标签：outer union all mit shanghai val 应该 union none number
一、多表连接查询: #重点：外链接语法
准备表
#建表 create table department( id int, name varchar(20) ); create table employee( id int primary key auto_increment, name varchar(20), sex enum(‘male‘,‘female‘) not null default ‘male‘, age int, dep_id int ); #插入数据 insert into department values (200,‘技术‘), (201,‘人力资源‘), (202,‘销售‘), (203,‘运营‘); insert into employee(name,sex,age,dep_id) values (‘egon‘,‘male‘,18,200), (‘alex‘,‘female‘,48,201), (‘wupeiqi‘,‘male‘,38,201), (‘yuanhao‘,‘female‘,28,202), (‘liwenzhou‘,‘male‘,18,200), (‘jingliyang‘,‘female‘,18,204) ; #查看表结构和数据 mysql> desc department; +-------+-------------+------+-----+---------+-------+ | Field | Type | Null | Key | Default | Extra | +-------+-------------+------+-----+---------+-------+ | id | int(11) | YES | | NULL | | | name | varchar(20) | YES | | NULL | | +-------+-------------+------+-----+---------+-------+ mysql> desc employee; +--------+-----------------------+------+-----+---------+----------------+ | Field | Type | Null | Key | Default | Extra | +--------+-----------------------+------+-----+---------+----------------+ | id | int(11) | NO | PRI | NULL | auto_increment | | name | varchar(20) | YES | | NULL | | | sex | enum(‘male‘,‘female‘) | NO | | male | | | age | int(11) | YES | | NULL | | | dep_id | int(11) | YES | | NULL | | +--------+-----------------------+------+-----+---------+----------------+ mysql> select * from department; +------+--------------+ | id | name | +------+--------------+ | 200 | 技术 | | 201 | 人力资源 | | 202 | 销售 | | 203 | 运营 | +------+--------------+ mysql> select * from employee; +----+------------+--------+------+--------+ | id | name | sex | age | dep_id | +----+------------+--------+------+--------+ | 1 | egon | male | 18 | 200 | | 2 | alex | female | 48 | 201 | | 3 | wupeiqi | male | 38 | 201 | | 4 | yuanhao | female | 28 | 202 | | 5 | liwenzhou | male | 18 | 200 | | 6 | jingliyang | female | 18 | 204 | +----+------------+--------+------+--------+
1 .交叉连接：不适用任何匹配条件(没意义)。生成笛卡尔积
select * from department,employee; #笛卡尔积
mysql> select * from employee,department; +----+------------+--------+------+--------+------+--------------+ | id | name | sex | age | dep_id | id | name | +----+------------+--------+------+--------+------+--------------+ | 1 | egon | male | 18 | 200 | 200 | 技术 | | 1 | egon | male | 18 | 200 | 201 | 人力资源 | | 1 | egon | male | 18 | 200 | 202 | 销售 | | 1 | egon | male | 18 | 200 | 203 | 运营 | | 2 | alex | female | 48 | 201 | 200 | 技术 | | 2 | alex | female | 48 | 201 | 201 | 人力资源 | | 2 | alex | female | 48 | 201 | 202 | 销售 | | 2 | alex | female | 48 | 201 | 203 | 运营 | | 3 | wupeiqi | male | 38 | 201 | 200 | 技术 | | 3 | wupeiqi | male | 38 | 201 | 201 | 人力资源 | | 3 | wupeiqi | male | 38 | 201 | 202 | 销售 | | 3 | wupeiqi | male | 38 | 201 | 203 | 运营 | | 4 | yuanhao | female | 28 | 202 | 200 | 技术 | | 4 | yuanhao | female | 28 | 202 | 201 | 人力资源 | | 4 | yuanhao | female | 28 | 202 | 202 | 销售 | | 4 | yuanhao | female | 28 | 202 | 203 | 运营 | | 5 | liwenzhou | male | 18 | 200 | 200 | 技术 | | 5 | liwenzhou | male | 18 | 200 | 201 | 人力资源 | | 5 | liwenzhou | male | 18 | 200 | 202 | 销售 | | 5 | liwenzhou | male | 18 | 200 | 203 | 运营 | | 6 | jingliyang | female | 18 | 204 | 200 | 技术 | | 6 | jingliyang | female | 18 | 204 | 201 | 人力资源 | | 6 | jingliyang | female | 18 | 204 | 202 | 销售 | | 6 | jingliyang | female | 18 | 204 | 203 | 运营 | +----+------------+--------+------+--------+------+--------------+
