in reply to Is it possible to create a sub exclusive to a sub?

Here's a way that avoids re-creating the local sub every invocation of the main, and avoids namespace pollution in main. It also allows as many (named) "local" subs as required. 

#! perl -slw
use strict;

sub test{
    ## create a 'local' sub the first time we're called
    ## As many as you like each with it's own name
    *test->{localsub} = sub{
        print 'localsub1';
        return 12345;
    } unless exists *test->{localsub};
    
    ## use it
    *test->{localsub}();
}

print test;

## It's not truly local, but no namespace pollution.
## Anyone doing this, is doing it deliberately
*test->{localsub}();

__END__
P:\test>test
localsub1
12345
localsub1

[download]
Examine what is said, not who speaks.
"Efficiency is intelligent laziness." -David Dunham
"Think for yourself!" - Abigail
"Memory, processor, disk in that order on the hardware side. Algorithm, algorithm, algorithm on the code side." - tachyon

Replies are listed 'Best First'.
Re^2: Is it possible to create a sub exclusive to a sub?
by Aristotle (Chancellor) on Sep 19, 2004 at 10:19 UTC

    That's just obfuscated syntax for

    sub test {
    $::test{localsub} = sub {
        print 'localsub1';
        return 12345;
    } unless exists $::test{localsub};
    $::test{localsub}->();
}

    [download]

    and implies all the usual ramifications of global variable use.

    Makeshifts last the longest.

[reply]
[d/l]

      *test->{localsub} isn't equivalent to $::test{localsub} in the general case. It just happens to be that in the current package in the code BrowserUk wrote. You probably know that, but I just want to point out that. It's important to realize that *foo doesn't by any means imply the main namespace.

      ihb

      Read argumentation in its context!

[reply]
[d/l]
[select]

        Yes, I know that. I guess I should have added a remark to the effect that I'm not implying that *test->{localsub} is shared across packages. I didn't want to formulate the code using vars or our because that wouldn't have been equivalent — the difference and key point of my argument is that *test->{localsub} is an unchecked use of a global that elides strictures. So I had to use a fully qualified global variable name instead; obviously, the package would have to change according to the context of the code.

        Makeshifts last the longest.

[reply]

      The only "global variable use", it *main::test, which is in use anyway and has several slots going free. And I guess obfuscated is in the eye of the beholder.

      Examine what is said, not who speaks.
      "Efficiency is intelligent laziness." -David Dunham
      "Think for yourself!" - Abigail
      "Memory, processor, disk in that order on the hardware side. Algorithm, algorithm, algorithm on the code side." - tachyon
[reply]
[d/l]

        You use %main::test. That that's the same as *main::test{HASH} and *main::test{CODE} happens to be defined doesn't make it any less of a global variable. At least if you're going to go that route, do it directly

        use vars qw( %test );

# ...

$test{localsub}->();

        [download]

        so that strict can catch your typos. (Though it still won't catch $test{lcoalsub}, as opposed to a lexical mistyped as $lcoalsub.) The way you wrote it deprives you of protection wholly.

        Makeshifts last the longest.

[reply]
[d/l]

In Section Seekers of Perl Wisdom