—

              
#
# This file is part of ElasticSearchX-Model
#
# This software is Copyright (c) 2019 by Moritz Onken.
#
# This is free software, licensed under:
#
#   The (three-clause) BSD License
#
package ElasticSearchX::Model::Document::Trait::Attribute;
$ElasticSearchX::Model::Document::Trait::Attribute::VERSION = '2.0.1';
# ABSTRACT: Trait that extends the meta class of a document class
use Moose::Role;
use ElasticSearchX::Model::Document::Mapping;
with 'MooseX::Attribute::LazyInflator::Meta::Role::Attribute';
use ElasticSearchX::Model::Document::Types qw(:all);
use MooseX::Types::Moose qw(ArrayRef);
has index   => ( is => 'ro' );
has boost   => ( is => 'ro', isa => 'Num' );
has store   => ( is => 'ro', isa => 'Str', default => 'yes' );
has type    => ( is => 'ro', isa => 'Str', default => 'string' );
has parent  => ( is => 'ro', isa => 'Bool', default => 0 );
has dynamic => ( is => 'ro', isa => 'Bool', default => 0 );
has analyzer =>
    ( is => 'ro', isa => ArrayRef, coerce => 1, default => sub { [] } );
has not_analyzed   => ( is => 'ro', isa => 'Bool', default => 1 );
has term_vector    => ( is => 'ro', isa => 'Str' );
has include_in_all => ( is => 'ro', isa => 'Bool', default => 1 );
has source_only    => ( is => 'ro', isa => 'Bool', default => 0 );
has include_in_root   => ( is => 'ro', isa => 'Bool' );
has include_in_parent => ( is => 'ro', isa => 'Bool' );
has property          => ( is => 'ro', isa => 'Bool', default => 1 );
has query_property    => ( is => 'ro', isa => 'Bool', default => 0 );
has field_name =>
    ( is => 'ro', isa => 'Str', lazy => 1, default => sub { shift->name } );
has isa_arrayref => (
    is      => 'ro',
    lazy    => 1,
    isa     => 'Bool',
    builder => '_build_isa_arrayref'
);
sub _build_isa_arrayref {
    my $self = shift;
    my $tc   = $self->type_constraint;
    return 0 unless $tc;
    return $tc->is_a_type_of("ArrayRef");
}
sub build_property {
    my $self = shift;
    return {
        ElasticSearchX::Model::Document::Mapping::maptc(
            $self, $self->type_constraint
        )
    };
}
before _process_options => sub {
    my ( $self, $name, $options ) = @_;
    %$options = ( builder => 'build_id', lazy => 1, %$options )
        if ( $options->{id} && ref $options->{id} eq 'ARRAY' );
    $options->{traits} ||= [];
    push(
        @{ $options->{traits} },
        'MooseX::Attribute::LazyInflator::Meta::Role::Attribute'
    ) if ( $options->{property} || !exists $options->{property} );
};
after _process_options => sub {
    my ( $class, $name, $options ) = @_;
    if (   $options->{required}
        && !$options->{builder}
        && !defined $options->{default} )
    {
        $options->{lazy}     = 1;
        $options->{required} = 1;
        $options->{default}  = sub {
            confess "Attribute $name is required";
        };
    }
};
sub mapping {
    my $self = shift;
    return ( $self->name => $self->build_property )
        unless ( $self->source_only || $self->parent );
    return ();
}
sub type_mapping { () }
after install_accessors => sub {
    my $self = shift;
    return
        unless (
        $self->associated_class->does_role(
            'ElasticSearchX::Model::Document::Role')
        );
    $self->associated_class->_add_field_alias(
        $self->name => $self->field_name );
    $self->associated_class->_add_reverse_field_alias(
        $self->field_name => $self->name );
};
1;
__END__
HideShow 150 lines of Pod
=pod
=encoding UTF-8
=head1 NAME
ElasticSearchX::Model::Document::Trait::Attribute - Trait that extends the meta class of a document class
=head1 VERSION
version 2.0.1
=head1 ATTRIBUTES
=head2 property
This defaults to C<1> and marks the attribute as ElasticSearch
property and thus will be added to the mapping. If you set this
to C<0> the attribute will act as a traditional Moose attribute.
