Embed URL and content like blockquote.

crawlquote.rb

# A Liquid tag for Jekyll sites that allows embedding URL.
#
# Usage: {% crawlquote http://www.google.co.jp/ %}

require 'cgi'
require 'digest/md5'
require 'net/https'
require 'uri'
require 'open-uri'
require 'nokogiri'
require 'digest/sha2'

module Jekyll
  class CrawlQuoteTag < Liquid::Tag
    def initialize(tag_name, text, token)
      super
      @text           = text
      @cache_disabled = false
      @cache_folder   = File.expand_path "../.crawl-cache", File.dirname(__FILE__)
      FileUtils.mkdir_p @cache_folder
    end

    def render(context)
      if parts = @text.match(/(https?:\/\/.*)/)
        url = parts[1].strip
        doc = Nokogiri::HTML(cache(url))
        title = doc.css('title').text
        description = doc.css('meta[name="description"]')
        if description
          description = description.attr('content')
        end
        "<blockquote><p>#{description}</p><cite><a href='#{url}'>#{title}</a></cite></blockquote>"
      else
        ""
      end
    end

    def cache(url)
      cp = cache_path(url)
      content = nil
      if File.exist? cp
        content = File.read(cp)
      else
        content = open(url).read
        File.open(cp, "w") do |io|
          io.write content
        end
      end
      content
    end
    
    def cache_path(url)
      dig = Digest::SHA256.hexdigest url
      File.join @cache_folder, "#{dig}.cache"
    end
  end
end

Liquid::Template.register_tag('crawlquote', Jekyll::CrawlQuoteTag)