Have a look at split, perlretut and perlre. It seems promising to split at the + and handle the different chunks separately, but you have to know a bit more about the structure of the data.

BTW you don't need to specify a doctype, header and body-tags here.