1.\" $OpenBSD: magic.5,v 1.14 2010/07/15 21:54:20 jmc Exp $ 2.\" 3.\" @(#)$FreeBSD: src/usr.bin/file/magic.5,v 1.11 2000/03/01 12:19:39 sheldonh Exp $ 4.\" 5.\" install as magic.4 on USG, magic.5 on V7 or Berkeley systems. 6.\" 7.\" Copyright (c) Ian F. Darwin 1986-1995. 8.\" Software written by Ian F. Darwin and others; 9.\" maintained 1995-present by Christos Zoulas and others. 10.\" 11.\" Redistribution and use in source and binary forms, with or without 12.\" modification, are permitted provided that the following conditions 13.\" are met: 14.\" 1. Redistributions of source code must retain the above copyright 15.\" notice immediately at the beginning of the file, without modification, 16.\" this list of conditions, and the following disclaimer. 17.\" 2. Redistributions in binary form must reproduce the above copyright 18.\" notice, this list of conditions and the following disclaimer in the 19.\" documentation and/or other materials provided with the distribution. 20.\" 21.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND 22.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 23.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 24.\" ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR 25.\" ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 26.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 27.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 28.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 29.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 30.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 31.\" SUCH DAMAGE. 32.\" 33.Dd $Mdocdate: July 15 2010 $ 34.Dt MAGIC 5 35.Os 36.\" install as magic.4 on USG, magic.5 on V7, Berkeley and Linux systems. 37.Sh NAME 38.Nm magic 39.Nd file command's magic pattern file 40.Sh DESCRIPTION 41This manual page documents the format of the magic file as 42used by the 43.Xr file 1 44command, version 4.24. 45The 46.Xr file 1 47command identifies the type of a file using, 48among other tests, 49a test for whether the file contains certain 50.Dq "magic patterns" . 51The file 52.Pa /etc/magic 53specifies what magic numbers are to be tested for, 54what message to print if a particular magic number is found, 55and additional information to extract from the file. 56.Pp 57Each line of the file specifies a test to be performed. 58A test compares the data starting at a particular offset 59in the file with a byte value, a string or a numeric value. 60If the test succeeds, a message is printed. 61The line consists of the following fields: 62.Bl -tag -width ".Dv message" 63.It Dv offset 64A number specifying the offset, in bytes, into the file of the data 65which is to be tested. 66.It Dv type 67The type of the data to be tested. 68The possible values are: 69.Bl -tag -width ".Dv lestring16" 70.It Dv byte 71A one-byte value. 72.It Dv short 73A two-byte value in this machine's native byte order. 74.It Dv long 75A four-byte value in this machine's native byte order. 76.It Dv quad 77An eight-byte value in this machine's native byte order. 78.It Dv float 79A 32-bit single precision IEEE floating point number in this machine's native byte order. 80.It Dv double 81A 64-bit double precision IEEE floating point number in this machine's native byte order. 82.It Dv string 83A string of bytes. 84The string type specification can be optionally followed 85by /[Bbc]*. 86The 87.Dq B 88flag compacts whitespace in the target, which must 89contain at least one whitespace character. 90If the magic has 91.Dv n 92consecutive blanks, the target needs at least 93.Dv n 94consecutive blanks to match. 95The 96.Dq b 97flag treats every blank in the target as an optional blank. 98Finally the 99.Dq c 100flag, specifies case insensitive matching: lowercase 101characters in the magic match both lower and upper case characters in the 102target, whereas upper case characters in the magic only match uppercase 103characters in the target. 104.It Dv pstring 105A Pascal-style string where the first byte is interpreted as the an 106unsigned length. 107The string is not NUL terminated. 108.It Dv date 109A four-byte value interpreted as a UNIX date. 110.It Dv qdate 111An eight-byte value interpreted as a UNIX date. 112.It Dv ldate 113A four-byte value interpreted as a UNIX-style date, but interpreted as 114local time rather than UTC. 115.It Dv qldate 116An eight-byte value interpreted as a UNIX-style date, but interpreted as 117local time rather than UTC. 118.It Dv beshort 119A two-byte value in big-endian byte order. 120.It Dv belong 121A four-byte value in big-endian byte order. 122.It Dv bequad 123An eight-byte value in big-endian byte order. 124.It Dv befloat 125A 32-bit single precision IEEE floating point number in big-endian byte order. 