Annotation of src/usr.bin/file/file.h, Revision 1.20
1.20 ! chl 1: /* $OpenBSD: file.h,v 1.19 2008/05/08 01:40:56 chl Exp $ */
1.1 deraadt 2: /*
1.12 ian 3: * Copyright (c) Ian F. Darwin 1986-1995.
4: * Software written by Ian F. Darwin and others;
5: * maintained 1995-present by Christos Zoulas and others.
6: *
7: * Redistribution and use in source and binary forms, with or without
8: * modification, are permitted provided that the following conditions
9: * are met:
10: * 1. Redistributions of source code must retain the above copyright
11: * notice immediately at the beginning of the file, without modification,
12: * this list of conditions, and the following disclaimer.
13: * 2. Redistributions in binary form must reproduce the above copyright
14: * notice, this list of conditions and the following disclaimer in the
15: * documentation and/or other materials provided with the distribution.
16: *
17: * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
18: * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19: * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20: * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR
21: * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22: * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23: * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24: * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25: * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26: * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27: * SUCH DAMAGE.
1.1 deraadt 28: */
1.16 tedu 29: /*
30: * file.h - definitions for file(1) program
1.20 ! chl 31: * @(#)$Id: file.h,v 1.19 2008/05/08 01:40:56 chl Exp $
1.16 tedu 32: */
1.1 deraadt 33:
1.3 millert 34: #ifndef __file_h__
35: #define __file_h__
36:
1.11 ian 37: #ifdef HAVE_CONFIG_H
38: #include "config.h"
39: #endif
40:
1.16 tedu 41: #include <stdio.h> /* Include that here, to make sure __P gets defined */
1.11 ian 42: #include <errno.h>
1.19 chl 43: #include <fcntl.h> /* For open and flags */
1.11 ian 44: #ifdef HAVE_STDINT_H
45: #include <stdint.h>
1.16 tedu 46: #endif
47: #ifdef HAVE_INTTYPES_H
1.11 ian 48: #include <inttypes.h>
49: #endif
1.19 chl 50: #include <regex.h>
51: #include <sys/types.h>
1.16 tedu 52: /* Do this here and now, because struct stat gets re-defined on solaris */
53: #include <sys/stat.h>
1.20 ! chl 54: #include <stdarg.h>
1.16 tedu 55:
1.19 chl 56: #define ENABLE_CONDITIONALS
57:
1.16 tedu 58: #ifndef MAGIC
59: #define MAGIC "/etc/magic"
60: #endif
61:
62: #ifdef __EMX__
63: #define PATHSEP ';'
64: #else
65: #define PATHSEP ':'
66: #endif
67:
68: #define private static
69: #ifndef protected
70: #define protected
71: #endif
72: #define public
1.11 ian 73:
1.19 chl 74: #ifndef __GNUC_PREREQ__
75: #ifdef __GNUC__
76: #define __GNUC_PREREQ__(x, y) \
77: ((__GNUC__ == (x) && __GNUC_MINOR__ >= (y)) || \
78: (__GNUC__ > (x)))
79: #else
80: #define __GNUC_PREREQ__(x, y) 0
81: #endif
82: #endif
83:
84: #ifndef MIN
85: #define MIN(a,b) (((a) < (b)) ? (a) : (b))
86: #endif
87:
1.20 ! chl 88: #ifndef MAX
! 89: #define MAX(a,b) (((a) > (b)) ? (a) : (b))
! 90: #endif
! 91:
1.1 deraadt 92: #ifndef HOWMANY
1.19 chl 93: # define HOWMANY (256 * 1024) /* how much of the file to look at */
1.1 deraadt 94: #endif
1.20 ! chl 95: #define MAXMAGIS 8192 /* max entries in any one magic file
! 96: or directory */
! 97: #define MAXDESC 64 /* max leng of text description/MIME type */
1.1 deraadt 98: #define MAXstring 32 /* max leng of "string" types */
99:
1.16 tedu 100: #define MAGICNO 0xF11E041C
1.20 ! chl 101: #define VERSIONNO 5
