review-retrovert 0.3.3 → 0.9.1

data/testdata/mybook/lib/ruby/review-epubmaker.rb

@@ -0,0 +1,606 @@
+# -*- coding: utf-8 -*-
+
+#
+# Copyright (c) 2010-2017 Kenshi Muto and Masayoshi Takahashi
+#
+# This program is free software.
+# You can distribute or modify this program under the terms of
+# the GNU LGPL, Lesser General Public License version 2.1.
+# For details of the GNU LGPL, see the file "COPYING".
+#
+
+require 'review/epubmaker'
+
+require_relative './review-maker'
+
+
+module ReVIEW
+
+
+  class EpubMaker < Maker
+
+    SCRIPT_NAME = "review-epubmaker"
+
+    def initialize(config_filename, optional_values={})
+      super
+      @htmltoc = nil
+      @buildlogtxt = 'build-log.txt'
+      @producer = ::EPUBMaker::Producer.new(@config)
+      @producer.load(config_filename)
+      @config = @producer.config
+      @config.maker = 'epubmaker'
+    end
+
+    def log(s)
+      puts s if debug?()
+    end
+
+    def prepare_build_dir()
+      dir = "#{@config['bookname']}-epub"
+      if debug?()
+        builddir = File.expand_path(dir, Dir.pwd)
+        FileUtils.rm_rf(builddir, secure: true) if File.exist?(builddir)
+        FileUtils.mkdir_p(builddir)
+      else
+        builddir = Dir.mktmpdir("#{dir}-")
+      end
+      #
+      if @config['imgmath']
+        math_dir = "./#{@config['imagedir']}/_review_math"
+        FileUtils.rm_rf(math_dir) if Dir.exist?(math_dir)
+      end
+      #
+      return builddir
+    end
+
+    def generate()
+      bookname = @config['bookname']
+      epubname = "#{bookname}.epub"
+      booktmpname = "#{bookname}-epub"
+
+      begin
+        @config.check_version(ReVIEW::VERSION)
+      rescue ReVIEW::ConfigError => e
+        warn e.message
+      end
+      log("I will produce #{epubname}.")
+
+      FileUtils.rm_f(epubname)
+
+      builddir = prepare_build_dir()
+      @builddir = builddir
+      begin
+        log("Created first temporary directory as #{builddir}.")
+
+        call_hook('hook_beforeprocess')
+
+        @htmltoc = ReVIEW::HTMLToc.new(builddir)
+        ## copy all files into builddir
+        copy_stylesheet()
+
+        copy_frontmatter()
+        call_hook('hook_afterfrontmatter')
+
+        build_body()
+        call_hook('hook_afterbody')
+
+        copy_backmatter()
+        call_hook('hook_afterbackmatter')
+
+        ## push contents in builddir into @producer
+        push_contents()
+
+        imagedir = @config['imagedir']
+        imagedir_dest = File.join(builddir, imagedir)
+        if @config['epubmaker']['verify_target_images'].present?
+          verify_target_images()
+          copy_images(imagedir, builddir)
+        else
+          copy_images(imagedir, imagedir_dest)
+        end
+
+        copy_resources('covers', imagedir_dest)
+        copy_resources('adv', imagedir_dest)
+        copy_resources(@config['fontdir'], "#{builddir}/fonts", @config['font_ext'])
+
+        call_hook('hook_aftercopyimage')
+
+        @producer.import_imageinfo(imagedir_dest, builddir)
+        @producer.import_imageinfo("#{builddir}/fonts", builddir, @config['font_ext'])
+
+        check_image_size(@config['image_maxpixels'], @config['image_ext'])
+
+        if debug?()
+          epubtmpdir = File.join(builddir, booktmpname)
+          FileUtils.mkdir(epubtmpdir)
+        else
+          epubtmpdir = nil
+        end
+        log('Call ePUB producer.')
+        @producer.produce(epubname, builddir, epubtmpdir)
+        log('Finished.')
+      rescue ApplicationError => e
+        raise if debug?()
+        error(e.message)
+      ensure
+        FileUtils.remove_entry_secure(builddir) unless debug?()
+      end
+    end
+
+    def call_hook(hook_name)
+      filename = @config['epubmaker'][hook_name]
+      log("Call #{hook_name}. (#{filename})")
+      return unless filename.present?
