Subquery in SELECT list vs join

Question

If I'm not mistaken, both queries below are functionally the same. Both return all rows from table1 and only show a value for table2.F7 if the first 4 characters of F7 equal F2.
My thoughts are this... It's nice that a subquery can be used in a select list, but it looks like, with all the sql I've written which uses that approach, I could have just as well used the join approach. Is this always true? Can someone show me as example where using a subquery in the select list is hands-down the better approach?
SELECT table1.F2,
(SELECT table2.F7 FROM table2 WHERE substr(table2.F7,1,4) = table1.F2)
FROM table1
SELECT table1.F2, table2.F7
FROM table1, table2
WHERE table1.F2 = substr(table2.F7,1,4)(+)
Let me add that, in my top example, multiple rows will never be returned by the subquery. If that were the case, then the join approach would be the only option.
Message was edited by:
Nick2012

Answer 1

Probably there is no general answer and you'll always have to check separately.
In the following example the scalar subquery turns out to be slightly more efficient than an outer join (15 gets versus 13 gets and compare the elapsed time):
SQL> SET autotrace traceonly explain statistics
SQL> SET timing on
SQL>
SQL> SELECT ename, (SELECT dname
  2                   FROM dept d
  3                  WHERE d.deptno = e.deptno) dname
  4    FROM emp e
  5  /
14 rows selected.
Elapsed: 00:00:00.72
Execution Plan
Plan hash value: 2981343222
| Id  | Operation                   | Name    | Rows  | Bytes | Cost (%CPU)| Time     |
|   0 | SELECT STATEMENT            |         |    15 |   135 |     3   (0)| 00:00:01 |
|   1 |  TABLE ACCESS BY INDEX ROWID| DEPT    |     1 |    11 |     1   (0)| 00:00:01 |
|*  2 |   INDEX UNIQUE SCAN         | PK_DEPT |     1 |       |     0   (0)| 00:00:01 |
|   3 |  TABLE ACCESS FULL          | EMP     |    15 |   135 |     3   (0)| 00:00:01 |
Predicate Information (identified by operation id):
   2 - access("D"."DEPTNO"=:B1)
Statistics
          0  recursive calls
          0  db block gets
         13  consistent gets
          0  physical reads
          0  redo size
        742  bytes sent via SQL*Net to client
        381  bytes received via SQL*Net from client
          2  SQL*Net roundtrips to/from client
          0  sorts (memory)
          0  sorts (disk)
         14  rows processed
SQL> SELECT ename, dname
  2    FROM emp e, dept d
  3   WHERE d.deptno(+) = e.deptno
  4  /
14 rows selected.
Elapsed: 00:00:00.74
Execution Plan
Plan hash value: 3387915970
| Id  | Operation          | Name | Rows  | Bytes | Cost (%CPU)| Time     |
|   0 | SELECT STATEMENT   |      |    15 |   300 |     7  (15)| 00:00:01 |
|*  1 |  HASH JOIN OUTER   |      |    15 |   300 |     7  (15)| 00:00:01 |
|   2 |   TABLE ACCESS FULL| EMP  |    15 |   135 |     3   (0)| 00:00:01 |
|   3 |   TABLE ACCESS FULL| DEPT |     5 |    55 |     3   (0)| 00:00:01 |
Predicate Information (identified by operation id):
   1 - access("D"."DEPTNO"(+)="E"."DEPTNO")
Statistics
          0  recursive calls
          0  db block gets
         15  consistent gets
          0  physical reads
          0  redo size
        666  bytes sent via SQL*Net to client
        381  bytes received via SQL*Net from client
          2  SQL*Net roundtrips to/from client
          0  sorts (memory)
          0  sorts (disk)
         14  rows processed