2.内链接（只连接匹配的行）：按照on条件只两张表的相同的部分，连接成一张虚拟的表
select * from employee inner join department on department.id=employee.dep_id;
select * from department inner join employee on department.id=employee.dep_id;
select * from employee,department where department.id=employee.dep_id;
在内连接两张表的时候不推荐使用where，推荐 inner join，因为join on 在where之前执行
mysql> select * from employee inner join department on department.id=employee.dep_id; +----+-----------+--------+------+--------+------+--------------+ | id | name | sex | age | dep_id | id | name | +----+-----------+--------+------+--------+------+--------------+ | 1 | egon | male | 18 | 200 | 200 | 技术 | | 2 | alex | female | 48 | 201 | 201 | 人力资源 | | 3 | wupeiqi | male | 38 | 201 | 201 | 人力资源 | | 4 | yuanhao | female | 28 | 202 | 202 | 销售 | | 5 | liwenzhou | male | 18 | 200 | 200 | 技术 | +----+-----------+--------+------+--------+------+--------------+ 5 rows in set (0.01 sec)
3.左链接（外链接之左连接：优先显示左表全部记录）：在按照on的条件取到两张表共同部分的基础上，保留左表的记录
select * from employee left join department on department.id=employee.dep_id;
mysql> select * from employee left join department on department.id=employee.dep_id; +----+------------+--------+------+--------+------+--------------+ | id | name | sex | age | dep_id | id | name | +----+------------+--------+------+--------+------+--------------+ | 1 | egon | male | 18 | 200 | 200 | 技术 | | 5 | liwenzhou | male | 18 | 200 | 200 | 技术 | | 2 | alex | female | 48 | 201 | 201 | 人力资源 | | 3 | wupeiqi | male | 38 | 201 | 201 | 人力资源 | | 4 | yuanhao | female | 28 | 202 | 202 | 销售 | | 6 | jingliyang | female | 18 | 204 | NULL | NULL | +----+------------+--------+------+--------+------+--------------+ 6 rows in set (0.00 sec)
4.右链接（外链接之右连接：优先显示右表全部记录）：在按照on的条件取到两张表共同部分的基础上，保留右表的记录
select * from employee right join department on department.id=employee.dep_id;
mysql> select * from employee right join department on department.id=employee.dep_id; +------+-----------+--------+------+--------+------+--------------+ | id | name | sex | age | dep_id | id | name | +------+-----------+--------+------+--------+------+--------------+ | 1 | egon | male | 18 | 200 | 200 | 技术 | | 2 | alex | female | 48 | 201 | 201 | 人力资源 | | 3 | wupeiqi | male | 38 | 201 | 201 | 人力资源 | | 4 | yuanhao | female | 28 | 202 | 202 | 销售 | | 5 | liwenzhou | male | 18 | 200 | 200 | 技术 | | NULL | NULL | NULL | NULL | NULL | 203 | 运营 | +------+-----------+--------+------+--------+------+--------------+ 6 rows in set (0.00 sec)
5.full join：（注意在sql语句中没有full join）
全链接（全外连接：显示左右两个表全部记录）：union
select * from employee left join department on department.id=employee.dep_id
union
select * from employee right join department on department.id=employee.dep_id;
