| 1 | # Magic data for mod_mime_magic Apache module (originally for file(1) command) |
|---|
| 2 | # The module is described in /manual/mod/mod_mime_magic.html |
|---|
| 3 | # |
|---|
| 4 | # The format is 4-5 columns: |
|---|
| 5 | # Column #1: byte number to begin checking from, ">" indicates continuation |
|---|
| 6 | # Column #2: type of data to match |
|---|
| 7 | # Column #3: contents of data to match |
|---|
| 8 | # Column #4: MIME type of result |
|---|
| 9 | # Column #5: MIME encoding of result (optional) |
|---|
| 10 | |
|---|
| 11 | #------------------------------------------------------------------------------ |
|---|
| 12 | # Localstuff: file(1) magic for locally observed files |
|---|
| 13 | # Add any locally observed files here. |
|---|
| 14 | |
|---|
| 15 | #------------------------------------------------------------------------------ |
|---|
| 16 | # end local stuff |
|---|
| 17 | #------------------------------------------------------------------------------ |
|---|
| 18 | |
|---|
| 19 | #------------------------------------------------------------------------------ |
|---|
| 20 | # Java |
|---|
| 21 | |
|---|
| 22 | 0 short 0xcafe |
|---|
| 23 | >2 short 0xbabe application/java |
|---|
| 24 | |
|---|
| 25 | #------------------------------------------------------------------------------ |
|---|
| 26 | # audio: file(1) magic for sound formats |
|---|
| 27 | # |
|---|
| 28 | # from Jan Nicolai Langfeldt <janl@ifi.uio.no>, |
|---|
| 29 | # |
|---|
| 30 | |
|---|
| 31 | # Sun/NeXT audio data |
|---|
| 32 | 0 string .snd |
|---|
| 33 | >12 belong 1 audio/basic |
|---|
| 34 | >12 belong 2 audio/basic |
|---|
| 35 | >12 belong 3 audio/basic |
|---|
| 36 | >12 belong 4 audio/basic |
|---|
| 37 | >12 belong 5 audio/basic |
|---|
| 38 | >12 belong 6 audio/basic |
|---|
| 39 | >12 belong 7 audio/basic |
|---|
| 40 | |
|---|
| 41 | >12 belong 23 audio/x-adpcm |
|---|
| 42 | |
|---|
| 43 | # DEC systems (e.g. DECstation 5000) use a variant of the Sun/NeXT format |
|---|
| 44 | # that uses little-endian encoding and has a different magic number |
|---|
| 45 | # (0x0064732E in little-endian encoding). |
|---|
| 46 | 0 lelong 0x0064732E |
|---|
| 47 | >12 lelong 1 audio/x-dec-basic |
|---|
| 48 | >12 lelong 2 audio/x-dec-basic |
|---|
| 49 | >12 lelong 3 audio/x-dec-basic |
|---|
| 50 | >12 lelong 4 audio/x-dec-basic |
|---|
| 51 | >12 lelong 5 audio/x-dec-basic |
|---|
| 52 | >12 lelong 6 audio/x-dec-basic |
|---|
| 53 | >12 lelong 7 audio/x-dec-basic |
|---|
| 54 | # compressed (G.721 ADPCM) |
|---|
| 55 | >12 lelong 23 audio/x-dec-adpcm |
|---|
| 56 | |
|---|
| 57 | # Bytes 0-3 of AIFF, AIFF-C, & 8SVX audio files are "FORM" |
|---|
| 58 | # AIFF audio data |
|---|
| 59 | 8 string AIFF audio/x-aiff |
|---|
| 60 | # AIFF-C audio data |
|---|
| 61 | 8 string AIFC audio/x-aiff |
|---|
| 62 | # IFF/8SVX audio data |
|---|
| 63 | 8 string 8SVX audio/x-aiff |
|---|
| 64 | |
|---|
| 65 | # Creative Labs AUDIO stuff |
|---|
| 66 | # Standard MIDI data |
|---|
| 67 | 0 string MThd audio/unknown |
