Joe McMahon
and 1 contributors

NAME

Tie::Hash::MultiValue - store multiple values per key

SYNOPSIS

  use Tie::Hash::MultiValue;
  my $controller = tie %hash, 'Tie::Hash::MultiValue';
  $hash{'foo'} = 'one';
  $hash{'bar'} = 'two';
  $hash{'bar'} = 'three';

  # Fetch the values as references to arrays.
  $controller->refs;
  my @values  = @{$hash{'foo'}};   # @values = ('one');
  my @more    = @{$hash{'bar'}};   # @more   = ('two', 'three');
  my @nothing = @{$hash{'baz'}};   # empty list if nothing there

  # You can tie an anonymous hash as well.
  my $hashref = {};
  tie %$hashref, 'Tie::Hash::MultiValue';
  $hashref->{'sample'} = 'one';
  $hashref->{'sample'} = 'two';
  # $hashref->{'sample'} now contains ['one','two']

  # Iterate over the items stored under a key.
  $controller->iterators;
  while(my $value = $hash{bar}) {
    print "bar: $value\n";
  }
  # prints
  #   bar: two
  #   bar: three

DESCRIPTION

Tie::Hash::MultiValue allows you to have hashes which store their values in anonymous arrays, appending any new value to the already-existing ones.

This means that you can store as many items as you like under a single key, and access them all at once by accessing the value stored under the key.

USAGE

See the synopsis for a typical usage.

BUGS

None currently known.

SUPPORT

Contact the author for support.

AUTHOR

        Joe McMahon
        CPAN ID: MCMAHON
        mcmahon@ibiblio.org
        http://ibiblio.org/mcmahon

COPYRIGHT

This program is free software; you can redistribute it and/or modify it under the same terms as Perl itself.

The full text of the license can be found in the LICENSE file included with this module.

SEE ALSO

Tie::Hash, perl(1), Perl Cookbook (1st version) recipe 13.15, program 13-5.

METHODS

This class is a subclass of Tie::ExtraHash; it needs to override the TIEHASH method to save the instance data (in $self->[1]), and the STORE method to actually save the values in an anonymous array.

TIEHASH

If the 'unique' argument is supplied, we check to see if it supplies a subroutine reference to be used to compare items. If it does, we store that reference in the object describing this tie; if not, we supply a function which simply uses 'eq' to test for equality.

The 'unique' function

This funtion will receive two scalar arguments. No assumption is made about whether or not either argument is defined, nor whether these are simple scalars or references. You can make any of these assumptions if you choose, but you are responsible for checking your input.

You can perform whatever tests you like in your routine; you should return a true value if the arguments are determined to be equal, and a false one if they are not.

STORE

Push the value(s) supplied onto the list of values stored here. The anonymous array is created automatically if it doesn't yet exist.

If the 'unique' argument was supplied at the time the hash was tied, we will use the associated function (either yours, if you supplied one; or ours, if you didn't) and only add the item or items that are not present.

FETCH

Fetches the current value(s) for a key, depending on the current mode we're in.

  • 'refs' mode

    Always returns an anonymous array containing the values stored under this key, or an empty anonymous array if there are none.

  • 'iterators' mode

    If there is a single entry, acts just like a normal hash fetch. If there are multiple entries for a key, we automatically iterate over the items stored under the key, returning undef when the last item under that key has been fetched.

    Storing more elements into a key while you're iterating over it will result in the new elements being returned at the end of the list. If you've turned on 'unique', remember that they won't be stored if they're already in the value list for the key.

      NOTE: If you store undef in your hash, and then store other values, the iterator will, when it sees your undef, return it as a normal value. This means that you won't be able to tell whether that's your undef, or the 'I have no more data here' undef. Using 'list' or 'refs' mode is strongly suggested if you need to store data that may include undefs.

    Note that every key has its own iterator, so you can mix accesses across keys and still get all the values:

      my $controller = tie %hash, 'Tie::Hash::MultiValue';
      $controller->iterators;
      $hash{x} = $_ for qw(a b c);
      $hash{y} = $_ for qw(d e f);
      while ( my($x, $y) = ($hash{x}, $hash{y}) {
         # gets (a,d) (b,e) (c,f)
      }

iterators

Called on the object returned from tie(). Tells FETCH to return elements one at a time each time the key is accessed until no more element remain.

refs

Tells FETCH to always return the reference associated with a key. (This allows you to, for instance, replace all of the values at once with different ones.)

mode

Tells you what mode you're currently in. Does not let you change it!