posterous_import.rb

# Import a posterous.com Blog into nanoc (http://nanoc.stoneship.org)
# Uses Nokogiri to parse the XML data from the posterous API

require 'rubygems'
require 'nanoc3'
require 'nokogiri'
require 'open-uri'

module Nanoc3::Extra
  module Importers
    class Posterous

      def initialize(hostname, user, password)
        @hostname = hostname
        @user = user
        @password = password
        @site = Nanoc3::Site.new('.')
      end

      def run
        page = 1
        sum = 0
        loop do
          puts "Processing page #{page}..."
          file = open("http://posterous.com/api/readposts?hostname=#{@hostname}&page=#{page}", :http_basic_authentication => [@user, @password])
          doc = Nokogiri::XML(file.read)
          count = process(doc)
          break if count == 0
          page += 1
          sum += count
        end
        puts "Finished. Imported #{sum} articles."
      end

      def process(xml)
        count = 0
        xml.xpath('//rsp/post').each do |post|
          count += 1
          
          date = post.xpath('date').text
          title = post.xpath('title').text
          author = post.xpath('author').text

          body = Nokogiri::HTML(post.xpath('body').text)
          media = []
          body.xpath("//a[contains(@href, 'getfile')]").each do |element|
            media << element.attributes['href'].text
          end

          puts "- #{title}"

          # Build post name
          post_name = title.downcase.
                            gsub(/[^a-z\-_]/, '-').
                            gsub(/^-+|-+|-+$/, '-')
                            
          post_date = Date.parse(date)
          
          content = body.text.gsub('See the full gallery on posterous','').
                              gsub('Watch on posterous','').
                              gsub(/\(\d+ KB\)/,'').
                              strip
                              
          # Create text item
          identifier = '/articles/' + post_date.year.to_s + '/' + '%02d' % post_date.month + '/' + '%02d' % post_date.day + '-' + post_name + '/index/'
          attributes = { :title => title, 
                         :author => author }
          @site.data_sources[0].create_item(content, attributes, identifier, :extension => '.txt')
          
          # Download media and create items
          media.each do |url|
            puts "Download #{url}"

            file = open(url)
            content = file.read
            
            file_name = url.split('/').last.gsub(/\.scaled.+\.jpg/i,'').gsub(/\.jpg/i,'')
            extension = File.extname(url)
            identifier = '/articles/' + post_date.year.to_s + '/' + '%02d' % post_date.month + '/' + '%02d' % post_date.day + '-' + post_name + '/' + file_name + '/'
            @site.data_sources[0].create_item(content, {}, identifier, :extension => extension)
          end
        end
        count
      end
    end
  end
end

begin
  importer = Nanoc3::Extra::Importers::Posterous.new('my_hostname', 'steve', 'secret')
  importer.run
end