Group
Extension

Mojo-PgX-Cursor/lib/Mojo/PgX/Cursor/Results.pm

package Mojo::PgX::Cursor::Results;

require Mojo::IOLoop;
require Mojo::PgX::Cursor::Cursor;

use Scalar::Util qw(weaken);
use Time::HiRes qw(time);

use Mojo::Base -base;

has [qw(seconds_blocked)];

sub array {
  my $self = shift->_fetch;
  $self->{remaining}--;
  return $self->{results}->array;
}

sub columns { shift->_results->columns }

sub cursor {
  my $self = shift;
  if (@_) {
    $self->wait;
    $self->{cursor} = shift;
    $self->{remaining} = 0;
    delete $self->{delay};
    delete $self->{next};
    delete $self->{results};
    return $self->_load_next;
  }
  return $self->{cursor};
}

sub expand {
  my $self = shift;
  $self->{expand}++;
  if ($self->{results}) {
      $self->{results}->expand;
  }
  return $self;
}

sub hash {
  my $self = shift->_fetch;
  $self->{remaining}--;
  return $self->{results}->hash;
}

sub new {
  my $self = shift->SUPER::new(@_);
  return $self->_load_next;
}

sub rows {
  my $self = shift;
  return $self->_results->rows unless @_;
  $self->{rows} = shift;
  return $self;
}

sub wait {
  my $self = shift;
  if ($self->{delay}) {
    $self->{delay}->wait;
  }
  return $self;
}

sub _fetch {
  my $self = shift;
  return $self if $self->{remaining};
  unless ($self->{next}) {
    my $start = time;
    $self->wait;
    $self->{seconds_blocked} += time - $start;
    delete $self->{delay};
  }
  $self->{results} = delete $self->{next};
  $self->{remaining} = $self->{results}->rows;
  return $self->_load_next;
}

sub _load_next {
  my $self = shift;
  my $cursor = $self->cursor;
  weaken($cursor);
  my $results = $self;
  weaken($results);
  $self->{delay} = Mojo::IOLoop->delay(
    sub {
      if ($cursor && $results) {
        $cursor->fetch($results->{rows}, shift->begin);
      }
    },
    sub {
      if ($results) {
        $results->{next} = $_[2];
        $results->{next}->expand if ($results->{expand});
      }
    },
  );
  return $self;
}

sub _results {
  my $self = shift;
  return $self->{results} if $self->{results};
  return $self->_fetch->{results};
}

1;
__END__

=encoding utf-8

=head1 NAME

Mojo::PgX::Cursor::Results

=head1 DESCRIPTION

L<Mojo::PgX::Cursor::Results> is a container for a L<Mojo::PgX::Cursor::Cursor>
like L<Mojo::Pg::Results> is for a statement handle.  Therefore it tries to
mimic the API of L<Mojo::Pg::Results> whereever it makes sense to do do.

This container should behave like L<Mojo::Pg::Results> for C<array>,
C<columns>, C<hash>, and C<expand>.  It intentionally does not support
C<hashes> or C<arrays> since if you wish to use those you should just use
C<query> instead.  One difference in behavior is C<rows> returns the number of
rows in the current iteration not the total rows for the query.

=head1 ATTRIBUTES

=head2 cursor

    my $cursor = $results->cursor;
    $results = $results->cursor($cursor);

The L<Mojo::PgX::Cursor::Cursor> rows are fetched from.

=head2 seconds_blocked

    my $time_wasted = $results->seconds_blocked;

The cumulative time the cursor has spent waiting for rows.

=head1 METHODS

=head2 array

    my $row = $results->array;

Return next row from L</"cursor"> as an array reference.

=head2 columns

    my $columns = $results->columns;

Return column names as an array reference.

=head2 expand

    $results = $results->expand;

Decode C<json> and C<jsonb> fields automatically for all rows.

=head2 hash

    my $row = $results->hash;

Return next row from L</"cursor"> as a hash reference.

=head2 new

    my $results = Mojo::PgX::Cursor::Results->new(cursor => $cursor);

Construct a new L<Mojo::PgX::Cursor::Results> object.

=head2 rows

    my $num = $results->rows;

Number of rows in current iteration; not the total for the original query.

    $results->rows(1000);

Set the number of rows to fetch in each iteration.  Since the next iteration is
always pre-fetched up to twice this many rows will be in memory at any given
time.  Set this to optimize for time or memory in your specific use case.

If not set then the L<Mojo::PgX::Cursor::Cursor> default fetch size is used.

=head2 wait

  $results->wait;

Wait for background query to finish.  This would be used to block before making
another background query on the same database handle.

=head1 LICENSE

Copyright (C) Nathaniel Nutter.

This library is free software; you can redistribute it and/or modify
it under the same terms as Perl itself.

=head1 AUTHOR

Nathaniel Nutter C<nnutter@cpan.org>

=head1 SEE ALSO

L<Mojo::PgX::Cursor>, L<Mojo::PgX::Cursor::Cursor>, L<Mojo::PgX::Cursor::Database>

=cut



Powered by Groonga
Maintained by Kenichi Ishigaki <ishigaki@cpan.org>. If you find anything, submit it on GitHub.