newstile 1.0.1

data/doc/tests.page

@@ -0,0 +1,51 @@
+---
+title: Tests and Benchmark
+---
+
+# Tests
+
+There exist several test suites for testing the correctness of a Markdown implementation. The
+original [Markdown Test Suite] is the standard which one needs to test against. The [PHP Markdown
+suite][MDTest] contains the original test suite and several more tests (some specifically geared
+towards the extension of the PHP Markdown Extra package). I have used the latter test tool to
+roughly verify that newstile is able to parse standard Markdown. However, since the syntax used by
+newstile varies slightly from standard Markdown most of the tests fail - which is fine. When looking
+at the differences one can see that the failures result from these differences.
+
+Besides using the above mentioned test suite newstile comes with its own set of tests which is used
+to verify that the implementation matches the newstile specification.
+
+If you believe you have found a bug in the implementation, please follow these steps:
+
+* Check the syntax page and see if the behaviour is not intended.
+
+* If the behaviour is not intended and it seems that newstile should parse some text in another
+  fashion, please open a bug report and attach two files: one with the text and one with the HTML
+  conversion you think is correct.
+
+
+# Benchmark
+
+newstile comes with a small benchmark to test how fast it is in regard to four other Ruby Markdown
+implementations: Maruku, BlueFeather, BlueCloth and RDiscount. The first two are written using only
+Ruby, the latter two use the C discount library for the actual hard work (which makes them really
+fast but they do not provide additional syntax elements). As one can see below, newstile is
+currently (June 2010) ~4x faster than Maruku, ~9x faster than BlueFeather but ~30x slower than
+BlueCloth and rdiscount:
+
+<pre><code>
+{execute_cmd: {command: "ruby -Ilib -rubygems benchmark/benchmark.rb", process_output: false, escape_html: true}}
+</code>
+</pre>
+
+And here are some graphs which show the execution times of the various newstile releases on
+different Ruby interpreters:
+
+![ruby 1.8.6]({relocatable: img/graph-ruby-1.8.6.png})
+![ruby 1.8.7]({relocatable: img/graph-ruby-1.8.7.png})
+![ruby 1.9.1p243]({relocatable: img/graph-ruby-1.9.1p243.png})
+![ruby 1.9.2dev]({relocatable: img/graph-ruby-1.9.2dev.png})
+![jruby 1.4.0]({relocatable: img/graph-jruby-1.4.0.png})
+
+[Markdown Test Suite]: http://daringfireball.net/projects/downloads/MarkdownTest_1.0.zip
+[MDTest]: http://www.michelf.com/docs/projets/mdtest-1.0.zip

data/doc/virtual

@@ -0,0 +1,2 @@
+rdoc/index.html:
+  title: API Documentation

data/lib/newstile.rb

@@ -0,0 +1,23 @@
+# -*- coding: utf-8 -*-
+#
+#--
+# Copyright (C) 2009-2010 Thomas Leitner <t_leitner@gmx.at>
+#
+# This file is part of newstile.
+#
+# newstile is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+#++
+#
+
+require 'newstile/document'

data/lib/newstile/compatibility.rb

@@ -0,0 +1,34 @@
+# -*- coding: utf-8 -*-
+#
+#--
+# Copyright (C) 2009-2010 Thomas Leitner <t_leitner@gmx.at>
+#
+# This file is part of newstile.
+#
+# newstile is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+#++
+#
+
+# All the code in this file is backported from Ruby 1.8.7 sothat newstile works under 1.8.5
+
+if RUBY_VERSION == '1.8.5'
+  require 'rexml/parsers/baseparser'
+  module REXML
+    module Parsers
+      class BaseParser
+        UNAME_STR= "(?:#{NCNAME_STR}:)?#{NCNAME_STR}"
+      end
+    end
+  end
+end

