图解 SQL 里的各种 JOIN

给定数据：

下文将使用两个数据库表 Table_A 和 Table_B 来进行示例讲解，其结构与数据分别如下：

mysql> SELECT * FROM Table_A ORDER BY PK ASC;
+----+---------+
| PK | Value   |
+----+---------+
|  1 | both ab |
|  2 | only a  |
+----+---------+
2 rows in set (0.00 sec)

mysql> SELECT * from Table_B ORDER BY PK ASC;
+----+---------+
| PK | Value   |
+----+---------+
|  1 | both ab |
|  3 | only b  |
+----+---------+
2 rows in set (0.00 sec)

其中 PK 为 1 的记录在 Table_A 和 Table_B 中都有，2 为 Table_A 特有，3 为 Table_B 特有。

常用的 JOIN

INNER JOIN

INNER JOIN 一般被译作内连接。内连接查询能将左表（表 A）和右表（表 B）中能关联起来的数据连接后返回。

     

示例查询：

SELECT A.PK AS A_PK, B.PK AS B_PK,
       A.Value AS A_Value, B.Value AS B_Value
FROM Table_A A
INNER JOIN Table_B B
ON A.PK = B.PK;

查询结果：

+------+------+---------+---------+
| A_PK | B_PK | A_Value | B_Value |
+------+------+---------+---------+
|    1 |    1 | both ab | both ab |
+------+------+---------+---------+
1 row in set (0.00 sec)

注：其中 A 为 Table_A 的别名，B 为 Table_B 的别名，下同。

LEFT JOIN

LEFT JOIN 一般被译作左连接，也写作 LEFT OUTER JOIN。左连接查询会返回左表（表 A）中所有记录，不管右表（表 B）中有没有关联的数据。在右表中找到的关联数据列也会被一起返回。

    

示例查询：

SELECT A.PK AS A_PK, B.PK AS B_PK,
       A.Value AS A_Value, B.Value AS B_Value
FROM Table_A A
LEFT JOIN Table_B B
ON A.PK = B.PK;

查询结果：

+------+------+---------+---------+
| A_PK | B_PK | A_Value | B_Value |
+------+------+---------+---------+
|    1 |    1 | both ab | both ba |
|    2 | NULL | only a  | NULL    |
+------+------+---------+---------+
2 rows in set (0.00 sec)

RIGHT JOIN

RIGHT JOIN 一般被译作右连接，也写作 RIGHT OUTER JOIN。右连接查询会返回右表（表 B）中所有记录，不管左表（表 A）中有没有关联的数据。在左表中找到的关联数据列也会被一起返回。

     

示例查询：

SELECT A.PK AS A_PK, B.PK AS B_PK,
       A.Value AS A_Value, B.Value AS B_Value
FROM Table_A A
RIGHT JOIN Table_B B
ON A.PK = B.PK;

查询结果：

+------+------+---------+---------+
| A_PK | B_PK | A_Value | B_Value |
+------+------+---------+---------+
|    1 |    1 | both ab | both ba |
| NULL |    3 | NULL    | only b  |
+------+------+---------+---------+
2 rows in set (0.00 sec)

FULL OUTER JOIN

FULL OUTER JOIN 一般被译作外连接、全连接，实际查询语句中可以写作 FULL OUTER JOIN 或 FULL JOIN。外连接查询能返回左右表里的所有记录，其中左右表里能关联起来的记录被连接后返回。

    

示例查询：

SELECT A.PK AS A_PK, B.PK AS B_PK,
       A.Value AS A_Value, B.Value AS B_Value
FROM Table_A A
FULL OUTER JOIN Table_B B
ON A.PK = B.PK;

查询结果：

mysql> SELECT * 
    -> FROM Table_A
    -> LEFT JOIN Table_B 
    -> ON Table_A.PK = Table_B.PK
    -> UNION ALL
    -> SELECT *
    -> FROM Table_A
    -> RIGHT JOIN Table_B 
    -> ON Table_A.PK = Table_B.PK
    -> WHERE Table_A.PK IS NULL;
+------+---------+------+---------+
| PK   | Value   | PK   | Value   |
+------+---------+------+---------+
|    1 | both ab |    1 | both ba |
|    2 | only a  | NULL | NULL    |
| NULL | NULL    |    3 | only b  |
+------+---------+------+---------+
3 rows in set (0.00 sec)

延伸用法

LEFT JOIN EXCLUDING INNER JOIN

返回左表有但右表没有关联数据的记录集。

    

示例查询：

SELECT A.PK AS A_PK, B.PK AS B_PK,
       A.Value AS A_Value, B.Value AS B_Value
FROM Table_A A
LEFT JOIN Table_B B
ON A.PK = B.PK
WHERE B.PK IS NULL;

查询结果：

+------+------+---------+---------+
| A_PK | B_PK | A_Value | B_Value |
+------+------+---------+---------+
|    2 | NULL | only a  | NULL    |
+------+------+---------+---------+
1 row in set (0.01 sec)

同理，还有RIGHT JOIN EXCLUDING INNER JOIN，FULL OUTER JOIN EXCLUDING INNER JOIN，不一一赘述。

---

CROSS JOIN

返回左表与右表之间符合条件的记录的迪卡尔集。

图示：

    

示例查询：

SELECT A.PK AS A_PK, B.PK AS B_PK,
       A.Value AS A_Value, B.Value AS B_Value
FROM Table_A A
CROSS JOIN Table_B B;

查询结果：

+------+------+---------+---------+
| A_PK | B_PK | A_Value | B_Value |
+------+------+---------+---------+
|    1 |    1 | both ab | both ba |
|    2 |    1 | only a  | both ba |
|    1 |    3 | both ab | only b  |
|    2 |    3 | only a  | only b  |
+------+------+---------+---------+
4 rows in set (0.00 sec)

上面讲过的几种 JOIN 查询的结果都可以用 CROSS JOIN 加条件模拟出来，比如 INNER JOIN 对应 CROSS JOIN ... WHERE A.PK = B.PK。

SELF JOIN

返回表与自己连接后符合条件的记录，一般用在表里有一个字段是用主键作为外键的情况。

比如 Table_C 的结构与数据如下：

+--------+----------+-------------+
| EMP_ID | EMP_NAME | EMP_SUPV_ID |
+--------+----------+-------------+
|   1001 | Ma       |        NULL |
|   1002 | Zhuang   |        1001 |
+--------+----------+-------------+
2 rows in set (0.00 sec)

EMP_ID 字段表示员工 ID，EMP_NAME 字段表示员工姓名，EMP_SUPV_ID 表示主管 ID。

示例查询：

现在我们想查询所有有主管的员工及其对应的主管 ID 和姓名，就可以用 SELF JOIN 来实现。

SELECT A.EMP_ID AS EMP_ID, A.EMP_NAME AS EMP_NAME, 
    B.EMP_ID AS EMP_SUPV_ID, B.EMP_NAME AS EMP_SUPV_NAME
FROM Table_C A, Table_C B
WHERE A.EMP_SUPV_ID = B.EMP_ID;

查询结果：

+--------+----------+-------------+---------------+
| EMP_ID | EMP_NAME | EMP_SUPV_ID | EMP_SUPV_NAME |
+--------+----------+-------------+---------------+
|   1002 | Zhuang   |        1001 | Ma            |
+--------+----------+-------------+---------------+
1 row in set (0.00 sec)