How can I join two hashes in Perl without using a loop?

Question

How do I append hash a to hash b in Perl without using a loop?

Why don't you want to use a loop? – brian d foy Aug 08 '09 at 01:12 — brian d foy, Aug 08 '09 at 01:12

score 32 · Accepted Answer · answered Aug 06 '09 at 23:26

32

If you mean take the union of their data, just do:

%c = (%a, %b);

answered Aug 06 '09 at 23:26

bdonlan

224,562
31
268
324

3

It works for me: "%a=('a' => 'b', 'c' => 1); %b=('c' => 'd'); %a = (%a, %b); print join(', ', map { "'$_' => \"${a{$_}}\"" } keys %a), "\n";" Items from %b with the same key take precedence, but other items from %a are present after the union. – outis Aug 06 '09 at 23:51
that might be the OP's problem, they might want to merge the values of the hash, so $a{'c'} ends up being [1, 'd'] – MkV Aug 07 '09 at 10:51
5

In which case they should look at Hash::Merge on the CPAN – MkV Aug 07 '09 at 10:53

score 24 · Answer 2 · answered Aug 06 '09 at 23:42

24

You can also use slices to merge one hash into another:

@a{keys %b} = values %b;

Note that items in %b will overwrite items in %a that have the same key.

answered Aug 06 '09 at 23:42

outis

75,655
22
151
221

5

For extra credit, with a hash ref in a and b, we do @{$a}{keys %$b} = values %$b; – daotoad Aug 07 '09 at 05:30
2

I expect this to be more efficient than the other reply, because that one recreates the hash also for the already existing items, while this one just adds keys for the new items. If there were many items in %a, compared to %b, the difference can be respectable. – bart Aug 10 '09 at 20:04
1

Maybe. My approach also walks %b twice. If %b is larger than %a, the other technique might be faster. Fastest of all might be a loop, but the OP didn't want that. – outis Aug 10 '09 at 20:49
1

@bart: looks like you were right. Some rather simple timing tests for perl 5.8.9 on OS X 10.4 place the slice approach about 2.4 times faster, even with %b much larger than %a. Your mileage may vary. – outis Aug 10 '09 at 21:01
1

This should also work for blessed hashrefs if I'm not severely mistaken. Trying it out now. Handy! – onitake Sep 25 '13 at 18:25

botkop · Answer 3 · 2012-05-02T10:33:26.433

2

my %c = %a;
map {$c{$_} = $b{$_}} keys %b;

edited May 02 '12 at 10:33

answered May 02 '12 at 09:22

botkop

934
1
8
17

score 2 · Answer 4 · answered Nov 18 '11 at 11:24

This will merge hashes and also take into account undefined entries, so they don't replace the content.

my %hash = merge(\%hash1, \%hash2, \%hash3);

sub merge {
    my %result;

    %result = %{ $_[0] };
    shift;

    foreach my $ref (@_) {
        for my $key ( keys %{$ref} ) {
            if ( defined $ref->{$key} ) {
                $result{$key} = $ref->{$key};
            }
        }
    }

    return %result;
}

How can I join two hashes in Perl without using a loop?

4 Answers4

Linked