iKnowledge 1.0 HTML格式自动转换成2.0 Markdown格式

html2md.rb

#!/usr/bin/ruby
# -*- coding: utf-8 -*-

require 'cgi'

module Html

  def self.tag(content)
    n, a = content.split /\s+/, 2
    {:tag => n}.update(attr: Hash[(a or '').scan(/(\w+)="([^"]*)"/)])
  end

  def self.yacc(token)
    doc = []
    loop do
      e = token.next
      break if e.start_with? '</'
      if !e.start_with? '<'
        doc << {
          :tag => '#text',
          :text => CGI.unescapeHTML(e)
            .gsub("&ldquo;", "“")
            .gsub("&rdquo;", "”")
            .gsub("&nbsp;", " ")
            .gsub("&hellip;", "…")
        }
      elsif e.end_with? '/>'
        doc << tag(e[1..-3])
      else
        doc << tag(e[1..-2]).update(node: yacc(token))
      end
    end
    doc
  end

  def self.lex(content)
    content.split(/(?:(?=<)|(?<=>))/).to_enum
  end

  def self.parse(xml)
    yacc lex xml
  end

end

class Markdown
  attr :date

  def initialize(html)
    @title = html[:attr]['title']
    @timestamp = html[:node][1][:node][0][:text]
    @date = @timestamp.split(/\s+/)[0]
    @tags = html[:node][2..-1]
  end

  def front_matter
    <<EOF
---
layout: article
title: #@title
date: #@timestamp
category: 
excerpt:
---
EOF
  end

  def md(tags, prefix = "")
    tags.map do |e|
      case e[:tag]
        when '#text'
          e[:text].strip

        when /h([1-6])/
          '#' * $~[1].to_i + ' ' + md(e[:node]) + "\n\n"

        when 'p'
          (prefix == "" ? '' : ' ' * 4) + md(e[:node]) + "\n\n"
        when 'a'
          '[' + md(e[:node]) + '](' + e[:attr]['href'] + ')'
        when 'span'
          '[' + md(e[:node]) + '](' + e[:attr]['title'] + ')'
        when 'i', 'b', 'strong'
          '*' + md(e[:node]) + '*'

        when 'pre'
          content = md(e[:node])
          if e[:node][0][:tag] == '#text'
            content.gsub!(/^/m, ' ' * (prefix == "" ? 4 : 8))
          end
          content + "\n\n"
        when 'blockquote'
          content = md(e[:node]).rstrip
          if e[:node][0][:tag] == '#text'
            content.gsub!(/^/m, '> ')
          end
          content + "\n\n"
        when 'code'
          if e[:attr] && e[:attr].key?('language')
            '```' + e[:attr]['language'].sub('shell', 'bash') + "\n" + md(e[:node]) + "\n```"
          else
            '`' + md(e[:node]) + '`'
          end

        when 'ol'
          md(e[:node], prefix + '1. ') + "\n"
        when 'ul'
          md(e[:node], prefix + '+ ') + "\n"
        when 'li'
          prefix + md(e[:node], '  ' + prefix) + "\n"

        when 'dl'
          "<dl>\n" + md(e[:node], '  ') + "</dl>\n\n"
        when 'dt', 'dd'
          "#{prefix}<#{e[:tag]}>#{CGI.escapeHTML(md(e[:node]))}</#{e[:tag]}>\n"

        when 'img'
          '{% img ' + File.basename(e[:attr]['src']) + " %}\n\n"
        when 'sub', 'sup'
          "<#{e[:tag]}>#{CGI.escapeHTML(md(e[:node]))}</#{e[:tag]}>"
        else
          puts "Unknow: #{e}"
          exit 1
      end
    end.join ''
  end

  def to_s
    front_matter + md(@tags)
  end
end

class IKnowledge
  def initialize(filename)
    @home = File.join("/", "Users", "redraiment", "Documents", "zzp.me")

    html = Html.parse File.read filename
    @md = Markdown.new html[0]
    @filename = @md.date + '-' + File.basename(filename).sub('.html', '.md')
  end

  def save
    fn = File.join(@home, "_posts", @filename)
    File.open(fn, 'w') do |f|
      f.write self.to_s
    end
    puts "save to #{fn}"
  end

  def to_s
    @md.to_s
  end
end

ikn = IKnowledge.new ARGV[0]
ikn.save