From Code to Community: Sponsoring The Perl and Raku Conference 2025 Learn more

use Moo;
has invalid_msg => (is=>'ro', required=>1, default=>sub {_t 'invalid'});
has validations => (is=>'ro', required=>1, default=>sub {0});
sub normalize_shortcut {
my ($class, $arg) = @_;
if(($arg eq '1') || ($arg eq 'nested')) {
return { validations => 1 };
}
}
sub validate_each {
my ($self, $record, $attribute, $value, $opts) = @_;
# If a row is marked to be deleted then don't bother to validate it
my @rows = grep { not $_->is_removed } @{$value->get_cache||[]};
# Ok, now run validations. If we find even one row is invalid, then we need
# to mark the attribute as invalid.
my $found_errors = 0;
foreach my $row (@rows) {
$row->validate(%$opts); #unless $row->validated
$found_errors = 1 if $row->errors->size;
}
# Ok, next if we are asking to aggregate the nested errors, do that
if($self->validations) {
my $rowidx = 0;
foreach my $row (@rows) {
foreach my $importable_error ($row->errors->errors->all) {
my $local_attribute = "${attribute}[${rowidx}].@{[ $importable_error->attribute||'*' ]}";
$record->errors->import_error($importable_error, +{attribute=>$local_attribute});
}
$rowidx++;
}
}
my $count = scalar(@rows) > 1 ? scalar(@rows) : 2;
$opts->{count} = $count; # Even if there are zero or one row since we use the plural form
# of the class name we need to set this to 2 so that the error message
# is correct grammatically.
$record->errors->add($attribute, $self->invalid_msg, $opts) if $found_errors;
}
1;
=head1 NAME
DBIx::Class::Valiant::Validator::ResultSet - Verify a DBIC related resultset
=head1 SYNOPSIS
package Example::Schema::Result::Person;
use base 'Example::Schema::Result';
__PACKAGE__->load_components(qw/
Valiant::Result
Core
/);
__PACKAGE__->table("person");
__PACKAGE__->add_columns(
id => { data_type => 'bigint', is_nullable => 0, is_auto_increment => 1 },
username => { data_type => 'varchar', is_nullable => 0, size => 48 },
first_name => { data_type => 'varchar', is_nullable => 0, size => 24 },
last_name => { data_type => 'varchar', is_nullable => 0, size => 48 },
password => {
data_type => 'varchar',
is_nullable => 0,
size => 64,
},
);
__PACKAGE__->has_many(
credit_cards =>
'Example::Schema::Result::CreditCard',
{ 'foreign.person_id' => 'self.id' }
);
__PACKAGE__->validates(
credit_cards => (
result_set=>+{ validations=>1, skip_if_empty=>1, min=>2, max=>4 },
)
);
=head1 DESCRIPTION
Validations on related resultsets. Used to apply constraints on the resultset as a whole
(such as total number of rows) or to trigger running validations on any related row objects.
Any errors from related resultsets will be added as sub errors on the parent result.
B<NOTE>: This gets added automatically for you if you setup C<accepts_nested> on the parent
object. So you shouldn't really ever need to use this code directly.
=head1 ATTRIBUTES
This validator supports the following attributes:
=head2 validations
Boolean. Default is 0 ('false'). Used to trigger validations on row objects found inside the
resultset. Please keep in mind this can be expensive if you have a lot of found rows (consider
using limits and validating in chunks).
Please keep in mind these errors will be localized to the associated object, not on the current
object.
=head2 invalid_msg
Error message returned on the current object if we find any errors inside related objects.
defaults to tag 'invalid_msg'.
=head1 SHORTCUT FORM
This validator supports the follow shortcut forms:
validates attribute => ( result_set => 1, ... );
Which is the same as:
validates attribute => (
result_set => {
validations => 1,
}
);
Which is a shortcut when you wish to run validations on the related rows
=head1 GLOBAL PARAMETERS
This validator supports all the standard shared parameters: C<if>, C<unless>,
C<message>, C<strict>, C<allow_undef>, C<allow_blank>.
=head1 SEE ALSO
L<Valiant>, L<Valiant::Validator>, L<Valiant::Validator::Each>.
=head1 AUTHOR
See L<Valiant>
=head1 COPYRIGHT & LICENSE
See L<Valiant>
=cut