file.h revision 34650d7da670a6653a5b0546b0656d1772779fd2
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 1.172 2015/09/11 17:24:09 christos Exp $
31 */
32
33#ifndef __file_h__
34#define __file_h__
35
36#ifdef HAVE_CONFIG_H
37#include <config.h>
38#endif
39
40#ifdef WIN32
41  #ifdef _WIN64
42    #define SIZE_T_FORMAT "I64"
43  #else
44    #define SIZE_T_FORMAT ""
45  #endif
46  #define INT64_T_FORMAT "I64"
47  #define INTMAX_T_FORMAT "I64"
48#else
49  #define SIZE_T_FORMAT "z"
50  #define INT64_T_FORMAT "ll"
51  #define INTMAX_T_FORMAT "j"
52#endif
53
54#include <stdio.h>	/* Include that here, to make sure __P gets defined */
55#include <errno.h>
56#include <fcntl.h>	/* For open and flags */
57#ifdef HAVE_STDINT_H
58#ifndef __STDC_LIMIT_MACROS
59#define __STDC_LIMIT_MACROS
60#endif
61#include <stdint.h>
62#endif
63#ifdef HAVE_INTTYPES_H
64#include <inttypes.h>
65#endif
66#include <regex.h>
67#include <time.h>
68#include <sys/types.h>
69#ifndef WIN32
70#include <sys/param.h>
71#endif
72/* Do this here and now, because struct stat gets re-defined on solaris */
73#include <sys/stat.h>
74#include <stdarg.h>
75
76#define ENABLE_CONDITIONALS
77
78#ifndef MAGIC
79#define MAGIC "/etc/magic"
80#endif
81
82#if defined(__EMX__) || defined (WIN32)
83#define PATHSEP	';'
84#else
85#define PATHSEP	':'
86#endif
87
88#define private static
89
90#if HAVE_VISIBILITY && !defined(WIN32)
91#define public  __attribute__ ((__visibility__("default")))
92#ifndef protected
93#define protected __attribute__ ((__visibility__("hidden")))
94#endif
95#else
96#define public
97#ifndef protected
98#define protected
99#endif
100#endif
101
102#ifndef __arraycount
103#define __arraycount(a) (sizeof(a) / sizeof(a[0]))
104#endif
105
106#ifndef __GNUC_PREREQ__
107#ifdef __GNUC__
108#define	__GNUC_PREREQ__(x, y)						\
109	((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) ||			\
110	 (__GNUC__ > (x)))
111#else
112#define	__GNUC_PREREQ__(x, y)	0
113#endif
114#endif
115
116#ifndef __GNUC__
117#ifndef __attribute__
118#define __attribute__(a)
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 HOWMANY
131# define HOWMANY (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 64		/* max len of "string" types */
138
139#define MAGICNO		0xF11E041C
140#define VERSIONNO	13
141#define FILE_MAGICSIZE	312
142
143#define	FILE_LOAD	0
144#define FILE_CHECK	1
145#define FILE_COMPILE	2
146#define FILE_LIST	3
147
148union VALUETYPE {
149	uint8_t b;
150	uint16_t h;
151	uint32_t l;
152	uint64_t q;
153	uint8_t hs[2];	/* 2 bytes of a fixed-endian "short" */
154	uint8_t hl[4];	/* 4 bytes of a fixed-endian "long" */
155	uint8_t hq[8];	/* 8 bytes of a fixed-endian "quad" */
156	char s[MAXstring];	/* the search string or regex pattern */
157	unsigned char us[MAXstring];
158	float f;
159	double d;
160};
161
162struct magic {
163	/* Word 1 */
164	uint16_t cont_level;	/* level of ">" */
165	uint8_t flag;
166#define INDIR		0x01	/* if '(...)' appears */
167#define OFFADD		0x02	/* if '>&' or '>...(&' appears */
168#define INDIROFFADD	0x04	/* if '>&(' appears */
169#define UNSIGNED	0x08	/* comparison is unsigned */
170#define NOSPACE		0x10	/* suppress space character before output */
171#define BINTEST		0x20	/* test is for a binary type (set only
172				   for top-level tests) */
173#define TEXTTEST	0x40	/* for passing to file_softmagic */
174
175	uint8_t factor;
176
177	/* Word 2 */
178	uint8_t reln;		/* relation (0=eq, '>'=gt, etc) */
179	uint8_t vallen;		/* length of string value, if any */
180	uint8_t type;		/* comparison type (FILE_*) */