data/lib/newstile/converter.rb

@@ -0,0 +1,43 @@
+# -*- coding: utf-8 -*-
+#
+#--
+# Copyright (C) 2009-2010 Thomas Leitner <t_leitner@gmx.at>
+#
+# This file is part of newstile.
+#
+# newstile is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+#++
+#
+
+module Newstile
+
+  # == Converter Module
+  #
+  # This module contains all available converters, i.e. classes that take a document and convert the
+  # document tree to a specific output format, normally a string. For example, the Html module
+  # converts the document tree into HTML.
+  #
+  # Converters use the Base class for common functionality (like applying a template to the output)-
+  # see its API documentation for how to create a converter class.
+  module Converter
+
+    autoload :Base, 'newstile/converter/base'
+    autoload :Html, 'newstile/converter/html'
+    autoload :Latex, 'newstile/converter/latex'
+    autoload :Newstile, 'newstile/converter/newstile'
+    autoload :Markdown, 'newstile/converter/markdown'
+
+  end
+
+end

data/lib/newstile/converter/base.rb

@@ -0,0 +1,111 @@
+# -*- coding: utf-8 -*-
+#
+#--
+# Copyright (C) 2009-2010 Thomas Leitner <t_leitner@gmx.at>
+#
+# This file is part of newstile.
+#
+# newstile is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+#++
+#
+
+require 'erb'
+
+module Newstile
+
+  module Converter
+
+    # == Base class for converters
+    #
+    # This class serves as base class for all converters. It provides methods that can/should be
+    # used by all converters (like #generate_id) as well as common functionality that is
+    # automatically applied to the result (for example, embedding the output into a template).
+    #
+    # == Implementing a converter
+    #
+    # Implementing a new converter is rather easy: just create a new sub class from this class and
+    # put it in the Newstile::Converter module (the latter is only needed if auto-detection should
+    # work properly). Then you need to implement the #convert(tree) method which takes a document
+    # tree and should return the converted output.
+    #
+    # The document instance is automatically set as @doc in Base#initialize. Furthermore, the
+    # document instance provides a hash called `conversion_infos` that is also automatically cleared
+    # and can be used to store information about the conversion process.
+    #
+    # The actual transformation of the document tree can be done in any way. However, writing one
+    # method per tree element type is a straight forward way to do it - this is how the Html and
+    # Latex converters do the transformation.
+    class Base
+
+      # Initialize the converter with the given Newstile document +doc+.
+      def initialize(doc)
+        @doc = doc
+        @doc.conversion_infos.clear
+      end
+      private_class_method(:new, :allocate)
+
+      # Convert the Newstile document +doc+ to the output format implemented by a subclass.
+      #
+      # Initializes a new instance of the calling class and then calls the #convert method that must
+      # be implemented by each subclass. If the +template+ option is specified and non-empty, the
+      # result is rendered into the specified template.
+      def self.convert(doc)
+        result = new(doc).convert(doc.tree)
+        result = apply_template(doc, result) if !doc.options[:template].empty?
+        result
+      end
+
+      # Apply the template specified in the +doc+ options, using +body+ as the body string.
+      def self.apply_template(doc, body)
+        erb = ERB.new(get_template(doc.options[:template]))
+        obj = Object.new
+        obj.instance_variable_set(:@doc, doc)
+        obj.instance_variable_set(:@body, body)
+        erb.result(obj.instance_eval{binding})
+      end
+
+      # Return the template specified by +template+.
+      def self.get_template(template)
+        format_ext = '.' + self.name.split(/::/).last.downcase
+        shipped = File.join(::Newstile.data_dir, template + format_ext)
+        if File.exist?(template)
+          File.read(template)
+        elsif File.exist?(template + format_ext)
+          File.read(template + format_ext)
+        elsif File.exist?(shipped)
+          File.read(shipped)
+        else
+          raise "The specified template file #{template} does not exist"
+        end
+      end
+
+
+      # Generate an unique alpha-numeric ID from the the string +str+ for use as header ID.
+      def generate_id(str)
+        gen_id = str.gsub(/[^a-zA-Z0-9 -]/, '').gsub(/^[^a-zA-Z]*/, '').gsub(' ', '-').downcase
+        gen_id = 'section' if gen_id.length == 0
+        @used_ids ||= {}
+        if @used_ids.has_key?(gen_id)
+          gen_id += '-' + (@used_ids[gen_id] += 1).to_s
+        else
+          @used_ids[gen_id] = 0
+        end
+        @doc.options[:auto_id_prefix] + gen_id
+      end
+
+    end
+
+  end
+
+end

