markdown_processor.rb

module Jekyll

  require 'html/pipeline'

  class LangFilter < HTML::Pipeline::Filter

    def call
        doc.css('pre[lang]').each do |node|
            node['data-lang'] = node['lang']
            node.remove_attribute('lang')
        end

        doc
    end
  end

    class ApiHeaderIdFilter < HTML::Pipeline::Filter

      def call
          # Headers in the new API design
          doc.css('div.api-article').each do |node|
            node = node.next_element
            until node.nil?
              if node.name == 'h3'
                node['id'] = node.text == 'Related Properties' ? 'related-properties' : sanitaze_id(node.text)
              end

              node = node.next_element
            end
          end

          # Headers in the old API design
          doc.css('h2').each do |node|
              text = node.text

              next unless text =~ /^(Configuration|Events|Properties|Methods|Class Methods|Fields|Constructor Parameters)$/

              prefix = text.downcase.gsub(' ', '-')

              node = node.next_element

              until node.nil?
                  break if node.name == 'h2'

                  if node.name == 'h3'
                      id = sanitaze_id(node.text)
                      node['id'] = "#{prefix}-#{id}"
                  end

                  node = node.next_element
              end

          end
          doc
      end

      def sanitaze_id(id)
        id.gsub!(/ .*/, '')
        id.gsub!(/`[^`]*`/, '')
        id.gsub!(/\\/,'')
        id.gsub!(/\*[^*]*\*/, '')
        id
      end
    end


  class RootRelativeFilter < HTML::Pipeline::Filter

    def call
        doc.search('a').each do |a|
            next if a['href'].nil?

            href = a['href'].strip

            if href.start_with? '/'
                a['href'] = context[:baseurl] + href unless href.start_with? context[:baseurl]
            end
        end

        doc.search('img').each do |img|
            next if img['src'].nil?

            src = img['src'].strip

            if src.start_with? '/'
                img['src'] = context[:baseurl] + src unless src.start_with? context[:baseurl]
            end
        end

        doc
      end
  end

  class HeaderLinkFilter < HTML::Pipeline::Filter

    PUNCTUATION_REGEXP = /[^\p{Word}\- ]/u

    def call()

        doc.css('h1, h2, h3').each do |heading|

            desc_node = heading.children.first()
            if desc_node
                id = desc_node.text
            else
                id = heading.text
            end

            id = id.downcase.strip
            id.gsub!(PUNCTUATION_REGEXP, '') # remove punctuation
            id.gsub!(' ', '-') # replace spaces with dash

            heading['id'] = id

            a = Nokogiri::XML::Node.new('a', doc)
            a['href'] = "##{id}"

            if desc_node
                a.children = desc_node
            end

            if next_child = heading.children.first()
                next_child.before(a)
            else
                heading.add_child a
            end

        end

        doc
    end
  end

  class Converters::Markdown::MarkdownProcessor
      def initialize(config)
          @config = config
          context = {
            :gfm => false,
            :baseurl => @config['baseurl'],
          }
          
          if @config['code_lang']
            @pipeline = HTML::Pipeline.new [ HTML::Pipeline::MarkdownFilter, LangFilter, RootRelativeFilter, HeaderLinkFilter ], context
          else 
            @pipeline = HTML::Pipeline.new [ HTML::Pipeline::MarkdownFilter, RootRelativeFilter, HeaderLinkFilter ], context
          end
      end

      def convert(content)
          @pipeline.call(content)[:output].to_s
      end
  end

end