mysql> select * from employee left join department on department.id=employee.dep_id -> union -> select * from employee right join department on department.id=employee.dep_id; +------+------------+--------+------+--------+------+--------------+ | id | name | sex | age | dep_id | id | name | +------+------------+--------+------+--------+------+--------------+ | 1 | egon | male | 18 | 200 | 200 | 技术 | | 5 | liwenzhou | male | 18 | 200 | 200 | 技术 | | 2 | alex | female | 48 | 201 | 201 | 人力资源 | | 3 | wupeiqi | male | 38 | 201 | 201 | 人力资源 | | 4 | yuanhao | female | 28 | 202 | 202 | 销售 | | 6 | jingliyang | female | 18 | 204 | NULL | NULL | | NULL | NULL | NULL | NULL | NULL | 203 | 运营 | +------+------------+--------+------+--------+------+--------------+ 7 rows in set (0.00 sec)
#注意 union与union all的区别：union会去掉相同的纪录
6.子查询：
1：子查询是将一个查询语句嵌套在另一个查询语句中。
2：内层查询语句的查询结果，可以为外层查询语句提供查询条件。
3：子查询中可以包含：IN、NOT IN、ANY、ALL、EXISTS 和 NOT EXISTS等关键字
4：还可以包含比较运算符：= 、 !=、> 、<等
1 带IN关键字的子查询： #查询employee表，但dep_id必须在department表中出现过 select * from employee where dep_id in (select id from department);
mysql> select * from employee where dep_id in (select id from department where name in (‘技术‘,‘销售‘)); +----+-----------+--------+------+--------+ | id | name | sex | age | dep_id | +----+-----------+--------+------+--------+ | 1 | egon | male | 18 | 200 | | 4 | yuanhao | female | 28 | 202 | | 5 | liwenzhou | male | 18 | 200 | +----+-----------+--------+------+--------+ 3 rows in set (0.02 sec)
2 带比较运算符的子查询： #比较运算符：=、!=、>、>=、<、<=、<>
#查询平均年龄在25岁以上的部门名 select name from department where id in ( select dep_id from employee group by dep_id having avg(age) > 25 ); +--------------+ | name | +--------------+ | 人力资源 | | 销售 | +--------------+ 2 rows in set (0.00 sec) #查看技术部员工姓名 select name from employee where dep_id = (select id from department where name=‘技术‘); +-----------+ | name | +-----------+ | egon | | liwenzhou | +-----------+ 2 rows in set (0.00 sec) #查看小于2人的部门名 select name from department where id in ( select dep_id from employee group by dep_id having count(id) < 2 ) union select name from department where id not in (select distinct dep_id from employee); +--------+ | name | +--------+ | 销售 | | 运营 | +--------+ 2 rows in set (0.01 sec) #提取空部门 #没有人的部门 select * from department where id not in (select distinct dep_id from employee); +------+--------+ | id | name | +------+--------+ | 203 | 运营 | +------+--------+ 1 row in set (0.00 sec) 或者： select name from department where id in ( select dep_id from employee group by dep_id having count(id) < 2 union select id from department where id not in (select distinct dep_id from employee) ); +--------+ | name | +--------+ | 销售 | | 运营 | +--------+ 2 rows in set (0.00 sec)
3 带EXISTS关键字的子查询：
exists关字键字表示存在。在使用EXISTS关键字时，内层查询语句不返回查询的记录。
而是返回一个真假值。True或False
当返回True时，外层查询语句将进行查询；当返回值为False时，外层查询语句不进行查询
mysql> select * from employee where exists (select id from department where name=‘hahahahah‘); Empty set (0.00 sec) mysql> select * from employee where exists (select id from department where name=‘技术‘); +----+------------+--------+------+--------+ | id | name | sex | age | dep_id | +----+------------+--------+------+--------+ | 1 | egon | male | 18 | 200 | | 2 | alex | female | 48 | 201 | | 3 | wupeiqi | male | 38 | 201 | | 4 | yuanhao | female | 28 | 202 | | 5 | liwenzhou | male | 18 | 200 | | 6 | jingliyang | female | 18 | 204 | +----+------------+--------+------+--------+ 6 rows in set (0.00 sec)
