Added Tizen formatter as a copy of the Android formatter.

lib/twine/formatters.rb

@@ -5,9 +5,10 @@ require 'twine/formatters/flash'
 require 'twine/formatters/gettext'
 require 'twine/formatters/jquery'
 require 'twine/formatters/django'
+require 'twine/formatters/tizen'
 
 module Twine
   module Formatters
-    FORMATTERS = [Formatters::Apple, Formatters::Android, Formatters::Gettext, Formatters::JQuery, Formatters::Flash, Formatters::Django]
+    FORMATTERS = [Formatters::Apple, Formatters::Android, Formatters::Gettext, Formatters::JQuery, Formatters::Flash, Formatters::Django, Formatters::Tizen]
   end
 end

lib/twine/formatters/tizen.rb

@@ -0,0 +1,155 @@
+# encoding: utf-8
+require 'cgi'
+require 'rexml/document'
+
+module Twine
+  module Formatters
+    class Tizen < Abstract
+      FORMAT_NAME = 'tizen'
+      EXTENSION = '.xml'
+      DEFAULT_FILE_NAME = 'strings.xml'
+      LANG_CODES = Hash[
+        'zh' => 'zh-Hans',
+        'zh-rCN' => 'zh-Hans',
+        'zh-rHK' => 'zh-Hant',
+        'en-rGB' => 'en-UK',
+        'in' => 'id',
+        'nb' => 'no'
+        # TODO: spanish
+      ]
+      DEFAULT_LANG_CODES = Hash[
+        'zh-TW' => 'zh-Hant' # if we don't have a zh-TW translation, try zh-Hant before en
+      ]
+
+      def self.can_handle_directory?(path)
+        Dir.entries(path).any? { |item| /^values.*$/.match(item) }
+      end
+
+      def default_file_name
+        return DEFAULT_FILE_NAME
+      end
+
+      def determine_language_given_path(path)
+        path_arr = path.split(File::SEPARATOR)
+        path_arr.each do |segment|
+          if segment == 'values'
+            return @strings.language_codes[0]
+          else
+            match = /^values-(.*)$/.match(segment)
+            if match
+              lang = match[1]
+              lang = LANG_CODES.fetch(lang, lang)
+              lang.sub!('-r', '-')
+              return lang
+            end
+          end
+        end
+
+        return
+      end
+
+      def read_file(path, lang)
+        resources_regex = /<resources(?:[^>]*)>(.*)<\/resources>/m
+        key_regex = /<string name="(\w+)">/
+        comment_regex = /<!-- (.*) -->/
+        value_regex = /<string name="\w+">(.*)<\/string>/
+        key = nil
+        value = nil
+        comment = nil
+
+        File.open(path, 'r:UTF-8') do |f|
+          content_match = resources_regex.match(f.read)
+          if content_match
+            for line in content_match[1].split(/\r?\n/)
+              key_match = key_regex.match(line)
+              if key_match
+                key = key_match[1]
+                value_match = value_regex.match(line)
+                if value_match
+                  value = value_match[1]
+                  value = CGI.unescapeHTML(value)
+                  value.gsub!('\\\'', '\'')
+                  value.gsub!('\\"', '"')
+                  value = iosify_substitutions(value)
+                  value.gsub!(/(\\u0020)*|(\\u0020)*\z/) { |spaces| ' ' * (spaces.length / 6) }
+                else
+                  value = ""
+                end
+                set_translation_for_key(key, lang, value)
+                if comment and comment.length > 0 and !comment.start_with?("SECTION:")
+                  set_comment_for_key(key, comment)
+                end
+                comment = nil
+              end
+
+              comment_match = comment_regex.match(line)
+              if comment_match
+                comment = comment_match[1]
+              end
+            end
+          end
+        end
+      end
+
+      def write_file(path, lang)
+        default_lang = nil
+        if DEFAULT_LANG_CODES.has_key?(lang)
+          default_lang = DEFAULT_LANG_CODES[lang]
+        end
+        File.open(path, 'w:UTF-8') do |f|
+          f.puts "<?xml version=\"1.0\" encoding=\"utf-8\"?>\n<!-- Tizen Strings File -->\n<!-- Generated by Twine #{Twine::VERSION} -->\n<!-- Language: #{lang} -->"
+          f.write '<resources>'
+          @strings.sections.each do |section|
+            printed_section = false
+            section.rows.each do |row|
+              if row.matches_tags?(@options[:tags], @options[:untagged])
+                if !printed_section
+                  f.puts ''
+                  if section.name && section.name.length > 0
+                    section_name = section.name.gsub('--', '—')
+                    f.puts "\t<!-- SECTION: #{section_name} -->"
+                  end
+                  printed_section = true
+                end
+
+                key = row.key
+
+                value = row.translated_string_for_lang(lang, default_lang)
+                if !value && @options[:include_untranslated]
+                  value = row.translated_string_for_lang(@strings.language_codes[0])
+                end
+
+                if value # if values is nil, there was no appropriate translation, so let Tizen handle the defaulting
+                  value = String.new(value) # use a copy to prevent modifying the original
+
+                  # Tizen enforces the following rules on the values
+                  #  1) apostrophes and quotes must be escaped with a backslash
+                  value.gsub!('\'', '\\\\\'')
+                  value.gsub!('"', '\\\\"')
+                  #  2) HTML escape the string
+                  value = CGI.escapeHTML(value)
+                  #  3) fix substitutions (e.g. %s/%@)
+                  value = androidify_substitutions(value)
+                  #  4) replace beginning and end spaces with \0020. Otherwise Tizen strips them.
+                  value.gsub!(/\A *| *\z/) { |spaces| '\u0020' * spaces.length }
+
+                  comment = row.comment
+                  if comment
+                    comment = comment.gsub('--', '—')
+                  end
+
+                  if comment && comment.length > 0
+                    f.puts "\t<!-- #{comment} -->\n"
+                  end
+                  f.puts "\t<string name=\"#{key}\">#{value}</string>"
+                end
+              end
+            end
+          end
+
+          f.puts '</resources>'
+        end
+      end
+    end
+  end
+end