how to use bulk collect instead of a loop in Oracle

Question

I have a sample query like below in my procedure:

result_rec              mypkg.mytype;

OPEN CUR1 FOR
  select col1, col2, col3 from table1 where something = 'a'; --rows will always be 50

      LOOP
         FETCH CUR1
          INTO myrectype;
         EXIT WHEN CUR1%NOTFOUND;
         result_rec.col1 := myrectype.col1;
         result_rec.col2 := myrectype.col2;
         result_rec.col3 := myrectype.col3;
         PIPE ROW (result_rec);
      END LOOP;

As you can see, every time I am looping 50 times. Is there a better way to do this? something like BULK COLLECT INTO? how would I implement that?

Answer 1

You can try the following.

DECLARE
  type tab_result_rec IS TABLE OF mypkg.mytype INDEX BY PLS_INTEGER;
  t_result_rec tab_result_rec;
BEGIN
  select col1, col2, col3 bulk collect into t_result_rec
  from table1 where something = 'a'; --rows will always be 50
  --
  for i in 1..t_result_rec.count LOOP
     PIPE ROW (t_result_rec(i));
  end loop;
END;

Answer 2

In Oracle 10g (possibly 9i), Oracle will automatically bulk collect implicit cursors. So code like:

DECLARE
  result_rec              mypkg.mytype;
BEGIN
  for i in (select col1, co2, col3 from table1 where something = 'a')
  loop
    result_rec.col1 := i.col1;
    result_rec.col2 := i.col2;
    result_rec.col3 := i.col3;
    pipe_row (result_rec);
  end loop;
END;

Will only make the context switch from the PL/SQL engine to the SQL engine for fetching records once every 100 rows. Run it under a SQL trace (dbms_monitor.session_trace_enable()) and see!