wikibacon.rb

require 'mechanize'

class BaconSeeker
  def queue; @queue ||= Queue.new; end
  def visited; @visited ||= []; end
  def agent; @agent ||= Mechanize.new; end

  attr_reader :target

  def initialize(url1, url2)
    @target = url2
    queue.push([url1, 1])
  end

  def parse(page)
    agent
      .get("http://en.wikipedia.org#{page}")
      .links
      .map(&:href)
      .select { |link| /^\/wiki\//.match(link) }
      .reject { |link| link.include?(':') }
      .map { |link| link.split(/#/).first }
      .uniq
  end

  def search
    new_links = []
    depth = nil
    until new_links.include?(target)
      url, depth = queue.pop
      puts "Processing #{url}, depth: #{depth}"
      unless visited.include?(url)
        self.visited << url
        new_links = parse(url)
        new_links.each do |link|
          self.queue << [link, depth + 1]
        end
      end
    end
    depth
  end
end

puts BaconSeeker.new("/wiki/Ruby", "/wiki/David_Bowie").search