Given an array of URLs, outputs an OPML file to STDOUT

opml_generator.rb

# blog post: http://blog.slashpoundbang.com/post/3385815540/how-to-generate-an-opml-file-with-ruby

require 'open-uri'
require 'builder'
require 'truffle-hog'
require 'nokogiri'

# CHANGE ME
urls = %w(http://example.com/ http://example.org/)

xml = Builder::XmlMarkup.new(:target => STDOUT, :indent => 2)
xml.instruct!
xml.opml(:version => 1.1) do
  xml.head do
    # CHANGE ME
    xml.title 'OPML TITLE'
    xml.dateCreated 'CREATION DATE'
    xml.dateModified Time.now.httpdate
    xml.ownerName 'YOUR NAME'
    xml.ownerEmail 'example@example.com'
    # ... and you're done!
  end
  xml.body do
    urls.each do |url|
      html = open(url).read
      title = Nokogiri::HTML(html).at_css('title').text.strip
      xml.outline({
        :type => 'rss',
        :version => 'RSS',
        :description => '',
        :title => title,
        :text => title,
        :htmlUrl => url,
        :xmlUrl => TruffleHog.parse_feed_urls(html, :atom).uniq.reject{|x| x =~ %r{/comments/}}.first,
      })
    end
  end
end