Question

I have defined the following query:

select count((select * from producers)) from producers;

Assuming a producers table with 3 columns (A, B and C) and 2 rows:

A B C
-----
0 1 2
3 4 5

I’d expect the following output:

2
2

It doesn’t work. While the query itself is basically useless (even if it worked, it wouldn’t yield any useful output), I’d like to try to understand why this doesn’t run.

(select * from producers)

This would yield a list of rows with information on all the attributes on the producers table.

select count((select * from producers)) from producers;

This one will for each row on producers, show up the number 2 (the number of elements in producers).

Why doesn’t it work? SQL limitation? Is there anything wrong with the logic I’m following here?

Thanks

Answer 1

It is a limitation of SQL, as far as I know. Subqueries are not allowed in the COUNT expression. Obviously (select * from producers) is a subquery, so it’s not allowed there.

I think your misunderstanding is that you’re thinking that you would call the function like COUNT(SELECT * FROM producers) whereas in SQL it’s like SELECT COUNT(*) FROM producers.

Functions like MAX, MIN, SUM, and COUNT are aggregate functions, meaning that they take a scalar argument but execute once for each row, accumulating results every iteration. So SELECT MAX(column) FROM table executes the MAX function once for each row in table, while you might be thinking that MAX executes once and gets passed in every row in table.

Contrast this with operators like IN, EXISTS, ANY, and ALL, which have a subquery as an argument. They are effectively passed all the results of their subquery every time they are invoked.