181	uint8_t in_type;	/* type of indirection */
182#define 			FILE_INVALID	0
183#define 			FILE_BYTE	1
184#define				FILE_SHORT	2
185#define				FILE_DEFAULT	3
186#define				FILE_LONG	4
187#define				FILE_STRING	5
188#define				FILE_DATE	6
189#define				FILE_BESHORT	7
190#define				FILE_BELONG	8
191#define				FILE_BEDATE	9
192#define				FILE_LESHORT	10
193#define				FILE_LELONG	11
194#define				FILE_LEDATE	12
195#define				FILE_PSTRING	13
196#define				FILE_LDATE	14
197#define				FILE_BELDATE	15
198#define				FILE_LELDATE	16
199#define				FILE_REGEX	17
200#define				FILE_BESTRING16	18
201#define				FILE_LESTRING16	19
202#define				FILE_SEARCH	20
203#define				FILE_MEDATE	21
204#define				FILE_MELDATE	22
205#define				FILE_MELONG	23
206#define				FILE_QUAD	24
207#define				FILE_LEQUAD	25
208#define				FILE_BEQUAD	26
209#define				FILE_QDATE	27
210#define				FILE_LEQDATE	28
211#define				FILE_BEQDATE	29
212#define				FILE_QLDATE	30
213#define				FILE_LEQLDATE	31
214#define				FILE_BEQLDATE	32
215#define				FILE_FLOAT	33
216#define				FILE_BEFLOAT	34
217#define				FILE_LEFLOAT	35
218#define				FILE_DOUBLE	36
219#define				FILE_BEDOUBLE	37
220#define				FILE_LEDOUBLE	38
221#define				FILE_BEID3	39
222#define				FILE_LEID3	40
223#define				FILE_INDIRECT	41
224#define				FILE_QWDATE	42
225#define				FILE_LEQWDATE	43
226#define				FILE_BEQWDATE	44
227#define				FILE_NAME	45
228#define				FILE_USE	46
229#define				FILE_CLEAR	47
230#define				FILE_NAMES_SIZE	48 /* size of array to contain all names */
231
232#define IS_STRING(t) \
233	((t) == FILE_STRING || \
234	 (t) == FILE_PSTRING || \
235	 (t) == FILE_BESTRING16 || \
236	 (t) == FILE_LESTRING16 || \
237	 (t) == FILE_REGEX || \
238	 (t) == FILE_SEARCH || \
239	 (t) == FILE_INDIRECT || \
240	 (t) == FILE_NAME || \
241	 (t) == FILE_USE)
242
243#define FILE_FMT_NONE 0
244#define FILE_FMT_NUM  1 /* "cduxXi" */
245#define FILE_FMT_STR  2 /* "s" */
246#define FILE_FMT_QUAD 3 /* "ll" */
247#define FILE_FMT_FLOAT 4 /* "eEfFgG" */
248#define FILE_FMT_DOUBLE 5 /* "eEfFgG" */
249
250	/* Word 3 */
251	uint8_t in_op;		/* operator for indirection */
252	uint8_t mask_op;	/* operator for mask */
253#ifdef ENABLE_CONDITIONALS
254	uint8_t cond;		/* conditional type */
255#else
256	uint8_t dummy;
257#endif
258	uint8_t factor_op;
259#define		FILE_FACTOR_OP_PLUS	'+'
260#define		FILE_FACTOR_OP_MINUS	'-'
261#define		FILE_FACTOR_OP_TIMES	'*'
262#define		FILE_FACTOR_OP_DIV	'/'
263#define		FILE_FACTOR_OP_NONE	'\0'
264
265#define				FILE_OPS	"&|^+-*/%"
266#define				FILE_OPAND	0
267#define				FILE_OPOR	1
268#define				FILE_OPXOR	2
269#define				FILE_OPADD	3
270#define				FILE_OPMINUS	4
271#define				FILE_OPMULTIPLY	5
272#define				FILE_OPDIVIDE	6
273#define				FILE_OPMODULO	7
274#define				FILE_OPS_MASK	0x07 /* mask for above ops */
275#define				FILE_UNUSED_1	0x08
276#define				FILE_UNUSED_2	0x10
277#define				FILE_UNUSED_3	0x20
278#define				FILE_OPINVERSE	0x40
279#define				FILE_OPINDIRECT	0x80
280
281#ifdef ENABLE_CONDITIONALS
282#define				COND_NONE	0
283#define				COND_IF		1
284#define				COND_ELIF	2
285#define				COND_ELSE	3
286#endif /* ENABLE_CONDITIONALS */
287
288	/* Word 4 */
289	uint32_t offset;	/* offset to magic number */
290	/* Word 5 */
291	int32_t in_offset;	/* offset from indirection */
292	/* Word 6 */
293	uint32_t lineno;	/* line number in magic file */
294	/* Word 7,8 */
295	union {
296		uint64_t _mask;	/* for use with numeric and date types */
297		struct {
298			uint32_t _count;	/* repeat/line count */
299			uint32_t _flags;	/* modifier flags */
