Asciidoctor manpage converter

manpage.rb

require 'asciidoctor'

# TODO
# * manually insert "NAME" section
# * still getting extra blank lines in definition lists
class ManpageConverter
  include Asciidoctor::Converter
  register_for 'manpage'
  EOL = "\n"

  def convert node, transform = nil
    transform ||= node.node_name
    send transform, node if respond_to? transform
  end

  def document node
    result = %['\\" t
.\\"     Title: #{node.attr 'mantitle'}
.\\"    Author: #{(node.attr? 'author') ? (node.attr 'author') : '[see the "AUTHORS" section]'}
.\\" Generator: Asciidoctor #{node.attr 'asciidoctor-version'}
.\\"      Date: #{node.attr 'docdate'}
.\\"    Manual: #{(manual = (node.attr? 'manmanual') ? (node.attr 'manmanual') : '\\ \\&')}
.\\"    Source: #{(source = (node.attr? 'mansource') ? (node.attr 'mansource') : '\\ \\&')}
.\\"  Language: English
.\\"
.TH "#{(node.attr 'manname').upcase}" "#{node.attr 'manvolnum'}" "#{node.attr 'docdate'}" "#{manify source, false}" "#{manify manual, false}"
.ie \\n(.g .ds Aq \\(aq
.el       .ds Aq '
.nh
.ad l]

    result = %(#{result}\n#{node.content})
    if node.attr? 'author'
      result = %(#{result}
.SH "AUTHOR"
.PP
\\fB#{node.attr 'author'}\\fR
.RE)
    end
    result
  end

  def section node
    title = node.title
    %(.S#{node.level > 1 ? 'S' : 'H'} "#{node.level > 1 ? title : title.upcase}"
#{title.upcase == 'NAME' ? node.content.gsub(/\A\.sp\n/, '') : node.content}) 
  end

  def paragraph node
    %(.sp
#{manify node.content})
  end

  # TODO needs cleanup
  def dlist node
    result = []
    node.items.each do |terms, dd|
      # TODO support multiple terms per item
      result << (manify [*terms].first.text, !!dd)
      if dd
        result << %(.RS 4
#{manify dd.text, false})
        if dd.blocks?
          result << dd.content
        else
          result << EOL
        end
        result << '.RE'
      end
    end
    result * EOL
  end

  def ulist node
    node.items.map {|li|
      %[.sp
.RS 4
.ie n \\{\\
\\h'-04'\\(bu\\h'+03'\\c
.\\}
.el \\{\\
.sp -1
.IP \\(bu 2.3
.\\}
#{manify li.text, false}#{li.blocks? ? li.content : nil}
.RE]
    } * EOL
  end

  def inline_anchor node
    target = node.target
    case node.type
    when :link
      if target.start_with? 'mailto:'
        target[7..-1]
      else
        target
      end
    else
      target
    end
  end

  def inline_quoted node
    case node.type
    when :emphasis
      %[\\fI#{node.text}\\fR]
    when :strong
      %[\\fB#{node.text}\\fR]
    when :single
      %[\\(oq#{node.text}\\(cq]
    when :double
      %[\\(lq#{node.text}\\(rq]
    else
      node.text
    end
  end

  # Folds each endline into a single space, escapes special man characters,
  # reverts HTML entity references back to their original form, strips trailing
  # whitespace and, optionally, appends a newline
  # see: man groff_char
  def manify str, append_newline = true, preserve_space = false
    if preserve_space
      str = str.gsub("\t", (' ' * 8))
    else
      str = str.tr_s("\n\t ", ' ')
    end
    str
        .gsub(/\./, '\\\&.')
        .gsub('-', '\\-')
        .gsub('&lt;', '<')
        .gsub('&gt;', '>')
        .gsub('&#169;', '\\(co')
        .gsub('&#174;', '\\(rg')
        .gsub('&#8482;', '\\(tm')
        .gsub('&#8201;', ' ') #.gsub('&#8201;&#8212;&#8201;', ' \\(em ')
        .gsub('&#8212;', '\\(em')
        .gsub('&#8230;', '...') #.gsub('&#8230;', '\\\&...')
        .gsub('&#8217;', '\\(cq')
        .gsub('&#8592;', '\\(<-')
        .gsub('&#8594;', '\\(->')
        .gsub('&#8656;', '\\(lA')
        .gsub('&#8658;', '\\(rA')
        .gsub('\'', '\\(aq')
        .rstrip + (append_newline ? EOL : '')
    end
end

#options = {}
#options[:header_footer] = true
#options[:backend] = :manpage
#Asciidoctor.convert_file ARGV.first, options