data/lib/newstile/converter/html.rb

@@ -0,0 +1,405 @@
+# -*- coding: utf-8 -*-
+#
+#--
+# Copyright (C) 2009-2010 Thomas Leitner <t_leitner@gmx.at>
+#
+# This file is part of newstile.
+#
+# newstile is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+#++
+#
+
+require 'rexml/parsers/baseparser'
+
+module Newstile
+
+  module Converter
+
+    # Converts a Newstile::Document to HTML.
+    class Html < Base
+
+      include ::Newstile::Utils::HTML
+
+      # :stopdoc:
+
+      # Defines the amount of indentation used when nesting HTML tags.
+      INDENTATION = 2
+
+      begin
+        require 'coderay'
+
+        # Highlighting via coderay is available if this constant is +true+.
+        HIGHLIGHTING_AVAILABLE = true
+      rescue LoadError => e
+        HIGHLIGHTING_AVAILABLE = false
+      end
+
+      # Initialize the HTML converter with the given Newstile document +doc+.
+      def initialize(doc)
+        super
+        @footnote_counter = @footnote_start = @doc.options[:footnote_nr]
+        @footnotes = []
+        @toc = []
+        @toc_code = nil
+      end
+
+      def convert(el, indent = -INDENTATION, opts = {})
+        send("convert_#{el.type}", el, indent, opts)
+      end
+
+      def inner(el, indent, opts)
+        result = ''
+        indent += INDENTATION
+        el.children.each do |inner_el|
+          opts[:parent] = el
+          result << send("convert_#{inner_el.type}", inner_el, indent, opts)
+        end
+        result
+      end
+
+      def convert_blank(el, indent, opts)
+        "\n"
+      end
+
+      def convert_text(el, indent, opts)
+        escape_html(el.value, :text)
+      end
+
+      def convert_p(el, indent, opts)
+        if el.options[:transparent]
+          "#{inner(el, indent, opts)}"
+        else
+          "#{' '*indent}<p#{html_attributes(el)}>#{inner(el, indent, opts)}</p>\n"
+        end
+      end
+
+      def convert_codeblock(el, indent, opts)
+        el = Marshal.load(Marshal.dump(el)) # so that the original is not changed
+        lang = el.attr.delete('lang')
+        if lang && HIGHLIGHTING_AVAILABLE
+          opts = {:wrap => @doc.options[:coderay_wrap], :line_numbers => @doc.options[:coderay_line_numbers],
+            :line_number_start => @doc.options[:coderay_line_number_start], :tab_width => @doc.options[:coderay_tab_width],
+            :bold_every => @doc.options[:coderay_bold_every], :css => @doc.options[:coderay_css]}
+          result = CodeRay.scan(el.value, lang.to_sym).html(opts).chomp + "\n"
+          "#{' '*indent}<div#{html_attributes(el)}>#{result}#{' '*indent}</div>\n"
+        else
+          result = escape_html(el.value)
+          if el.attr['class'].to_s =~ /\bshow-whitespaces\b/
+            result.gsub!(/(?:(^[ \t]+)|([ \t]+$)|([ \t]+))/) do |m|
+              suffix = ($1 ? '-l' : ($2 ? '-r' : ''))
+              m.scan(/./).map do |c|
+                case c
+                when "\t" then "<span class=\"ws-tab#{suffix}\">\t</span>"
+                when " " then "<span class=\"ws-space#{suffix}\">&#8901;</span>"
+                end
+              end.join('')
+            end
+          end
+          "#{' '*indent}<pre#{html_attributes(el)}><code>#{result}#{result =~ /\n\Z/ ? '' : "\n"}</code></pre>\n"
+        end
+      end
+
+      def convert_blockquote(el, indent, opts)
+        "#{' '*indent}<blockquote#{html_attributes(el)}>\n#{inner(el, indent, opts)}#{' '*indent}</blockquote>\n"
+      end
+
+      def convert_summary(el, indent, opts)
+        "#{' '*indent}<p#{html_attributes(el)}><b>#{inner(el, indent, opts)}#{' '*indent}</b></p>\n"
+      end
+
+      def convert_header(el, indent, opts)
+        el = Marshal.load(Marshal.dump(el)) # so that the original is not changed
