Yeah, I've read it. In the last chapters there's a kind of statistical/mathematical approach
in order to explain the possibilities of various matchings. Very hard&cool!

IMHO: perldoc perlre && perldoc perlretut