126.It Dv bedouble 127A 64-bit double precision IEEE floating point number in big-endian byte order. 128.It Dv bedate 129A four-byte value in big-endian byte order, 130interpreted as a Unix date. 131.It Dv beqdate 132An eight-byte value in big-endian byte order, 133interpreted as a Unix date. 134.It Dv beldate 135A four-byte value in big-endian byte order, 136interpreted as a UNIX-style date, but interpreted as local time rather 137than UTC. 138.It Dv beqldate 139An eight-byte value in big-endian byte order, 140interpreted as a UNIX-style date, but interpreted as local time rather 141than UTC. 142.It Dv bestring16 143A two-byte unicode (UCS16) string in big-endian byte order. 144.It Dv leshort 145A two-byte value in little-endian byte order. 146.It Dv lelong 147A four-byte value in little-endian byte order. 148.It Dv lequad 149An eight-byte value in little-endian byte order. 150.It Dv lefloat 151A 32-bit single precision IEEE floating point number in little-endian byte order. 152.It Dv ledouble 153A 64-bit double precision IEEE floating point number in little-endian byte order. 154.It Dv ledate 155A four-byte value in little-endian byte order, 156interpreted as a UNIX date. 157.It Dv leqdate 158An eight-byte value in little-endian byte order, 159interpreted as a UNIX date. 160.It Dv leldate 161A four-byte value in little-endian byte order, 162interpreted as a UNIX-style date, but interpreted as local time rather 163than UTC. 164.It Dv leqldate 165An eight-byte value in little-endian byte order, 166interpreted as a UNIX-style date, but interpreted as local time rather 167than UTC. 168.It Dv lestring16 169A two-byte unicode (UCS16) string in little-endian byte order. 170.It Dv melong 171A four-byte value in middle-endian (PDP-11) byte order. 172.It Dv medate 173A four-byte value in middle-endian (PDP-11) byte order, 174interpreted as a UNIX date. 175.It Dv meldate 176A four-byte value in middle-endian (PDP-11) byte order, 177interpreted as a UNIX-style date, but interpreted as local time rather 178than UTC. 179.It Dv regex 180A regular expression match in extended POSIX regular expression syntax 181(like egrep). 182Regular expressions can take exponential time to process, 183and their performance is hard to predict, so their use is discouraged. 184When used in production environments, 185their performance should be carefully checked. 186The type specification can be optionally followed by 187.Dv /[c][s] . 188The 189.Dq c 190flag makes the match case insensitive, while the 191.Dq s 192flag update the offset to the start offset of the match, rather than the end. 193The regular expression is tested against line 194.Dv N + 1 195onwards, where 196.Dv N 197is the given offset. 198Line endings are assumed to be in the machine's native format. 199.Dv ^ 200and 201.Dv $ 202match the beginning and end of individual lines, respectively, 203not beginning and end of file. 204.It Dv search 205A literal string search starting at the given offset. 206The same modifier flags can be used as for string patterns. 207The modifier flags (if any) must be followed by 208.Dv /number 209the range, that is, the number of positions at which the match will be 210attempted, starting from the start offset. 211This is suitable for searching larger binary expressions 212with variable offsets, using 213.Dv \e 214escapes for special characters. 215The offset works as for regex. 216.It Dv default 217This is intended to be used with the test 218.Em x 219(which is always true) and a message that is to be used if there are 220no other matches. 221.El 222.Pp 223Each top-level magic pattern (see below for an explanation of levels) 224is classified as text or binary according to the types used. 225Types 226.Dq regex 227and 228.Dq search 229are classified as text tests, unless non-printable characters are used 230in the pattern. 231All other tests are classified as binary. 232A top-level pattern is considered to be a test text 233when all its patterns are text 234patterns; otherwise, it is considered to be a binary pattern. 235When matching a file, binary patterns are tried first; if no match is 236found, and the file looks like text, then its encoding is determined 237and the text patterns are tried. 238.Pp 239The numeric types may optionally be followed by 240.Dv & 241and a numeric value, 242to specify that the value is to be AND'ed with the 243numeric value before any comparisons are done. 244Prepending a 245.Dv u 246to the type indicates that ordered comparisons should be unsigned. 247.It Dv test 248The value to be compared with the value from the file. 249If the type is 250numeric, this value 251is specified in C form; if it is a string, it is specified as a C string 252with the usual escapes permitted (e.g. \en for new-line). 