+      return unless File.exist?(filename)
+      return unless FileTest.executable?(filename)
+      if ENV['REVIEW_SAFE_MODE'].to_i & 0x1 != 0
+        warn 'hook is prohibited in safe mode. ignored.'
+      else
+        system(filename, @builddir)
+      end
+    end
+
+    def verify_target_images()
+      ## なんでメソッド名が「verify_xxx」なのに副作用があるの？アホ？
+      items = @config['epubmaker']['force_include_images']
+      @producer.contents.each do |content|
+        File.open("#{@builddir}/#{content.file}") do |f|
+          if content.media == 'application/xhtml+xml'
+            REXML::Document.new(File.new(f)).each_element('//img') do |e|
+              items << e.attributes['src']
+              content.properties << 'svg' if e.attributes['src'] =~ /svg\Z/i
+            end
+          elsif content.media == 'text/css'
+            f.each_line do |l|
+              l.scan(/url\((.+?)\)/) { items << $1.strip }
+            end
+          else
+          end
+        end
+      end
+      @config['epubmaker']['force_include_images'] = items.compact.sort.uniq
+    end
+
+    def copy_images(resdir, destdir, allow_exts=nil)
+      return nil unless File.exist?(resdir)
+      allow_exts ||= @config['image_ext']
+      FileUtils.mkdir_p(destdir)
+      if @config['epubmaker']['verify_target_images'].present?
+        @config['epubmaker']['force_include_images'].each do |file|
+          unless File.exist?(file)
+            warn "#{file} is not found, skip." if file !~ /\Ahttp[s]?:/
+            next
+          end
+          basedir = File.dirname(file)
+          dirpath = "#{destdir}/#{basedir}"
+          FileUtils.mkdir_p(dirpath)
+          log("Copy #{file} to the temporary directory.")
+          FileUtils.cp(file, dirpath)
+        end
+      else
+        recursive_copy_files(resdir, destdir, allow_exts)
+      end
+    end
+
+    def copy_resources(resdir, destdir, allow_exts=nil)
+      return nil unless File.exist?(resdir)
+      allow_exts ||= @config['image_ext']
+      FileUtils.mkdir_p(destdir)
+      recursive_copy_files(resdir, destdir, allow_exts)
+    end
+
+    def recursive_copy_files(resdir, destdir, allow_exts)
+      allow_exts_rexp = /\.(#{allow_exts.join('|')})\Z/i
+      Dir.open(resdir) do |dir|
+        dir.each do |fname|
+          srcfile = "#{resdir}/#{fname}"
+          next if fname.start_with?('.')
+          if File.directory?(srcfile)
+            recursive_copy_files(srcfile, "#{destdir}/#{fname}", allow_exts)
+          elsif fname =~ allow_exts_rexp
+            FileUtils.mkdir_p(destdir)
+            log("Copy #{srcfile} to the temporary directory.")
+            FileUtils.cp(srcfile, destdir)
+          end
+        end
+      end
+    end
+
+    def check_compile_status
+      return unless @compile_errors
+
+      $stderr.puts 'compile error, No EPUB file output.'
+      exit 1
+    end
+
+    def new_builder()
+      builder = ReVIEW::EPUBBuilder.new()
+      builder.starter_config = @starter_config
+      return builder
+    end
+
+    def build_body()
+      @precount = 0
+      @bodycount = 0
+      @postcount = 0
+
+      @manifeststr = ''
+      @ncxstr = ''
+      @tocdesc = []
+
+      basedir = @basedir
+      base_path = Pathname.new(basedir)
+      book = ReVIEW::Book.load(basedir)
+      book.config = @config
+      @book = book
+      @converter = ReVIEW::Converter.new(book, new_builder())
+      @compile_errors = nil
+      book.parts.each do |part|
+        if part.name.present?
+          if part.file?
+            build_chap(part, base_path, 'part')
+          else
+            htmlfile = "part_#{part.number}.#{@config['htmlext']}"
+            build_part(part, htmlfile)
+            title = ReVIEW::I18n.t('part', part.number)
+            title += ReVIEW::I18n.t('chapter_postfix') + part.name.strip if part.name.strip.present?
