On Tue, Jun 28, 2005 at 01:19:20PM +1000, Adam Kennedy wrote: > >Yes... quasiquoting in Perl 5 is currently crudely emulated > >by feeding things to PPI::Tokenizer and PPI::Transform. :-) > > PPI is not a code parser. By code parser I mean taking a string and > turning it into working code. PPI is a document parser. It parses ONLY > on syntax to form a model of a document. Any conclusions it might make > on the structure of the resulting code are secondary.
Right. Hence it guarantees at most well-formness, instead of semantic correctness. Hence, "crudely emulated", because we don't have anything better at this moment. > So to summarise, PPI is of limited use when it comes to working with > bytecode or something that will be executed. It is for working with > documents, not code. Yup. But PPI is still much better than consuming and concatenating raw strings. :-) Thanks, /Autrijus/
pgp21Qkcvnkj2.pgp
Description: PGP signature
