If your data is key driven then you can use a hash to find the unique keys and then grep them from the list.

my @wanted;

my %keys;
$keys{$_->{flag}}++ foreach (@array1, @array2);

my @unique = grep {$keys{$_->{flag}} == 1} @array2;
print Dumper(@unique);
[download]

This is exactly what I was looking for. How do I get a list of the keys that are common to both ?

I'm going to solve the problem by creating a rudimentary "digest" of each anonymous hash in each array. Hash %structures will hold the "digests" as keys to anonymous arrays that will hold where the "digest" came from, what array, and what position.

This will be ugly- the output will be mainly for you- it wouldn't be so useful for code, I suppose- but I really have no idea what you expect as output- it's too vague.

#!/usr/bin/perl -w
use strict;
use Data::Dumper;

my @array1 = 
(
  {
    hour => "1",
    minute => "30",
    flag => "cmd1"
  },

  {
    hour => "1",
    minute => "30",
    flag => "cmd2"
  },

  {
    hour => "1",
    minute => "30",
    flag => "cmd3"
  }
);

my @array2 = (
  {
    hour => "1",
    minute => "30",
    flag => "cmd1"
  },

  {
    hour => "1",
    minute => "30",
    flag => "cmd2"
  },

  {
    hour => "1",
    minute => "30",
    flag => "cmd3"
  },

  {
    iamdifferent =>'because',
    hour => "1",
    minute => "30",
    flag => "cmd4"
  },
  
  {
    iamdifferent =>'because',
    hour => "1",
    minute => "40",
    flag => "cmd6"
  },
  
  {
    iamdifferent =>'because',
    hour => "1",
    minute => "40",
    flag => "cmd6",
    minute => "40"
     
  },
  
  {
    a =>'because',
    b => "1",
    c => "40",
    d => "cmd6",
    minute => "40"
     
  }



);


# this will hold my hash "digests" as keys to lists of
# what array and position it happened in.
my %structures=();

# the arrays with anon hashes we will analize
my @arrays=(\@array1, \@array2);




my $which_array=0;

for (@arrays){        
    my $a=$_;

    my $anon_hash_position_in_array=0;

    for (@{$a}){
    
        my $structure;
    
        for my $k(sort (keys %{$_})){
            $structure.=":$k:";
        }
        my $val= "array$which_array"."_position$anon_hash_position_in_
+array";
    
        push @{$structures{$structure}}, $val;
        $anon_hash_position_in_array++;    
    }

    $which_array++;

}


print Dumper(\%structures);
exit;
[download]

Output:

$VAR1 = {
          ':flag::hour::minute:' => [
                                      'array0_position0',
                                      'array0_position1',
                                      'array0_position2',
                                      'array1_position0',
                                      'array1_position1',
                                      'array1_position2'
                                    ],
          ':flag::hour::iamdifferent::minute:' => [
                                                    'array1_position3'
+,
                                                    'array1_position4'
+,
                                                    'array1_position5'
                                                  ],
          ':a::b::c::d::minute:' => [
                                      'array1_position6'
                                    ]
        };
[download]