/ 
Switch-Perlish-1.0.5
River stage zero No dependents
++
 / Switch::Perlish
Dan Brook

NAME

Switch::Perlish - A Perlish implementation of the switch statement.

VERSION

1.0.5 - Mostly cosmetic changes for this release.

SYNOPSIS

  use Switch::Perlish;

  switch $var, sub {
    case 'foo',
      sub { print "$var is equal to 'foo'\n" };
    case 42,
      sub { print "$var is equal to 42\n";
            fallthrough };
    case [qw/ foo bar baz /],
      sub { print "$var found in list\n" };
    case { foo => 'bar' },
      sub { print "$var key found in hash\n" };
    case \&func,
      sub { print "$var as arg to func() returned true\n" };
    case $obj,
      sub { print "$var is method in $obj and returned true\n" };
    case qr/\bfoo\b/,
      sub { print "$var matched against foo\n" };
    default
      sub { print "$var did not find a match\n" };
  };

BACKGROUND

If you're unfamiliar with switch then this is the best place to start. A switch statement is essentially syntactic sugar for an if/elsif/else chain where the same $variable is tested in every conditional e.g:

  my $foo = 'a string';
  if($foo eq 'something') {
    print '$foo matched "something"';
  } elsif($foo eq 'a string') {
    print '$foo matched "a string"';
  } else {
    print '$foo matched nothing';
  }

This simply matches $foo against a series of strings, then defaulting to the last else block if nothing matched. An equivalent switch statement (using this module) would be:

  use Switch::Perlish;
  my $foo = 'a string';
  switch $foo, sub {
    case 'something',
      sub { print '$foo matched "something"' };
    case 'a string',
      sub { print '$foo matched "a string"'  };
    default
      sub { print '$foo matched nothing' };
  };

So the first argument to switch is the thing to be tested (in the code above, $foo), and the second argument is the block of tests. Each case statement matches its first argument against $foo, and if the match is successful, the associated block is executed, so running the above code outputs: $foo matched "a string". Note the use of semi-colon at the end of the switch, case and default calls - they're just simple subroutine calls.

DESCRIPTION

This is a Perl-oriented implementation of the switch statement. It uses smart-matching in cases which can be configured and extended by the user. There is no magical syntax so switch/case/default expect coderefs, which are most simply provided by anonymous subroutines. By default successful case statements do not fall through[1]. To fall through a case block call the fallthrough subroutine explicitly. For C style switch behaviour[2] simply call the module with an upper-case C i.e

  use Switch::Perlish 'C';

[1] To 'fall through' in a case block means that the switch block isn't exited upon success.

[2] upon a case succesfully matching all subsequent cases succeed; to break out from the current switch completely use stop.

Smart Matching

The idea behind smart matching is that the given values are matched in an intelligent manner, so as to get a meaningful result regardless of the values' types. This allows for flexible code and a certain amount of "just do it" when using smart matching. Below is a basic example using smart matching (which is done implictly in case) where a simple value is being matched against an array e.g

  use Switch::Perlish;

  my $num = $ARGV[0];
  
  switch $num, sub {
    case undef,
      sub { die "Usage: $0 NUM\n" };
    case [0 .. 10],
      sub { print "Your number was between 0 and 10" };
    case [11 .. 100],
      sub { print "Your number was between 11 and 100" };
    case [101 .. 1000],
      sub { print "Your number was between 101 and 1000" };
    default
      sub { print "Your number was less than 0 or greater than 1000" };
  };

So here the smart matching is checking for the existence of $num in the provided arrays. In the above code ranges happen to be used, but any array would suffice. To see how different value types compare with each other see. Switch::Perlish::Smatch::Comparators, which provides descriptions for all the default comparators.

The code behind this smart matching can be found in Switch::Perlish::Smatch which itself delegates to the appropriate comparator subroutine depending on the value types. See Switch::Perlish::Smatch for more details on the smart matching implementation and how it can be extended.

COMPARISON

Because there is an existing module which implements switch this section intends to provide clarification of the differences that module, Switch, and this one.

Native vs. New

To create a more natural switch syntax, Switch uses source filters[3], which facilitate the creation of this natural syntax. Switch::Perlish however uses the native syntax of perl, so what you code is what you see. The big advantage of source filtering is the ability to create new syntax, but it has several disadvantages - the new syntax can conflict with, and break, existing code, the filtered code can be difficult to debug and because you can't easily see the post-filtered code it can be difficult to integrate into production code. The raison d'être for this module is to have the syntax of switch without the baggage that goes with filtered code.

Extensibility

The Switch module deals with the Perl's types superbly, however, that is all, so there is no extensibility as such. This module was designed from the outset to allow an extensibilty of how types are dealt with, i.e how they are compared, and this is done through the companion module Switch::Perlish::Smatch.

The sub keyword

Unlike Switch, Switch::Perlish requires the use of the the sub keyword when creating blocks. This is because there is no standard way of magically coercing bare blocks into closures, unless one uses the (&) prototype, and that is only applicable where it is the first argument. Also, prototypes are too restrictive for what is intended as a very perlish module e.g

  $ perl -e 'sub f(&) { print $_[0]->() } sub g{'foo'} my $r = \&g; f $r'
  Type of arg 1 to main::f must be block or sub {} (not private variable)
  at -e line 1, at EOF
  Execution of -e aborted due to compilation errors.

So, for now, 3 extra keystrokes are necessary when using blocks with Switch::Perlish.

[3] see. Filter::Simple for more info on source filters.

SUBROUTINES

switch( $topic, $block )

Execute the given $block allowing case statements to access the $topic. This, along with case and default, will also attempt to return in the same manner as normal subroutines e.g you can assign to the result of them.

case( $match, $block )

If the current $topic successfully smart matches against $match then execute $block and exit from current switch, but if using C style switch behaviour, then continue executing the block and all subsequent case $blocks until the end of the current switch or a call to stop. Also, if using C style switch behaviour then $block is optional. NB: this subroutine cannot be called outside of switch, if you want to use smart matching functionality, see. Switch::Perlish::Smatch.

default( $block )

Execute $block and exit from switch. NB: this subroutine cannot be called outside of switch.

fallthrough()

Fall through the the current case block i.e continue to evaluate the rest of the switch block. NB: this subroutine cannot be called outside of case.

stop()

Use in case blocks to exit the current switch block, ideally when used with the C style behaviour as it mimics C's break. NB: this subroutine cannot be called outside of case.

Globals

$SWITCH

The current switch block.

$CASE

The current case block.

$TOPIC

The current topic block, also aliased to $_.

$MATCH

The current thing being matched against.

$CSTYLE

If Switch::Perlish is called with the C argument, this is set to true and C style switch behaviour is enabled.

$FALLING

Set to true when falling through the current switch block i.e set to true when fallthrough has been called.

SEE. ALSO

Switch

How do I create a switch or case statement?

Basic_BLOCKs_and_Switch_Statements

Switch::Perlish::Smatch

Switch::Perlish::Smatch::Comparators

TODO

  • Implement localizing comparators

  • Test with earlier versions of perl

  • Drop warnings for compatibility with older perls?

  • Allow lists as the topic and/or cases to match against

AUTHOR

Dan Brook <mr.daniel.brook@gmail.com>

COPYRIGHT

Copyright (c) 2006, Dan Brook. All Rights Reserved. This module is free software. It may be used, redistributed and/or modified under the same terms as Perl itself.