|---|
| 68 | #>9 byte >0 (format %d) |
|---|
| 69 | #>11 byte >1 using %d channels |
|---|
| 70 | # Creative Music (CMF) data |
|---|
| 71 | 0 string CTMF audio/unknown |
|---|
| 72 | # SoundBlaster instrument data |
|---|
| 73 | 0 string SBI audio/unknown |
|---|
| 74 | # Creative Labs voice data |
|---|
| 75 | 0 string Creative\ Voice\ File audio/unknown |
|---|
| 76 | ## is this next line right? it came this way... |
|---|
| 77 | #>19 byte 0x1A |
|---|
| 78 | #>23 byte >0 - version %d |
|---|
| 79 | #>22 byte >0 \b.%d |
|---|
| 80 | |
|---|
| 81 | # [GRR 950115: is this also Creative Labs? Guessing that first line |
|---|
| 82 | # should be string instead of unknown-endian long...] |
|---|
| 83 | #0 long 0x4e54524b MultiTrack sound data |
|---|
| 84 | #0 string NTRK MultiTrack sound data |
|---|
| 85 | #>4 long x - version %ld |
|---|
| 86 | |
|---|
| 87 | # Microsoft WAVE format (*.wav) |
|---|
| 88 | # [GRR 950115: probably all of the shorts and longs should be leshort/lelong] |
|---|
| 89 | # Microsoft RIFF |
|---|
| 90 | 0 string RIFF audio/unknown |
|---|
| 91 | # - WAVE format |
|---|
| 92 | >8 string WAVE audio/x-wav |
|---|
| 93 | # MPEG audio. |
|---|
| 94 | 0 beshort&0xfff0 0xfff0 audio/mpeg |
|---|
| 95 | # C64 SID Music files, from Linus Walleij <triad@df.lth.se> |
|---|
| 96 | 0 string PSID audio/prs.sid |
|---|
| 97 | |
|---|
| 98 | #------------------------------------------------------------------------------ |
|---|
| 99 | # c-lang: file(1) magic for C programs or various scripts |
|---|
| 100 | # |
|---|
| 101 | |
|---|
| 102 | # XPM icons (Greg Roelofs, newt@uchicago.edu) |
|---|
| 103 | # ideally should go into "images", but entries below would tag XPM as C source |
|---|
| 104 | 0 string /*\ XPM image/x-xbm 7bit |
|---|
| 105 | |
|---|
| 106 | # this first will upset you if you're a PL/1 shop... (are there any left?) |
|---|
| 107 | # in which case rm it; ascmagic will catch real C programs |
|---|
| 108 | # C or REXX program text |
|---|
| 109 | 0 string /* text/plain |
|---|
| 110 | # C++ program text |
|---|
| 111 | 0 string // text/plain |
|---|
| 112 | |
|---|
| 113 | #------------------------------------------------------------------------------ |
|---|
| 114 | # compress: file(1) magic for pure-compression formats (no archives) |
|---|
| 115 | # |
|---|
| 116 | # compress, gzip, pack, compact, huf, squeeze, crunch, freeze, yabba, whap, etc. |
|---|
| 117 | # |
|---|
| 118 | # Formats for various forms of compressed data |
|---|
| 119 | # Formats for "compress" proper have been moved into "compress.c", |
|---|
| 120 | # because it tries to uncompress it to figure out what's inside. |
|---|
| 121 | |
|---|
| 122 | # standard unix compress |
|---|
| 123 | 0 string \037\235 application/octet-stream x-compress |
|---|
| 124 | |
|---|
| 125 | # gzip (GNU zip, not to be confused with [Info-ZIP/PKWARE] zip archiver) |
|---|
| 126 | 0 string \037\213 application/octet-stream x-gzip |
|---|
| 127 | |
|---|
| 128 | # According to gzip.h, this is the correct byte order for packed data. |
