lib/kramdown/utils/html.rb



# -*- coding: utf-8; frozen_string_literal: true -*-
#
#--
# Copyright (C) 2009-2019 Thomas Leitner <t_leitner@gmx.at>
#
# This file is part of kramdown which is licensed under the MIT.
#++
#

require 'rexml/parsers/baseparser'

module Kramdown

  module Utils

    # Provides convenience methods for HTML related tasks.
    #
    # *Note* that this module has to be mixed into a class that has a @root (containing an element
    # of type :root) and an @options (containing an options hash) instance variable so that some of
    # the methods can work correctly.
    module Html

      # Convert the entity +e+ to a string. The optional parameter +original+ may contain the
      # original representation of the entity.
      #
      # This method uses the option +entity_output+ to determine the output form for the entity.
      def entity_to_str(e, original = nil)
        entity_output = @options[:entity_output]

        if entity_output == :as_char &&
            (c = e.char.encode(@root.options[:encoding]) rescue nil) &&
            ((c = e.char) == '"' || !ESCAPE_MAP.key?(c))
          c
        elsif (entity_output == :as_input || entity_output == :as_char) && original
          original
        elsif (entity_output == :symbolic || ESCAPE_MAP.key?(e.char)) && !e.name.nil?
          "&#{e.name};"
        else # default to :numeric
          "&##{e.code_point};"
        end
      end

      # Return the HTML representation of the attributes +attr+.
      def html_attributes(attr)
        return '' if attr.empty?

        attr.map do |k, v|
          v.nil? || (k == 'id' && v.strip.empty?) ? '' : " #{k}=\"#{escape_html(v.to_s, :attribute)}\""
        end.join
      end

      # :stopdoc:
      ESCAPE_MAP = {
        '<' => '&lt;',
        '>' => '&gt;',
        '&' => '&amp;',
        '"' => '&quot;',
      }
      ESCAPE_ALL_RE = /<|>|&/
      ESCAPE_TEXT_RE = Regexp.union(REXML::Parsers::BaseParser::REFERENCE_RE, /<|>|&/)
      ESCAPE_ATTRIBUTE_RE = Regexp.union(REXML::Parsers::BaseParser::REFERENCE_RE, /<|>|&|"/)
      ESCAPE_RE_FROM_TYPE = {all: ESCAPE_ALL_RE, text: ESCAPE_TEXT_RE, attribute: ESCAPE_ATTRIBUTE_RE}
      # :startdoc:

      # Escape the special HTML characters in the string +str+. The parameter +type+ specifies what
      # is escaped: :all - all special HTML characters except the quotation mark as well as
      # entities, :text - all special HTML characters except the quotation mark but no entities and
      # :attribute - all special HTML characters including the quotation mark but no entities.
      def escape_html(str, type = :all)
        str.gsub(ESCAPE_RE_FROM_TYPE[type]) {|m| ESCAPE_MAP[m] || m }
      end

      REDUNDANT_LINE_BREAK_REGEX = /([\p{Han}\p{Hiragana}\p{Katakana}]+)\n([\p{Han}\p{Hiragana}\p{Katakana}]+)/u
      def fix_cjk_line_break(str)
        while str.gsub!(REDUNDANT_LINE_BREAK_REGEX, '\1\2')
        end
        str
      end

    end

  end

end