最近一位朋友提了这个问题，MySQL中一条SQL执行计划，如下所示，其中有PRIMARY、<derived2>、DERIVED这些内容，他怎么和SQL对应上？

小白学习MySQL - Derived Table_sql

MySQL确实有些和Oracle不同的专业术语，但是背后的原理机制是相通的。

为了说明，模拟创建测试表，

create table t01(
  id int,
  code varchar(10),
  start_date datetime,
  title varchar(10),
  content varchar(30),
  class int,
  end_date datetime
);

插入一些数据，

bisal@mysqldb:  [test]> select * from t01;
+------+--------+---------------------+---------+------------+-------+---------------------+
| id   | code   | start_date          | title   | content    | class | end_date            |
+------+--------+---------------------+---------+------------+-------+---------------------+
|    1 | code1  | 2022-01-01 00:00:00 | title1  | content_1  |     1 | 2022-01-03 00:00:00 |
|    2 | code2  | 2022-01-02 00:00:00 | title2  | content_2  |     3 | 2022-01-03 00:00:00 |
|    3 | code3  | 2022-01-03 00:00:00 | title3  | content_3  |     2 | 2022-01-03 00:00:00 |
|    4 | code4  | 2022-01-04 00:00:00 | title4  | content_4  |     1 | 2022-01-06 00:00:00 |
|    5 | code5  | 2022-01-05 00:00:00 | title5  | content_5  |     1 | 2022-01-07 00:00:00 |
|    6 | code6  | 2022-01-06 00:00:00 | title6  | content_6  |     2 | 2022-01-10 00:00:00 |
|    7 | code7  | 2022-01-07 00:00:00 | title7  | content_7  |     1 | 2022-01-11 00:00:00 |
|    8 | code8  | 2022-01-08 00:00:00 | title8  | content_8  |     1 | 2022-01-12 00:00:00 |
|    9 | code9  | 2022-01-09 00:00:00 | title9  | content_9  |     3 | 2022-01-10 00:00:00 |
|   10 | code10 | 2022-01-10 00:00:00 | title10 | content_10 |     1 | 2022-01-13 00:00:00 |
+------+--------+---------------------+---------+------------+-------+---------------------+
10 rows in set (0.00 sec)

执行的是这条SQL，

bisal@mysqldb:  [test]>
    -> select * from (
    -> (select id, code, start_date, title, content, concat('测试1:', start_date), class, end_date
    -> from t01 t
    -> where id=1 and code='code1' and title='title1' and start_date>='2022-01-01' and start_date<='2022-01-05'
    -> order by end_date desc limit 1)
    -> union all
    -> (select id, code, start_date, title, content, concat('测试2:', start_date), class, end_date
    -> from t01 t
    -> where id=2 and code='code2' and title='title2' and start_date>='2022-01-01' and start_date<='2022-01-05'
    -> order by end_date desc limit 1)
    -> union all
    -> (select id, code, start_date, title, content, concat('测试3:', start_date), class, end_date
    -> from t01 t
    -> where id=3 and code='code3' and title='title3' and start_date>='2022-01-01' and start_date<='2022-01-05'
    -> order by end_date desc limit 1)
    -> union all
    -> (select id, code, start_date, title, content, concat('测试4:', title), class, end_date
    -> from t01 t
    -> where id=4 and code='code4' and title='title4' and start_date>='2022-01-01' and start_date<='2022-01-05' and end_date<='2022-01-05')
    -> union all
    -> (select id, code, start_date, title, content, concat('测试5:', content), class, end_date
    -> from t01 t
    -> where id=5 and code='code5' and title='title5' and start_date>='2022-01-01' and start_date<='2022-01-05'and end_date<='2022-01-05')
    -> ) tt order by class, end_date desc;

他的执行计划，就和文章开始提到的基本一致，

小白学习MySQL - Derived Table_数据库_02

从官方文档中，我们可以了解到，什么是Derived Tables？简单来讲，就是会将FROM子句中出现的检索结果集当做一张表，例如FROM中的SELECT子查询就是一张derived table，而且每张FROM子句中的表都需要一个表别名，任何来自derived table的列必须有唯一的名称，其他要求和示例，可以参考链接，

https://dev.mysql.com/doc/refman/5.7/en/derived-tables.html

A derived table is an expression that generates a table within the scope of a query FROM clause. For example, a subquery in a SELECT statement FROM clause is a derived table:
SELECT … FROM (subquery) [AS] tbl_name 

The [AS] tbl_name clause is mandatory because every table in a FROM clause must have a name. Any columns in the derived table must have unique names

执行计划中的第一行<derived2>这张"表"是全表扫描，

小白学习MySQL - Derived Table_mysql_03

原因就是上述SQL，实际可以理解为，对derived  table的检索，实际上是没有任何检索条件的，

select * from ( ... ) tt order by class, end_date desc;

其实仔细观察上面这个SQL，derived table中union all连接的前三个SQL检索条件基本是一致的，而union all连接的后两个SQL检索条件基本是一致的，只是SELECT中concat内容不同，因此能做改写。

以前三个SQL为例，concat通过case when判断不同的id和title条件下，应该输出的内容，where条件中带上之前所有的字段，改造完这就是独立的一条SQL，不存在子查询，

bisal@mysqldb:  [test]> explain
    -> select id, code, start_date, content,
    ->        (case when id=1 and code='code1' and title='title1' then concat('测试1:', start_date)
    ->              when id=2 and code='code2' and title='title2' then concat('测试2:', start_date)
    ->              when id=3 and code='code3' and title='title3' then concat('测试3:', start_date) end) c,
    ->        class, end_date
    -> from t01
    -> where id in (1, 2, 3) and title in ('title1', 'title2', 'title3')
    -> and start_date>='2022-01-01' and start_date<='2022-01-05'
    -> order by class, end_date desc;
+----+-------------+-------+------------+-------+---------------+------------+---------+------+------+----------+----------------------------------------------------+
| id | select_type | table | partitions | type  | possible_keys | key        | key_len | ref  | rows | filtered | Extra                                              |
+----+-------------+-------+------------+-------+---------------+------------+---------+------+------+----------+----------------------------------------------------+
|  1 | SIMPLE      | t01   | NULL       | range | idx_t01_01    | idx_t01_01 | 5       | NULL |    3 |    10.00 | Using index condition; Using where; Using filesort |
+----+-------------+-------+------------+-------+---------------+------------+---------+------+------+----------+----------------------------------------------------+
1 row in set, 1 warning (0.00 sec)

优化法则之一，就是少做事儿。如果改写成这个，最大的优势，就是同一张表只需要读取一次，而之前derived table中每个union all的子查询都需要读取一次表t01。