juretta.com

require 'win32ole'

# Path to directory with Word Files.
PATH = "file:///c|/path/to/doc/files/"

# converts a word file to docbook XML. 
# The XML file is named after the original file
# e.g.: ABC.doc -> ABC.xml
def convert_word_to_docbook(file, path)
  serviceManager = WIN32OLE.new("com.sun.star.ServiceManager")
  desktop = serviceManager.createInstance("com.sun.star.frame.Desktop")

  url = path + file
  document = desktop.loadComponentFromURL(url, "_blank", 0, [])
  url_to = path + file.gsub(/\.doc/, ".xml")
  fprops = []
  property = serviceManager.Bridge_GetStruct("com.sun.star.beans.PropertyValue")
  property["Name"] = "FilterName"
  property["Value"] = "DocBook File"  
  fprops << property
  begin
    document.storeToUrl(url_to, fprops) # this line works!
  ensure
    document.close true
  end
end

# convert all ".doc" files to DocBook XML
Dir.glob("*.doc").each do |file|
  print "converting #{file}...\n"
  $stdout.flush
  convert_word_to_docbook file, PATH
end