Mastering Perl, 2nd Edition

Chapter 8. Dynamic Subroutines

For the purposes of this chapter, I’m going to label as “dynamic subroutines” anything I don’t explicitly name by typing sub some_name or that doesn’t exist until runtime. Perl is extremely flexible in letting me figure out the code as I go along, and I can even have code that writes code. I’m going to lump a bunch of different subroutine topics in this chapter just because there’s no good home for them apart from each other.

We first showed anonymous subroutines in Learning Perl when we showed user-defined sorting, although we didn’t tell you that they were anonymous subroutines. In Intermediate Perl we used them to create closures, work with map and grep, and do a few other things. I’ll pick up where Intermediate Perl left off to show just how powerful they can be. With any of these tricks, not knowing everything ahead of time can be very liberating.

Subroutines as Data

I can store anonymous subroutines in variables. They don’t actually execute until I tell them to. Instead of storing values, I store behavior. This anonymous subroutine adds its first two arguments and returns the result, but it won’t do that until I execute it. I merely define the subroutine and store it in $add_sub:

my $add_sub = sub { $_[0] + $_[1] };

This way, I can decide what to do simply by choosing the variable that has the behavior that I want. A simple-minded program might do this with a series of if-elsif tests and branches because it needs to hardcode ...