exports tDiary data ad HTML text

tdiary-export.rb

#!/usr/bin/ruby
#
# this script dumps diary data from tDiary 2.2.x.
#
# usage: $0 /path/to/tdiary/dir [/path/to/tdiary.conf/dir] > tdiary.dat
#
# Author: akira at arika.org
# License: tDiary's

config_override = proc do |config|
=begin
  if sp = config.options['sp.selected']
    config.options['sp.selected'] = 
      (sp.split(/\n/) - %w(hatena_star.rb add_bookmark.rb)).join("\n")
  end
  config.options['image.dir'] = '/path/to/image/dir'
=end
end

plugin_override = proc do
  def subtitle_link(date, index, subtitle)
    subtitle
  end

=begin
  def my(url, text = '', title = '')
    ret = %Q!<typo:tdiarycompat method="my" url="#{h url}"!
    ret << %Q! text="#{title}"! unless title.empty?
    ret << %Q!>#{text}</typo:tdiarycompat>!
  end
=end

=begin
  alias :orig_image :image
  def image(*arg)
    html = orig_image(*arg)
    m = %r!\A(?:<a href="([^"]+)">)?<img class="([^"]+)" src="([^"]+)" alt="([^"]*)" title="[^"]*"(?: width="([^"]+)"(?: height="([^"]+)")?)?>!.match(html)
    unless m
      warn "unexpected html by image: #{html}"
      return html
    end

    patt = /\A#{Regexp.quote(h(@image_url))}/
    ret = %Q!<typo:tdiarycompat method="image"!
    ret << %Q! image="#{m[3].sub(patt, '')}"!
    ret << %Q! link="#{m[1].sub(patt, '')}"! if m[1]
    ret << %Q! place="#{m[2]}"! if m[2]
    ret << %Q! title="#{m[4]}"! if m[4]
    ret << %Q! width="#{m[5]}"! if m[5]
    ret << %Q! height="#{m[6]}"! if m[6]
    ret << '/>'

    ret
  end
=end

=begin
  def amazon_get(asin, show_image = true, label = nil, position = 'amazon')
    position = 'box' if position == 'detail'
    ret = %Q!<typo:asin asin="#{h asin}" style="#{position}"!
    if show_image
      size = case @conf['amazon.imgsize']
             when 0
               'large'
             when 2
               'small'
             else
               'medium'
             end
      ret << %Q! image="#{size}"}!
    end
    if label
      ret << %Q!>#{label}</typo:asin>!
    else
      ret << '/>'
    end
    ret
  end
=end

=begin
  def flickr(pid, size = nil, place = nil)
    size ||= @conf['flickr.default_size'] || 'small'
    ret = %Q!<typo:tdiarycompat method="flickr" img="#{h pid.to_s}" size="#{h size}"!
    ret << %Q! place="#{h place}"! if place
    ret << '/>'
  end
=end
end

section_modify = proc do |section|
  section.body.gsub!(%r!^\s*<%=\s*a_plugin_method .*%>\s*!, '')
end

td_dir = ARGV.shift
td_conf_dir = ARGV.shift

td_dir = File.expand_path(td_dir)
td_dir.untaint
$LOAD_PATH.unshift(td_dir)

td_conf_dir = td_dir unless td_conf_dir
td_conf_dir.untaint
save_cwd = Dir.pwd
Dir.chdir(td_conf_dir)

require 'cgi'
require 'tdiary'

class TDiaryProxy < TDiary::TDiaryBase
  def initialize(*args, &block)
    super
    @ignore_parser_cache = true # don't use parser cache
    @plugin_override = block
    @_processed_subtitle = nil
    calendar # initialize @years
  end
  attr_accessor :_processed_subtitle

  def clear_cache(*args); end
  def store_cache(*args); end

  def transaction(date)
    @io.transaction(date) do |*arg|
      yield(*arg)
      DIRTY_NONE
    end
  end

