Reputation: 858
I have the following relations in my database:
Invoice InvoiceMeal
--------------------- ---------------------------
| InvoiceId | Total | | Id | InvoiceId | MealId |
--------------------- ---------------------------
| 1 | 22.32 | | 1 | 1 | 3 |
--------------------- ---------------------------
| 2 | 12.18 | | 2 | 1 | 2 |
--------------------- ---------------------------
| 3 | 27.76 | | 3 | 2 | 2 |
--------------------- ---------------------------
Meal Type
----------------------------------- -------------------
| Id | Name | TypeId | | Id | Name |
----------------------------------- -------------------
| 1 | Hamburger | 1 | | 1 | Meat |
----------------------------------- -------------------
| 2 | Soja Beans | 2 | | 2 | Vegetarian |
----------------------------------- -------------------
| 3 | Chicken | 2 |
-----------------------------------
What I want to query from the database is InvoiceId
and Total
of all Invoices
which consist of at least two Meals
where at least one of the Meals
is of Type
Vegetarian. I have the following SQL query and it works:
SELECT
i."Id", i."Total"
FROM
public."Invoice" i
WHERE
(SELECT COUNT(*)
FROM public."InvoiceMeal" im
WHERE im."InvoiceId" = i."Id" AND
(SELECT COUNT(*)
FROM public."Meal" m, public."Type" t
WHERE im."MealId" = m."Id" AND
m."TypeId" = t."Id" AND
g."Name" = 'Vegetarian') > 0
) >= 2;
My problem with this query is that I can not easily modify the condition that there must at least one vegetarien Meal
. I want to be able, for example, to change it to at least two vegetarian meals. How can I achieve this with my query?
Upvotes: 1
Views: 79
Reputation: 1269543
I would approach this by joining the tables together and using aggregation. The having
clause can handle the conditions:
select i.Id, i.Total
from InvoiceMeal im join
Invoice i
on i.InvoiceId = im.InvoiceId join
Meal m
on im.mealid = m.mealid join
Type t
on m.typeid = t.typeid
group by i.Id, i.Total
having count(distinct im.mealid) >= 2 and
sum(case when t.name = 'Vegetarian' then 1 else 0 end) > 0;
I also see no reason to put double quotes around column names. That just makes the query harder to write and read.
Upvotes: 3