They mean something that can parse the language. You could adapt B::Deparse to output something that produces the output requested by CPD. The thing is though... the work I suggested in the previous node is very do-able. That's a relatively short script. I have no need for this and so won't write it but I could see this being a relatively minor thing.
Re: Re: Re: Finding duplicated code in Perl