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 1.208 2019/06/26 20:31:31 christos Exp $
40 #ifndef __STDC_LIMIT_MACROS
41 #define __STDC_LIMIT_MACROS
43 #ifndef __STDC_FORMAT_MACROS
44 #define __STDC_FORMAT_MACROS
49 #define SIZE_T_FORMAT "I64"
51 #define SIZE_T_FORMAT ""
53 #define INT64_T_FORMAT "I64"
54 #define INTMAX_T_FORMAT "I64"
56 #define SIZE_T_FORMAT "z"
57 #define INT64_T_FORMAT "ll"
58 #define INTMAX_T_FORMAT "j"
63 #include <stdio.h> /* Include that here, to make sure __P gets defined */
65 #include <fcntl.h> /* For open and flags */
66 #ifdef HAVE_INTTYPES_H
71 #include <sys/types.h>
73 #include <sys/param.h>
75 /* Do this here and now, because struct stat gets re-defined on solaris */
80 #define ENABLE_CONDITIONALS
83 #define MAGIC "/etc/magic"
86 #if defined(__EMX__) || defined (WIN32)
92 #define private static
94 #if HAVE_VISIBILITY && !defined(WIN32)
95 #define public __attribute__ ((__visibility__("default")))
97 #define protected __attribute__ ((__visibility__("hidden")))
107 #define __arraycount(a) (sizeof(a) / sizeof(a[0]))
110 #ifndef __GNUC_PREREQ__
112 #define __GNUC_PREREQ__(x, y) \
113 ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \
116 #define __GNUC_PREREQ__(x, y) 0
121 #ifndef __attribute__
122 #define __attribute__(a)
127 #define MIN(a,b) (((a) < (b)) ? (a) : (b))
131 #define MAX(a,b) (((a) > (b)) ? (a) : (b))
134 #ifndef FILE_BYTES_MAX
135 # define FILE_BYTES_MAX (1024 * 1024) /* how much of the file to look at */
137 #define MAXMAGIS 8192 /* max entries in any one magic file
139 #define MAXDESC 64 /* max len of text description/MIME type */
140 #define MAXMIME 80 /* max len of text MIME type */
141 #define MAXstring 96 /* max len of "string" types */
143 #define MAGICNO 0xF11E041C
145 #define FILE_MAGICSIZE 344
149 #define FILE_COMPILE 2
167 uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */
168 uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */
169 uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */
170 char s[MAXstring]; /* the search string or regex pattern */
171 unsigned char us[MAXstring];
178 uint16_t cont_level; /* level of ">" */
180 #define INDIR 0x01 /* if '(...)' appears */
181 #define OFFADD 0x02 /* if '>&' or '>...(&' appears */
182 #define INDIROFFADD 0x04 /* if '>&(' appears */
183 #define UNSIGNED 0x08 /* comparison is unsigned */
184 #define NOSPACE 0x10 /* suppress space character before output */
185 #define BINTEST 0x20 /* test is for a binary type (set only
186 for top-level tests) */
187 #define TEXTTEST 0x40 /* for passing to file_softmagic */
192 uint8_t reln; /* relation (0=eq, '>'=gt, etc) */
193 uint8_t vallen; /* length of string value, if any */
194 uint8_t type; /* comparison type (FILE_*) */
195 uint8_t in_type; /* type of indirection */
196 #define FILE_INVALID 0
199 #define FILE_DEFAULT 3
201 #define FILE_STRING 5
203 #define FILE_BESHORT 7
204 #define FILE_BELONG 8
205 #define FILE_BEDATE 9
206 #define FILE_LESHORT 10
207 #define FILE_LELONG 11
208 #define FILE_LEDATE 12
209 #define FILE_PSTRING 13
210 #define FILE_LDATE 14
211 #define FILE_BELDATE 15
212 #define FILE_LELDATE 16
213 #define FILE_REGEX 17
214 #define FILE_BESTRING16 18
215 #define FILE_LESTRING16 19
216 #define FILE_SEARCH 20
217 #define FILE_MEDATE 21
218 #define FILE_MELDATE 22
219 #define FILE_MELONG 23
221 #define FILE_LEQUAD 25
222 #define FILE_BEQUAD 26
223 #define FILE_QDATE 27
224 #define FILE_LEQDATE 28
225 #define FILE_BEQDATE 29
226 #define FILE_QLDATE 30
227 #define FILE_LEQLDATE 31
228 #define FILE_BEQLDATE 32
229 #define FILE_FLOAT 33
230 #define FILE_BEFLOAT 34
231 #define FILE_LEFLOAT 35
232 #define FILE_DOUBLE 36
233 #define FILE_BEDOUBLE 37
234 #define FILE_LEDOUBLE 38
235 #define FILE_BEID3 39
236 #define FILE_LEID3 40
237 #define FILE_INDIRECT 41
238 #define FILE_QWDATE 42
239 #define FILE_LEQWDATE 43
240 #define FILE_BEQWDATE 44
243 #define FILE_CLEAR 47