=head2 id
Usually there is one property which also acts as id for the whole
document. If there is no attribute with the C<id> option defined
ElasticSearch will assign a random id. This option can either
be set to a true value or an arrayref. The former will make the
value of the attribute the id. The latter will generate a SHA1
digest of the concatenated values of the attributes listed in 
the arrayref.
Only one attribute with the C<id> option set can be present in 
a document.
=head2 type
Most of the time L<ElasticSearchX::Model::Document::Mapping> will take 
care of this option and set the correct value based on the
type constriant. In case it doesn't know what to do, this 
value will be used as the type for the attribute. Defaults
to C<string>.
=head2 parent
The value of this property will be used as C<parent> id.
Since the parent id is stored in the C<_parent> field, it
is adviced to set L</source_only> to C<1> to prevent the
field from being stored redundantly.
=head2 source_only
A C<source_only> attribute is not added to the type mapping,
but its value is included in the C<_source> of a document.
This is helpful if you don't want to index the value of this
attribute in Elasticsearch, but still want to be able to access
its value.
=head2 timestamp
 has timestamp => ( timestamp => 1, is => 'ro' );
 has timestamp => ( timestamp => { store => 1 }, is => 'ro' );
The attribute using this option will become the timestamp field
(L<http://www.elasticsearch.org/guide/reference/mapping/timestamp-field.html>).
=head2 ttl
 has expire => ( ttl => 1, is => 'ro' );
 $document->expire(86400) # document will expire in one day
Documents with a ttl field will set the _ttl property on the document in
ElasticSearch. This will cause ElasticSearch to remove the document
from the index automatically once the time to live has expired.
See L<http://www.elasticsearch.org/guide/reference/mapping/ttl-field.html>.
=head1 PASS THROUGH ATTRIBUTES
The following attributes are passed through - as is - to the
type mapping.
=head2 store
Defaults to C<yes>.
=head2 boost
=head2 index
=head2 dynamic
By default, properties are not dynamic. That means that fields that
are unknown to the ElasticSearch mapping will not be indexed and
no dynamic mapping will be auto generated. Since the mapping is
generated by this module, there is no need to have the dynamic
mapping enabled. If you want to enable it anyway, set this attribute
to C<1> value.
This will change the behaviour of C<HashRef> attributes.
Instead of deflating to a JSON string they will be stored as object
in ElasticSearch.
=head2 analyzer
If this attribute or L</index> is set to C<analyzed>, the field is stored
both analyzed and not analyzed. The not analyzed field can be accessed in
queries by the field name. The analyzed field will have a C<.analyzed>
postfix if there is only one analyzer. If there are more than one analyzer,
the C<.analyzed> postfix refers to the first analyzer. For all other
analyzers, its name is prefixed. This is done by using the C<multi_field>
feature of ElasticSearch.
=head2 include_in_all
=head2 include_in_parent
=head2 include_in_root
=head2 term_vector
=head2 not_analyzed
By default, a field is not analyzed. That means the raw value is stored
in the field. To add analyzed fields, either set L</index> to C<analyzed>
or add an L</analyzer>. Set this to C<0> if you don't want to store
the not analyzed version of this field along with the analyzed.
=head1 METHODS
=head2 build_id
=head2 build_property
This method is called by L<ElasticSearchX::Model::Document::Trait::Class/mapping>
and returns the mapping for this property as a HashRef.
=head1 AUTHOR
Moritz Onken
=head1 COPYRIGHT AND LICENSE
This software is Copyright (c) 2019 by Moritz Onken.
This is free software, licensed under:
  The (three-clause) BSD License
=cut