300		} _s;		/* for use with string types */
301	} _u;
302#define num_mask _u._mask
303#define str_range _u._s._count
304#define str_flags _u._s._flags
305	/* Words 9-24 */
306	union VALUETYPE value;	/* either number or string */
307	/* Words 25-40 */
308	char desc[MAXDESC];	/* description */
309	/* Words 41-60 */
310	char mimetype[MAXMIME]; /* MIME type */
311	/* Words 61-62 */
312	char apple[8];		/* APPLE CREATOR/TYPE */
313	/* Words 63-78 */
314	char ext[64];		/* Popular extensions */
315};
316
317#define BIT(A)   (1 << (A))
318#define STRING_COMPACT_WHITESPACE		BIT(0)
319#define STRING_COMPACT_OPTIONAL_WHITESPACE	BIT(1)
320#define STRING_IGNORE_LOWERCASE			BIT(2)
321#define STRING_IGNORE_UPPERCASE			BIT(3)
322#define REGEX_OFFSET_START			BIT(4)
323#define STRING_TEXTTEST				BIT(5)
324#define STRING_BINTEST				BIT(6)
325#define PSTRING_1_BE				BIT(7)
326#define PSTRING_1_LE				BIT(7)
327#define PSTRING_2_BE				BIT(8)
328#define PSTRING_2_LE				BIT(9)
329#define PSTRING_4_BE				BIT(10)
330#define PSTRING_4_LE				BIT(11)
331#define REGEX_LINE_COUNT			BIT(11)
332#define PSTRING_LEN	\
333    (PSTRING_1_BE|PSTRING_2_LE|PSTRING_2_BE|PSTRING_4_LE|PSTRING_4_BE)
334#define PSTRING_LENGTH_INCLUDES_ITSELF		BIT(12)
335#define	STRING_TRIM				BIT(13)
336#define CHAR_COMPACT_WHITESPACE			'W'
337#define CHAR_COMPACT_OPTIONAL_WHITESPACE	'w'
338#define CHAR_IGNORE_LOWERCASE			'c'
339#define CHAR_IGNORE_UPPERCASE			'C'
340#define CHAR_REGEX_OFFSET_START			's'
341#define CHAR_TEXTTEST				't'
342#define	CHAR_TRIM				'T'
343#define CHAR_BINTEST				'b'
344#define CHAR_PSTRING_1_BE			'B'
345#define CHAR_PSTRING_1_LE			'B'
346#define CHAR_PSTRING_2_BE			'H'
347#define CHAR_PSTRING_2_LE			'h'
348#define CHAR_PSTRING_4_BE			'L'
349#define CHAR_PSTRING_4_LE			'l'
350#define CHAR_PSTRING_LENGTH_INCLUDES_ITSELF     'J'
351#define STRING_IGNORE_CASE		(STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE)
352#define STRING_DEFAULT_RANGE		100
353
354#define	INDIRECT_RELATIVE			BIT(0)
355#define	CHAR_INDIRECT_RELATIVE			'r'
356
357/* list of magic entries */
358struct mlist {
359	struct magic *magic;		/* array of magic entries */
360	uint32_t nmagic;		/* number of entries in array */
361	void *map;			/* internal resources used by entry */
362	struct mlist *next, *prev;
363};
364
365#ifdef __cplusplus
366#define CAST(T, b)	static_cast<T>(b)
367#define RCAST(T, b)	reinterpret_cast<T>(b)
368#else
369#define CAST(T, b)	(T)(b)
370#define RCAST(T, b)	(T)(b)
371#endif
372
373struct level_info {
374	int32_t off;
375	int got_match;
376#ifdef ENABLE_CONDITIONALS
377	int last_match;
378	int last_cond;	/* used for error checking by parse() */
379#endif
380};
381
382#define MAGIC_SETS	2
383
384struct magic_set {
385	struct mlist *mlist[MAGIC_SETS];	/* list of regular entries */
386	struct cont {
387		size_t len;
388		struct level_info *li;
389	} c;
390	struct out {
391		char *buf;		/* Accumulation buffer */
392		char *pbuf;		/* Printable buffer */
393	} o;
394	uint32_t offset;
395	int error;
396	int flags;			/* Control magic tests. */
397	int event_flags;		/* Note things that happened. */
398#define 		EVENT_HAD_ERR		0x01
399	const char *file;
400	size_t line;			/* current magic line number */
401
402	/* data for searches */
403	struct {
404		const char *s;		/* start of search in original source */
405		size_t s_len;		/* length of search region */
406		size_t offset;		/* starting offset in source: XXX - should this be off_t? */
407		size_t rm_len;		/* match length */
408	} search;
409
410	/* FIXME: Make the string dynamically allocated so that e.g.
