Question

I am trying to get data from three different tables at once, and I don’t think I quite understand joins correctly, as I’m getting nowhere fast.
For example, lets say the tables are houses, sellers, and selling_details. houses and sellers are linked by the selling_details: it has a seller_id and a house_id, plus more information such as price, and a link to a user.

I want to build a query that returns all houses in the system, matched up to all sellers, and list the selling details if present. For example:

+------------+-------------+-------------------------+----------------------+-----------+
| house.name | seller.name | selling_details.details | selling_details.date | user.name |
+------------+-------------+-------------------------+----------------------+-----------+
| One        | One         | details                 | 2011-02-18           | bobby     |
| One        | Two         | details                 | 2011-02-24           | frank     |
| One        | Three       | NULL                    | NULL                 | NULL      |
| One        | Four        | NULL                    | NULL                 | NULL      |
| Two        | One         | details                 | 2011-01-16           | ned       |
| Two        | Two         | NULL                    | NULL                 | NULL      |
| Two        | Three       | details                 | 2011-02-12           | will      |
| Two        | Four        | NULL                    | NULL                 | NULL      |
+------------+-------------+-------------------------+----------------------+-----------+

What would be the easiest way to go about this?

Edit: It seems I’m trying to oversimplify the problem so here’s some more detail:

Here’s a fraction of the schema I’m using:

create table `house` (`id` int not null auto_increment, `name` varchar(255) null, primary key (`id`))
create table `seller` (`id` int not null auto_increment, `name` varchar(255) null, primary key (`id`))
create table `user` (`id` int not null auto_increment, `name` varchar(255) null, primary key (`id`))
create table `selling_details` (`id` int not null auto_increment, `details` varchar(255) not null, date datetime not null, `house_id` int null, `seller_id` int null, `user_id` int not null, primary key (`id`))

alter table `selling_details` add index `FK_selling_details_user` (`user_id`), add constraint `FK_selling_details_user` foreign key (`user_id`) references `user` (`id`)
alter table `selling_details` add index `FK_selling_details_house` (`house_id`), add constraint `FK_selling_details_house` foreign key (`house_id`) references `house` (`id`)
alter table `selling_details` add index `FK_selling_details_seller` (`seller_id`), add constraint `FK_selling_details_seller` foreign key (`seller_id`) references `seller` (`id`)

Now, just to make things really complicated too, there may be many rows in the selling_details table linking house and seller. If there is one or more of these rows present, I only want the one with the most recent date; if there’s no such row, still return the house and seller combination as in the example results above.

Answer 1

Cross join the house and sellers, then left join to find any details for the combination

select h.house_id, s.seller_id, sd.details
from houses h
cross join sellers s
left join selling_details sd
    on sd.seller_id = s.seller_id
    and sd.house_id = h.house_id
order by h.house_id, s.seller_id