245 #define FILE_NAMES_SIZE 49 /* size of array to contain all names */
247 #define IS_STRING(t) \
248 ((t) == FILE_STRING || \
249 (t) == FILE_PSTRING || \
250 (t) == FILE_BESTRING16 || \
251 (t) == FILE_LESTRING16 || \
252 (t) == FILE_REGEX || \
253 (t) == FILE_SEARCH || \
254 (t) == FILE_INDIRECT || \
255 (t) == FILE_NAME || \
258 #define FILE_FMT_NONE 0
259 #define FILE_FMT_NUM 1 /* "cduxXi" */
260 #define FILE_FMT_STR 2 /* "s" */
261 #define FILE_FMT_QUAD 3 /* "ll" */
262 #define FILE_FMT_FLOAT 4 /* "eEfFgG" */
263 #define FILE_FMT_DOUBLE 5 /* "eEfFgG" */
266 uint8_t in_op; /* operator for indirection */
267 uint8_t mask_op; /* operator for mask */
268 #ifdef ENABLE_CONDITIONALS
269 uint8_t cond; /* conditional type */
274 #define FILE_FACTOR_OP_PLUS '+'
275 #define FILE_FACTOR_OP_MINUS '-'
276 #define FILE_FACTOR_OP_TIMES '*'
277 #define FILE_FACTOR_OP_DIV '/'
278 #define FILE_FACTOR_OP_NONE '\0'
280 #define FILE_OPS "&|^+-*/%"
285 #define FILE_OPMINUS 4
286 #define FILE_OPMULTIPLY 5
287 #define FILE_OPDIVIDE 6
288 #define FILE_OPMODULO 7
289 #define FILE_OPS_MASK 0x07 /* mask for above ops */
290 #define FILE_UNUSED_1 0x08
291 #define FILE_UNUSED_2 0x10
292 #define FILE_OPSIGNED 0x20
293 #define FILE_OPINVERSE 0x40
294 #define FILE_OPINDIRECT 0x80
296 #ifdef ENABLE_CONDITIONALS
301 #endif /* ENABLE_CONDITIONALS */
304 int32_t offset; /* offset to magic number */
306 int32_t in_offset; /* offset from indirection */
308 uint32_t lineno; /* line number in magic file */
311 uint64_t _mask; /* for use with numeric and date types */
313 uint32_t _count; /* repeat/line count */
314 uint32_t _flags; /* modifier flags */
315 } _s; /* for use with string types */
317 #define num_mask _u._mask
318 #define str_range _u._s._count
319 #define str_flags _u._s._flags
321 union VALUETYPE value; /* either number or string */
323 char desc[MAXDESC]; /* description */
325 char mimetype[MAXMIME]; /* MIME type */
327 char apple[8]; /* APPLE CREATOR/TYPE */
329 char ext[64]; /* Popular extensions */
332 #define BIT(A) (1 << (A))
333 #define STRING_COMPACT_WHITESPACE BIT(0)
334 #define STRING_COMPACT_OPTIONAL_WHITESPACE BIT(1)
335 #define STRING_IGNORE_LOWERCASE BIT(2)
336 #define STRING_IGNORE_UPPERCASE BIT(3)
337 #define REGEX_OFFSET_START BIT(4)
338 #define STRING_TEXTTEST BIT(5)
339 #define STRING_BINTEST BIT(6)
340 #define PSTRING_1_BE BIT(7)
341 #define PSTRING_1_LE BIT(7)
342 #define PSTRING_2_BE BIT(8)
343 #define PSTRING_2_LE BIT(9)
344 #define PSTRING_4_BE BIT(10)
345 #define PSTRING_4_LE BIT(11)
346 #define REGEX_LINE_COUNT BIT(11)
347 #define PSTRING_LEN \
348 (PSTRING_1_BE|PSTRING_2_LE|PSTRING_2_BE|PSTRING_4_LE|PSTRING_4_BE)
349 #define PSTRING_LENGTH_INCLUDES_ITSELF BIT(12)
350 #define STRING_TRIM BIT(13)
351 #define CHAR_COMPACT_WHITESPACE 'W'
352 #define CHAR_COMPACT_OPTIONAL_WHITESPACE 'w'
353 #define CHAR_IGNORE_LOWERCASE 'c'
354 #define CHAR_IGNORE_UPPERCASE 'C'
355 #define CHAR_REGEX_OFFSET_START 's'
356 #define CHAR_TEXTTEST 't'
357 #define CHAR_TRIM 'T'
358 #define CHAR_BINTEST 'b'
359 #define CHAR_PSTRING_1_BE 'B'
360 #define CHAR_PSTRING_1_LE 'B'
361 #define CHAR_PSTRING_2_BE 'H'
362 #define CHAR_PSTRING_2_LE 'h'
363 #define CHAR_PSTRING_4_BE 'L'
364 #define CHAR_PSTRING_4_LE 'l'
365 #define CHAR_PSTRING_LENGTH_INCLUDES_ITSELF 'J'
366 #define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE)
367 #define STRING_DEFAULT_RANGE 100
369 #define INDIRECT_RELATIVE BIT(0)
370 #define CHAR_INDIRECT_RELATIVE 'r'
372 /* list of magic entries */
374 struct magic *magic; /* array of magic entries */
375 uint32_t nmagic; /* number of entries in array */
376 void *map; /* internal resources used by entry */
377 struct mlist *next, *prev;
381 #define CAST(T, b) static_cast<T>(b)
382 #define RCAST(T, b) reinterpret_cast<T>(b)
383 #define CCAST(T, b) const_cast<T>(b)
385 #define CAST(T, b) ((T)(b))
386 #define RCAST(T, b) ((T)(uintptr_t)(b))