253.Pp 254Numeric values 255may be preceded by a character indicating the operation to be performed. 256It may be 257.Dv = , 258to specify that the value from the file must equal the specified value, 259.Dv \*(Lt , 260to specify that the value from the file must be less than the specified 261value, 262.Dv \*(Gt , 263to specify that the value from the file must be greater than the specified 264value, 265.Dv & , 266to specify that the value from the file must have set all of the bits 267that are set in the specified value, 268.Dv ^ , 269to specify that the value from the file must have clear any of the bits 270that are set in the specified value, or 271.Dv ~ , 272the value specified after is negated before tested. 273.Dv x , 274to specify that any value will match. 275If the character is omitted, it is assumed to be 276.Dv = . 277Operators 278.Dv & , 279.Dv ^ , 280and 281.Dv ~ 282don't work with floats and doubles. 283The operator 284.Dv !\& 285specifies that the line matches if the test does 286.Em not 287succeed. 288.Pp 289Numeric values are specified in C form; e.g. 290.Dv 13 291is decimal, 292.Dv 013 293is octal, and 294.Dv 0x13 295is hexadecimal. 296.Pp 297For string values, the string from the 298file must match the specified string. 299The operators 300.Dv = , 301.Dv \*(Lt 302and 303.Dv \*(Gt 304(but not 305.Dv & ) 306can be applied to strings. 307The length used for matching is that of the string argument 308in the magic file. 309This means that a line can match any non-empty string (usually used to 310then print the string), with 311.Em \*(Gt\e0 312(because all non-empty strings are greater than the empty string). 313.Pp 314The special test 315.Em x 316always evaluates to true. 317.Dv message 318The message to be printed if the comparison succeeds. 319If the string contains a 320.Xr printf 3 321format specification, the value from the file (with any specified masking 322performed) is printed using the message as the format string. 323If the string begins with 324.Dq \eb , 325the message printed is the remainder of the string with no whitespace 326added before it: multiple matches are normally separated by a single 327space. 328.El 329.Pp 330A MIME type is given on a separate line, which must be the next 331non-blank or comment line after the magic line that identifies the 332file type, and has the following format: 333.Bd -literal -offset indent 334!:mime MIMETYPE 335.Ed 336.Pp 337i.e. the literal string 338.Dq !:mime 339followed by the MIME type. 340.Pp 341Some file formats contain additional information which is to be printed 342along with the file type or need additional tests to determine the true 343file type. 344These additional tests are introduced by one or more 345.Em \*(Gt 346characters preceding the offset. 347The number of 348.Em \*(Gt 349on the line indicates the level of the test; a line with no 350.Em \*(Gt 351at the beginning is considered to be at level 0. 352Tests are arranged in a tree-like hierarchy: 353If a the test on a line at level 354.Em n 355succeeds, all following tests at level 356.Em n+1 357are performed, and the messages printed if the tests succeed, untile a line 358with level 359.Em n 360(or less) appears. 361For more complex files, one can use empty messages to get just the 362"if/then" effect, in the following way: 363.Bd -literal -offset indent 3640 string MZ 365\*(Gt0x18 leshort \*(Lt0x40 MS-DOS executable 366\*(Gt0x18 leshort \*(Gt0x3f extended PC executable (e.g., MS Windows) 367.Ed 368.Pp 369Offsets do not need to be constant, but can also be read from the file 370being examined. 371If the first character following the last 372.Em \*(Gt 373is a 374.Em \&( 375then the string after the parenthesis is interpreted as an indirect offset. 376That means that the number after the parenthesis is used as an offset in 377the file. 378The value at that offset is read, and is used again as an offset 379in the file. 380Indirect offsets are of the form: 381.Em (( x [.[bslBSL]][+\-][ y ]) . 382The value of 383.Em x 384is used as an offset in the file. 385A byte, short or long is read at that offset depending on the 386.Op bslBSLm 387type specifier. 388The capitalized types interpret the number as a big endian 389value, whereas the small letter versions interpret the number as a little 390endian value; 391the 392.Em m 393type interprets the number as a middle endian (PDP-11) value. 