converter.rb in asciidoctor-epub3-1.0.0.alpha.2

- old
+ new

@@ -2,12 +2,13 @@
 require_relative 'spine_item_processor'
 require_relative 'font_icon_map'
 
 module Asciidoctor
 module Epub3
-#WordJoiner = [8288].pack 'U*'
+# tried 8288, but it didn't work in older readers
 WordJoiner = [65279].pack 'U*'
+WordJoinerRx = RUBY_ENGINE_JRUBY ? /uFEFF/ : WordJoiner
 
 # Public: The main converter for the epub3 backend that handles packaging the
 # EPUB3 or KF8 publication file.
 class Converter
   include ::Asciidoctor::Converter
@@ -92,11 +93,11 @@
     end
   end
 
   # TODO aggregate authors of spine document into authors attribute(s) on main document
   def navigation_document node, spine
-    doctitle_sanitized = ((node.doctitle sanitize: true) || (node.attr 'untitled-label')).gsub WordJoiner, ''
+    doctitle_sanitized = (node.doctitle sanitize: true, use_fallback: true).gsub WordJoinerRx, ''
     lines = [%(<!DOCTYPE html>
 <html xmlns="http://www.w3.org/1999/xhtml" xmlns:epub="http://www.idpf.org/2007/ops" xml:lang="#{lang = (node.attr 'lang', 'en')}" lang="#{lang}">
 <head>
 <meta charset="UTF-8"/>
 <title>#{doctitle_sanitized}</title>
@@ -107,38 +108,34 @@
 <h1>#{doctitle_sanitized}</h1>
 <nav epub:type="toc" id="toc">
 <h2>#{node.attr 'toc-title'}</h2>
 <ol>)]
     spine.each do |item|
-      lines << %(<li><a href="#{item.id || (item.attr 'docname')}.xhtml">#{((item.doctitle sanitize: true) || (item.attr 'untitled-label')).gsub WordJoiner, ''}</a></li>)
+      lines << %(<li><a href="#{item.id || (item.attr 'docname')}.xhtml">#{(item.doctitle sanitize: true, use_fallback: true).gsub WordJoinerRx, ''}</a></li>)
     end
     lines << %(</ol>
 </nav>
 </body>
 </html>)
     lines * EOL
   end
 
   def document node
     docid = node.id
-    if (doctitle = node.doctitle)
-      doctitle_sanitized = (node.doctitle sanitize: :sgml).gsub WordJoiner, ''
-      if doctitle.include? ': '
-        title, _, subtitle = doctitle.rpartition ': '
-      else
-        # HACK until we get proper handling of title-only in CSS
-        title = ''
-        subtitle = doctitle
-      end
+
+    if (doctitle = node.doctitle partition: true, sanitize: true, use_fallback: true).subtitle?
+      title = doctitle.main
+      title_upper = title.upcase
+      subtitle = doctitle.subtitle
     else
       # HACK until we get proper handling of title-only in CSS
-      title = ''
-      subtitle = node.attr 'untitled-label'
+      title = title_upper = ''
+      subtitle = doctitle.combined
     end
-    subtitle_formatted = subtitle.gsub(WordJoiner, '').split(' ').map {|w| %(<b>#{w}</b>) } * ' '
 
-    title_upper = title.upcase
+    doctitle_sanitized = (node.doctitle sanitize: true, use_fallback: true).gsub WordJoinerRx, ''
+    subtitle_formatted = subtitle.gsub(WordJoinerRx, '').split(' ').map {|w| %(<b>#{w}</b>) } * ' '
     # FIXME make this uppercase routine more intelligent, less fragile
     subtitle_formatted_upper = subtitle_formatted.upcase
         .gsub(UppercaseTagRx) { %(<#{$1}#{$2.downcase}>) }
         .gsub(NamedEntityRx) { %(&#{$1.downcase};) }
 
@@ -792,11 +789,11 @@
   def xml_sanitize value, target = :attribute
     sanitized = (value.include? '<') ? value.gsub(XmlElementRx, '').tr_s(' ', ' ').strip : value
     if target == :plain && (sanitized.include? ';')
       sanitized = sanitized.gsub(CharEntityRx) { [$1.to_i].pack('U*') }.gsub(FromHtmlSpecialCharsRx, FromHtmlSpecialCharsMap)
     elsif target == :attribute
-      sanitized = sanitized.gsub(WordJoiner, '').gsub('"', '&quot;')
+      sanitized = sanitized.gsub(WordJoinerRx, '').gsub('"', '&quot;')
     end
     sanitized
   end
 
   # TODO make check for last content paragraph a feature of Asciidoctor
@@ -815,10 +812,10 @@
 class DocumentIdGenerator
   class << self
     def generate_id doc
       unless (id = doc.id)
         id = if doc.header?
-          doc.doctitle(sanitize: :sgml).gsub(WordJoiner, '').downcase.delete(':').tr_s(' ', '-').tr_s('-', '-')
+          doc.doctitle(sanitize: true).gsub(WordJoinerRx, '').downcase.delete(':').tr_s(' ', '-').tr_s('-', '-')
         elsif (first_section = doc.first_section)
           first_section.id
         else
           %(document-#{doc.object_id})
         end