my $self = shift not only defines $self; it also modifies @_ (removing the first array item). The difference is only significant if you intend on using @_ or goto later on in the sub.
Update: Personally I've recently decided to adopt a coding style where I use shift for the invocant (i.e. $self or $class; but also "conceptual invocants" like $orig in Moose around method modifiers, or $ctx in Catalyst actions) and list assignment for other parameters.
For example:
use v5.14;
package Person {
use Moose;
has first_name => (is => 'rw', isa => 'Str');
has last_name => (is => 'rw', isa => 'Str');
sub set_name {
my $self = shift;
my ($first, $last) = @_;
$self->first_name($first);
$self->last_name($last);
}
}
package LoudNameChanges {
use Moose::Role;
around set_name => sub {
my $orig = shift;
my $self = shift;
say "Changing name to: @_";
$self->$orig(@_);
};
}
package Person::LoudNameChanges {
use Moose;
extends 'Person'; with 'LoudNameChanges';
}
my $bob = Person::LoudNameChanges->new(
first_name => 'Robert',
last_name => 'Smith',
);
$bob->set_name('David', 'Jones');
print $bob->dump;
For regular subs which are not intended to be called as methods, there is no invocant, so I just use list assignment to unpack all the arguments.
Why did I decide to adopt this style? I find it works nicely in terms of delegating method calls and so forth where you often need to pass on @_ to other methods and functions, but without passing on the invocant.
I'm not yet very strict with myself about this style, but hopefully it will become habit soon.
package Cow { use Moo; has name => (is => 'lazy', default => sub { 'Mooington' }) } say Cow->new->name
|