+            @htmltoc.add_item(0, htmlfile, title, chaptype: 'part')
+            write_buildlogtxt(htmlfile, '')
+          end
+        end
+        #
+        part.chapters.each do |chap|
+          build_chap(chap, base_path, nil)
+        end
+      end
+      check_compile_status
+    end
+
+    def new_renderer()
+      return EPUBRenderer.new(@config, @book, @basedir, @starter_config)
+    end
+
+    def build_part(part, htmlfile)
+      log("Create #{htmlfile} from a template.")
+      new_renderer().generate_part_page(part, File.join(@builddir, htmlfile))
+    end
+
+    def build_chap(chap, base_path, chaptype)
+      chaptype ||= _chapter_type(chap)
+      filename = chaptype == 'part' ? chap.path : \
+                 Pathname.new(chap.path).relative_path_from(base_path).to_s
+      chap_id = _chapter_id(chap, chaptype)
+      #
+      htmlfile = "#{chap_id}.#{@config['htmlext']}"
+      write_buildlogtxt(htmlfile, filename)
+      log("Create #{htmlfile} from #{filename}.")
+      #
+      if (params_str = @config['params']).present?
+        warn "'params:' in config.yml is obsoleted."
+        if params_str =~ /stylesheet=/
+          warn "stylesheets should be defined in 'stylesheet:', not in 'params:'"
+        end
+      end
+      #
+      begin
+        @converter.convert(filename, File.join(@builddir, htmlfile))
+        write_info_body(htmlfile, chaptype)
+        remove_hidden_title(htmlfile)
+      rescue => e
+        @compile_errors = true
+        warn "compile error in #{filename} (#{e.class})"
+        warn e.message
+      end
+    end
+
+    def _chapter_type(chap)
+      return 'pre'  if chap.on_predef?
+      return 'post' if chap.on_appendix?
+      return 'body'
+    end
+
+    def _chapter_id(chap, chaptype)
+      id = File.basename(chap.path).sub(/\.re\z/, '')
+      if @config['epubmaker']['rename_for_legacy']
+        if    chaptype == 'part';
+        elsif chap.on_predef?   ; id = 'pre%02d'  % (@precount  += 1)
+        elsif chap.on_appendix? ; id = 'post%02d' % (@postcount += 1)
+        else                    ; id = 'chap%02d' % (@bodycount += 1)
+        end
+      end
+      return id
+    end
+
+    def remove_hidden_title(htmlfile)
+      File.open("#{@builddir}/#{htmlfile}", 'r+') do |f|
+        html = _remove_hidden_title(f.read())
+        f.rewind()
+        f.print(html)
+        f.truncate(f.tell)
+      end
+    end
+
+    def _remove_hidden_title(html)
+      html = html.gsub(/<h\d .*?hidden=['"]true['"].*?>.*?<\/h\d>\n/, '')
+      html = html.gsub(/(<h\d .*?)\s*notoc=['"]true['"]\s*(.*?>.*?<\/h\d>\n)/, '\1\2')
+      return html
+    end
+
+    def detect_properties(path)
+      doc = File.open(path) {|f| REXML::Document.new(f) }
+      has_math = REXML::XPath.first(doc, '//m:math', 'm' => 'http://www.w3.org/1998/Math/MathML')
+      has_svg  = REXML::XPath.first(doc, '//s:svg', 's' => 'http://www.w3.org/2000/svg')
+      properties = []
+      properties << 'mathml' if has_math
+      properties << 'svg'    if has_svg
+      return properties
+    end
+
+    def write_info_body(filename, chaptype)
+      headlines = []
+      path = File.join(@builddir, filename)
+      File.open(path) do |htmlio|
+        REXML::Document.parse_stream(htmlio, ReVIEWHeaderListener.new(headlines))
+      end
+      properties = detect_properties(path)
+      prop_str = ''
+      prop_str = ',properties=' + properties.join(' ') if properties.present?