394To that number the value of 395.Em y 396is added and the result is used as an offset in the file. 397The default type if one is not specified is long. 398.Pp 399That way variable length structures can be examined: 400.Bd -literal -offset indent 401# MS Windows executables are also valid MS-DOS executables 4020 string MZ 403\*(Gt0x18 leshort \*(Lt0x40 MZ executable (MS-DOS) 404# skip the whole block below if it is not an extended executable 405\*(Gt0x18 leshort \*(Gt0x3f 406\*(Gt\*(Gt(0x3c.l) string PE\e0\e0 PE executable (MS-Windows) 407\*(Gt\*(Gt(0x3c.l) string LX\e0\e0 LX executable (OS/2) 408.Ed 409.Pp 410This strategy of examining has a drawback: You must make sure that 411you eventually print something, or users may get empty output (like, when 412there is neither PE\e0\e0 nor LE\e0\e0 in the above example) 413.Pp 414If this indirect offset cannot be used directly, simple calculations are 415possible: appending 416.Em [+-*/%&|^]number 417inside parentheses allows one to modify 418the value read from the file before it is used as an offset: 419.Bd -literal -offset indent 420# MS Windows executables are also valid MS-DOS executables 4210 string MZ 422# sometimes, the value at 0x18 is less that 0x40 but there's still an 423# extended executable, simply appended to the file 424\*(Gt0x18 leshort \*(Lt0x40 425\*(Gt\*(Gt(4.s*512) leshort 0x014c COFF executable (MS-DOS, DJGPP) 426\*(Gt\*(Gt(4.s*512) leshort !0x014c MZ executable (MS-DOS) 427.Ed 428.Pp 429Sometimes you do not know the exact offset as this depends on the length or 430position (when indirection was used before) of preceding fields. 431You can specify an offset relative to the end of the last up-level 432field using 433.Sq & 434as a prefix to the offset: 435.Bd -literal -offset indent 4360 string MZ 437\*(Gt0x18 leshort \*(Gt0x3f 438\*(Gt\*(Gt(0x3c.l) string PE\e0\e0 PE executable (MS-Windows) 439# immediately following the PE signature is the CPU type 440\*(Gt\*(Gt\*(Gt&0 leshort 0x14c for Intel 80386 441\*(Gt\*(Gt\*(Gt&0 leshort 0x184 for DEC Alpha 442.Ed 443.Pp 444Indirect and relative offsets can be combined: 445.Bd -literal -offset indent 4460 string MZ 447\*(Gt0x18 leshort \*(Lt0x40 448\*(Gt\*(Gt(4.s*512) leshort !0x014c MZ executable (MS-DOS) 449# if it's not COFF, go back 512 bytes and add the offset taken 450# from byte 2/3, which is yet another way of finding the start 451# of the extended executable 452\*(Gt\*(Gt\*(Gt&(2.s-514) string LE LE executable (MS Windows VxD driver) 453.Ed 454.Pp 455Or the other way around: 456.Bd -literal -offset indent 4570 string MZ 458\*(Gt0x18 leshort \*(Gt0x3f 459\*(Gt\*(Gt(0x3c.l) string LE\e0\e0 LE executable (MS-Windows) 460# at offset 0x80 (-4, since relative offsets start at the end 461# of the up-level match) inside the LE header, we find the absolute 462# offset to the code area, where we look for a specific signature 463\*(Gt\*(Gt\*(Gt(&0x7c.l+0x26) string UPX \eb, UPX compressed 464.Ed 465.Pp 466Or even both! 467.Bd -literal -offset indent 4680 string MZ 469\*(Gt0x18 leshort \*(Gt0x3f 470\*(Gt\*(Gt(0x3c.l) string LE\e0\e0 LE executable (MS-Windows) 471# at offset 0x58 inside the LE header, we find the relative offset 472# to a data area where we look for a specific signature 473\*(Gt\*(Gt\*(Gt&(&0x54.l-3) string UNACE \eb, ACE self-extracting archive 474.Ed 475.Pp 476Finally, if you have to deal with offset/length pairs in your file, even the 477second value in a parenthesized expression can be taken from the file itself, 478using another set of parentheses. 479Note that this additional indirect offset is always relative to the 480start of the main indirect offset. 481.Bd -literal -offset indent 4820 string MZ 483\*(Gt0x18 leshort \*(Gt0x3f 484\*(Gt\*(Gt(0x3c.l) string PE\e0\e0 PE executable (MS-Windows) 485# search for the PE section called ".idata"... 486\*(Gt\*(Gt\*(Gt&0xf4 search/0x140 .idata 487# ...and go to the end of it, calculated from start+length; 488# these are located 14 and 10 bytes after the section name 489\*(Gt\*(Gt\*(Gt\*(Gt(&0xe.l+(-4)) string PK\e3\e4 \eb, ZIP self-extracting archive 490.Ed 491.Sh SEE ALSO 492.Xr file 1 493\- the command that reads this file. 494.Sh BUGS 495The formats 496.Dv long , 497.Dv belong , 498.Dv lelong , 499.Dv melong , 500.Dv short , 501.Dv beshort , 502.Dv leshort , 503.Dv date , 504.Dv bedate , 505.Dv medate , 506.Dv ledate , 507.Dv beldate , 508.Dv leldate , 509and 510.Dv meldate 511are system-dependent; perhaps they should be specified as a number 512of bytes (2B, 4B, etc), 513since the files being recognized typically come from 514a system on which the lengths are invariant. 515.\" 516.\" From: guy@sun.uucp (Guy Harris) 517.\" Newsgroups: net.bugs.usg 518.\" Subject: /etc/magic's format isn't well documented 519.\" Message-ID: <2752@sun.uucp> 520.\" Date: 3 Sep 85 08:19:07 GMT 521.\" Organization: Sun Microsystems, Inc. 522.\" Lines: 136 523.\" 524.\" Here's a manual page for the format accepted by the "file" made by adding 525.\" the changes I posted to the S5R2 version. 526.\" 527.\" Modified for Ian Darwin's version of the file command. 528