3

I would like to return a set of records from a pl/pgsql. Is there a way to do that without using the "for" construct and cursors? For instance when I compared these two stored procedures:

create or replace function sqlProc () returns setof integer as $$                                                                                                                                              
  select * from foo;
$$ language sql;


create or replace function plpgSqlProc () returns setof integer as $$
declare
  c integer;
begin
  for c in select * from foo loop
    return next c;
end loop;

end;
$$ language plpgsql;

The pure SQL version has 2x better time performance than the pl/pgsql one! Unfortunately I have other logic that cannot be expressed in pure SQL so I am wondering how I should write my stored procedure?

4

There are a few ways of doing this in PostgreSQL depending on the versions you have to support. The simplest is:

create or replace function plpgSqlProc () returns setof integer as $$
BEGIN

RETURN QUERY select * from foo;
END;
$$ LANGUAGE PLPGSQL;

This being said, I find where you don't need helper procedural code, plain SQL functions are usually better. They are usually fast and sometimes they are even planner-transparent.

| improve this answer | |
  • Thanks. But I see a lot of stored procedures that uses the for loop construct that only does return next inside the loop to iterate through query results. Is there any advantage to using that as opposed to using return query? – JRR Mar 3 '13 at 5:16
  • yes. Backward compatibility. RETURN QUERY is fairly new and early versions had significant corner case bugs. For new versions of PostgreSQL, those aren't issues. – Chris Travers Mar 3 '13 at 5:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.