compile errors

pglenski has asked for the wisdom of the Perl Monks concerning the following question:

I found this code on your site I copied to my system and I get these errors. This was to return a unique string. Any ideas?

syntax error at 3.pl line 5, near "%tmp{"
syntax error at 3.pl line 7, near "}"
Execution of 3.pl aborted due to compilation errors.
[download]

#!/usr/bin/perl
sub unique {
  my $self=shift;
  my %tmp;
  map { %tmp{$_}=1 } @_;
  return keys %tmp;
}

@array = qw(apples cheeries apples bananas apples);
$o->unique(@array);
[download]

Comment on compile errors Select or Download Code

Replies are listed 'Best First'.
Re: compile errors by tirwhan (Abbot) on Feb 24, 2006 at 23:23 UTC
To access a hash element you need to use `$hash{key}`, not `%hash{key}`. So change `map { %tmp{$_}=1 } @_;` to `map { $tmp{$_}=1 } @_;` Or consider using the `uniq` function from List::MoreUtils (which has the added benefit of preserving order in your array). All dogma is stupid.	[reply] [d/l] [select]
Re: compile errors by InfiniteSilence (Curate) on Feb 24, 2006 at 23:31 UTC
"on your site"... Who's code was it? Code corrections: `#!/usr/bin/perl -w use strict; my $o; sub unique { my $self=shift; my %tmp; map { $tmp{$_}=1 } @_; return keys %tmp; } my @array = qw(apples cheeries apples bananas apples); #$o->unique(@array); print unique(@array); 1;` [download] $o->unique means you are calling the unique method of the $o object, which your code doesn't show being instantiated anywhere. The problem you listed was with the %tmp, which is basically incorrect syntax when you are trying to assign a value to a hash element. Celebrate Intellectual Diversity	[reply] [d/l]
Re: compile errors by ayrnieu (Beadle) on Feb 25, 2006 at 01:05 UTC
`#! /usr/bin/env perl package Foo; use strict; use warnings; sub new { bless {}, shift } sub uniq { shift; my %h; $h{$_}++ for @_; grep { $h{$_} == 1 } keys %h + } package main; print "@{[Foo->new->uniq(@ARGV)]}\n";` [download] ... `./11911 hi hi hi hey lo hey may ':-)' may lo :-)` [download]	[reply] [d/l] [select]
Re: compile errors by jonadab (Parson) on Feb 25, 2006 at 14:16 UTC
If I am not mistaken, that removes even the first instance of any entry that has a duplicate. Usually one wants to remove only subsequent entries, leaving the first. For that, cache the count on the fly: `sub uniq { my %cache; return grep { not $cache{$_}++ } @_ }` `print join " ", uniq qw(foo bar baz foo quux foo wibble bar quux wibble bongo foo dazzle);` Results: `foo bar baz quux wibble bongo dazzle`	[reply] [d/l] [select]