Yes, I guess you're most probably right that %+ hash is probably lost when you enter a new regex. Although my initial point was really about using the named capture within the same regex.

Besides the fact that I wrote my previous post on a mobile device while commuting to work, with no chance to test, I guess I got carried away by the fact that I am working much more on Perl 6 than on Perl 5 these days when it comes to these matters.

Having said that, I still believe that named captures can bring you something (limited) in that direction. But too late now, I'll try to come up with something tomorrow.