  def each_month
    @years.keys.sort.each do |year|
      @years[year].sort.each do |month|
        yield(Time.local(year.to_i, month.to_i))
      end
    end
  end

  def _eval_rhtml(rhtml)
    r = ERB.new('<%= rhtml %>').result(binding)
    r = ERB.new(r).src
    if @plugin
      @plugin.eval_src(r.untaint, @conf.secure)
    else
      r
    end
  end

  def _body_enter_proc(date)
    _eval_rhtml("<% body_enter_proc(Time.at(#{date.to_i})) %>")
  end
  def _body_leave_proc(date)
    _eval_rhtml("<% body_leave_proc(Time.at(#{date.to_i})) %>")
  end

  def _dump_each_section
    each_month do |month|
      transaction(month) do |@diaries|
        @diaries.each do |dstr, diary|
          @date = diary.date
          load_plugins
          @plugin.instance_eval(&@plugin_override) if @plugin_override
          _body_enter_proc(@date)
          yield(diary)
          _body_leave_proc(@date)
        end
      end
    end
  end

  def _section_rhtml(sec)
    @_current_rhtml = sec.body_to_html
  end

  def _section_html(rhtml, sec)
    sep = "\n__tdiary_dump_processed_subtitle__#{rand(9999)}\n"
    eval_body =
      "<%= section_enter_proc(Time.at(#{@date.to_i})) %>" + 
      "<% _tdiary_dump_processed_subtitle = subtitle_proc(Time.at(#{@date.to_i}), #{(sec.subtitle || '').dump.gsub(/%/, '\\\\045')}) %>" +
      rhtml +
      "<%= section_leave_proc(Time.at(#{@date.to_i})) %>" +
      "#{sep}<%= _tdiary_dump_processed_subtitle %>"
    html, @_processed_subtitle = _eval_rhtml(eval_body).split(/#{Regexp.quote(sep)}/, 2)
    html
  end
end

ENV['REQUEST_METHOD'] = 'GET'
cgi = CGI.new
conf = TDiary::Config.new(cgi)
config_override.call(conf) if config_override

articles = []

tdiary = TDiaryProxy.new(cgi, nil, conf, &plugin_override)
tdiary._dump_each_section do |diary|
  diary_hash = {
    :title => diary.title,
    :date => diary.date,
    :last_modified => diary.last_modified,
    :article => [],
    :comment => [],
    :trackback => [],
    :visible => diary.visible?,
  }
  diary.each_section do |sec|
    if section_modify
      section_modify.call(sec)
    end
    rhtml = tdiary._section_rhtml(sec)
    rhtml = '<a name="p<%= "%02d"%@section_index[@date] %>"></a>' + rhtml
    html = tdiary._section_html(rhtml, sec)
    html.sub!(%r{\A(<a name="p\d+"></a>)(.*?<p>)}) { "#{$2}#{$1}" }

    diary_hash[:article] << {
      :title => tdiary._processed_subtitle.sub(/^(\[[^\[\]]+\]\s*)+/, ''),
      :body => html,
      :category => sec.categories,
    }
  end

  diary.each_comment do |com|
    comment_hash = {
      :time => com.date,
      :visible => com.instance_eval { @show }, # avoid overriding of "visible?" on tb-show.rb, etc.
    }
    if com.name == 'TrackBack'
      turl, tblogname, ttitle, tbody = com.body.split(/\n/, 4)
      comment_hash.merge!({
        :blog_name => tblogname,
        :title => ttitle,
        :url => turl,
        :body => tbody,
      })
      type = :trackback
    else
      comment_hash.merge!({
        :author => com.name,
        :mail => com.mail,
        :body => "<p>#{com.body.make_link.gsub(/\n/, '<br>').gsub(/<br><br>\Z/, '')}</p>",
      })
      type = :comment
    end
    diary_hash[type] << comment_hash
  end
  articles << diary_hash
#  break if articles.size > 10
end

Marshal.dump(articles, $stdout)