wordpress.rb

require 'nokogiri'

module Wordpress
  class Post
    attr_reader :title, :post_date, :body, :description, :tags, :author
    WP_NS = {
      'wp'      => 'http://wordpress.org/export/1.0/',
      'content' => 'http://purl.org/rss/1.0/modules/content/',
      'dc'      => 'http://purl.org/dc/elements/1.1/'
    }

    def initialize(node)
      @node = node
    end

    def title
      @title ||= @node.xpath('title').first.content
    end

    def author
      @author ||= @node.xpath('dc:creator', WP_NS).first.content
    end

    def post_date
      @post_date ||= DateTime.parse(@node.xpath('wp:post_date', WP_NS).first.content)
    end

    def body
      @body ||= @node.xpath('content:encoded').first.content
    end

    def description
      @description ||= postmeta['description']
    end

    def tags
      @tags ||= postmeta['keywords']
    end

    private
    def postmeta
      @postmeta ||= begin
        meta_data = {}
        @node.xpath('wp:postmeta').each do |pm|
          key = pm.children[1].content
          case key
          when 'keywords'
            meta_data['keywords'] = pm.children[3].content.split(',').map {|e| e.strip}
          when 'description'
            meta_data['description'] = pm.children[3].content
          end
        end
        meta_data
      end
    end
  end

  def self.import(path, options={}, &block)
    return if not File.exists?(path)

    doc = Nokogiri::XML(File.read(path))

    doc.xpath('//item').each do |post|
      status    = post.xpath('wp:status', Post::WP_NS).first.content
      next if status != 'publish'

      yield Post.new(post)

      #content.scan(/\<img.+?src="(.+?)"/m).collect {|x| x.first}.each do |image|
      #path = File.basename(image).gsub(/\?.*$/, "")
      #download_path = File.join(images_path, path)

      #if ! File.exists? download_path
      #File.open(download_path, "w") do |f|
      #f.write(open(image).read)
      #end
      #end
      #end
    end
  end
end