387 #define CCAST(T, b) ((T)(uintptr_t)(b))
393 #ifdef ENABLE_CONDITIONALS
395 int last_cond; /* used for error checking by parse() */
402 struct mlist *mlist[MAGIC_SETS]; /* list of regular entries */
405 struct level_info *li;
408 char *buf; /* Accumulation buffer */
409 char *pbuf; /* Printable buffer */
411 uint32_t offset; /* a copy of m->offset while we */
412 /* are working on the magic entry */
413 uint32_t eoffset; /* offset from end of file */
415 int flags; /* Control magic tests. */
416 int event_flags; /* Note things that happened. */
417 #define EVENT_HAD_ERR 0x01
419 size_t line; /* current magic line number */
420 mode_t mode; /* copy of current stat mode */
422 /* data for searches */
424 const char *s; /* start of search in original source */
425 size_t s_len; /* length of search region */
426 size_t offset; /* starting offset in source: XXX - should this be off_t? */
427 size_t rm_len; /* match length */
430 /* FIXME: Make the string dynamically allocated so that e.g.
431 strings matched in files can be longer than MAXstring */
432 union VALUETYPE ms_value; /* either number or string */
435 uint16_t elf_shnum_max;
436 uint16_t elf_phnum_max;
437 uint16_t elf_notes_max;
439 size_t bytes_max; /* number of bytes to read from file */
440 #define FILE_INDIR_MAX 50
441 #define FILE_NAME_MAX 30
442 #define FILE_ELF_SHNUM_MAX 32768
443 #define FILE_ELF_PHNUM_MAX 2048
444 #define FILE_ELF_NOTES_MAX 256
445 #define FILE_REGEX_MAX 8192
448 /* Type for Unicode characters */
449 typedef unsigned long unichar;
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_default(struct magic_set *, size_t);
458 protected int file_buffer(struct magic_set *, int, struct stat *, const char *,
459 const void *, size_t);
460 protected int file_fsmagic(struct magic_set *, const char *, struct stat *);
461 protected int file_pipe2file(struct magic_set *, int, const void *, size_t);
462 protected int file_vprintf(struct magic_set *, const char *, va_list)
463 __attribute__((__format__(__printf__, 2, 0)));
464 protected int file_separator(struct magic_set *);
465 protected size_t file_printedlen(const struct magic_set *);
466 protected int file_replace(struct magic_set *, const char *, const char *);
467 protected int file_printf(struct magic_set *, const char *, ...)
468 __attribute__((__format__(__printf__, 2, 3)));
469 protected int file_reset(struct magic_set *, int);
470 protected int file_tryelf(struct magic_set *, const struct buffer *);
471 protected int file_trycdf(struct magic_set *, const struct buffer *);
473 protected int file_zmagic(struct magic_set *, const struct buffer *,
476 protected int file_ascmagic(struct magic_set *, const struct buffer *,
478 protected int file_ascmagic_with_encoding(struct magic_set *,
479 const struct buffer *, unichar *, size_t, const char *, const char *, int);
480 protected int file_encoding(struct magic_set *, const struct buffer *,
481 unichar **, size_t *, const char **, const char **, const char **);
482 protected int file_is_json(struct magic_set *, const struct buffer *);
483 protected int file_is_csv(struct magic_set *, const struct buffer *, int);
484 protected int file_is_tar(struct magic_set *, const struct buffer *);
485 protected int file_softmagic(struct magic_set *, const struct buffer *,
486 uint16_t *, uint16_t *, int, int);
487 protected int file_apprentice(struct magic_set *, const char *, int);
488 protected int buffer_apprentice(struct magic_set *, struct magic **,
490 protected int file_magicfind(struct magic_set *, const char *, struct mlist *);
491 protected uint64_t file_signextend(struct magic_set *, struct magic *,
493 protected void file_badread(struct magic_set *);
494 protected void file_badseek(struct magic_set *);
495 protected void file_oomem(struct magic_set *, size_t);
496 protected void file_error(struct magic_set *, int, const char *, ...)
