一、描述

SQL连接用于根据join criteria组合来自两个关系的行。以下部分描述了整个join语法，子部分介绍了不同类型的连接以及示例。

二、语法

relation { [ join_type ] JOIN [ LATERAL ] relation [ join_criteria ] | NATURAL join_type JOIN [ LATERAL ] relation }

三、参数

• relation
  指定要join的关系。
• join_type
  指定join类型。
  语法：

[ INNER ] | CROSS | LEFT [ OUTER ] | [ LEFT ] SEMI | RIGHT [ OUTER ] | FULL [ OUTER ] | [ LEFT ] ANTI

• join_criteria
  指定如何将一个关系中的行与另一个关系的行组合。
  语法：

ON boolean_expression | USING ( column_name [ , ... ] ) boolean_expression

指定返回类型为布尔值的表达式。

四、Join类型

4.1 Inner Join

inner join是Spark SQL的默认join。它选择在两个关系中具有匹配值的行。
语法:

relation [ INNER ] JOIN relation [ join_criteria ]

4.2 Left Join

左连接返回左关系中的所有值和右关系中的匹配值，如果没有匹配，则追加NULL。它也被称为左外连接。
语法:

relation LEFT [ OUTER ] JOIN relation [ join_criteria ]

4.3 Right Join

右连接返回来自右关系的所有值和来自左关系的匹配值，如果没有匹配，则追加NULL。它也被称为右外连接。
语法:

relation RIGHT [ OUTER ] JOIN relation [ join_criteria ]

4.4 Full Join

全连接返回两个关系中的所有值，并在没有匹配的一侧附加NULL值。它也被称为全外连接。
语法:

relation FULL [ OUTER ] JOIN relation [ join_criteria ]

4.5 Cross Join

交叉连接返回两个关系的笛卡尔积。
语法:

relation CROSS JOIN relation [ join_criteria ]

4.6 Semi Join

半连接从关系的左侧返回与右侧匹配的值。它也被称为左半连接。
语法:

relation [ LEFT ] SEMI JOIN relation [ join_criteria ]

4.7 Anti Join

反连接返回与右关系不匹配的左关系中的值。它也被称为左反连接。
语法:

relation [ LEFT ] ANTI JOIN relation [ join_criteria ]

五、例子

-- Use employee and department tables to demonstrate different type of joins.
SELECT * FROM employee;
+---+-----+------+
| id| name|deptno|
+---+-----+------+
|105|Chloe|     5|
|103| Paul|     3|
|101| John|     1|
|102| Lisa|     2|
|104| Evan|     4|
|106|  Amy|     6|
+---+-----+------+SELECT * FROM department;
+------+-----------+
|deptno|   deptname|
+------+-----------+
|     3|Engineering|
|     2|      Sales|
|     1|  Marketing|
+------+-----------+-- Use employee and department tables to demonstrate inner join.
SELECT id, name, employee.deptno, deptnameFROM employee INNER JOIN department ON employee.deptno = department.deptno;
+---+-----+------+-----------|
| id| name|deptno|   deptname|
+---+-----+------+-----------|
|103| Paul|     3|Engineering|
|101| John|     1|  Marketing|
|102| Lisa|     2|      Sales|
+---+-----+------+-----------|-- Use employee and department tables to demonstrate left join.
SELECT id, name, employee.deptno, deptnameFROM employee LEFT JOIN department ON employee.deptno = department.deptno;
+---+-----+------+-----------|
| id| name|deptno|   deptname|
+---+-----+------+-----------|
|105|Chloe|     5|       NULL|
|103| Paul|     3|Engineering|
|101| John|     1|  Marketing|
|102| Lisa|     2|      Sales|
|104| Evan|     4|       NULL|
|106|  Amy|     6|       NULL|
+---+-----+------+-----------|-- Use employee and department tables to demonstrate right join.
SELECT id, name, employee.deptno, deptnameFROM employee RIGHT JOIN department ON employee.deptno = department.deptno;
+---+-----+------+-----------|
| id| name|deptno|   deptname|
+---+-----+------+-----------|
|103| Paul|     3|Engineering|
|101| John|     1|  Marketing|
|102| Lisa|     2|      Sales|
+---+-----+------+-----------|-- Use employee and department tables to demonstrate full join.
SELECT id, name, employee.deptno, deptnameFROM employee FULL JOIN department ON employee.deptno = department.deptno;
+---+-----+------+-----------|
| id| name|deptno|   deptname|
+---+-----+------+-----------|
|101| John|     1|  Marketing|
|106|  Amy|     6|       NULL|
|103| Paul|     3|Engineering|
|105|Chloe|     5|       NULL|
|104| Evan|     4|       NULL|
|102| Lisa|     2|      Sales|
+---+-----+------+-----------|-- Use employee and department tables to demonstrate cross join.
SELECT id, name, employee.deptno, deptname FROM employee CROSS JOIN department;
+---+-----+------+-----------|
| id| name|deptno|   deptname|
+---+-----+------+-----------|
|105|Chloe|     5|Engineering|
|105|Chloe|     5|  Marketing|
|105|Chloe|     5|      Sales|
|103| Paul|     3|Engineering|
|103| Paul|     3|  Marketing|
|103| Paul|     3|      Sales|
|101| John|     1|Engineering|
|101| John|     1|  Marketing|
|101| John|     1|      Sales|
|102| Lisa|     2|Engineering|
|102| Lisa|     2|  Marketing|
|102| Lisa|     2|      Sales|
|104| Evan|     4|Engineering|
|104| Evan|     4|  Marketing|
|104| Evan|     4|      Sales|
|106|  Amy|     4|Engineering|
|106|  Amy|     4|  Marketing|
|106|  Amy|     4|      Sales|
+---+-----+------+-----------|-- Use employee and department tables to demonstrate semi join.
SELECT * FROM employee SEMI JOIN department ON employee.deptno = department.deptno;
+---+-----+------+
| id| name|deptno|
+---+-----+------+
|103| Paul|     3|
|101| John|     1|
|102| Lisa|     2|
+---+-----+------+-- Use employee and department tables to demonstrate anti join.
SELECT * FROM employee ANTI JOIN department ON employee.deptno = department.deptno;
+---+-----+------+
| id| name|deptno|
+---+-----+------+
|105|Chloe|     5|
|104| Evan|     4|
|106|  Amy|     6|
+---+-----+------+