! 102: #define FILE_MAGICSIZE (32 * 6)
1.16 tedu 103:
104: #define FILE_LOAD 0
105: #define FILE_CHECK 1
106: #define FILE_COMPILE 2
107:
1.1 deraadt 108: struct magic {
1.16 tedu 109: /* Word 1 */
110: uint16_t cont_level; /* level of ">" */
111: uint8_t flag;
1.20 ! chl 112: #define INDIR 0x01 /* if '(...)' appears */
! 113: #define OFFADD 0x02 /* if '>&' or '>...(&' appears */
! 114: #define INDIROFFADD 0x04 /* if '>&(' appears */
! 115: #define UNSIGNED 0x08 /* comparison is unsigned */
! 116: #define NOSPACE 0x10 /* suppress space character before output */
! 117: #define BINTEST 0x20 /* test is for a binary type (set only
! 118: for top-level tests) */
! 119: #define TEXTTEST 0 /* for passing to file_softmagic */
! 120:
! 121: uint8_t dummy1;
1.19 chl 122:
1.16 tedu 123: /* Word 2 */
124: uint8_t reln; /* relation (0=eq, '>'=gt, etc) */
125: uint8_t vallen; /* length of string value, if any */
1.20 ! chl 126: uint8_t type; /* comparison type (FILE_*) */
! 127: uint8_t in_type; /* type of indirection */
1.19 chl 128: #define FILE_INVALID 0
1.16 tedu 129: #define FILE_BYTE 1
130: #define FILE_SHORT 2
1.19 chl 131: #define FILE_DEFAULT 3
1.16 tedu 132: #define FILE_LONG 4
133: #define FILE_STRING 5
134: #define FILE_DATE 6
135: #define FILE_BESHORT 7
136: #define FILE_BELONG 8
137: #define FILE_BEDATE 9
138: #define FILE_LESHORT 10
139: #define FILE_LELONG 11
140: #define FILE_LEDATE 12
141: #define FILE_PSTRING 13
142: #define FILE_LDATE 14
143: #define FILE_BELDATE 15
144: #define FILE_LELDATE 16
145: #define FILE_REGEX 17
1.19 chl 146: #define FILE_BESTRING16 18
147: #define FILE_LESTRING16 19
148: #define FILE_SEARCH 20
149: #define FILE_MEDATE 21
150: #define FILE_MELDATE 22
151: #define FILE_MELONG 23
152: #define FILE_QUAD 24
153: #define FILE_LEQUAD 25
154: #define FILE_BEQUAD 26
155: #define FILE_QDATE 27
156: #define FILE_LEQDATE 28
157: #define FILE_BEQDATE 29
158: #define FILE_QLDATE 30
159: #define FILE_LEQLDATE 31
160: #define FILE_BEQLDATE 32
1.20 ! chl 161: #define FILE_FLOAT 33
! 162: #define FILE_BEFLOAT 34
! 163: #define FILE_LEFLOAT 35
! 164: #define FILE_DOUBLE 36
! 165: #define FILE_BEDOUBLE 37
! 166: #define FILE_LEDOUBLE 38
! 167: #define FILE_NAMES_SIZE 39/* size of array to contain all names */
1.19 chl 168:
169: #define IS_STRING(t) \
170: ((t) == FILE_STRING || \
171: (t) == FILE_PSTRING || \
172: (t) == FILE_BESTRING16 || \
173: (t) == FILE_LESTRING16 || \
174: (t) == FILE_REGEX || \
175: (t) == FILE_SEARCH || \
176: (t) == FILE_DEFAULT)
177:
178: #define FILE_FMT_NONE 0
179: #define FILE_FMT_NUM 1 /* "cduxXi" */
180: #define FILE_FMT_STR 2 /* "s" */
181: #define FILE_FMT_QUAD 3 /* "ll" */
1.20 ! chl 182: #define FILE_FMT_FLOAT 4 /* "eEfFgG" */
! 183: #define FILE_FMT_DOUBLE 5 /* "eEfFgG" */
1.19 chl 184:
1.16 tedu 185: /* Word 3 */
186: uint8_t in_op; /* operator for indirection */
187: uint8_t mask_op; /* operator for mask */
1.19 chl 188: #ifdef ENABLE_CONDITIONALS
189: uint8_t cond; /* conditional type */
1.20 ! chl 190: uint8_t dummy2;
1.19 chl 191: #else
1.16 tedu 192: uint8_t dummy2;
1.20 ! chl 193: uint8_t dummy3;
1.19 chl 194: #endif
195:
1.16 tedu 196: #define FILE_OPS "&|^+-*/%"
197: #define FILE_OPAND 0
198: #define FILE_OPOR 1
199: #define FILE_OPXOR 2
200: #define FILE_OPADD 3
201: #define FILE_OPMINUS 4
202: #define FILE_OPMULTIPLY 5
203: #define FILE_OPDIVIDE 6
204: #define FILE_OPMODULO 7
1.19 chl 205: #define FILE_OPS_MASK 0x07 /* mask for above ops */
206: #define FILE_UNUSED_1 0x08
207: #define FILE_UNUSED_2 0x10
208: #define FILE_UNUSED_3 0x20