7.！！！重中之重：务必搞清楚sql逻辑查询语句的执行顺序
SELECT语句关键字的定义顺序 SELECT DISTINCT <select_list> FROM <left_table> <join_type> JOIN <right_table> ON <join_condition> WHERE <where_condition> GROUP BY <group_by_list> HAVING <having_condition> ORDER BY <order_by_condition> LIMIT <limit_number>
SELECT语句关键字的执行顺序 (7) SELECT (8) DISTINCT <select_list> (1) FROM <left_table> (3) <join_type> JOIN <right_table> (2) ON <join_condition> (4) WHERE <where_condition> (5) GROUP BY <group_by_list> (6) HAVING <having_condition> (9) ORDER BY <order_by_condition> (10) LIMIT <limit_number>
三 准备表和数据 1. 新建一个测试数据库TestDB； create database TestDB; 2.创建测试表table1和table2； CREATE TABLE table1 ( customer_id VARCHAR(10) NOT NULL, city VARCHAR(10) NOT NULL, PRIMARY KEY(customer_id) )ENGINE=INNODB DEFAULT CHARSET=UTF8; CREATE TABLE table2 ( order_id INT NOT NULL auto_increment, customer_id VARCHAR(10), PRIMARY KEY(order_id) )ENGINE=INNODB DEFAULT CHARSET=UTF8; 3.插入测试数据； INSERT INTO table1(customer_id,city) VALUES(‘163‘,‘hangzhou‘); INSERT INTO table1(customer_id,city) VALUES(‘9you‘,‘shanghai‘); INSERT INTO table1(customer_id,city) VALUES(‘tx‘,‘hangzhou‘); INSERT INTO table1(customer_id,city) VALUES(‘baidu‘,‘hangzhou‘); INSERT INTO table2(customer_id) VALUES(‘163‘); INSERT INTO table2(customer_id) VALUES(‘163‘); INSERT INTO table2(customer_id) VALUES(‘9you‘); INSERT INTO table2(customer_id) VALUES(‘9you‘); INSERT INTO table2(customer_id) VALUES(‘9you‘); INSERT INTO table2(customer_id) VALUES(‘tx‘); INSERT INTO table2(customer_id) VALUES(NULL); 准备工作做完以后，table1和table2看起来应该像下面这样： mysql> select * from table1; +-------------+----------+ | customer_id | city | +-------------+----------+ | 163 | hangzhou | | 9you | shanghai | | baidu | hangzhou | | tx | hangzhou | +-------------+----------+ 4 rows in set (0.00 sec) mysql> select * from table2; +----------+-------------+ | order_id | customer_id | +----------+-------------+ | 1 | 163 | | 2 | 163 | | 3 | 9you | | 4 | 9you | | 5 | 9you | | 6 | tx | | 7 | NULL | +----------+-------------+ 7 rows in set (0.00 sec) 四 准备SQL逻辑查询测试语句 #查询来自杭州，并且订单数少于2的客户。 SELECT a.customer_id, COUNT(b.order_id) as total_orders FROM table1 AS a LEFT JOIN table2 AS b ON a.customer_id = b.customer_id WHERE a.city = ‘hangzhou‘ GROUP BY a.customer_id HAVING count(b.order_id) < 2 ORDER BY total_orders DESC; 五 执行顺序分析 在这些SQL语句的执行过程中，都会产生一个虚拟表，用来保存SQL语句的执行结果（这是重点），我现在就来跟踪这个虚拟表的变化，得到最终的查询结果的过程，来分析整个SQL逻辑查询的执行顺序和过程。 执行FROM语句 第一步，执行FROM语句。我们首先需要知道最开始从哪个表开始的，这就是FROM告诉我们的。现在有了<left_table>和<right_table>两个表，我们到底从哪个表开始，还是从两个表进行某种联系以后再开始呢？它们之间如何产生联系呢？——笛卡尔积 关于什么是笛卡尔积，请自行Google补脑。