in reply to Re^6: Shifty Politics
in thread stoping and restarting a loop

I'd rather have two dozen shifts than a single goto.

I wonder if that particular form of goto were renamed to

forward_my_args_without_the_need_to_build_a_new_stackframe_or_replicate_the_args_array()

would you still be so reluctant to use it? (Apart from the stupid long name of course:)

moving the @_ check to the front of the subroutine might not be viable

An example of when this might be the case?

I still can't see why using shift is better in any respect

I don't think I ever said it was better. I just fail to see why one peice of perl syntax should be eshewed simply on the basis of prejudice.

My use in the original code was simple, clear, precise and correct. The only example you have given as to why it should not be used that (to me) holds water, is your example of making it easier to debug if @_ is left intact, but even then, placing a breakpoint on the initial list assignment goes a long way to negating that arguement. Especially as you are advocating using shift to obtain the first arguement where that happens to be a class handle.

As for where I would draw the line, a guess that would be wherever it makes most sense given the particular example I am coding. Given time and more experience with perl, this could change.

Well It's better than the Abottoire, but Yorkshire!

Replies are listed 'Best First'.
Re^8: Shifty Politics
by tadman (Prior) on Sep 10, 2002 at 22:23 UTC
    My argument is simply that using shift to extract several arguments is usually inefficient and, by induction, not very desirable: 
    my ($foo_a) = @_;                       # Simple
my $foo_a = shift;                      # Comparable
my ($foo_a) = (shift);                  # Odd

my ($foo_a, $foo_b) = @_;               # Typical
my ($foo_a, $foo_b) = (shift, shift);   # Odder

my ($foo_a, $foo_b, $foo_c, $foo_d) = @_;
my ($foo_a, $foo_b, $foo_c, $foo_d) = (shift, shift, shift, shift);

    [download]
    The last one is, I hope you'll agree, outrageous, and is best reserved for those sitations where you'd explain it, but it's a "long story".

    I'm not saying that you shouldn't use shift. You can program any way you like. You can also use goto with impunity, variable names that make no sense, and use substr and unpack instead of regular expressions. What I'm suggesting is that using multiple shifts in a function is poor form and is a bad habit to develop.

    Oh, and a goto by any other name still smells as rotten. I'd like to see any Benchmark that could show that using that technique you describe could provide measurable performance gains. As far as I can tell, passing the whole args array is fairly efficient. Pulling them apart and repacking them isn't.
[reply]
[d/l]

      I'd like to see any Benchmark that could show that using that technique you describe could provide measurable performance gains.

      #! perl -w
use strict;
use Benchmark qw(cmpthese);

sub parent {
    return ++$_[$#_]; # minimal use of entire passed array.
}

sub child1 {
    my ($self, @rest) = @_;
    return parent(@rest);
}

sub child2 {
    my $self = shift;
    goto &parent;
}

print 'child1: ', child1( "fred", 1..1000 ), $/;
print 'child2: ', child2( "bill", 1..1000 ), $/;

cmpthese( 10000, {
    noShiftOrGoto    => sub { child1( "fred", 1..1000 ); },
    shiftAndGoto    => sub { child2( "bill", 1..1000 ); },
});

__DATA__
C:\test>196823
child1: 1001
child2: 1001
Benchmark: timing 10000 iterations of noShiftOrGoto, shiftAndGoto...
noShiftOrGoto: 33 wallclock secs (32.33 usr +  0.00 sys = 32.33 CPU) @
+ 309.35/s (n=10000)
shiftAndGoto:  6 wallclock secs ( 5.85 usr +  0.00 sys =  5.85 CPU) @ 
+1709.69/s (n=10000)
                Rate noShiftOrGoto  shiftAndGoto
noShiftOrGoto  309/s            --          -82%
shiftAndGoto  1710/s          453%            --

C:\test>

      [download]
      Well It's better than the Abottoire, but Yorkshire!
[reply]
[d/l]
        That's not precisely a fair test. How about this? 
        #!/usr/bin/perl

use Benchmark 'cmpthese';

package Foo;

sub new
{
        my $class = shift;
        bless({@_}, $class);
}

sub foo
{
        my ($self, @stuff) = @_;

        return map { $_.$_ } @stuff;
}

sub foo_pipe
{
        my $self = shift;
        $self->foo(@_);
}

sub foo_my
{
        my ($self,@stuff) = @_;
        $self->foo(@stuff);
}

sub foo_goto
{
        goto &foo;
}

package main;

my $foo = Foo->new();
my @stuff = 0..100;
my @foo;

cmpthese(500, {
        pipe => sub { $foo->foo_pipe(@foo=@stuff) },
        goto => sub { $foo->foo_goto(@foo=@stuff) },
        my   => sub { $foo->foo_my(@foo=@stuff) },
});

        [download]
        This actually uses the entire array. The difference is rather minor, easily buried by the subroutine overhead. 
        Benchmark: timing 5000 iterations of goto, my, pipe...
      goto:  3 wallclock secs ( 2.96 usr +  0.00 sys =  2.96 CPU) @ 16
+89.19/s (n=5000)
        my:  4 wallclock secs ( 3.18 usr +  0.01 sys =  3.19 CPU) @ 15
+67.40/s (n=5000)
      pipe:  3 wallclock secs ( 2.89 usr +  0.00 sys =  2.89 CPU) @ 17
+30.10/s (n=5000)
       Rate   my goto pipe
my   1567/s   --  -7%  -9%
goto 1689/s   8%   --  -2%
pipe 1730/s  10%   2%   --

        [download]
[reply]
[d/l]
[select]

In Section Seekers of Perl Wisdom