209: #define FILE_OPINVERSE 0x40
210: #define FILE_OPINDIRECT 0x80
211:
212: #ifdef ENABLE_CONDITIONALS
213: #define COND_NONE 0
214: #define COND_IF 1
215: #define COND_ELIF 2
216: #define COND_ELSE 3
217: #endif /* ENABLE_CONDITIONALS */
218:
1.16 tedu 219: /* Word 4 */
220: uint32_t offset; /* offset to magic number */
221: /* Word 5 */
1.19 chl 222: int32_t in_offset; /* offset from indirection */
1.16 tedu 223: /* Word 6 */
1.19 chl 224: uint32_t lineno; /* line number in magic file */
225: /* Word 7,8 */
226: union {
227: uint64_t _mask; /* for use with numeric and date types */
228: struct {
229: uint32_t _count; /* repeat/line count */
230: uint32_t _flags; /* modifier flags */
231: } _s; /* for use with string types */
232: } _u;
233: #define num_mask _u._mask
1.20 ! chl 234: #define str_range _u._s._count
1.19 chl 235: #define str_flags _u._s._flags
236:
1.16 tedu 237: /* Words 9-16 */
1.1 deraadt 238: union VALUETYPE {
1.16 tedu 239: uint8_t b;
240: uint16_t h;
1.10 itojun 241: uint32_t l;
1.19 chl 242: uint64_t q;
1.16 tedu 243: uint8_t hs[2]; /* 2 bytes of a fixed-endian "short" */
244: uint8_t hl[4]; /* 4 bytes of a fixed-endian "long" */
1.19 chl 245: uint8_t hq[8]; /* 8 bytes of a fixed-endian "quad" */
246: char s[MAXstring]; /* the search string or regex pattern */
1.20 ! chl 247: float f;
! 248: double d;
1.1 deraadt 249: } value; /* either number or string */
1.16 tedu 250: /* Words 17..31 */
1.1 deraadt 251: char desc[MAXDESC]; /* description */
1.20 ! chl 252: /* Words 32..47 */
! 253: char mimetype[MAXDESC]; /* MIME type */
1.1 deraadt 254: };
255:
1.16 tedu 256: #define BIT(A) (1 << (A))
1.19 chl 257: #define STRING_COMPACT_BLANK BIT(0)
258: #define STRING_COMPACT_OPTIONAL_BLANK BIT(1)
259: #define STRING_IGNORE_LOWERCASE BIT(2)
260: #define STRING_IGNORE_UPPERCASE BIT(3)
261: #define REGEX_OFFSET_START BIT(4)
1.16 tedu 262: #define CHAR_COMPACT_BLANK 'B'
263: #define CHAR_COMPACT_OPTIONAL_BLANK 'b'
1.19 chl 264: #define CHAR_IGNORE_LOWERCASE 'c'
265: #define CHAR_IGNORE_UPPERCASE 'C'
266: #define CHAR_REGEX_OFFSET_START 's'
267: #define STRING_IGNORE_CASE (STRING_IGNORE_LOWERCASE|STRING_IGNORE_UPPERCASE)
1.20 ! chl 268: #define STRING_DEFAULT_RANGE 100
1.16 tedu 269:
270:
271: /* list of magic entries */
272: struct mlist {
273: struct magic *magic; /* array of magic entries */
274: uint32_t nmagic; /* number of entries in array */
275: int mapped; /* allocation type: 0 => apprentice_file
276: * 1 => apprentice_map + malloc
277: * 2 => apprentice_map + mmap */
278: struct mlist *next, *prev;
279: };
1.1 deraadt 280:
1.16 tedu 281: struct magic_set {
1.19 chl 282: struct mlist *mlist;
283: struct cont {
284: size_t len;
285: struct level_info {
286: int32_t off;
287: int got_match;
288: #ifdef ENABLE_CONDITIONALS
289: int last_match;
290: int last_cond; /* used for error checking by parse() */
291: #endif
292: } *li;
293: } c;
294: struct out {
1.20 ! chl 295: char *buf; /* Accumulation buffer */
! 296: char *pbuf; /* Printable buffer */
1.19 chl 297: } o;
298: uint32_t offset;
299: int error;
300: int flags;
301: int haderr;
302: const char *file;
303: size_t line; /* current magic line number */
304:
305: /* data for searches */
306: struct {
307: const char *s; /* start of search in original source */
308: size_t s_len; /* length of search region */
309: size_t offset; /* starting offset in source: XXX - should this be off_t? */
310: size_t rm_len; /* match length */
311: } search;
312:
1.20 ! chl 313: /* FIXME: Make the string dynamically allocated so that e.g.
