view release on metacpan or  search on metacpan

lib/Apache/FakeTable.pm  view on Meta::CPAN

package Apache::FakeTable;
use strict;
use vars qw($VERSION);
$VERSION = '0.06';

=head1 Name

Apache::FakeTable - Pure Perl implementation of the Apache::Table interface

=head1 Synopsis

  use Apache::FakeTable;

  my $table = Apache::FakeTable->new($r);

  $table->set(From => 'david@example.com');

  $table->add(Cookie => 'One Cookie');
  $table->add(Cookie => 'Another Cookie');

  while(my($key, $val) = each %$table) {
      print "$key: $val\n";
  }

=head1 Description

This class emulates the behavior of the L<Apache::Table> class, and is
designed to behave exactly like Apache::Table. This means that all keys are
case-insensitive and may have multiple values. As a drop-in substitute for
Apache::Table, you should be able to use it exactly like Apache::Table.

You can treat an Apache::FakeTable object much like any other hash. However,
like Apache Table, those keys that contain multiple values will trigger
slightly different behavior than a traditional hash. The variations in
behavior are as follows:

=over

=item keys

Will return the same key multiple times, once for each value stored for that
key.

=item values

Will return the first value multiple times, once for each value stored for a
given key. It'd be nice if it returned all the values for a given key, instead
of the first value C<*> the number of values, but that's not the way
Apache::Table works, and I'm not sure I'd know how to implement it even if it
did!

=item each

Will return the same key multiple times, pairing it with each of its values
in turn.

=back

Otherwise, things should be quite hash-like, particularly when a key has only
a single value.

=head1 Interface

=head3 new()

  my $table = Apache::FakeTable->new($r);
  $table = Apache::FakeTable->new($r, $initial_size);

Returns a new C<Apache::FakeTable> object. An L<Apache> object is required as
the first argument. An optional second argument sets the initial size of the
table for storing values.

=cut

sub new {
    # We actually ignore the optional initial size argument.
    my ($class, $r) = @_;
    unless (UNIVERSAL::isa($r, 'Apache')) {
        require Carp;