1 2#------------------------------------------------------------------------------ 3# $File: msooxml,v 1.7 2018/03/12 12:38:59 christos Exp $ 4# msooxml: file(1) magic for Microsoft Office XML 5# From: Ralf Brown <ralf.brown@gmail.com> 6 7# .docx, .pptx, and .xlsx are XML plus other files inside a ZIP 8# archive. The first member file is normally "[Content_Types].xml". 9# but some libreoffice generated files put this later. Perhaps skip 10# the "[Content_Types].xml" test? 11# Since MSOOXML doesn't have anything like the uncompressed "mimetype" 12# file of ePub or OpenDocument, we'll have to scan for a filename 13# which can distinguish between the three types 14 150 name msooxml 16>0 string word/ Microsoft Word 2007+ 17!:mime application/vnd.openxmlformats-officedocument.wordprocessingml.document 18>0 string ppt/ Microsoft PowerPoint 2007+ 19!:mime application/vnd.openxmlformats-officedocument.presentationml.presentation 20>0 string xl/ Microsoft Excel 2007+ 21!:mime application/vnd.openxmlformats-officedocument.spreadsheetml.sheet 22 23# start by checking for ZIP local file header signature 240 string PK\003\004 25!:strength +10 26# make sure the first file is correct 27>0x1E regex \\[Content_Types\\]\\.xml|_rels/\\.rels 28# skip to the second local file header 29# since some documents include a 520-byte extra field following the file 30# header, we need to scan for the next header 31>>(18.l+49) search/2000 PK\003\004 32# now skip to the *third* local file header; again, we need to scan due to a 33# 520-byte extra field following the file header 34>>>&26 search/1000 PK\003\004 35# and check the subdirectory name to determine which type of OOXML 36# file we have. Correct the mimetype with the registered ones: 37# http://technet.microsoft.com/en-us/library/cc179224.aspx 38>>>>&26 use msooxml 39>>>>&26 default x 40# OpenOffice/Libreoffice orders ZIP entry differently, so check the 4th file 41>>>>>&26 search/1000 PK\003\004 42>>>>>>&26 use msooxml 43>>>>>>&26 default x Microsoft OOXML 44