! 314: strings matched in files can be longer than MAXstring */
1.19 chl 315: union VALUETYPE ms_value; /* either number or string */
1.16 tedu 316: };
1.1 deraadt 317:
1.20 ! chl 318: /* Type for Unicode characters */
! 319: typedef unsigned long unichar;
! 320:
1.16 tedu 321: struct stat;
1.19 chl 322: protected const char *file_fmttime(uint32_t, int);
323: protected int file_buffer(struct magic_set *, int, const char *, const void *,
324: size_t);
1.16 tedu 325: protected int file_fsmagic(struct magic_set *, const char *, struct stat *);
326: protected int file_pipe2file(struct magic_set *, int, const void *, size_t);
327: protected int file_printf(struct magic_set *, const char *, ...);
328: protected int file_reset(struct magic_set *);
1.19 chl 329: protected int file_tryelf(struct magic_set *, int, const unsigned char *,
330: size_t);
331: protected int file_zmagic(struct magic_set *, int, const char *,
332: const unsigned char *, size_t);
1.16 tedu 333: protected int file_ascmagic(struct magic_set *, const unsigned char *, size_t);
334: protected int file_is_tar(struct magic_set *, const unsigned char *, size_t);
1.20 ! chl 335: protected int file_softmagic(struct magic_set *, const unsigned char *, size_t, int);
1.16 tedu 336: protected struct mlist *file_apprentice(struct magic_set *, const char *, int);
1.19 chl 337: protected uint64_t file_signextend(struct magic_set *, struct magic *,
338: uint64_t);
1.16 tedu 339: protected void file_delmagic(struct magic *, int type, size_t entries);
340: protected void file_badread(struct magic_set *);
341: protected void file_badseek(struct magic_set *);
1.19 chl 342: protected void file_oomem(struct magic_set *, size_t);
1.16 tedu 343: protected void file_error(struct magic_set *, int, const char *, ...);
1.19 chl 344: protected void file_magerror(struct magic_set *, const char *, ...);
345: protected void file_magwarn(struct magic_set *, const char *, ...);
1.16 tedu 346: protected void file_mdump(struct magic *);
347: protected void file_showstr(FILE *, const char *, size_t);
348: protected size_t file_mbswidth(const char *);
349: protected const char *file_getbuffer(struct magic_set *);
1.19 chl 350: protected ssize_t sread(int, void *, size_t, int);
351: protected int file_check_mem(struct magic_set *, unsigned int);
1.20 ! chl 352: protected int file_looks_utf8(const unsigned char *, size_t, unichar *, size_t *);
1.19 chl 353:
354: #ifndef COMPILE_ONLY
355: extern const char *file_names[];
356: extern const size_t file_nnames;
357: #endif
358:
359: #ifndef HAVE_STRERROR
360: extern int sys_nerr;
361: extern char *sys_errlist[];
362: #define strerror(e) \
363: (((e) >= 0 && (e) < sys_nerr) ? sys_errlist[(e)] : "Unknown error")
364: #endif
365:
366: #ifndef HAVE_STRTOUL
367: #define strtoul(a, b, c) strtol(a, b, c)
368: #endif
369:
1.20 ! chl 370: #ifndef HAVE_VASPRINTF
! 371: int vasprintf(char **ptr, const char *format_string, va_list vargs);
! 372: #endif
! 373: #ifndef HAVE_ASPRINTF
! 374: int asprintf(char **ptr, const char *format_string, ...);
1.19 chl 375: #endif
1.16 tedu 376:
377: #if defined(HAVE_MMAP) && defined(HAVE_SYS_MMAN_H) && !defined(QUICK)
378: #define QUICK
379: #endif
380:
1.19 chl 381: #ifndef O_BINARY
382: #define O_BINARY 0
383: #endif
384:
1.20 ! chl 385: #ifdef __GNUC__
! 386: static const char *rcsid(const char *) __attribute__((__used__));
! 387: #endif
1.16 tedu 388: #define FILE_RCSID(id) \
389: static const char *rcsid(const char *p) { \
390: return rcsid(p = id); \
391: }
1.3 millert 392:
393: #endif /* __file_h__ */