[Dbix-class] Using joins and aggregate functions with DBIC

[Nilson Santos Figueiredo Junior]

On 5/26/06, Matt S Trout <dbix-class at trout.me.uk> wrote:
> > sub search {
> >     my ($self, $cond, $attrs) = @_;
>        my $alias = $self->{attrs}{alias}
> >     $self->next::method($cond, $attrs)->next::method(
> >         {},
> >         {
> >             join     => [qw/logs/],
>                select   => [(map {"${alias}.${_}"}
> > $self->result_source->columns), {sum => 'logs.hours_spent'} ],
> >             as       => [$self->result_source->columns, qw/hours_spent/],
>                group_by => [ "${alias}.id" ]
> >         }
> >     );
> > }

This sort of worked. However, it didn't perfectly DWIM.
When I use the relationship accessors ($user->tasks), I expected it to
sum only the specific user task logs. Instead, it summed everything
up.

There's another problem, take a look at the query produced:

SELECT
  task.id,
  task.project_id,
  task.date,
  task.description,
  task.is_closed,
  task.comments,
SUM( logs.hours_spent )
FROM task_user me
JOIN task task
ON ( task.id = me.task_id )
LEFT JOIN task_log logs ON ( logs.task_id = task.id )
LEFT JOIN task_log logs_2 ON ( logs_2.task_id = task.id )
WHERE ( ( ( ( ( me.user_id = ? ) ) ) ) ) GROUP BY task.id;

Why there are two references to task_log when only of them is used? I
called $user->tasks without any additional parameters so I can't
imagine where this came from.

When I search() without using the relationship accessor magic (i.e.
$model->search), this duplication does not occur. Might this be a bug?

-Nilson Santos F. Jr.