in reply to Can this code be optimized further?

Another way. 

## I'd personally use qw( ) here, but this is probalby just sample dat
+a
my @temp = ("a_1","b_1","a_2","a_3","a_4","b_2","a_5","b_3","a_6","b_4
+");

my( @a, @b );
my %arrays;
eval qq{\$arrays{ "$_" } = \\\@$_} for qw( a b );
for( @temp ) {
  die "Invalid prefix on element '$_'\n" unless /^([ab])_(.*)/;
  push @{ $arrays{ $1 } }, $2;
}

[download]

Update: Gah, left off a couple of \\ to get a reference in the eval.

Replies are listed 'Best First'.
Re^2: Can this code be optimized further?
by dragonchild (Archbishop) on Feb 10, 2005 at 14:19 UTC
    eval qq{\$arrays{ "$_" } = \\\@$_} for qw( a b ); ?!?!?

    Am I the only one that thinks unncessary evals are just ugly? While I commend you for trying to keep the surrounding code unchanged (by creating @a and @b), I think the better solution is to improve the surrounding code by making sure the appropriate data structures are used - in this case, a HoA is appropriate.

    Being right, does not endow the right to be rude; politeness costs nothing.
    Being unknowing, is not the same as being stupid.
    Expressing a contrary opinion, whether to the individual or the group, is more often a sign of deeper thought than of cantankerous belligerence.
    Do not mistake your goals as the only goals; your opinion as the only opinion; your confidence as correctness. Saying you know better is not the same as explaining you know better.

[reply]
[d/l]

      If you can think of a better way to get a reference to an arbitrary lexical without eval sure. The original question was how to get things into arrays @a and @b; this gets things into the lexical arrays.

      Granted they may really need a HoA to begin with, but this technique (or the equivalent $arrays{ $_ } = do { no strict 'refs'; \@{ $_ } } for a package variable) sometimes is useful.

[reply]
[d/l]
[select]
Re^2: Can this code be optimized further?
by Roy Johnson (Monsignor) on Feb 10, 2005 at 15:16 UTC
    You can use a hash to reroute things to the lexical arrays so you don't have to eval: 
    my (@a, @b);
my %router = (a => \@a, b => \@b);
foreach my $value (@temp) {
    push @{$router{$1}}, $2 if $value =~ /([ab])_(.*)/;
}
print "@a <==> @b";

    [download]
    Caution: Contents may have been coded under pressure.
[reply]
[d/l]

      And if you had 10 arrays you'd want to keep that hash up to date? 20? 30?

      (Of course not, you'd use a HoA and not screw with this in the first place. :)

      That being said, in this specific case of just a and b I'd personally have probably used something like BrowserUK's @{ $1 eq 'a' ? \@a : \@b} below; but the eval's more flexible if you have a large number of destinations (then again with a large number of destinations you'd probably want an HoA). Of course at the moment I'd be more likely to just use Enumerable#partition and be done with it, but that's another language all together.

[reply]
[d/l]
        And if you had 10 arrays you'd want to keep that hash up to date? 20? 30?
        The effort is linear with declaring them in the first place: one entry in the hash for each new array.

        And if your names are simply consecutive letters, the additional effort is constant: just change the ending character in the slice below. 

        my @temp=qw(a_1 b_1 a_2 a_3 a_4 e_2 a_5 b_3 a_6 b_4);

my %router;
@router{'a'..'e'} = \(my (@a, @b, @c, @d, @e));
my $keys = join '', keys %router;

foreach my $value (@temp) {
    push @{$router{$1}}, $2 if $value =~ /([$keys])_(.*)/;
}
print "@a <==> @b, @e\n";

        [download]
        Caution: Contents may have been coded under pressure.
[reply]
[d/l]

In Section Seekers of Perl Wisdom