|---|
| 129 | 0 string \037\036 application/octet-stream |
|---|
| 130 | # |
|---|
| 131 | # This magic number is byte-order-independent. |
|---|
| 132 | # |
|---|
| 133 | 0 short 017437 application/octet-stream |
|---|
| 134 | |
|---|
| 135 | # XXX - why *two* entries for "compacted data", one of which is |
|---|
| 136 | # byte-order independent, and one of which is byte-order dependent? |
|---|
| 137 | # |
|---|
| 138 | # compacted data |
|---|
| 139 | 0 short 0x1fff application/octet-stream |
|---|
| 140 | 0 string \377\037 application/octet-stream |
|---|
| 141 | # huf output |
|---|
| 142 | 0 short 0145405 application/octet-stream |
|---|
| 143 | |
|---|
| 144 | # Squeeze and Crunch... |
|---|
| 145 | # These numbers were gleaned from the Unix versions of the programs to |
|---|
| 146 | # handle these formats. Note that I can only uncrunch, not crunch, and |
|---|
| 147 | # I didn't have a crunched file handy, so the crunch number is untested. |
|---|
| 148 | # Keith Waclena <keith@cerberus.uchicago.edu> |
|---|
| 149 | #0 leshort 0x76FF squeezed data (CP/M, DOS) |
|---|
| 150 | #0 leshort 0x76FE crunched data (CP/M, DOS) |
|---|
| 151 | |
|---|
| 152 | # Freeze |
|---|
| 153 | #0 string \037\237 Frozen file 2.1 |
|---|
| 154 | #0 string \037\236 Frozen file 1.0 (or gzip 0.5) |
|---|
| 155 | |
|---|
| 156 | # lzh? |
|---|
| 157 | #0 string \037\240 LZH compressed data |
|---|
| 158 | |
|---|
| 159 | #------------------------------------------------------------------------------ |
|---|
| 160 | # frame: file(1) magic for FrameMaker files |
|---|
| 161 | # |
|---|
| 162 | # This stuff came on a FrameMaker demo tape, most of which is |
|---|
| 163 | # copyright, but this file is "published" as witness the following: |
|---|
| 164 | # |
|---|
| 165 | 0 string \<MakerFile application/x-frame |
|---|
| 166 | 0 string \<MIFFile application/x-frame |
|---|
| 167 | 0 string \<MakerDictionary application/x-frame |
|---|
| 168 | 0 string \<MakerScreenFon application/x-frame |
|---|
| 169 | 0 string \<MML application/x-frame |
|---|
| 170 | 0 string \<Book application/x-frame |
|---|
| 171 | 0 string \<Maker application/x-frame |
|---|
| 172 | |
|---|
| 173 | #------------------------------------------------------------------------------ |
|---|
| 174 | # html: file(1) magic for HTML (HyperText Markup Language) docs |
|---|
| 175 | # |
|---|
| 176 | # from Daniel Quinlan <quinlan@yggdrasil.com> |
|---|
| 177 | # and Anna Shergold <anna@inext.co.uk> |
|---|
| 178 | # |
|---|
| 179 | 0 string \<!DOCTYPE\ HTML text/html |
|---|
| 180 | 0 string \<!doctype\ html text/html |
|---|
| 181 | 0 string \<HEAD text/html |
|---|
| 182 | 0 string \<head text/html |
|---|
| 183 | 0 string \<TITLE text/html |
|---|
| 184 | 0 string \<title text/html |
|---|
| 185 | 0 string \<html text/html |
|---|
| 186 | 0 string \<HTML text/html |
|---|
| 187 | 0 string \<!-- text/html |
|---|
| 188 | 0 string \<h1 text/html |
|---|
| 189 | 0 string \<H1 text/html |
|---|
| 190 | |
|---|
