 const struct magic *, const char *);
549 protected char * file_printable(char *, size_t, const char *, size_t);
551 protected int file_os2_apptype(struct magic_set *, const char *, const void *,
554 protected int file_pipe_closexec(int *);
555 protected int file_clear_closexec(int);
556 protected char *file_strtrim(char *);
558 protected void buffer_init(struct buffer *, int, const struct stat *,
559 const void *, size_t);
560 protected void buffer_fini(struct buffer *);
561 protected int buffer_fill(const struct buffer *);
564 #if defined(HAVE_XLOCALE_H)
570 #if defined(HAVE_NEWLOCALE) && defined(HAVE_USELOCALE) && defined(HAVE_FREELOCALE)
572 locale_t old_lc_ctype;
581 protected int file_regcomp(file_regex_t *, const char *, int);
582 protected int file_regexec(file_regex_t *, const char *, size_t, regmatch_t *,
584 protected void file_regfree(file_regex_t *);
585 protected void file_regerror(file_regex_t *, int, struct magic_set *);
593 protected file_pushbuf_t *file_push_buffer(struct magic_set *);
594 protected char *file_pop_buffer(struct magic_set *, file_pushbuf_t *);
597 extern const char *file_names[];
598 extern const size_t file_nnames;
602 ssize_t pread(int, void *, size_t, off_t);
604 #ifndef HAVE_VASPRINTF
605 int vasprintf(char **, const char *, va_list);
607 #ifndef HAVE_ASPRINTF
608 int asprintf(char **, const char *, ...);
611 int dprintf(int, const char *, ...);
615 size_t strlcpy(char *, const char *, size_t);
618 size_t strlcat(char *, const char *, size_t);
620 #ifndef HAVE_STRCASESTR
621 char *strcasestr(const char *, const char *);
624 ssize_t getline(char **, size_t *, FILE *);
625 ssize_t getdelim(char **, size_t *, int, FILE *);
628 char *ctime_r(const time_t *, char *);
630 #ifndef HAVE_ASCTIME_R
631 char *asctime_r(const struct tm *, char *);
633 #ifndef HAVE_GMTIME_R
634 struct tm *gmtime_r(const time_t *, struct tm *);
636 #ifndef HAVE_LOCALTIME_R
637 struct tm *localtime_r(const time_t *, struct tm *);
639 #ifndef HAVE_FMTCHECK
640 const char *fmtcheck(const char *, const char *)
641 __attribute__((__format_arg__(2)));
644 #ifdef HAVE_LIBSECCOMP
646 // this mode should not interfere with normal operations
647 // only some dangerous syscalls are blacklisted
648 int enable_sandbox_basic(void);
651 // this mode allows only the necessary syscalls used during normal operation
652 // extensive testing required !!!
653 int enable_sandbox_full(void);
656 protected const char *file_getprogname(void);
657 protected void file_setprogname(const char *);
658 protected void file_err(int, const char *, ...)
659 __attribute__((__format__(__printf__, 2, 3), __noreturn__));
660 protected void file_errx(int, const char *, ...)
661 __attribute__((__format__(__printf__, 2, 3), __noreturn__));
662 protected void file_warn(const char *, ...)
663 __attribute__((__format__(__printf__, 1, 2)));
664 protected void file_warnx(const char *, ...)
665 __attribute__((__format__(__printf__, 1, 2)));
667 #if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK)
679 #if defined(__GNUC__) && (__GNUC__ >= 3)
680 #define FILE_RCSID(id) \
681 static const char rcsid[] __attribute__((__used__)) = id;
683 #define FILE_RCSID(id) \
684 static const char *rcsid(const char *p) { \
685 return rcsid(p = id); \
689 #define FILE_RCSID(id)
695 #endif /* __file_h__ */