package DBIx::Class::Schema::PopulateMore::Visitor;
use Moo;
use Scalar::Util qw/refaddr/;
use Type::Library -base;
use Types::Standard -types;
use namespace::clean;
=head1 NAME
DBIx::Class::Schema::PopulateMore::Visitor - Visitor for the Populate Data
=head1 SYNOPSIS
##Example Usage
See Tests for more example usage.
=head1 DESCRIPTION
When populating a table, sometimes we need to inflate values that we won't
know of in advance. For example we might have a column that is FK to another
column in another table. We want to make it easy to 'tag' a value as something
other than a real value to be inserted into the table.
Right now we only have one substitution to do, which is the FK one mentioned
above, but we might eventually create other substitution types so we've broken
this out to make it neat and easy to do so.
=head1 ATTRIBUTES
This class defines the following attributes.
=head2 update_callback
The coderef to be execute should the match condition succeed
=cut
has 'update_callback' => (
is=>'rw',
required=>1,
lazy=>1,
isa=>CodeRef,
default=> sub {
return sub {
return shift;
};
},
);
=head2 match_condition
How we know the value is really something to inflate or perform a substitution
on. This get's the namespace of the substitution plugin and it's other data.
The default behavior (where there is no substitution namespace, is to do the
inflate to resultset. This is the most common usecase.
=cut
has 'match_condition' => (
is=>'ro',
required=>1,
isa=>RegexpRef,
);
=head2 seen
Used to collect up ref addresses of arrays/hashes we have already seen
=cut
has seen => (
is => 'rw',
isa => HashRef,
default => sub { {} },
);
=head1 METHODS
This module defines the following methods.
=head2 callback
Given a coderef, sets the current callback and returns self so that we can chain
=cut
sub callback
{
my $self = shift @_;
$self->update_callback(shift @_);
return $self;
}
=head2 visit
A simple visitor that only expects to perform replacements on values
=cut
sub visit
{
my ( $self, $target ) = @_;
if ( ref $target eq 'ARRAY' ) {
my $addr = refaddr $target;
return $self->seen->{$addr} if defined $self->seen->{$addr};
my $new_array = $self->seen->{$addr} = [];
@$new_array = map { $self->visit($_) } @$target;
return $new_array;
}
elsif ( ref $target eq 'HASH' ) {
my $addr = refaddr $target;
return $self->seen->{$addr} if defined $self->seen->{$addr};
my $new_hash = $self->seen->{$addr} = {};
%$new_hash = map { $_ => $self->visit( $target->{$_} ) } keys %$target;
return $new_hash;
}
else {
$self->visit_value($target);
}
}
=head2 visit_value
Here is where we make the choice as to if this value needs to be inflated via a plugin
=cut
sub visit_value
{
my ($self, $data) = @_;
if(my $item = $self->match_or_not($data))
{
return $self->update_callback->($item);
}
return $data;
}
=head2 match_or_not
We break this out to handle the ugliness surrounding dealing with undef values
and also to make it easier on subclassers.
=cut
sub match_or_not
{
my ($self, $data) = @_;
my $match_condition = $self->match_condition;
if( !defined $data )
{
return;
}
elsif(my ($item) = ($data=~m/$match_condition/))
{
return $item;
}
return;
}
=head1 AUTHOR
Please see L<DBIx::Class::Schema::PopulateMore> For authorship information
visit method culled from code in L<Data::Visitor::Lite> which is copyright 2011 Daichi Hiroki <hirokidaichi {at} gmail.com>
=head1 LICENSE
Please see L<DBIx::Class::Schema::PopulateMore> For licensing terms.
=cut
1;