497 __attribute__((__format__(__printf__, 3, 4)));
498 protected void file_magerror(struct magic_set *, const char *, ...)
499 __attribute__((__format__(__printf__, 2, 3)));
500 protected void file_magwarn(struct magic_set *, const char *, ...)
501 __attribute__((__format__(__printf__, 2, 3)));
502 protected void file_mdump(struct magic *);
503 protected void file_showstr(FILE *, const char *, size_t);
504 protected size_t file_mbswidth(const char *);
505 protected const char *file_getbuffer(struct magic_set *);
506 protected ssize_t sread(int, void *, size_t, int);
507 protected int file_check_mem(struct magic_set *, unsigned int);
508 protected int file_looks_utf8(const unsigned char *, size_t, unichar *,
510 protected size_t file_pstring_length_size(const struct magic *);
511 protected size_t file_pstring_get_length(const struct magic *, const char *);
512 protected char * file_printable(char *, size_t, const char *, size_t);
514 protected int file_os2_apptype(struct magic_set *, const char *, const void *,
518 protected void buffer_init(struct buffer *, int, const struct stat *,
519 const void *, size_t);
520 protected void buffer_fini(struct buffer *);
521 protected int buffer_fill(const struct buffer *);
524 #if defined(HAVE_XLOCALE_H)
530 #if defined(HAVE_NEWLOCALE) && defined(HAVE_USELOCALE) && defined(HAVE_FREELOCALE)
532 locale_t old_lc_ctype;
541 protected int file_regcomp(file_regex_t *, const char *, int);
542 protected int file_regexec(file_regex_t *, const char *, size_t, regmatch_t *,
544 protected void file_regfree(file_regex_t *);
545 protected void file_regerror(file_regex_t *, int, struct magic_set *);
552 protected file_pushbuf_t *file_push_buffer(struct magic_set *);
553 protected char *file_pop_buffer(struct magic_set *, file_pushbuf_t *);
556 extern const char *file_names[];
557 extern const size_t file_nnames;
561 ssize_t pread(int, void *, size_t, off_t);
563 #ifndef HAVE_VASPRINTF
564 int vasprintf(char **, const char *, va_list);
566 #ifndef HAVE_ASPRINTF
567 int asprintf(char **, const char *, ...);
570 int dprintf(int, const char *, ...);
574 size_t strlcpy(char *, const char *, size_t);
577 size_t strlcat(char *, const char *, size_t);
579 #ifndef HAVE_STRCASESTR
580 char *strcasestr(const char *, const char *);
583 ssize_t getline(char **, size_t *, FILE *);
584 ssize_t getdelim(char **, size_t *, int, FILE *);
587 char *ctime_r(const time_t *, char *);
589 #ifndef HAVE_ASCTIME_R
590 char *asctime_r(const struct tm *, char *);
592 #ifndef HAVE_GMTIME_R
593 struct tm *gmtime_r(const time_t *, struct tm *);
595 #ifndef HAVE_LOCALTIME_R
596 struct tm *localtime_r(const time_t *, struct tm *);
598 #ifndef HAVE_FMTCHECK
599 const char *fmtcheck(const char *, const char *)
600 __attribute__((__format_arg__(2)));
603 #ifdef HAVE_LIBSECCOMP
605 // this mode should not interfere with normal operations
606 // only some dangerous syscalls are blacklisted
607 int enable_sandbox_basic(void);
610 // this mode allows only the necessary syscalls used during normal operation
611 // extensive testing required !!!
612 int enable_sandbox_full(void);
615 protected const char *file_getprogname(void);
616 protected void file_setprogname(const char *);
617 protected void file_err(int, const char *, ...)
618 __attribute__((__format__(__printf__, 2, 3), __noreturn__));
619 protected void file_errx(int, const char *, ...)
620 __attribute__((__format__(__printf__, 2, 3), __noreturn__));
621 protected void file_warn(const char *, ...)
622 __attribute__((__format__(__printf__, 1, 2)));
623 protected void file_warnx(const char *, ...)
624 __attribute__((__format__(__printf__, 1, 2)));
626 #if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK)
638 #if defined(__GNUC__) && (__GNUC__ >= 3)
639 #define FILE_RCSID(id) \
640 static const char rcsid[] __attribute__((__used__)) = id;
642 #define FILE_RCSID(id) \
643 static const char *rcsid(const char *p) { \
644 return rcsid(p = id); \
648 #define FILE_RCSID(id)
654 #endif /* __file_h__ */