NAME - Provide a pseudo-class NEXT that allows method redispatch


    use NEXT;
    package A;
    sub A::method   { print "$_[0]: A method\n";   $_[0]->NEXT::method() }
    sub A::DESTROY  { print "$_[0]: A dtor\n";     $_[0]->NEXT::DESTROY() }
    package B;
    use base qw( A );
    sub B::AUTOLOAD { print "$_[0]: B AUTOLOAD\n"; $_[0]->NEXT::AUTOLOAD() }
    sub B::DESTROY  { print "$_[0]: B dtor\n";     $_[0]->NEXT::DESTROY() }
    package C;
    sub C::method   { print "$_[0]: C method\n";   $_[0]->NEXT::method() }
    sub C::AUTOLOAD { print "$_[0]: C AUTOLOAD\n"; $_[0]->NEXT::AUTOLOAD() }
    sub C::DESTROY  { print "$_[0]: C dtor\n";     $_[0]->NEXT::DESTROY() }
    package D;
    use base qw( B C );
    sub D::method   { print "$_[0]: D method\n";   $_[0]->NEXT::method() }
    sub D::AUTOLOAD { print "$_[0]: D AUTOLOAD\n"; $_[0]->NEXT::AUTOLOAD() }
    sub D::DESTROY  { print "$_[0]: D dtor\n";     $_[0]->NEXT::DESTROY() }
    package main;
    my $obj = bless {}, "D";
    $obj->method();		# Calls D::method, A::method, C::method
    $obj->missing_method(); # Calls D::AUTOLOAD, B::AUTOLOAD, C::AUTOLOAD
    # Clean-up calls D::DESTROY, B::DESTROY, A::DESTROY, C::DESTROY

DESCRIPTION adds a pseudoclass named NEXT to any program that uses it. If a method m calls $self-NEXT::m()>, the call to m is redispatched as if the calling method had not originally been found.

In other words, a call to $self-NEXT::m()> resumes the depth-first, left-to-right search of $self's class hierarchy that resulted in the original call to m.

Note that this is not the same thing as $self-SUPER::m()>, which begins a new dispatch that is restricted to searching the ancestors of the current class. $self-NEXT::m()> can backtrack past the current class -- to look for a suitable method in other ancestors of $self -- whereas $self-SUPER::m()> cannot.

A typical use would be in the destructors of a class hierarchy, as illustrated in the synopsis above. Each class in the hierarchy has a DESTROY method that performs some class-specific action and then redispatches the call up the hierarchy. As a result, when an object of class D is destroyed, the destructors of all its parent classes are called (in depth-first, left-to-right order).

Another typical use of redispatch would be in AUTOLOAD'ed methods. If such a method determined that it was not able to handle a particular call, it might choose to redispatch that call, in the hope that some other AUTOLOAD (above it, or to its left) might do better.

By default, if a redispatch attempt fails to find another method elsewhere in the objects class hierarchy, it quietly gives up and does nothing (but see "Enforcing redispatch"). This gracious acquiesence is also unlike the (generally annoying) behaviour of SUPER, which throws an exception if it cannot redispatch.

Note that it is a fatal error for any method (including AUTOLOAD) to attempt to redispatch any method that does not have the same name. For example:

        sub D::oops { print "oops!\n"; $_[0]->NEXT::other_method() }

Enforcing redispatch

It is possible to make NEXT redispatch more demandingly (i.e. like SUPER does), so that the redispatch throws an exception if it cannot find a "next" method to call.

To do this, simple invoke the redispatch as:


rather than:


The ACTUAL tells NEXT that there must actually be a next method to call, or it should throw an exception.

NEXT::ACTUAL is most commonly used in AUTOLOAD methods, as a means to decline an AUTOLOAD request, but preserve the normal exception-on-failure semantics:

		if ($AUTOLOAD =~ /foo|bar/) {
			# handle here
		else {  # try elsewhere

By using NEXT::ACTUAL, if there is no other AUTOLOAD to handle the method call, an exception will be thrown (as usually happens in the absence of a suitable AUTOLOAD).

Avoiding repetitions

If NEXT redispatching is used in the methods of a "diamond" class hierarchy:

	#     A   B
	#    / \ /
	#   C   D
	#    \ /
	#     E
	use NEXT;
	package A;                 
	sub foo { print "called A::foo\n"; shift->NEXT::foo() }
	package B;                 
	sub foo { print "called B::foo\n"; shift->NEXT::foo() }
	package C; @ISA = qw( A );
	sub foo { print "called C::foo\n"; shift->NEXT::foo() }
	package D; @ISA = qw(A B);
	sub foo { print "called D::foo\n"; shift->NEXT::foo() }
	package E; @ISA = qw(C D);
	sub foo { print "called E::foo\n"; shift->NEXT::foo() }

then derived classes may (re-)inherit base-class methods through two or more distinct paths (e.g. in the way E inherits A::foo twice -- through C and D). In such cases, a sequence of NEXT redispatches will invoke the multiply inherited method as many times as it is inherited. For example, the above code prints:

        called E::foo
        called C::foo
        called A::foo
        called D::foo
        called A::foo
        called B::foo

(i.e. A::foo is called twice).

In some cases this may be the desired effect within a diamond hierarchy, but in others (e.g. for destructors) it may be more appropriate to call each method only once during a sequence of redispatches.

To cover such cases, you can redispatch methods via:


rather than:


This causes the redispatcher to skip any classes in the hierarchy that it has already visited in an earlier redispatch. So, for example, if the previous example were rewritten:

        package A;                 
        sub foo { print "called A::foo\n"; shift->NEXT::UNSEEN::foo() }
        package B;                 
        sub foo { print "called B::foo\n"; shift->NEXT::UNSEEN::foo() }
        package C; @ISA = qw( A );
        sub foo { print "called C::foo\n"; shift->NEXT::UNSEEN::foo() }
        package D; @ISA = qw(A B);
        sub foo { print "called D::foo\n"; shift->NEXT::UNSEEN::foo() }
        package E; @ISA = qw(C D);
        sub foo { print "called E::foo\n"; shift->NEXT::UNSEEN::foo() }

then it would print: called E::foo called C::foo called A::foo called D::foo called B::foo

and omit the second call to A::foo.

Note that you can also use:




to get both unique invocation and exception-on-failure.


Damian Conway (


Because it's a module, not an integral part of the interpreter, has to guess where the surrounding call was found in the method look-up sequence. In the presence of diamond inheritance patterns it occasionally guesses wrong.

It's also too slow (despite caching).

Comment, suggestions, and patches welcome.


 Copyright (c) 2000-2001, Damian Conway. All Rights Reserved.
 This module is free software. It may be used, redistributed
    and/or modified under the same terms as Perl itself.