| 191 | # XML eXtensible Markup Language, from Linus Walleij <triad@df.lth.se> |
|---|
| 192 | 0 string \<?xml text/xml |
|---|
| 193 | |
|---|
| 194 | #------------------------------------------------------------------------------ |
|---|
| 195 | # images: file(1) magic for image formats (see also "c-lang" for XPM bitmaps) |
|---|
| 196 | # |
|---|
| 197 | # originally from jef@helios.ee.lbl.gov (Jef Poskanzer), |
|---|
| 198 | # additions by janl@ifi.uio.no as well as others. Jan also suggested |
|---|
| 199 | # merging several one- and two-line files into here. |
|---|
| 200 | # |
|---|
| 201 | # XXX - byte order for GIF and TIFF fields? |
|---|
| 202 | # [GRR: TIFF allows both byte orders; GIF is probably little-endian] |
|---|
| 203 | # |
|---|
| 204 | |
|---|
| 205 | # [GRR: what the hell is this doing in here?] |
|---|
| 206 | #0 string xbtoa btoa'd file |
|---|
| 207 | |
|---|
| 208 | # PBMPLUS |
|---|
| 209 | # PBM file |
|---|
| 210 | 0 string P1 image/x-portable-bitmap 7bit |
|---|
| 211 | # PGM file |
|---|
| 212 | 0 string P2 image/x-portable-greymap 7bit |
|---|
| 213 | # PPM file |
|---|
| 214 | 0 string P3 image/x-portable-pixmap 7bit |
|---|
| 215 | # PBM "rawbits" file |
|---|
| 216 | 0 string P4 image/x-portable-bitmap |
|---|
| 217 | # PGM "rawbits" file |
|---|
| 218 | 0 string P5 image/x-portable-greymap |
|---|
| 219 | # PPM "rawbits" file |
|---|
| 220 | 0 string P6 image/x-portable-pixmap |
|---|
| 221 | |
|---|
| 222 | # NIFF (Navy Interchange File Format, a modification of TIFF) |
|---|
| 223 | # [GRR: this *must* go before TIFF] |
|---|
| 224 | 0 string IIN1 image/x-niff |
|---|
| 225 | |
|---|
| 226 | # TIFF and friends |
|---|
| 227 | # TIFF file, big-endian |
|---|
| 228 | 0 string MM image/tiff |
|---|
| 229 | # TIFF file, little-endian |
|---|
| 230 | 0 string II image/tiff |
|---|
| 231 | |
|---|
| 232 | # possible GIF replacements; none yet released! |
|---|
| 233 | # (Greg Roelofs, newt@uchicago.edu) |
|---|
| 234 | # |
|---|
| 235 | # GRR 950115: this was mine ("Zip GIF"): |
|---|
| 236 | # ZIF image (GIF+deflate alpha) |
|---|
| 237 | 0 string GIF94z image/unknown |
|---|
| 238 | # |
|---|
| 239 | # GRR 950115: this is Jeremy Wohl's Free Graphics Format (better): |
|---|
| 240 | # FGF image (GIF+deflate beta) |
|---|
| 241 | 0 string FGF95a image/unknown |
|---|
| 242 | # |
|---|
| 243 | # GRR 950115: this is Thomas Boutell's Portable Bitmap Format proposal |
|---|
| 244 | # (best; not yet implemented): |
|---|
| 245 | # PBF image (deflate compression) |
|---|
| 246 | 0 string PBF image/unknown |
|---|
| 247 | |
|---|
| 248 | # GIF |
|---|
| 249 | 0 string GIF image/gif |
|---|
| 250 | |
|---|
| 251 | # JPEG images |
|---|
| 252 | 0 beshort 0xffd8 image/jpeg |
|---|
| 253 | |
|---|
| 254 | # PC bitmaps (OS/2, Windoze BMP files) (Greg Roelofs, newt@uchicago.edu) |
|---|
| 255 | 0 string BM image/bmp |
|---|
| 256 | #>14 byte 12 (OS/2 1.x format) |
|---|
| 257 | #>14 byte 64 (OS/2 2.x format) |
|---|
| 258 | #>14 byte 40 (Windows 3.x format) |
|---|
| 259 | #0 string IC icon |
|---|