+      headlines.each_with_index do |headline, i|
+        headline['level'] = 0 if chaptype == 'part' && headline['level'] == 1
+        if i == 0
+          opts = {chaptype: chaptype + prop_str, force_include: true}
+          _add_headline_item(headline, filename, nil, opts)
+        else
+          opts = {chaptype: chaptype}
+          _add_headline_item(headline, filename, headline['id'], opts)
+        end
+      end
+    end
+
+    def _add_headline_item(headline, filename, id, opts)
+      opts[:notoc] = headline['notoc']
+      entry = id ? "#{filename}\##{id}" : filename
+      @htmltoc.add_item(headline['level'], entry, headline['title'], opts)
+    end
+    private :_add_headline_item
+
+    def push_contents()
+      toclevel = @config['toclevel']
+      @htmltoc.each_item do |level, file, title, args|
+        if level.to_i <= toclevel || args[:force_include]
+          log("Push #{file} to ePUB contents.")
+          x = nil
+          hash = {'file'=>file, 'level'=>level.to_i, 'title'=>title, 'chaptype'=>args[:chaptype]}
+          hash['id']         = x if (x = args[:id]).present?
+          hash['properties'] = x.split(' ') if (x = args[:properties]).present?
+          hash['notoc']      = x if (x = args[:notoc]).present?
+          @producer.contents.push(::EPUBMaker::Content.new(hash))
+        end
+      end
+    end
+
+    def copy_stylesheet()
+      return if @config['stylesheet'].empty?
+      @config['stylesheet'].each do |sfile|
+        FileUtils.cp(sfile, @builddir)
+        @producer.contents.push(::EPUBMaker::Content.new('file' => sfile))
+      end
+    end
+
+    def copy_frontmatter()
+      cover = @config['cover']
+      if cover.present?
+        FileUtils.cp(cover, @builddir) if File.exist?(cover)
+      end
+      #
+      if @config['titlepage']
+        titlepage = "titlepage.#{@config['htmlext']}"
+        titlefile = @config['titlefile']
+        FileUtils.cp(titlefile, @builddir) if titlefile.present?
+        build_titlepage(titlepage)     unless titlefile.present?
+        _add_pre_item(titlepage, 'titlepagetitle')
+      end
+      #
+      original_titlefile = @config['originaltitlefile']
+      if original_titlefile.present? && File.exist?(original_titlefile)
+        FileUtils.cp(original_titlefile, @builddir)
+        _add_pre_item(original_titlefile, 'originaltitle')
+      end
+      #
+      creditfile = @config['creditfile']
+      if creditfile.present? && File.exist?(creditfile)
+        FileUtils.cp(creditfile, @builddir)
+        _add_pre_item(creditfile, 'credittitle')
+      end
+      #
+      true
+    end
+
+    def _add_pre_item(filename, key)
+      @htmltoc.add_item(1, File.basename(filename), @producer.res.v(key), chaptype: 'pre')
+    end
+
+    def build_titlepage(htmlfile)
+      # TODO: should be created via epubcommon
+      new_renderer().generate_title_page("#{@builddir}/#{htmlfile}")
+    end
+
+    def _join_names(names)
+      sep = ReVIEW::I18n.t('names_splitter')
+      return [names].flatten.join(sep)
+    end
+
+    def copy_backmatter()
+      if (profile = @config['profile'])
+        FileUtils.cp(profile, @builddir)
+        _add_post_item(profile, 'profiletitle')
+      end
+      #
+      if (advfile = @config['advfile'])
+        FileUtils.cp(advfile, @builddir)
+        _add_post_item(advfile, 'advtitle')
+      end
+      #
+      if (colophon = @config['colophon'])
+        colophon_page = File.join(@builddir, "colophon.#{@config['htmlext']}")
+        if colophon.is_a?(String) # FIXME: should let obsolete this style?