411	   strings matched in files can be longer than MAXstring */
412	union VALUETYPE ms_value;	/* either number or string */
413	uint16_t indir_max;
414	uint16_t name_max;
415	uint16_t elf_shnum_max;
416	uint16_t elf_phnum_max;
417	uint16_t elf_notes_max;
418	uint16_t regex_max;
419#define	FILE_INDIR_MAX			15
420#define	FILE_NAME_MAX			30
421#define	FILE_ELF_SHNUM_MAX		32768
422#define	FILE_ELF_PHNUM_MAX		2048
423#define	FILE_ELF_NOTES_MAX		256
424#define	FILE_REGEX_MAX			8192
425};
426
427/* Type for Unicode characters */
428typedef unsigned long unichar;
429
430struct stat;
431#define FILE_T_LOCAL	1
432#define FILE_T_WINDOWS	2
433protected const char *file_fmttime(uint64_t, int, char *);
434protected struct magic_set *file_ms_alloc(int);
435protected void file_ms_free(struct magic_set *);
436protected int file_buffer(struct magic_set *, int, const char *, const void *,
437    size_t);
438protected int file_fsmagic(struct magic_set *, const char *, struct stat *);
439protected int file_pipe2file(struct magic_set *, int, const void *, size_t);
440protected int file_vprintf(struct magic_set *, const char *, va_list)
441    __attribute__((__format__(__printf__, 2, 0)));
442protected size_t file_printedlen(const struct magic_set *);
443protected int file_replace(struct magic_set *, const char *, const char *);
444protected int file_printf(struct magic_set *, const char *, ...)
445    __attribute__((__format__(__printf__, 2, 3)));
446protected int file_reset(struct magic_set *);
447protected int file_tryelf(struct magic_set *, int, const unsigned char *,
448    size_t);
449protected int file_trycdf(struct magic_set *, int, const unsigned char *,
450    size_t);
451#if HAVE_FORK
452protected int file_zmagic(struct magic_set *, int, const char *,
453    const unsigned char *, size_t);
454#endif
455protected int file_ascmagic(struct magic_set *, const unsigned char *, size_t,
456    int);
457protected int file_ascmagic_with_encoding(struct magic_set *,
458    const unsigned char *, size_t, unichar *, size_t, const char *,
459    const char *, int);
460protected int file_encoding(struct magic_set *, const unsigned char *, size_t,
461    unichar **, size_t *, const char **, const char **, const char **);
462protected int file_is_tar(struct magic_set *, const unsigned char *, size_t);
463protected int file_softmagic(struct magic_set *, const unsigned char *, size_t,
464    uint16_t, uint16_t *, int, int);
465protected int file_apprentice(struct magic_set *, const char *, int);
466protected int buffer_apprentice(struct magic_set *, struct magic **,
467    size_t *, size_t);
468protected int file_magicfind(struct magic_set *, const char *, struct mlist *);
469protected uint64_t file_signextend(struct magic_set *, struct magic *,
470    uint64_t);
471protected void file_badread(struct magic_set *);
472protected void file_badseek(struct magic_set *);
473protected void file_oomem(struct magic_set *, size_t);
474protected void file_error(struct magic_set *, int, const char *, ...)
475    __attribute__((__format__(__printf__, 3, 4)));
476protected void file_magerror(struct magic_set *, const char *, ...)
477    __attribute__((__format__(__printf__, 2, 3)));
478protected void file_magwarn(struct magic_set *, const char *, ...)
