reproducer project for a codetrans issue, when a string with \u-escaped values is used the files generated by codetrans for js and groovy have the unquoted representation of the string in the source. this causes a parse error in asciidoctor with invalid utf8 sequence or leaves binary chars in the file it is valid utf8.

the same error was fixed for examples includes before, looks like the code is converted differently for code in the asciidoc source.