Contents

module Janis 

  module Parsing

    module SpecificParsers      
      
      class ProxyWebsiteParser
        
        attr_reader :url

        @@subclasses = []

        def self.subclasses
          @@subclasses
        end

        def self.inherited(subclass)
          add_subclass(subclass)
        end

        def initialize
          @url = self.class.url
        end

        def self.url
          raise "Subclass Responsibility!"
        end

        # It should return an array of strings. Each string should have the format "IP:PORT"
        def parse
          raise "Subclass Responsibility!"
        end

        private

        def obtain_html_doc
          if self.url.include?("http://")
            Nokogiri::HTML(get_content_by_http(self.url))
          elsif url.include?("file://")
            Nokogiri::HTML(read_content_from_file(self.url))
          else
            raise "#{self.url} is not a supported URL!"
          end  
        end

        def get_content_by_http(url)
          open(self.url)
        end

        def read_content_from_file(url)
          File::open(self.url.gsub('file://',''), 'r').read
        end

        def self.add_subclass(subclass)
          @@subclasses << subclass
        end

      end

    end

  end

end

Version data entries

3 entries across 3 versions & 1 rubygems

Version	Path
janis-0.1.2	lib/janis/proxy_website_parser.rb
janis-0.1.1	lib/janis/proxy_website_parser.rb
janis-0.1.0	lib/janis/proxy_website_parser.rb