@@ -115,6 +115,7 @@ class Message @mail.subject end + def body @body ||= begin if @mail.multipart? @@ -123,8 +124,9 @@ class Message scan_parts(@mail) @body = @body.join("\n") else - @mail.body + @body = @mail.body end + @mail.charset ? @body : convert_to_utf8(@body) end end @@ -224,4 +226,19 @@ protected } s end + + # Attempts to run iconv conversions in common charsets to UTF-8. Needed for + # those crappy emails that don't properly specify a charset in the headers. + ICONV_CONVERSIONS = %w(utf-8 ISO-8859-1 ISO-8859-2 ISO-8859-3 ISO-8859-4 ISO-8859-5 ISO-8859-6 ISO-8859-7 ISO-8859-8 ISO-8859-9 + ISO-8859-15 GB2312) + def convert_to_utf8(s) + ICONV_CONVERSIONS.each do |from| + begin + return Iconv.iconv(ICONV_CONVERSIONS[0], from, s).to_s + rescue Iconv::IllegalSequence + ensure + s + end + end + end end \ No newline at end of file app/models/message.rb @@ -138,7 +138,7 @@ describe Message do @message.body.should == "Dear Sirs, \r\nWe are given to understand that you are Manufacturer of plstic Bottles\r\nAdd： blah China" end end - + describe "utf-8 encoded headers" do before :all do @raw = mail("utf-8") spec/models/message_spec.rb 827e038468ebff6f936430f2aa9e315ad302304c rick technoweenie@gmail.com http://github.com/entp/astrotrain/commit/3df35c6580bcb4c84848c5b1115e8928ea3da288 3df35c6580bcb4c84848c5b1115e8928ea3da288 2009-07-06T15:54:32-07:00 2009-07-06T15:54:32-07:00 attempt to convert mail bodies with bad/unspecified charsets into utf-8 3e981228515f68988b671db1e4b4a096f25eb446 rick technoweenie@gmail.com