Working in progress script used to create ePub xml files from CHM hhc file.

hhc2xml.rb

#!/usr/bin/env ruby
$KCODE = 'UTF8'
require 'rubygems'
require 'hpricot'
require 'fileutils'
require 'builder'

=begin
def print_tree(ul, level)
  ul.search("/li").each do |li|
    puts "\t" * level + li.search("object/param[@name='Name']").first.attributes["value"]
    inner_ul = li.at("/ul")
    unless inner_ul.nil?
      inner_level = level + 1
      print_tree(inner_ul, inner_level)
    end    
  end
end
=end

if ARGV.size != 2
  puts "Usage: #{File.basename(__FILE__)} filename.hhc epub_data_dir"
  exit 1
end

begin
  hhcfile = open ARGV[0]
rescue Exception => e
  puts "File: \"#{ARGV[0]}\" is not exist."
  exit 1
end

ARGV[1].nil? ? epub_data_dir = FileUtils.pwd : epub_data_dir = ARGV[1]

FileUtils.cd epub_data_dir do
  # Common Variables
  uuid = `uuidgen`.strip.downcase
  book_title = "Advanced Programming in the UNIX Environment, 2nd Edition"
  language = "en-US"
  rights = "Commercial"
  isbn = "0-20-143307-9"
  subjects = "C, Unix, gcc, Linux"
  author = "W. Richard Stevens, Stephen A. Rago"
  publisher = "Addison Wesley Professional"
  html_dir = "OEBPS"
  cover = "cover.png"

  # toc.ncx
  doc = Hpricot(hhcfile)
  @@order = 0
  xml = Builder::XmlMarkup.new(:indent => 2)
  xml.instruct!(:xml, :encoding => "UTF-8")
  xml.ncx :xmlns => "http://www.daisy.org/z3986/2005/ncx/", :version => "2005-1" do |ncx|
    ncx.head do |head|
      head.meta :name => "dtb:uid", :content => uuid # Change later
      head.meta :name => "dtb:depth", :content => "-1" # Change later
      head.meta :name => "dtb:totalPageCount", :content => "0"
      head.meta :name => "dtb:maxPageNumber", :content => "0"
    end  
    ncx.docTitle do |docTitle|
      docTitle.text book_title # Doc title
    end
    @@order = 1
    # Inner function to do recursive
    def build_struct(builder, ul)
      ul.search("/li").each do |li|
        name = li.search("object/param[@name='Name']").first.attributes["value"]
        src = li.search("object/param[@name='Local']").first.attributes["value"]
        # Make a navPoint Anyway
        builder.navPoint(:id => File.basename(src).gsub(".", "_"), :playOrder => "#{@@order}") do |navPoint|
          navPoint.navLabel do |navLabel|
            navLabel.text name
          end
          navPoint.content(:src => src.split("/").last) # Fixme detecting folder structure!!!
          # Increment the navPoint number
          @@order += 1
          # Do inner ul generation
          inner_ul = li.at("/ul")
          unless inner_ul.nil?
            build_struct(navPoint, inner_ul)
          end
        end
      end
    end
    # Generate Table of Content
    ncx.navMap do |navMap|
      build_struct(navMap, doc.at("body/ul"))
    end
  end

  # Write to a file
  ncxfile = open "OEBPS/toc.ncx", "w"
  ncxfile.puts xml.target!
  ncxfile.close

  #content.opf
  mimetypes = {
    ".ncx" => "application/x-dtbncx+xml",
    ".html" => "application/xhtml+xml",
    ".xhtml" => "application/xhtml+xml",
    ".htm" => "application/xhtml+xml",
    ".css" => "text/css",
    ".jpg" => "image/jpeg",
    ".jpeg" => "image/jpeg",
    ".jpe" => "image/jpeg",
    ".png" => "image/png",
    ".gif" => "image/gif",
    ".svg" => "image/svg+xml"
  }

  xml = Builder::XmlMarkup.new(:indent => 2)
  xml.instruct!(:xml, :encoding => "UTF-8")
  xml.package :version => "2.0", :xmlns => "http://www.idpf.org/2007/opf",
              :"unique-identifier" => "BookId" do |package|
    package.metadata :"xmlns:dc" => "http://purl.org/dc/elements/1.1/", 
                     :"xmlns:opf" => "http://www.idpf.org/2007/opf" do |metadata|
      metadata.__send__ "dc:title", book_title
      metadata.__send__ "dc:language", language
      metadata.__send__ "dc:rights", rights
      metadata.__send__ "dc:identifier", uuid, {:id => "BookId"}
      metadata.__send__ "dc:identifier", isbn, {:"opf:scheme" => "ISBN"}
      subjects.split(",").each do |sub|
        metadata.__send__ "dc:subject", sub.strip
      end
      metadata.__send__ "dc:creator", author, {:"opf:role" => "aut"}
      metadata.__send__ "dc:publisher", publisher
      # Cover "content" should be the same to the cover image id 
      metadata.meta :name => "cover", :content => cover.gsub(".", "_")
    end
    package.manifest do |manifest|
      # If ncx file not generated when script created, then, uncomment following line
      manifest.item :id => "ncx", :href => "toc.ncx", :"media-type" => mimetypes[".ncx"]
      FileUtils.cd(html_dir) do
        Dir["**/*"].each do |filename|
          ext = File.extname(filename).downcase # Downcase it to mach the dict key
          next if !File.file?(filename)
          next if ext == "opf"
          next if ext == "ncx"
          next if ext == "" || mimetypes[ext].nil?
          # generate ID with the file path to ensure uniqness
          manifest.item :id => filename.gsub(/[\/.\[\]]/, "_").reverse , :href => filename, :"media-type" => mimetypes[ext]
        end
      end
    end
    package.spine :toc => "ncx" do |spine|
      doc.search("li").each do |li|
        src = li.search("object/param[@name='Local']").first.attributes["value"]
        # idref should get rid of the abs path from the hhc file, so use basename to precess
        spine.itemref :idref => File.basename(src).gsub(/[\/.]/, "_").reverse
      end
    end
  end

  #puts xml.target!

  # Write to a file
  opffile = open "OEBPS/content.opf", "w"
  opffile.puts xml.target!
  opffile.close

end