Evolution of a solution
Combing code, I found I have used these 5 different solutions to a common task for my work:
A command returns an unordered list of integers, with some duplicates.
Remove duplicates (after sorting the list).
(Listed below from older to more recent):
mkmcconn
#!/usr/bin/perl -w
use strict;
use Benchmark;
my $n = 10000;
my $o = rand($n);
my @rray = (1..$n,$o..$n); # make an array and pollute it
# with duplicates
my @array = ();
@array = sort {$a <=> $b} @rray;
my %tests = (
totemp => sub {
my @newarray = ();
my $i = 0;
for ( @array ){
my $temp = $array[ ++$i ] ? $array[ $i ]: "";
push @newarray, $_ unless $_ eq $temp;
}
},
ternary => sub {
my @newarray = ();
my $i = 0;
for ( @array ){
push @newarray, $_ unless $_ eq $array[ $array[ $i+1 ] ? ++$
+i : 0 ];
}
},
toend => sub {
my @newarray = ();
my $i = 0;
for my $list (@array){
if ( $array[ ++$i ] ){
push @newarray, $list unless $list eq $array[ $i ];
}else{
push @newarray,$array[-1];
}
}
},
cookbook => sub {
my @newarray = ();
my %duplicates = ();
@newarray = grep { ! $duplicates{$_} ++ } @array;
},
modulus => sub {
my @newarray = ();
for (0..(@array-2)){
$array[$_] % $array[$_+1] and push @newarray, $array[$_];
}
push @newarray, $array[-1];
},
);
timethese ( shift @ARGV || -5, \%tests);
Update:
Thanks chipmunk
ternary is a bizarre solution - and as you discovered (and
I did not realize, but it's obvious looking at it now), it breaks completely when there's a 0 in the array! The
original was echo'd from the commandline, and appears to have been used to remove duplicates from a list of words.
modulus also works as long as there is no zero in the list. If there is more than one zero, it will not run. If there is one zero, it will be skipped.
Thanks for taking time to comment, chipmunk. Some small adjustments made in the post, to prevent complete failure of the tests.
