# 多表连接查询 #### 连接类型介绍 通过SQL完成各种复杂的查询，多表之间的连接是必不可少的。连接分为：内连接和外连接两大类，每大类中还可进行细分。 - 内连接：标准内连接（INNER JOIN），交叉连接（CROSS JOIN）和自然连接（NATURAL JOIN）。 - 外连接：左外连接（LEFT OUTER JOIN），右外连接（RIGHT OUTER JOIN）和全外连接（FULL JOIN）。 为了能更好的说明各种连接之间的区别，下面通过具体示例进行详细的阐述。 创建示例表student和math_score，并插入数据。 ``` CREATE TABLE student( id INTEGER, name varchar(50) ) STORE AS orc; CREATE TABLE math_score( id INTEGER, score INTEGER ) STORE AS orc; INSERT INTO student VALUES(1, 'Tom'); INSERT INTO student VALUES(2, 'Lily'); INSERT INTO student VALUES(3, 'Tina'); INSERT INTO student VALUES(4, 'Perry'); INSERT INTO math_score VALUES(1, 80); INSERT INTO math_score VALUES(2, 75); INSERT INTO math_score VALUES(4, 95); INSERT INTO math_score VALUES(6, NULL); ``` #### 内连接 - 标准内连接（INNER JOIN） 语法： ``` left_table [INNER] JOIN right_table [ ON join_condition | USING ( join_column )] ``` 说明：表示left_table和right_table中满足join_condition的行拼接在一起作为结果输出，不满足条件的元组不会输出。 示例1：查询学生的数学成绩。 ``` SELECT s.id, s.name, ms.score FROM student s JOIN math_score ms on s.id = ms.id; id | name | score ----+-------+------- 2 | Lily | 75 1 | Tom | 80 4 | Perry | 95 (3 rows) EXPLAIN SELECT s.id, s.name, ms.score FROM student s JOIN math_score ms on s.id = ms.id; QUERY PLAN -------------------------------------------------------------------------------------------------- id | operation | E-rows | E-memory | E-width | E-costs ----+---------------------------------------------------+--------+----------+---------+--------- 1 | -> Row Adapter | 5000 | | 126 | 435.27 2 | -> Vector Streaming (type: GATHER) | 5000 | | 126 | 435.27 3 | -> Vector Sonic Hash Join (4,6) | 5000 | | 126 | 200.89 4 | -> Vector Streaming(type: BROADCAST) | 2000 | | 122 | 104.64 5 | -> Vector Foreign Scan on student s | 1000 | | 122 | 60.00 6 | -> Vector Foreign Scan on math_score ms | 1000 | | 8 | 60.00 Predicate Information (identified by plan id) --------------------------------------------- 3 --Vector Sonic Hash Join (4,6) Hash Cond: (s.id = ms.id) 5 --Vector Foreign Scan on student s Server Type: lf Total files left: 4 6 --Vector Foreign Scan on math_score ms Server Type: lf Total files left: 4 ====== Query Summary ===== -------------------------------------------------------- System available mem: 1992294KB Query Max mem: 2097152KB Query estimated mem: 2097152KB LakeFormation request time: 290.913 ms, request count: 5 Total billed bytes: 0 bytes Turbo Engine: true (28 rows) ``` - 交叉连接（CROSS JOIN） 语法： ``` left_table CROSS JOIN right_table ``` 说明：表示left_table中所有行和right_table中的所有行分别进行连接，最终结果行数等于两边行数的乘积。又称笛卡尔积。 示例2：学生表和数学成绩表的交叉连接。 ``` SELECT s.id, s.name, ms.score FROM student s CROSS JOIN math_score ms order by id; id | name | score ----+-------+------- 1 | Tom | 80 1 | Tom | 95 1 | Tom | 75 1 | Tom | 2 | Lily | 80 2 | Lily | 95 2 | Lily | 75 2 | Lily | 3 | Tina | 80 3 | Tina | 95 3 | Tina | 75 3 | Tina | 4 | Perry | 80 4 | Perry | 95 4 | Perry | 75 4 | Perry | (16 rows) EXPLAIN SELECT s.id, s.name, ms.score FROM student s CROSS JOIN math_score ms order by id; QUERY PLAN --------------------------------------------------------------------------------------------------------------- id | operation | E-rows | E-memory | E-width | E-costs ----+---------------------------------------------------------+---------+---------------+---------+---------- 1 | -> Row Adapter | 1000000 | | 126 | 55463.56 2 | -> Vector Streaming (type: GATHER) | 1000000 | | 126 | 55463.56 3 | -> Vector Sort | 1000000 | 518MB(2043MB) | 126 | 54994.81 4 | -> Vector Nest Loop (5,7) | 1000000 | 1MB | 126 | 6415.89 5 | -> Vector Streaming(type: BROADCAST) | 2000 | 2MB | 122 | 104.64 6 | -> Vector Foreign Scan on student s | 1000 | 1MB | 122 | 60.00 7 | -> Vector Materialize | 1000 | 16MB | 4 | 62.50 8 | -> Vector Foreign Scan on math_score ms | 1000 | 1MB | 4 | 60.00 Predicate Information (identified by plan id) --------------------------------------------- 6 --Vector Foreign Scan on student s Server Type: lf Total files left: 4 8 --Vector Foreign Scan on math_score ms Server Type: lf Total files left: 4 ====== Query Summary ===== -------------------------------------------------------- System available mem: 1992294KB Query Max mem: 2097152KB Query estimated mem: 2097152KB LakeFormation request time: 210.501 ms, request count: 5 Total billed bytes: 0 bytes Turbo Engine: true (28 rows) ``` - 自然连接（NATURAL JOIN） 语法： ``` left_table NATURAL JOIN right_table ``` 说明：表示left_table和right_table中列名相同的列进行等值连接，且自动将同名列只保留一份。 示例3：学生表和数学成绩表的自然连接，两表的同名列为id列，将按照id列进行等值连接。 ``` SELECT * FROM student s NATURAL JOIN math_score ms order by id; id | name | score ----+-------+------- 1 | Tom | 80 2 | Lily | 75 4 | Perry | 95 (3 rows) EXPLAIN SELECT * FROM student s NATURAL JOIN math_score ms order by id; QUERY PLAN ----------------------------------------------------------------------------------------------------- id | operation | E-rows | E-memory | E-width | E-costs ----+------------------------------------------------------+--------+----------+---------+--------- 1 | -> Row Adapter | 5000 | | 126 | 582.61 2 | -> Vector Streaming (type: GATHER) | 5000 | | 126 | 582.61 3 | -> Vector Sort | 5000 | | 126 | 348.24 4 | -> Vector Sonic Hash Join (5,7) | 5000 | | 126 | 200.89 5 | -> Vector Streaming(type: BROADCAST) | 2000 | | 122 | 104.64 6 | -> Vector Foreign Scan on student s | 1000 | | 122 | 60.00 7 | -> Vector Foreign Scan on math_score ms | 1000 | | 8 | 60.00 Predicate Information (identified by plan id) --------------------------------------------- 4 --Vector Sonic Hash Join (5,7) Hash Cond: (s.id = ms.id) 6 --Vector Foreign Scan on student s Server Type: lf Total files left: 4 7 --Vector Foreign Scan on math_score ms Server Type: lf Total files left: 4 ====== Query Summary ===== -------------------------------------------------------- System available mem: 1992294KB Query Max mem: 2097152KB Query estimated mem: 2097152KB LakeFormation request time: 377.624 ms, request count: 5 Total billed bytes: 0 bytes Turbo Engine: true (29 rows) ``` #### 外连接 - 左外连接（LEFT JOIN） 语法： ``` left_table LEFT [OUTER] JOIN right_table [ ON join_condition | USING ( join_column )] ``` 说明：左外连接的结果集包括left_table的所有行，而不仅是连接列所匹配的行。