479    __attribute__((__format__(__printf__, 2, 3)));
480protected void file_mdump(struct magic *);
481protected void file_showstr(FILE *, const char *, size_t);
482protected size_t file_mbswidth(const char *);
483protected const char *file_getbuffer(struct magic_set *);
484protected ssize_t sread(int, void *, size_t, int);
485protected int file_check_mem(struct magic_set *, unsigned int);
486protected int file_looks_utf8(const unsigned char *, size_t, unichar *,
487    size_t *);
488protected size_t file_pstring_length_size(const struct magic *);
489protected size_t file_pstring_get_length(const struct magic *, const char *);
490protected char * file_printable(char *, size_t, const char *);
491#ifdef __EMX__
492protected int file_os2_apptype(struct magic_set *, const char *, const void *,
493    size_t);
494#endif /* __EMX__ */
495
496#if defined(HAVE_LOCALE_H)
497#include <locale.h>
498#endif
499#if defined(HAVE_XLOCALE_H)
500#include <xlocale.h>
501#endif
502
503typedef struct {
504	const char *pat;
505#if defined(HAVE_NEWLOCALE) && defined(HAVE_USELOCALE) && defined(HAVE_FREELOCALE)
506#define USE_C_LOCALE
507	locale_t old_lc_ctype;
508	locale_t c_lc_ctype;
509#endif
510	int rc;
511	regex_t rx;
512} file_regex_t;
513
514protected int file_regcomp(file_regex_t *, const char *, int);
515protected int file_regexec(file_regex_t *, const char *, size_t, regmatch_t *,
516    int);
517protected void file_regfree(file_regex_t *);
518protected void file_regerror(file_regex_t *, int, struct magic_set *);
519
520typedef struct {
521	char *buf;
522	uint32_t offset;
523} file_pushbuf_t;
524
525protected file_pushbuf_t *file_push_buffer(struct magic_set *);
526protected char  *file_pop_buffer(struct magic_set *, file_pushbuf_t *);
527
528#ifndef COMPILE_ONLY
529extern const char *file_names[];
530extern const size_t file_nnames;
531#endif
532
533#ifndef HAVE_STRERROR
534extern int sys_nerr;
535extern char *sys_errlist[];
536#define strerror(e) \
537	(((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error")
538#endif
539
540#ifndef HAVE_STRTOUL
541#define strtoul(a, b, c)	strtol(a, b, c)
542#endif
543
544#ifndef HAVE_PREAD
545ssize_t pread(int, void *, size_t, off_t);
546#endif
547#ifndef HAVE_VASPRINTF
548int vasprintf(char **, const char *, va_list);
549#endif
550#ifndef HAVE_ASPRINTF
551int asprintf(char **, const char *, ...);
552#endif
553
554#ifndef HAVE_STRLCPY
555size_t strlcpy(char *, const char *, size_t);
556#endif
557#ifndef HAVE_STRLCAT
558size_t strlcat(char *, const char *, size_t);
559#endif
560#ifndef HAVE_STRCASESTR
561char *strcasestr(const char *, const char *);
562#endif
563#ifndef HAVE_GETLINE
564ssize_t getline(char **, size_t *, FILE *);
565ssize_t getdelim(char **, size_t *, int, FILE *);
566#endif
567#ifndef HAVE_CTIME_R
568char   *ctime_r(const time_t *, char *);
569#endif
570#ifndef HAVE_ASCTIME_R
571char   *asctime_r(const struct tm *, char *);
572#endif
573#ifndef HAVE_GMTIME_R
574struct tm *gmtime_r(const time_t *, struct tm *);
575#endif
576#ifndef HAVE_LOCALTIME_R
577struct tm *localtime_r(const time_t *, struct tm *);
578#endif
579#ifndef HAVE_FMTCHECK
580const char *fmtcheck(const char *, const char *)
581     __attribute__((__format_arg__(2)));
582#endif
583
584#if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK)
585#define QUICK
586#endif
587
588#ifndef O_BINARY
589#define O_BINARY	0
590#endif
591
592#ifndef __cplusplus
593#if defined(__GNUC__) && (__GNUC__ >= 3)
594#define FILE_RCSID(id) \
595static const char rcsid[] __attribute__((__used__)) = id;
596#else
597#define FILE_RCSID(id) \
598static const char *rcsid(const char *p) { \
599	return rcsid(p = id); \
600}
601#endif
602#else
603#define FILE_RCSID(id)
604#endif
605#ifndef __RCSID
606#define __RCSID(a)
607#endif
608
609#endif /* __file_h__ */
610