+        if @doc.options[:auto_ids] && !el.attr['id']
+          el.attr['id'] = generate_id(el.options[:raw_text])
+        end
+        @toc << [el.options[:level], el.attr['id'], el.children] if el.attr['id'] && within_toc_depth?(el)
+        "#{' '*indent}<h#{el.options[:level]}#{html_attributes(el)}>#{inner(el, indent, opts)}</h#{el.options[:level]}>\n"
+      end
+
+      def within_toc_depth?(el)
+        @doc.options[:toc_depth] <= 0 || el.options[:level] <= @doc.options[:toc_depth]
+      end
+
+      def convert_hr(el, indent, opts)
+        "#{' '*indent}<hr />\n"
+      end
+
+      def convert_ul(el, indent, opts)
+        if !@toc_code && (el.options[:ial][:refs].include?('toc') rescue nil) && (el.type == :ul || el.type == :ol)
+          @toc_code = [el.type, el.attr, (0..128).to_a.map{|a| rand(36).to_s(36)}.join]
+          @toc_code.last
+        else
+          "#{' '*indent}<#{el.type}#{html_attributes(el)}>\n#{inner(el, indent, opts)}#{' '*indent}</#{el.type}>\n"
+        end
+      end
+      alias :convert_ol :convert_ul
+      alias :convert_dl :convert_ul
+
+      def convert_li(el, indent, opts)
+        output = ' '*indent << "<#{el.type}" << html_attributes(el) << ">"
+        res = inner(el, indent, opts)
+        if el.children.empty? || (el.children.first.type == :p && el.children.first.options[:transparent])
+          output << res << (res =~ /\n\Z/ ? ' '*indent : '')
+        else
+          output << "\n" << res << ' '*indent
+        end
+        output << "</#{el.type}>\n"
+      end
+      alias :convert_dd :convert_li
+
+      def convert_dt(el, indent, opts)
+        "#{' '*indent}<dt#{html_attributes(el)}>#{inner(el, indent, opts)}</dt>\n"
+      end
+
+      HTML_TAGS_WITH_BODY=['div', 'script', 'iframe', 'textarea']
+
+      def convert_html_element(el, indent, opts)
+        parent = opts[:parent]
+        res = inner(el, indent, opts)
+        if el.options[:category] == :span
+          "<#{el.value}#{html_attributes(el)}" << (!res.empty? || HTML_TAGS_WITH_BODY.include?(el.value) ? ">#{res}</#{el.value}>" : " />")
+        else
+          output = ''
+          output << ' '*indent if parent.type != :html_element || parent.options[:parse_type] != :raw
+          output << "<#{el.value}#{html_attributes(el)}"
+          if !res.empty? && el.options[:parse_type] != :block
+            output << ">#{res}</#{el.value}>"
+          elsif !res.empty?
+            output << ">\n#{res.chomp}\n"  << ' '*indent << "</#{el.value}>"
+          elsif HTML_TAGS_WITH_BODY.include?(el.value)
+            output << "></#{el.value}>"
+          else
+            output << " />"
+          end
+          output << "\n" if parent.type != :html_element || parent.options[:parse_type] != :raw
+          output
+        end
+      end
+
+      def convert_xml_comment(el, indent, opts)
+        if el.options[:category] == :block && (opts[:parent].type != :html_element || opts[:parent].options[:parse_type] != :raw)
+          ' '*indent + el.value + "\n"
+        else
+          el.value
+        end
+      end
+      alias :convert_xml_pi :convert_xml_comment
+      alias :convert_html_doctype :convert_xml_comment
+
+      def convert_table(el, indent, opts)
+        if el.options[:alignment].all? {|a| a == :default}
+          alignment = ''
+        else
+          alignment = el.options[:alignment].map do |a|
+            "#{' '*(indent + INDENTATION)}" + (a == :default ? "<col />" : "<col align=\"#{a}\" />") + "\n"
+          end.join('')
+        end
+        "#{' '*indent}<table#{html_attributes(el)}>\n#{alignment}#{inner(el, indent, opts)}#{' '*indent}</table>\n"
+      end
+
