Question

I’m sure this has a very simple answer, but I can’t seem to find it (not sure what to search on!). A standard count / group by query may look like this:

SELECT COUNT(`t2`.`name`) 
FROM `table_1` `t1` 
    LEFT JOIN `table_2` `t2` ON `t1`.`key_id` = `t2`.`key_id` 
WHERE `t1`.`another_column` = 123 

and this works as expected, returning 0 if no rows are found. However:

SELECT COUNT(`t2`.`name`) 
FROM `table_1` `t1` 
    LEFT JOIN `table_2` `t2` ON `t1`.`key_id` = `t2`.`key_id` 
WHERE `t1`.`another_column` = 123 
GROUP BY `t1`.`any_col` 

only works if there is at least one row in table_1 and fails miserably returning an empty result set if there are zero rows. I would really like this to return 0! Anyone enlighten me on this? Beer can be provided in exchange if you are in London 😉

Answer 1

The reason it returns zero rows is that you are grouping on a value in table_1. SInce there are no values in table_1, there are no rows to return. Said another way, if you returned t1.any_col in your query from the GROUP BY like so:

SELECT `t1`.`any_col`, COUNT(`t2`.`name`) 
FROM `table_1` `t1` 
    LEFT JOIN `table_2` `t2` ON `t1`.`key_id` = `t2`.`key_id` 
WHERE `t1`.`another_column` = 123 
GROUP BY `t1`.`any_col` 

What would display for t1.any_col when there were no rows? The only way to achieve what you want is to union your results with another query that checks for no rows in table_1. In this example, I’m using the INFORMATION_SCHEMA view simply to have something against which I can query.

SELECT COUNT(`t2`.`name`) 
FROM `table_1` `t1` 
    LEFT JOIN `table_2` `t2` ON `t1`.`key_id` = `t2`.`key_id` 
WHERE `t1`.`another_column` = 123 
GROUP BY `t1`.`any_col` 
UNION ALL
SELECT 0
FROM INFORMATION_SCHEMA.TABLES
Where Not Exists( Select 1 From `table_1` )
LIMIT 1