quickly extract the text from a MOOX ("OOXML") .docx file

catdocx.rb

#!/opt/local/bin/ruby1.9
require 'zip/zipfilesystem'
require 'nokogiri'

MS_S = "http://schemas.openxmlformats.org/"
MS_W = MS_S + "wordprocessingml/2006/main"
MS_W_OD = MS_S + "officeDocument/2006/relationships/officeDocument"

ARGV.each do |fn|
  Zip::ZipFile.open(fn) do |zf|
    docrels = Nokogiri::XML(zf.read("_rels/.rels"))/"Relationship"
    docrels.each do |rel|
      if rel["Type"] == MS_W_OD
        doc = Nokogiri::XML(zf.read(rel["Target"]))
        puts doc.xpath(".//w:t", "w" => MS_W).map(&:text)
      end
    end
  end
end