msooxml 2.5 KB

12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758
  1. #------------------------------------------------------------------------------
  2. # $File: msooxml,v 1.18 2022/08/16 11:16:39 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. !:ext docx
  16. >0 string ppt/ Microsoft PowerPoint 2007+
  17. !:mime application/vnd.openxmlformats-officedocument.presentationml.presentation
  18. !:ext pptx
  19. >0 string xl/ Microsoft Excel 2007+
  20. !:mime application/vnd.openxmlformats-officedocument.spreadsheetml.sheet
  21. !:ext xlsx
  22. >0 string visio/ Microsoft Visio 2013+
  23. !:mime application/vnd.ms-visio.drawing.main+xml
  24. >0 string AppManifest.xaml Microsoft Silverlight Application
  25. !:mime application/x-silverlight-app
  26. # start by checking for ZIP local file header signature
  27. 0 string PK\003\004
  28. !:strength +10
  29. # make sure the first file is correct
  30. >0x1E use msooxml
  31. >0x1E default x
  32. >>0x1E regex \\[Content_Types\\]\\.xml|_rels/\\.rels|docProps|customXml
  33. # skip to the second local file header
  34. # since some documents include a 520-byte extra field following the file
  35. # header, we need to scan for the next header
  36. >>>(18.l+49) search/6000 PK\003\004
  37. # now skip to the *third* local file header; again, we need to scan due to a
  38. # 520-byte extra field following the file header
  39. >>>>&26 search/6000 PK\003\004
  40. # and check the subdirectory name to determine which type of OOXML
  41. # file we have. Correct the mimetype with the registered ones:
  42. # https://technet.microsoft.com/en-us/library/cc179224.aspx
  43. >>>>>&26 use msooxml
  44. >>>>>&26 default x
  45. # OpenOffice/Libreoffice orders ZIP entry differently, so check the 4th file
  46. >>>>>>&26 search/6000 PK\003\004
  47. >>>>>>>&26 use msooxml
  48. # Some OOXML generators add an extra customXml directory. Check another file.
  49. >>>>>>>&26 default x
  50. >>>>>>>>&26 search/6000 PK\003\004
  51. >>>>>>>>>&26 use msooxml
  52. >>>>>>>>>&26 default x Microsoft OOXML
  53. >>>>>>>&26 default x Microsoft OOXML
  54. >>>>>&26 default x Microsoft OOXML