msooxml 2.0 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445
  1. #------------------------------------------------------------------------------
  2. # $File: msooxml,v 1.12 2019/04/19 00:42:27 christos Exp $
  3. # msooxml: file(1) magic for Microsoft Office XML
  4. # From: Ralf Brown <ralf.brown@gmail.com>
  5. # .docx, .pptx, and .xlsx are XML plus other files inside a ZIP
  6. # archive. The first member file is normally "[Content_Types].xml".
  7. # but some libreoffice generated files put this later. Perhaps skip
  8. # the "[Content_Types].xml" test?
  9. # Since MSOOXML doesn't have anything like the uncompressed "mimetype"
  10. # file of ePub or OpenDocument, we'll have to scan for a filename
  11. # which can distinguish between the three types
  12. 0 name msooxml
  13. >0 string word/ Microsoft Word 2007+
  14. !:mime application/vnd.openxmlformats-officedocument.wordprocessingml.document
  15. >0 string ppt/ Microsoft PowerPoint 2007+
  16. !:mime application/vnd.openxmlformats-officedocument.presentationml.presentation
  17. >0 string xl/ Microsoft Excel 2007+
  18. !:mime application/vnd.openxmlformats-officedocument.spreadsheetml.sheet
  19. # start by checking for ZIP local file header signature
  20. 0 string PK\003\004
  21. !:strength +10
  22. # make sure the first file is correct
  23. >0x1E use msooxml
  24. >0x1E regex \\[Content_Types\\]\\.xml|_rels/\\.rels|docProps
  25. # skip to the second local file header
  26. # since some documents include a 520-byte extra field following the file
  27. # header, we need to scan for the next header
  28. >>(18.l+49) search/6000 PK\003\004
  29. # now skip to the *third* local file header; again, we need to scan due to a
  30. # 520-byte extra field following the file header
  31. >>>&26 search/6000 PK\003\004
  32. # and check the subdirectory name to determine which type of OOXML
  33. # file we have. Correct the mimetype with the registered ones:
  34. # https://technet.microsoft.com/en-us/library/cc179224.aspx
  35. >>>>&26 use msooxml
  36. >>>>&26 default x
  37. # OpenOffice/Libreoffice orders ZIP entry differently, so check the 4th file
  38. >>>>>&26 search/6000 PK\003\004
  39. >>>>>>&26 use msooxml
  40. >>>>>>&26 default x Microsoft OOXML
  41. >>>>>&26 default x Microsoft OOXML