首页 > 代码库 > 数据库外连接及MySQL实现
数据库外连接及MySQL实现
MySQL查询分为内连接查询和外连接查询,他们的区别在于:内连接查询的两个表示对等关系,根据条件进行匹配;外连接是以某一个表为主,两一个表根据条件进行关联。外连接分为左外连接、右外连接和全外连接。本文重点介绍各外连接的思想,以及如何实现全外连接,并举例。
左外连接
左外连接以左边表为基础,根据条件,将右边表附属到左边表,语法:SELECT * FROM A LEFT JOIN B ON condition。几何图形关系如下图,即查询结果集除了A表所有数据外,还包含满足条件的B表数据:
右外连接
右外连接以右边表为基础,根据条件,将左边表附属到右边表,语法:SELECT * FROM A RIGHT JOIN B ON condition。几何图形关系如下图,即查询结果集除了B表所有数据外,还包含满足条件的A表数据:
全外连接
全外连接是除了能够根据条件匹配得到的数据,还包含左右两表中都不匹配的数据(默认应为null),应用全外连接的情况一般都有一个联系左右两表的主线。几何关系如下图所示,对应A和B的并集(去重):
但不幸的是MySQL不支持全外连接,那在需要全外连接查询的情况下,如何实现呢?最常见的是左连接与右连接合并。
实例
项目中存在这样的场景:某项任务task具有2种不同的状态todo和done,分别存储在todolist和donelist表中,任务存储在task表中,现在需要统计每个task的已处理和未处理情况。首先先到了全外连接,那么如何实现呢?
举例实现表结构如下:
实现四种方法:
1、左连接,右连接,合并;(需保持两个结果集结构一致)
- 首先是左连接:
1 SELECT 2 A.id AS Aid, 3 B.id AS Bid, 4 A.taskid tid 5 FROM 6 ( 7 SELECT 8 * 9 FROM 10 todolist 11 WHERE 12 todolist.user = ‘张三‘ 13 ) A 14 LEFT JOIN ( 15 SELECT 16 * 17 FROM 18 donelist 19 WHERE 20 donelist.user = ‘张三‘ 21 ) B ON A.taskid = B.taskid
查询结果:
- 其次是右连接(注意由于需要合并,故左右连接的结果集结构需一致):
1 SELECT 2 A.id AS Aid, 3 B.id AS Bid, 4 A.taskid tid 5 FROM 6 ( 7 SELECT 8 * 9 FROM 10 todolist 11 WHERE 12 todolist.user = ‘张三‘ 13 ) A 14 RIGHT JOIN ( 15 SELECT 16 * 17 FROM 18 donelist 19 WHERE 20 donelist.user = ‘张三‘ 21 ) B ON A.taskid = B.taskid
查询结果:
- 最后进行合并,并与task表进行内连接:
1 SELECT 2 SUM(IF(Aid IS NOT NULL, 1, 0)) todo, 3 SUM(IF(Bid IS NOT NULL, 1, 0)) done, 4 task.name 5 FROM 6 ( 7 SELECT 8 A.id AS Aid, 9 B.id AS Bid, 10 A.taskid tid 11 FROM 12 ( 13 SELECT 14 * 15 FROM 16 todolist 17 WHERE 18 todolist.user = ‘张三‘ 19 ) A 20 LEFT JOIN ( 21 SELECT 22 * 23 FROM 24 donelist 25 WHERE 26 donelist.user = ‘张三‘ 27 ) B ON A.taskid = B.taskid 28 UNION 29 SELECT 30 A.id AS Aid, 31 B.id AS Bid, 32 B.taskid tid 33 FROM 34 ( 35 SELECT 36 * 37 FROM 38 todolist 39 WHERE 40 todolist.user = ‘张三‘ 41 ) A 42 RIGHT JOIN ( 43 SELECT 44 * 45 FROM 46 donelist 47 WHERE 48 donelist.user = ‘张三‘ 49 ) B ON A.taskid = B.taskid 50 ) AS AB 51 INNER JOIN task ON task.id = AB.tid 52 GROUP BY 53 task.name
运行结果如下表,实现全外连接:
2、A+B左连接,B-A去除左连接到A的记录,然后合并两个结果集;(需保持两个结果集结构一致)
这是另一种实现全外连接的方式,即先查询A B的左连接,然后查询B中去除左连接到A的记录,最后合并(A代表todolist,B代表donelist):
- A+B左连接
1 SELECT 2 1 AS todo, 3 CASE 4 WHEN B.id IS NOT NULL THEN 5 1 6 ELSE 7 0 8 END AS done, 9 A.taskid tid 10 FROM 11 ( 12 SELECT 13 * 14 FROM 15 todolist 16 WHERE 17 todolist.user = ‘张三‘ 18 ) A 19 LEFT JOIN ( 20 SELECT 21 * 22 FROM 23 donelist 24 WHERE 25 donelist.user = ‘张三‘ 26 ) B ON A.taskid = B.taskid