+          FileUtils.cp(colophon, colophon_page)
+        else
+          File.open(colophon_page, 'w') {|f| @producer.colophon(f) }
+        end
+        _add_post_item(colophon_page, 'colophontitle')
+      end
+      #
+      if (backcover = @config['backcover'])
+        FileUtils.cp(backcover, @builddir)
+        _add_post_item(backcover, 'backcovertitle')
+      end
+      #
+      true
+    end
+
+    def _add_post_item(filename, key)
+      @htmltoc.add_item(1, File.basename(filename), @producer.res.v(key), chaptype: 'post')
+    end
+
+    def write_buildlogtxt(htmlfile, reviewfile)
+      File.open("#{@builddir}/#{@buildlogtxt}", 'a') { |f| f.puts "#{htmlfile},#{reviewfile}" }
+    end
+
+    def check_image_size(maxpixels, allow_exts = nil)
+      begin
+        require 'image_size'
+      rescue LoadError
+        return nil
+      end
+      require 'find'
+      allow_exts ||= @config['image_ext']
+
+      extre = Regexp.new('\\.(' + allow_exts.delete_if { |t| %w[ttf woff otf].include?(t) }.join('|') + ')', Regexp::IGNORECASE)
+      Find.find(@builddir) do |fname|
+        next unless fname.match(extre)
+        img = ImageSize.path(fname)
+        next if img.width.nil? || img.width * img.height <= maxpixels
+        h = Math.sqrt(img.height * maxpixels / img.width)
+        w = maxpixels / h
+        fname.sub!("#{@builddir}/", '')
+        warn "#{fname}: #{img.width}x#{img.height} exceeds a limit. suggeted value is #{w.to_i}x#{h.to_i}"
+      end
+
+      true
+    end
+
+
+    class EPUBRenderer < BaseRenderer
+
+      def initialize(config, book, basedir, starter_config)
+        super
+        @language    = @config['language']
+        @stylesheets = @config['stylesheet']
+      end
+
+      protected
+
+      def layout_template_name()
+        #if @config['htmlversion'].to_i == 5
+        #  './html/layout-html5.html.erb'
+        #else
+        #  './html/layout-xhtml1.html.erb'
+        #end
+        "layout.epub.erb"
+      end
+
+      def escape(s)
+        return CGI.escapeHTML(s)
+      end
+
+      public
+
+      def generate_title_page(filepath)
+        c = @config
+        title = escape(c.name_of('booktitle'))
+        body = []
+        body << "<div class=\"titlepage\">\n"
+        body << "<h1 class=\"tp-title\">#{escape c.name_of('booktitle')}</h1>\n"
+        body << "<h2 class=\"tp-subtitle\">#{escape c.name_of('subtitle')}</h2>\n" if c['subtitle']
+        body << "<h2 class=\"tp-author\">#{escape _join_names(c.names_of('aut'))}</h2>\n" if c['aut']
+        body << "<h3 class=\"tp-publisher\">#{escape _join_names(c.names_of('pbl'))}</h3>\n" if c['pbl']
+        body << "</div>"
+        generate_file(filepath, {title: title, body: body.join()})
+      end
+
+      def generate_part_page(part, filepath)
+        part_number = i18n('part', part.number)
+        part_name   = part.name.strip()
+        body = []
+        body << "<div class=\"part\">\n"
+        body << "<h1 class=\"part-number\">#{escape part_number}</h1>\n"
+        body << "<h2 class=\"part-title\">#{escape part_name}</h2>\n" if part_name.present?
+        body << "</div>\n"
+        generate_file(filepath, {body: body.join()})
+      end
+
+      private
+
+      def _join_names(names)
+        sep = i18n('names_splitter')
+        return [names].flatten.join(sep)
+      end
+
+    end
+
+
+    class ReVIEWHeaderListener
+      include REXML::StreamListener
+
+      def initialize(headlines)
+        @headlines = headlines
+        _clear()
+      end
+
+      def _clear()
+        @content = ''
+        @level   = nil
+        @id      = nil
+        @notoc   = nil
+      end
+      private :_clear
+
+      HEADING_REXP = /\Ah(\d+)\z/
+
+      def tag_start(name, attrs)
+        v = nil
+        if name =~ HEADING_REXP
+          raise "#{name}, #{attrs}" if @level.present?
+          @level = $1.to_i
+          @id    = v if (v = attrs['id']).present?
+          @notoc = v if (v = attrs['notoc']).present?
+        elsif @level
+          if name == 'img'
+            @content << v if (v = attrs['alt']).present?
+          elsif name == 'a'
+            @id  = v if (v = attrs['id']).present?
+          end
+        end
+      end
+
+      def tag_end(name)
+        if name =~ HEADING_REXP
+          if @id
+            attrs = {'level'=>@level, 'id'=>@id, 'title'=>@content, 'notoc'=>@notoc}
+            @headlines.push(attrs)
+          end
+          _clear()
+        end
+        true
+      end
+
+      def text(text)
+        @content << text.gsub("\t", '　') if @level
+      end
+
+    end
+
+  end
+
+end