如果left_table的某行在right_table中没有匹配行，则在相关联的结果集行中输出right_table的列均为空值。 示例4：学生表和数学成绩表进行左外连接，学生表中id为3的行在结果集中对应的右表数据用NULL填充。 ``` SELECT s.id, s.name, ms.score FROM student s LEFT JOIN math_score ms on (s.id = ms.id) order by id; id | name | score ----+-------+------- 1 | Tom | 80 2 | Lily | 75 3 | Tina | 4 | Perry | 95 (4 rows) EXPLAIN SELECT s.id, s.name, ms.score FROM student s LEFT JOIN math_score ms on (s.id = ms.id) order by id; QUERY PLAN -------------------------------------------------------------------------------------------------------- id | operation | E-rows | E-memory | E-width | E-costs ----+---------------------------------------------------------+--------+----------+---------+--------- 1 | -> Row Adapter | 5000 | | 126 | 582.61 2 | -> Vector Streaming (type: GATHER) | 5000 | | 126 | 582.61 3 | -> Vector Sort | 5000 | | 126 | 348.24 4 | -> Vector Sonic Hash Right Join (5, 7) | 5000 | | 126 | 200.89 5 | -> Vector Streaming(type: BROADCAST) | 2000 | | 8 | 104.64 6 | -> Vector Foreign Scan on math_score ms | 1000 | | 8 | 60.00 7 | -> Vector Foreign Scan on student s | 1000 | | 122 | 60.00 Predicate Information (identified by plan id) --------------------------------------------- 4 --Vector Sonic Hash Right Join (5, 7) Hash Cond: (ms.id = s.id) 6 --Vector Foreign Scan on math_score ms Server Type: lf Total files left: 4 7 --Vector Foreign Scan on student s Server Type: lf Total files left: 4 ====== Query Summary ===== --------------------------------------------------------- System available mem: 1992294KB Query Max mem: 2097152KB Query estimated mem: 2097152KB LakeFormation request time: 1253.487 ms, request count: 5 Total billed bytes: 0 bytes Turbo Engine: true (29 rows) ``` - 右外连接（RIGHT JOIN） 语法： ``` left_table RIGHT [OUTER] JOIN right_table [ ON join_condition | USING ( join_column )] ``` 说明：与左外连接相反，右外连接的结果集包括right_table的所有行，而不仅是连接列所匹配的行。如果right_table的某行在left_table中没有匹配行，则在相关联的结果集行中left_table的列均为空值。 示例5：学生表和数学成绩表进行右外连接，数学成绩表中id为6的行在结果集中对应的左表数据用NULL填充。 ``` SELECT ms.id, s.name, ms.score FROM student s RIGHT JOIN math_score ms on (s.id = ms.id) order by id; id | name | score ----+-------+------- 1 | Tom | 80 2 | Lily | 75 4 | Perry | 95 6 | | (4 rows) EXPLAIN SELECT ms.id, s.name, ms.score FROM student s RIGHT JOIN math_score ms on (s.id = ms.id) order by id; QUERY PLAN ----------------------------------------------------------------------------------------------------- id | operation | E-rows | E-memory | E-width | E-costs ----+------------------------------------------------------+--------+----------+---------+--------- 1 | -> Row Adapter | 5000 | | 126 | 582.61 2 | -> Vector Streaming (type: GATHER) | 5000 | | 126 | 582.61 3 | -> Vector Sort | 5000 | | 126 | 348.24 4 | -> Vector Sonic Hash Right Join (5, 7) | 5000 | | 126 | 200.89 5 | -> Vector Streaming(type: BROADCAST) | 2000 | | 122 | 104.64 6 | -> Vector Foreign Scan on student s | 1000 | | 122 | 60.00 7 | -> Vector Foreign Scan on math_score ms | 1000 | | 8 | 60.00 Predicate Information (identified by plan id) --------------------------------------------- 4 --Vector Sonic Hash Right Join (5, 7) Hash Cond: (s.id = ms.id) 6 --Vector Foreign Scan on student s Server Type: lf Total files left: 4 7 --Vector Foreign Scan on math_score ms Server Type: lf Total files left: 4 ====== Query Summary ===== -------------------------------------------------------- System available mem: 1992294KB Query Max mem: 2097152KB Query estimated mem: 2097152KB LakeFormation request time: 214.389 ms, request count: 5 Total billed bytes: 0 bytes Turbo Engine: true (29 rows) ``` 对于右外连接，在join算子中却显示的是Left。这是因为，右外连接其实就是将左右表进行交互后的左外连接，所以数据库内部实现为了减少处理逻辑，会将右外连接转为左外连接。 - 全外连接（FULL JOIN） 语法： ``` left_table FULL [OUTER] JOIN right_table [ ON join_condition | USING ( join_column )] ``` 说明：全外连接是左外连接和右外连接的综合。全外连接的结果集包括left_table和right_table的所有行，而不仅是连接列所匹配的行。如果left_table的某行在right_table中没有匹配行，则在相关联的结果集行中right_table的列均为空值。如果right_table的某行在left_table中没有匹配行，则在相关联的结果集行中left_table的列均为空值。 示例6：学生表和数学成绩表进行全外连接，学生表中id为3的行在结果集中对应的右表数据用NULL填充，数学成绩表中id为6的行在结果集中对应的左表数据用NULL填充。 ``` SELECT s.id, s.name, ms.id, ms.score FROM student s FULL JOIN math_score ms ON (s.id = ms.id) order by s.id; id | name | id | score ----+-------+----+------- 1 | Tom | 1 | 80 2 | Lily | 2 | 75 3 | Tina | | 4 | Perry | 4 | 95 | | 6 | (5 rows) EXPLAIN SELECT s.id, s.name, ms.id, ms.score FROM student s FULL JOIN math_score ms ON (s.id = ms.id) order by s.id; QUERY PLAN -------------------------------------------------------------------------------------------------------- id | operation | E-rows | E-memory | E-width | E-costs ----+---------------------------------------------------------+--------+----------+---------+--------- 1 | -> Row Adapter | 5000 | | 130 | 734.79 2 | -> Vector Streaming (type: GATHER) | 5000 | | 130 | 734.79 3 | -> Vector Sort | 5000 | | 130 | 434.86 4 | -> Vector Sonic Hash Full Join (5, 7) | 5000 | | 130 | 243.95 5 | -> Vector Streaming(type: REDISTRIBUTE) | 1000 | | 122 | 103.57 6 | -> Vector Foreign Scan on student s | 1000 | | 122 | 60.00 7 | -> Vector Streaming(type: REDISTRIBUTE) | 1000 | | 8 | 103.57 8 | -> Vector Foreign Scan on math_score ms | 1000 | | 8 | 60.00 Predicate Information (identified by plan id) --------------------------------------------- 4 --Vector Sonic Hash Full Join (5, 7) Hash Cond: (s.id = ms.id) 6 --Vector Foreign Scan on student s Server Type: lf Total files left: 4 8 --Vector Foreign Scan on math_score ms Server Type: lf Total files left: 4 ====== Query Summary ===== -------------------------------------------------------- System available mem: 1992294KB Query Max mem: 2097152KB Query estimated mem: 2097152KB LakeFormation request time: 206.619 ms, request count: 5 Total billed bytes: 0 bytes Turbo Engine: true (30 rows) ``` #### 多表查询中on条件和where条件的区别 从上面各种连接语法中可见，除自然连接和交叉连接外，其他都需要有on条件（using在查询解析过程中会被转为on条件）来限制两表连接的结果。通常在查询的语句中也都会有where条件限制查询结果。这里说的on连接条件和where过滤条件是指不含可以下推到表上的过滤条件。on和where的区别是： - on条件是两表连接的约束条件。 - where是对两表连接后产生的结果集再次进行过滤。 简单总结就是：on条件优先于where条件，在两表进行连接时被应用；生成两表连接结果集后，再应用where条件。