SQL语句之on子句过滤和where子句过滤区别

1、测试数据：

SQL> select * from dept;
 
DEPTNO DNAME          LOC
------ -------------- -------------
     10 ACCOUNTING     NEW YORK
     20 RESEARCH       DALLAS
     30 SALES          CHICAGO
     40 OPERATIONS     BOSTON
 
SQL> select * from emp;
 
EMPNO ENAME      JOB         MGR HIREDATE          SAL      COMM DEPTNO
----- ---------- --------- ----- ----------- --------- --------- ------
  7369 SMITH      CLERK      7902 1980/12/17     800.00               20
  7499 ALLEN      SALESMAN   7698 1981/2/20     1600.00    300.00     30
  7521 WARD       SALESMAN   7698 1981/2/22     1250.00    500.00     30
  7566 JONES      MANAGER    7839 1981/4/2      2975.00               20
  7654 MARTIN     SALESMAN   7698 1981/9/28     1250.00   1400.00     30
  7698 BLAKE      MANAGER    7839 1981/5/1      2850.00               30
  7782 CLARK      MANAGER    7839 1981/6/9      2450.00               10
  7788 SCOTT      ANALYST    7566 1987/4/19     3000.00               20
  7839 KING       PRESIDENT       1981/11/17    5000.00               10
  7844 TURNER     SALESMAN   7698 1981/9/8      1500.00      0.00     30
  7876 ADAMS      CLERK      7788 1987/5/23     1100.00               20
  7900 JAMES      CLERK      7698 1981/12/3      950.00               30
  7902 FORD       ANALYST    7566 1981/12/3     3000.00               20
  7934 MILLER     CLERK      7782 1982/1/23     1300.00               10
 
14 rows selected

2、左连接测试

SQL> select
e.empno, e.ename, e.deptno, d.deptno, d.dname
   3  from scott.emp e left join scott.dept d on e.deptno = d.deptno and e.deptno != 20 and d.deptno != 30;

     EMPNO ENAME                  DEPTNO     DEPTNO DNAME
---------- ------------------------------ ---------- ---------- ------------------------------------------
       7934 MILLER                  10         10 ACCOUNTING
       7839 KING                   10         10 ACCOUNTING
       7782 CLARK                  10         10 ACCOUNTING
       7900 JAMES                  30
       7844 TURNER                  30
       7698 BLAKE                  30
       7654 MARTIN                  30
       7521 WARD                   30
       7499 ALLEN                  30
       7902 FORD                   20
       7876 ADAMS                  20
       7788 SCOTT                  20
       7566 JONES                  20
       7369 SMITH                  20

14 rows selected.

Execution Plan
----------------------------------------------------------
Plan hash value: 3387915970

---------------------------------------------------------------------------
| Id  | Operation       | Name | Rows  | Bytes | Cost (%CPU)| Time      |
---------------------------------------------------------------------------
|   0 | SELECT STATEMENT   |      |    14 |   364 |    7  (15)| 00:00:01 |
|*  1 |  HASH JOIN OUTER   |      |    14 |   364 |    7  (15)| 00:00:01 |
|   2 |   TABLE ACCESS FULL| EMP  |    14 |   182 |    3   (0)| 00:00:01 |
|*  3 |   TABLE ACCESS FULL| DEPT |    3 |    39 |    3   (0)| 00:00:01 |
---------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------

   1 - access("E"."DEPTNO"="D"."DEPTNO"(+))
        filter("E"."DEPTNO"<>CASE  WHEN ("D"."DEPTNO"(+) IS NOT NULL)
           THEN 20 ELSE 20 END )
    3 - filter("D"."DEPTNO"(+)<>30)

Statistics
----------------------------------------------------------
       0  recursive calls
       0  db block gets
      13  consistent gets
       0  physical reads
       0  redo size
        1082  bytes sent via SQL*Net to client
     524  bytes received via SQL*Net from client
       2  SQL*Net roundtrips to/from client
       0  sorts (memory)
       0  sorts (disk)
      14  rows processed