| 260 | #0 string PI pointer |
|---|
| 261 | #0 string CI color icon |
|---|
| 262 | #0 string CP color pointer |
|---|
| 263 | #0 string BA bitmap array |
|---|
| 264 | |
|---|
| 265 | |
|---|
| 266 | #------------------------------------------------------------------------------ |
|---|
| 267 | # lisp: file(1) magic for lisp programs |
|---|
| 268 | # |
|---|
| 269 | # various lisp types, from Daniel Quinlan (quinlan@yggdrasil.com) |
|---|
| 270 | 0 string ;; text/plain 8bit |
|---|
| 271 | # Emacs 18 - this is always correct, but not very magical. |
|---|
| 272 | 0 string \012( application/x-elc |
|---|
| 273 | # Emacs 19 |
|---|
| 274 | 0 string ;ELC\023\000\000\000 application/x-elc |
|---|
| 275 | |
|---|
| 276 | #------------------------------------------------------------------------------ |
|---|
| 277 | # mail.news: file(1) magic for mail and news |
|---|
| 278 | # |
|---|
| 279 | # There are tests to ascmagic.c to cope with mail and news. |
|---|
| 280 | 0 string Relay-Version: message/rfc822 7bit |
|---|
| 281 | 0 string #!\ rnews message/rfc822 7bit |
|---|
| 282 | 0 string N#!\ rnews message/rfc822 7bit |
|---|
| 283 | 0 string Forward\ to message/rfc822 7bit |
|---|
| 284 | 0 string Pipe\ to message/rfc822 7bit |
|---|
| 285 | 0 string Return-Path: message/rfc822 7bit |
|---|
| 286 | 0 string Path: message/news 8bit |
|---|
| 287 | 0 string Xref: message/news 8bit |
|---|
| 288 | 0 string From: message/rfc822 7bit |
|---|
| 289 | 0 string Article message/news 8bit |
|---|
| 290 | #------------------------------------------------------------------------------ |
|---|
| 291 | # msword: file(1) magic for MS Word files |
|---|
| 292 | # |
|---|
| 293 | # Contributor claims: |
|---|
| 294 | # Reversed-engineered MS Word magic numbers |
|---|
| 295 | # |
|---|
| 296 | |
|---|
| 297 | 0 string \376\067\0\043 application/msword |
|---|
| 298 | 0 string \333\245-\0\0\0 application/msword |
|---|
| 299 | |
|---|
| 300 | # disable this one because it applies also to other |
|---|
| 301 | # Office/OLE documents for which msword is not correct. See PR#2608. |
|---|
| 302 | #0 string \320\317\021\340\241\261 application/msword |
|---|
| 303 | |
|---|
| 304 | |
|---|
| 305 | |
|---|
| 306 | #------------------------------------------------------------------------------ |
|---|
| 307 | # printer: file(1) magic for printer-formatted files |
|---|
| 308 | # |
|---|
| 309 | |
|---|
| 310 | # PostScript |
|---|
| 311 | 0 string %! application/postscript |
|---|
| 312 | 0 string \004%! application/postscript |
|---|
| 313 | |
|---|
| 314 | # Acrobat |
|---|
| 315 | # (due to clamen@cs.cmu.edu) |
|---|
| 316 | 0 string %PDF- application/pdf |
|---|
| 317 | |
|---|
| 318 | #------------------------------------------------------------------------------ |
|---|
| 319 | # sc: file(1) magic for "sc" spreadsheet |
|---|
| 320 | # |
|---|
| 321 | 38 string Spreadsheet application/x-sc |
|---|
| 322 | |
|---|
| 323 | #------------------------------------------------------------------------------ |
|---|