+      def convert_thead(el, indent, opts)
+        "#{' '*indent}<#{el.type}#{html_attributes(el)}>\n#{inner(el, indent, opts)}#{' '*indent}</#{el.type}>\n"
+      end
+      alias :convert_tbody :convert_thead
+      alias :convert_tfoot :convert_thead
+      alias :convert_tr  :convert_thead
+
+      def convert_td(el, indent, opts)
+        res = inner(el, indent, opts)
+        "#{' '*indent}<#{el.type}#{html_attributes(el)}>#{res.empty? ? "&nbsp;" : res}</#{el.type}>\n"
+      end
+      alias :convert_th :convert_td
+
+      def convert_comment(el, indent, opts)
+        if el.options[:category] == :block
+          "#{' '*indent}<!-- #{el.value} -->\n"
+        else
+          "<!-- #{el.value} -->"
+        end
+      end
+
+      def convert_br(el, indent, opts)
+        "<br />"
+      end
+
+      def convert_a(el, indent, opts)
+        do_obfuscation = el.attr['href'] =~ /^mailto:/
+        if do_obfuscation
+          el = Marshal.load(Marshal.dump(el)) # so that the original is not changed
+          href = obfuscate(el.attr['href'].sub(/^mailto:/, ''))
+          mailto = obfuscate('mailto')
+          el.attr['href'] = "#{mailto}:#{href}"
+        end
+        res = inner(el, indent, opts)
+        res = obfuscate(res) if do_obfuscation
+        "<a#{html_attributes(el)}>#{res}</a>"
+      end
+
+      def convert_img(el, indent, opts)
+        "<img#{html_attributes(el)} />"
+      end
+
+      def convert_codespan(el, indent, opts)
+        el = Marshal.load(Marshal.dump(el)) # so that the original is not changed
+        lang = el.attr.delete('lang')
+        if lang && HIGHLIGHTING_AVAILABLE
+          result = CodeRay.scan(el.value, lang.to_sym).html(:wrap => :span, :css => @doc.options[:coderay_css]).chomp
+          "<code#{html_attributes(el)}>#{result}</code>"
+        else
+          "<code#{html_attributes(el)}>#{escape_html(el.value)}</code>"
+        end
+      end
+
+      def convert_footnote(el, indent, opts)
+        number = @footnote_counter
+        @footnote_counter += 1
+        @footnotes << [el.options[:name], @doc.parse_infos[:footnotes][el.options[:name]]]
+        "<sup id=\"fnref:#{el.options[:name]}\"><a href=\"#fn:#{el.options[:name]}\" rel=\"footnote\">#{number}</a></sup>"
+      end
+
+      def convert_raw(el, indent, opts)
+        if !el.options[:type] || el.options[:type].empty? || el.options[:type].include?('html')
+          el.value + (el.options[:category] == :block ? "\n" : '')
+        else
+          ''
+        end
+      end
+
+      def convert_em(el, indent, opts)
+        "<#{el.type}#{html_attributes(el)}>#{inner(el, indent, opts)}</#{el.type}>"
+      end
+      alias :convert_strong :convert_em
+
+      def convert_entity(el, indent, opts)
+        entity_to_str(el.value, el.options[:original])
+      end
+
+      TYPOGRAPHIC_SYMS = {
+        :mdash => [::Newstile::Utils::Entities.entity('mdash')],
+        :ndash => [::Newstile::Utils::Entities.entity('ndash')],
+        :hellip => [::Newstile::Utils::Entities.entity('hellip')],
+        :laquo_space => [::Newstile::Utils::Entities.entity('laquo'), ::Newstile::Utils::Entities.entity('nbsp')],
+        :raquo_space => [::Newstile::Utils::Entities.entity('nbsp'), ::Newstile::Utils::Entities.entity('raquo')],
+        :laquo => [::Newstile::Utils::Entities.entity('laquo')],
+        :raquo => [::Newstile::Utils::Entities.entity('raquo')],
+        :qdash => [::Newstile::Utils::Entities.entity('8213')],
+        :qdash_space => [::Newstile::Utils::Entities.entity('8213'), ::Newstile::Utils::Entities.entity('nbsp')]
+      }
+      def convert_typographic_sym(el, indent, opts)
+        TYPOGRAPHIC_SYMS[el.value].map {|e| entity_to_str(e)}.join('')