经过FROM语句对两个表执行笛卡尔积，会得到一个虚拟表，暂且叫VT1（vitual table 1），内容如下： +-------------+----------+----------+-------------+ | customer_id | city | order_id | customer_id | +-------------+----------+----------+-------------+ | 163 | hangzhou | 1 | 163 | | 9you | shanghai | 1 | 163 | | baidu | hangzhou | 1 | 163 | | tx | hangzhou | 1 | 163 | | 163 | hangzhou | 2 | 163 | | 9you | shanghai | 2 | 163 | | baidu | hangzhou | 2 | 163 | | tx | hangzhou | 2 | 163 | | 163 | hangzhou | 3 | 9you | | 9you | shanghai | 3 | 9you | | baidu | hangzhou | 3 | 9you | | tx | hangzhou | 3 | 9you | | 163 | hangzhou | 4 | 9you | | 9you | shanghai | 4 | 9you | | baidu | hangzhou | 4 | 9you | | tx | hangzhou | 4 | 9you | | 163 | hangzhou | 5 | 9you | | 9you | shanghai | 5 | 9you | | baidu | hangzhou | 5 | 9you | | tx | hangzhou | 5 | 9you | | 163 | hangzhou | 6 | tx | | 9you | shanghai | 6 | tx | | baidu | hangzhou | 6 | tx | | tx | hangzhou | 6 | tx | | 163 | hangzhou | 7 | NULL | | 9you | shanghai | 7 | NULL | | baidu | hangzhou | 7 | NULL | | tx | hangzhou | 7 | NULL | +-------------+----------+----------+-------------+ 总共有28（table1的记录条数 * table2的记录条数）条记录。这就是VT1的结果，接下来的操作就在VT1的基础上进行。 执行ON过滤 执行完笛卡尔积以后，接着就进行ON a.customer_id = b.customer_id条件过滤，根据ON中指定的条件，去掉那些不符合条件的数据，得到VT2表，内容如下： +-------------+----------+----------+-------------+ | customer_id | city | order_id | customer_id | +-------------+----------+----------+-------------+ | 163 | hangzhou | 1 | 163 | | 163 | hangzhou | 2 | 163 | | 9you | shanghai | 3 | 9you | | 9you | shanghai | 4 | 9you | | 9you | shanghai | 5 | 9you | | tx | hangzhou | 6 | tx | +-------------+----------+----------+-------------+ VT2就是经过ON条件筛选以后得到的有用数据，而接下来的操作将在VT2的基础上继续进行。 添加外部行 这一步只有在连接类型为OUTER JOIN时才发生，如LEFT OUTER JOIN、RIGHT OUTER JOIN和FULL OUTER JOIN。在大多数的时候，我们都是会省略掉OUTER关键字的，但OUTER表示的就是外部行的概念。 LEFT OUTER JOIN把左表记为保留表，得到的结果为： +-------------+----------+----------+-------------+ | customer_id | city | order_id | customer_id | +-------------+----------+----------+-------------+ | 163 | hangzhou | 1 | 163 | | 163 | hangzhou | 2 | 163 | | 9you | shanghai | 3 | 9you | | 9you | shanghai | 4 | 9you | | 9you | shanghai | 5 | 9you | | tx | hangzhou | 6 | tx | | baidu | hangzhou | NULL | NULL | +-------------+----------+----------+-------------+ RIGHT OUTER JOIN把右表记为保留表，得到的结果为： +-------------+----------+----------+-------------+ | customer_id | city | order_id | customer_id | +-------------+----------+----------+-------------+ | 163 | hangzhou | 1 | 163 | | 163 | hangzhou | 2 | 163 | | 9you | shanghai | 3 | 9you | | 9you | shanghai | 4 | 9you | | 9you | shanghai | 5 | 9you | | tx | hangzhou | 6 | tx | | NULL | NULL | 7 | NULL | +-------------+----------+----------+-------------+ FULL OUTER JOIN把左右表都作为保留表，得到的结果为： +-------------+----------+----------+-------------+ | customer_id | city | order_id | customer_id | +-------------+----------+----------+-------------+ | 163 | hangzhou | 1 | 163 | | 163 | hangzhou | 2 | 163 | | 9you | shanghai | 3 | 9you | | 9you | shanghai | 4 | 9you | | 9you | shanghai | 5 | 9you | | tx | hangzhou | 6 | tx | | baidu | hangzhou | NULL | NULL | | NULL | NULL | 7 | NULL | +-------------+----------+----------+-------------+ 添加外部行的工作就是在VT2表的基础上添加保留表中被过滤条件过滤掉的数据，非保留表中的数据被赋予NULL值，最后生成虚拟表VT3。 