| 324 | # tex: file(1) magic for TeX files |
|---|
| 325 | # |
|---|
| 326 | # XXX - needs byte-endian stuff (big-endian and little-endian DVI?) |
|---|
| 327 | # |
|---|
| 328 | # From <conklin@talisman.kaleida.com> |
|---|
| 329 | |
|---|
| 330 | # Although we may know the offset of certain text fields in TeX DVI |
|---|
| 331 | # and font files, we can't use them reliably because they are not |
|---|
| 332 | # zero terminated. [but we do anyway, christos] |
|---|
| 333 | 0 string \367\002 application/x-dvi |
|---|
| 334 | #0 string \367\203 TeX generic font data |
|---|
| 335 | #0 string \367\131 TeX packed font data |
|---|
| 336 | #0 string \367\312 TeX virtual font data |
|---|
| 337 | #0 string This\ is\ TeX, TeX transcript text |
|---|
| 338 | #0 string This\ is\ METAFONT, METAFONT transcript text |
|---|
| 339 | |
|---|
| 340 | # There is no way to detect TeX Font Metric (*.tfm) files without |
|---|
| 341 | # breaking them apart and reading the data. The following patterns |
|---|
| 342 | # match most *.tfm files generated by METAFONT or afm2tfm. |
|---|
| 343 | #2 string \000\021 TeX font metric data |
|---|
| 344 | #2 string \000\022 TeX font metric data |
|---|
| 345 | #>34 string >\0 (%s) |
|---|
| 346 | |
|---|
| 347 | # Texinfo and GNU Info, from Daniel Quinlan (quinlan@yggdrasil.com) |
|---|
| 348 | #0 string \\input\ texinfo Texinfo source text |
|---|
| 349 | #0 string This\ is\ Info\ file GNU Info text |
|---|
| 350 | |
|---|
| 351 | # correct TeX magic for Linux (and maybe more) |
|---|
| 352 | # from Peter Tobias (tobias@server.et-inf.fho-emden.de) |
|---|
| 353 | # |
|---|
| 354 | 0 leshort 0x02f7 application/x-dvi |
|---|
| 355 | |
|---|
| 356 | # RTF - Rich Text Format |
|---|
| 357 | 0 string {\\rtf application/rtf |
|---|
| 358 | |
|---|
| 359 | #------------------------------------------------------------------------------ |
|---|
| 360 | # animation: file(1) magic for animation/movie formats |
|---|
| 361 | # |
|---|
| 362 | # animation formats, originally from vax@ccwf.cc.utexas.edu (VaX#n8) |
|---|
| 363 | # MPEG file |
|---|
| 364 | 0 string \000\000\001\263 video/mpeg |
|---|
| 365 | # |
|---|
| 366 | # The contributor claims: |
|---|
| 367 | # I couldn't find a real magic number for these, however, this |
|---|
| 368 | # -appears- to work. Note that it might catch other files, too, |
|---|
| 369 | # so BE CAREFUL! |
|---|
| 370 | # |
|---|
| 371 | # Note that title and author appear in the two 20-byte chunks |
|---|
| 372 | # at decimal offsets 2 and 22, respectively, but they are XOR'ed with |
|---|
| 373 | # 255 (hex FF)! DL format SUCKS BIG ROCKS. |
|---|
| 374 | # |
|---|
| 375 | # DL file version 1 , medium format (160x100, 4 images/screen) |
|---|
| 376 | 0 byte 1 video/unknown |
|---|
| 377 | 0 byte 2 video/unknown |
|---|
| 378 | # Quicktime video, from Linus Walleij <triad@df.lth.se> |
|---|
| 379 | # from Apple quicktime file format documentation. |
|---|
| 380 | 4 string moov video/quicktime |
|---|
| 381 | 4 string mdat video/quicktime |
|---|
| 382 | |
|---|
