[Dbix-class] Automatically counting rows of another table - and how to disable temporarily

[Bernhard Graf]

Hello,

following the examples in the DBIx::Class distribution I have 2 tables:

package My::Schema::Artist;
use base qw/DBIx::Class/;
__PACKAGE__->load_components(qw/PK::Auto Core/);
__PACKAGE__->table('artist');
__PACKAGE__->add_columns(qw/id name albums/);
__PACKAGE__->set_primary_key('id');
__PACKAGE__->has_many(albums => 'My::Schema::Album', 'artist');


package My::Schema::Album;
use base qw/DBIx::Class/;
__PACKAGE__->load_components(qw/PK::Auto Core/);
__PACKAGE__->table('album');
__PACKAGE__->add_columns(qw/id artist title/);
__PACKAGE__->set_primary_key('id');
__PACKAGE__->belongs_to(artist => 'My::Schema::Artist');

sub insert {
    my $self = shift;
    $self->next::method(@_);
    $self->artist->update({'albums' => \'albums+1'});
    $self;
}

sub delete {
    my $self = shift;
    $self->next::method(@_);
    $self->artist->update({'albums' => \'albums-1'});
    $self;
}


This means, that "artist.albums" keeps track of the number of rows in 
table "album".

So far, so good.

For cases where I want to create many albums in a row, it would be 
desirable to disable this automatism for effeciency reasons and adjust 
"artist.albums" manually afterwards.

I suppose I would need an accessor in the result set of 
My::Schema::Album that reflects this state, like:

  $artist = $schema->resultset('My::Schema::Artist')->find($artistid);
  $rs = $schema->resultset('My::Schema::Album');
  $rs->disable_count(1);
  for ($i = 0; $name = get_albumname(); ++$i) {
    $rs->create({artist => $artist, name => $name});
  }
  $artist->update(albums => \"albums+$i");
  $rs->disable_count(0);

Could someone give advice if this concept is correct and how to define 
such a ResultSet accessor?
-- 
Bernhard Graf