查询结果:
- B-A去除左连接到A的记录
1 SELECT 2 0 AS todo, 3 1 AS done, 4 donelist.taskid tid 5 FROM 6 donelist 7 WHERE 8 donelist.user = ‘张三‘ 9 AND NOT EXISTS ( 10 SELECT 11 * 12 FROM 13 todolist 14 WHERE 15 todolist.taskid = donelist.taskid 16 AND donelist.user = ‘张三‘ 17 AND odolist.user = donelist.user 18 )
查询结果:
- 合并
1 SELECT 2 SUM(AB.todo) todo, 3 SUM(AB.done) done, 4 task.name 5 FROM 6 ( 7 SELECT 8 1 AS todo, 9 CASE 10 WHEN B.id IS NOT NULL THEN 11 1 12 ELSE 13 0 14 END AS done, 15 A.taskid tid 16 FROM 17 ( 18 SELECT 19 * 20 FROM 21 todolist 22 WHERE 23 todolist.user = ‘张三‘ 24 ) A 25 LEFT JOIN ( 26 SELECT 27 * 28 FROM 29 donelist 30 WHERE 31 donelist.user = ‘张三‘ 32 ) B ON A.taskid = B.taskid 33 UNION 34 SELECT 35 0 AS todo, 36 1 AS done, 37 donelist.taskid tid 38 FROM 39 donelist 40 WHERE 41 donelist.user = ‘张三‘ 42 AND NOT EXISTS ( 43 SELECT 44 * 45 FROM 46 todolist 47 WHERE 48 todolist.taskid = donelist.taskid 49 AND donelist.user = ‘张三‘ 50 AND odolist.user = donelist.user 51 ) 52 ) AB 53 INNER JOIN task ON task.id = AB.tid 54 GROUP BY 55 task.name
结果同上
3、以task表为根本,将A和B表左连接,实现查询;
该方法的思想是,不管A和B表有什么关系,他们都跟作为主线的表task相关,只需要将A和B表与task表进行左连接,得到连接后的数据集,即为最后需要查询的结果集。SQL代码如下:
1 SELECT 2 SUM(AB.todo) AS todo, 3 SUM(AB.done) AS done, 4 task.name 5 FROM 6 ( 7 SELECT 8 task.name, 9 CASE 10 WHEN A.id IS NULL THEN 11 0 12 ELSE 13 1 14 END AS todo, 15 CASE 16 WHEN B.id IS NULL THEN 17 0 18 ELSE 19 1 20 END AS done 21 FROM 22 task 23 LEFT JOIN ( 24 SELECT 25 * 26 FROM 27 todolist 28 WHERE 29 todolist.user = ‘张三‘ 30 ) A ON A.taskid = task.id 31 LEFT JOIN ( 32 SELECT 33 * 34 FROM 35 donelist 36 WHERE 37 donelist.user = ‘张三‘ 38 ) B ON B.taskid = task.id 39 WHERE 40 A.id IS NOT NULL 41 OR B.id IS NOT NULL 42 ) AB 43 GROUP BY 44 task.name
查询结果同上,但这种方法存在一定的缺陷,即当主线表(task表)特别大的时候,性能会比较差。
4、A表查a状态,B表查b状态,然后合并;(需保持两个结果集结构一致)
该方法是不管A和B表的关系,现根据条件查询,然后在合并。SQL语句如下:
1 SELECT 2 SUM(A.todo) todo, 3 SUM(A.done) done, 4 task.name 5 FROM 6 ( 7 SELECT 8 1 todo, 9 0 done, 10 todolist.taskid tid 11 FROM 12 todolist 13 WHERE 14 todolist.user = ‘张三‘ 15 UNION ALL 16 SELECT 17 0 todo, 18 1 done, 19 donelist.taskid tid 20 FROM 21 donelist 22 WHERE 23 donelist.user = ‘张三‘ 24 ) A 25 INNER JOIN task ON task.id = A.tid 26 GROUP BY 27 task.name
查询结果同上。
四种方式只是实现功能,具体性能没有考虑,待后续学习。希望看客们多提意见,共同学习。
总结:
- 理解左连接、右连接和全外连接的思想
- 四种实现全外连接效果的方法
- 语法上注意:UNION可以去重,UNION ALL不去重
数据库外连接及MySQL实现