由于我在准备的测试SQL查询逻辑语句中使用的是LEFT JOIN，过滤掉了以下这条数据： | baidu | hangzhou | NULL | NULL | 现在就把这条数据添加到VT2表中，得到的VT3表如下： +-------------+----------+----------+-------------+ | customer_id | city | order_id | customer_id | +-------------+----------+----------+-------------+ | 163 | hangzhou | 1 | 163 | | 163 | hangzhou | 2 | 163 | | 9you | shanghai | 3 | 9you | | 9you | shanghai | 4 | 9you | | 9you | shanghai | 5 | 9you | | tx | hangzhou | 6 | tx | | baidu | hangzhou | NULL | NULL | +-------------+----------+----------+-------------+ 接下来的操作都会在该VT3表上进行。 执行WHERE过滤 对添加外部行得到的VT3进行WHERE过滤，只有符合<where_condition>的记录才会输出到虚拟表VT4中。当我们执行WHERE a.city = ‘hangzhou‘的时候，就会得到以下内容，并存在虚拟表VT4中： +-------------+----------+----------+-------------+ | customer_id | city | order_id | customer_id | +-------------+----------+----------+-------------+ | 163 | hangzhou | 1 | 163 | | 163 | hangzhou | 2 | 163 | | tx | hangzhou | 6 | tx | | baidu | hangzhou | NULL | NULL | +-------------+----------+----------+-------------+ 但是在使用WHERE子句时，需要注意以下两点： 由于数据还没有分组，因此现在还不能在WHERE过滤器中使用where_condition=MIN(col)这类对分组统计的过滤； 由于还没有进行列的选取操作，因此在SELECT中使用列的别名也是不被允许的，如：SELECT city as c FROM t WHERE c=‘shanghai‘;是不允许出现的。 执行GROUP BY分组 GROU BY子句主要是对使用WHERE子句得到的虚拟表进行分组操作。我们执行测试语句中的GROUP BY a.customer_id，就会得到以下内容(默认只显示组内第一条)： +-------------+----------+----------+-------------+ | customer_id | city | order_id | customer_id | +-------------+----------+----------+-------------+ | 163 | hangzhou | 1 | 163 | | baidu | hangzhou | NULL | NULL | | tx | hangzhou | 6 | tx | +-------------+----------+----------+-------------+ 得到的内容会存入虚拟表VT5中，此时，我们就得到了一个VT5虚拟表，接下来的操作都会在该表上完成。 执行HAVING过滤 HAVING子句主要和GROUP BY子句配合使用，对分组得到的VT5虚拟表进行条件过滤。当我执行测试语句中的HAVING count(b.order_id) < 2时，将得到以下内容： +-------------+----------+----------+-------------+ | customer_id | city | order_id | customer_id | +-------------+----------+----------+-------------+ | baidu | hangzhou | NULL | NULL | | tx | hangzhou | 6 | tx | +-------------+----------+----------+-------------+ 这就是虚拟表VT6。 SELECT列表 现在才会执行到SELECT子句，不要以为SELECT子句被写在第一行，就是第一个被执行的。 我们执行测试语句中的SELECT a.customer_id, COUNT(b.order_id) as total_orders，从虚拟表VT6中选择出我们需要的内容。我们将得到以下内容： +-------------+--------------+ | customer_id | total_orders | +-------------+--------------+ | baidu | 0 | | tx | 1 | +-------------+--------------+ 还没有完，这只是虚拟表VT7。 执行DISTINCT子句 如果在查询中指定了DISTINCT子句，则会创建一张内存临时表（如果内存放不下，就需要存放在硬盘了）。这张临时表的表结构和上一步产生的虚拟表VT7是一样的，不同的是对进行DISTINCT操作的列增加了一个唯一索引，以此来除重复数据。 由于我的测试SQL语句中并没有使用DISTINCT，所以，在该查询中，这一步不会生成一个虚拟表。 执行ORDER BY子句 对虚拟表中的内容按照指定的列进行排序，然后返回一个新的虚拟表，我们执行测试SQL语句中的ORDER BY total_orders DESC，就会得到以下内容： +-------------+--------------+ | customer_id | total_orders | +-------------+--------------+ | tx | 1 | | baidu | 0 | +-------------+--------------+ 可以看到这是对total_orders列进行降序排列的。上述结果会存储在VT8中。 执行LIMIT子句 LIMIT子句从上一步得到的VT8虚拟表中选出从指定位置开始的指定行数据。对于没有应用ORDER BY的LIMIT子句，得到的结果同样是无序的，所以，很多时候，我们都会看到LIMIT子句会和ORDER BY子句一起使用。 MySQL数据库的LIMIT支持如下形式的选择： LIMIT n, m 表示从第n条记录开始选择m条记录。而很多开发人员喜欢使用该语句来解决分页问题。对于小数据，使用LIMIT子句没有任何问题，当数据量非常大的时候，使用LIMIT n, m是非常低效的。因为LIMIT的机制是每次都是从头开始扫描，如果需要从第60万行开始，读取3条数据，就需要先扫描定位到60万行，然后再进行读取，而扫描的过程是一个非常低效的过程。所以，对于大数据处理时，是非常有必要在应用层建立一定的缓存机制（现在的大数据处理，大都使用缓存）
标签：outer union all mit shanghai val 应该 union none number
原文地址：http://www.cnblogs.com/liuwei0824/p/7498048.html
