This is probably more than you wanted to know, but it does answer your question directly (how to sort and unique in one step).
It's a suggestion MeowChow posted last summer in a discussion of sorts.  It uses a hash slice to store the "transform" with the array index and an array slice to return the elements sorted on the transform key.  It does not scale well(!).

  my %by;
  @by{ map $_->{num}, @$data } = 0 .. $#$data;

  @sorted_unique = @{$data}[ @by{ sort keys %by } ];
[download]

update:   Actually, I now see that the indexing is totally superfluous, so it would be:

  my %by = map{ $_->{num} => $_ } @$data;

  @sorted_unique = @by{sort keys %by};
[download]

which actually begins to look reasonable.  On the other hand, the sure enough do-it-all-in-one-line with an anonymous hash:

  @sorted_unique =
    @{{ map{ $_->{num} => $_ } @$d }}{ sort map $_->{num}, @$d };
[download]

goes through the array twice and also builds a hash.  And it even sorts the whole array before selecting unique elements which is the worst way to do it as khkramer demonstrates above.  It is kinda neat though.

  p