Contents

require 'content_urls'

# Alternative to Anemone::Page which handles HTML pages and other ccontent types.

module Anemone
  class Resource < Anemone::Page

    #
    # Array of distinct URLs from the resource
    #
    def links
      return @links unless @links.nil?
      @links = []
      ContentUrls.urls(body, content_type).each do |u|
        next if u.nil? or u.empty?
        abs = to_absolute(u) rescue next
        @links << abs if in_domain?(abs)
      end
      @links.uniq!
      @links
    end

    #
    # Base URI from the HTML doc head element
    #
    # Anemone::Base#to_absolute expects an instance of URI or nil
    #
    def base
      return @base unless @body_parsed.nil?
      @body_parsed = true

      base = ContentUrls.base_url(body, content_type)

      return @base if base.nil?
      base = URI(base) unless base.nil? rescue nil
      @base = base unless base.to_s.empty?

      @base
    end

  end
end

Version data entries

3 entries across 3 versions & 1 rubygems

Version	Path
sutch-anemone-0.7.2.2	lib/anemone/resource.rb
sutch-anemone-0.7.2.1	lib/anemone/resource.rb
sutch-anemone-0.7.2	lib/anemone/resource.rb