    
-- 结论：left join 仅有on子句
-- 过滤条件对左表无效；

-- on子句过滤条件仅对右表生效；
-- 对于右表，过滤条件在连接之前生效（即先过滤，后连接）；
-- 对于左表，过滤条件在连接之后生效（即先连接，后过滤）。
     
SQL> select
e.empno, e.ename, e.deptno, d.deptno, d.dname
   3  from scott.emp e left join scott.dept d on e.deptno = d.deptno where e.deptno != 20 and d.deptno != 30;

     EMPNO ENAME                  DEPTNO     DEPTNO DNAME
---------- ------------------------------ ---------- ---------- ------------------------------------------
       7782 CLARK                  10         10 ACCOUNTING
       7934 MILLER                  10         10 ACCOUNTING
       7839 KING                   10         10 ACCOUNTING

Execution Plan
----------------------------------------------------------
Plan hash value: 844388907

----------------------------------------------------------------------------------------
| Id  | Operation             | Name    | Rows  | Bytes | Cost (%CPU)| Time     |
----------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT         |           |     9 |   234 |     6    (17)| 00:00:01 |
|   1 |  MERGE JOIN             |           |     9 |   234 |     6    (17)| 00:00:01 |
|   2 |   TABLE ACCESS BY INDEX ROWID| DEPT    |     3 |    39 |     2     (0)| 00:00:01 |
|*  3 |    INDEX FULL SCAN         | PK_DEPT |     3 |       |     1     (0)| 00:00:01 |
|*  4 |   SORT JOIN             |           |     9 |   117 |     4    (25)| 00:00:01 |
|*  5 |    TABLE ACCESS FULL         | EMP     |     9 |   117 |     3     (0)| 00:00:01 |
----------------------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------

   3 - filter("D"."DEPTNO"<>30 AND "D"."DEPTNO"<>20)
    4 - access("E"."DEPTNO"="D"."DEPTNO")
        filter("E"."DEPTNO"="D"."DEPTNO")
    5 - filter("E"."DEPTNO"<>20 AND "E"."DEPTNO"<>30)

Statistics
----------------------------------------------------------
       0  recursive calls
       0  db block gets
      10  consistent gets
       0  physical reads
       0  redo size
     916  bytes sent via SQL*Net to client
     524  bytes received via SQL*Net from client
       2  SQL*Net roundtrips to/from client
       1  sorts (memory)
       0  sorts (disk)
       3  rows processed

-- 结论：where子句过滤
-- 过滤条件对于左右表都有效；
-- 对于右表，过滤条件在连接之前生效（即先过滤，后连接）；
-- 对于左表，过滤条件在连接之前生效（即先过滤，后连接）；
-- 对于左连接，右表过滤字段出现在连接条件中，左连解会变成内连接。

3、以下SQL自行测试，可验证以上结论：

select
e.empno, e.ename, e.deptno, d.deptno, d.dname
from scott.dept d join scott.emp e on e.deptno = d.deptno where e.deptno != 20 and d.deptno != 30;
  
select
e.empno, e.ename, e.deptno, d.deptno, d.dname
from scott.dept d join scott.emp e on e.deptno = d.deptno where d.deptno != 30;
  
select
e.empno, e.ename, e.deptno, d.deptno, d.dname
from scott.dept d join scott.emp e on e.deptno = d.deptno where e.deptno != 20;

select
e.empno, e.ename, e.deptno, d.deptno, d.dname
from scott.dept d left join scott.emp e on e.deptno = d.deptno where e.deptno != 20 and d.deptno != 30;
  
select
e.empno, e.ename, e.deptno, d.deptno, d.dname
from scott.dept d left join scott.emp e on e.deptno = d.deptno where d.deptno != 30;
  
select
e.empno, e.ename, e.deptno, d.deptno, d.dname
from scott.dept d left join scott.emp e on e.deptno = d.deptno where e.deptno != 20;