I would like to write a Postgres function which takes various stats from other tables and combines the results into a single row.
e.g.
select count(*) as acount from tableA
select count(*) as bcount from tableB
select count(*) as ccount from tableC
returning
acount, bcount, ccount as a single record
Obviously the "count" functions listed above are a lot more complex than that, and may even be different amounts of counts for each table.
So ideally is it possible to store them as variables and then build a record from the variables ?