+      end
+
+      def convert_smart_quote(el, indent, opts)
+        entity_to_str(::Newstile::Utils::Entities.entity(el.value.to_s))
+      end
+
+      def convert_math(el, indent, opts)
+        el = Marshal.load(Marshal.dump(el)) # so that the original is not changed
+        el.attr['class'] ||= ''
+        el.attr['class'] += (el.attr['class'].empty? ? '' : ' ') + 'math'
+        type = 'span'
+        type = 'div' if el.options[:category] == :block
+        "<#{type}#{html_attributes(el)}>#{escape_html(el.value)}</#{type}>#{type == 'div' ? "\n" : ''}"
+      end
+
+      def convert_abbreviation(el, indent, opts)
+        title = @doc.parse_infos[:abbrev_defs][el.value]
+        title = nil if title.empty?
+        "<abbr#{title ? " title=\"#{title}\"" : ''}>#{el.value}</abbr>"
+      end
+
+      def convert_root(el, indent, opts)
+        result = inner(el, indent, opts)
+        result << footnote_content
+        if @toc_code
+          toc_tree = generate_toc_tree(@toc, @toc_code[0], @toc_code[1] || {})
+          text = if toc_tree.children.size > 0
+                   convert(toc_tree, 0)
+                 else
+                   ''
+                 end
+          result.sub!(/#{@toc_code.last}/, text)
+        end
+        result
+      end
+
+      def generate_toc_tree(toc, type, attr)
+        sections = Element.new(type, nil, attr)
+        sections.attr['id'] ||= 'markdown-toc'
+        stack = []
+        toc.each do |level, id, children|
+          li = Element.new(:li, nil, nil, {:level => level})
+          li.children << Element.new(:p, nil, nil, {:transparent => true})
+          a = Element.new(:a, nil, {'href' => "##{id}"})
+          a.children += children
+          li.children.last.children << a
+          li.children << Element.new(type)
+
+          success = false
+          while !success
+            if stack.empty?
+              sections.children << li
+              stack << li
+              success = true
+            elsif stack.last.options[:level] < li.options[:level]
+              stack.last.children.last.children << li
+              stack << li
+              success = true
+            else
+              item = stack.pop
+              item.children.pop unless item.children.last.children.size > 0
+            end
+          end
+        end
+        while !stack.empty?
+          item = stack.pop
+          item.children.pop unless item.children.last.children.size > 0
+        end
+        sections
+      end
+
+      # Helper method for obfuscating the +text+ by using HTML entities.
+      def obfuscate(text)
+        result = ""
+        text.each_byte do |b|
+          result += (b > 128 ? b.chr : "&#%03d;" % b)
+        end
+        result.force_encoding(text.encoding) if RUBY_VERSION >= '1.9'
+        result
+      end
+
+      # Return a HTML list with the footnote content for the used footnotes.
+      def footnote_content
+        ol = Element.new(:ol)
+        ol.attr['start'] = @footnote_start if @footnote_start != 1
+        @footnotes.each do |name, data|
+          li = Element.new(:li, nil, {'id' => "fn:#{name}"}, {:first_is_block => true})
+          li.children = Marshal.load(Marshal.dump(data[:content].children))
+          ol.children << li
+
+          ref = Element.new(:raw, "<a href=\"#fnref:#{name}\" rev=\"footnote\">&#8617;</a>")
+          if li.children.last.type == :p
+            para = li.children.last
+          else
+            li.children << (para = Element.new(:p))
+          end
+          para.children << ref
+        end
+        (ol.children.empty? ? '' : "<div class=\"footnotes\">\n#{convert(ol, 2)}</div>\n")
+      end
+
+    end
+
+  end
+end