SQL 查询基础(2)-实例

上次说到SQL查询语句的逻辑执行过程，现在来用一个实例说明一下逻辑执行的过程。

前提：我们有三个表，分别记住客户信息、订单信息和产品信息

客户信息表：Customer

ID，Name，Adress，PhoneNumber

ID	Name	Adress	PhoneNumber
1	CompanyA	No.1 Street	123456
2	CompanyB	No.2 Street	23453
3	CompanyC	No.3 Street	45321
4	CompanyD	No.4 Street	4567890
5	CompanyE	No.5 Street	123890

订单信息表：OrderInfo 

ID，Cus_ID，Product_ID，OrderDate，Qua

ID	Cus_ID	Product_ID	OrderDate	Qua
1	2	2	1/1/2012	1000
2	4	1	5/2/2012	500
3	1	5	9/10/2013	3069

产品信息表：Product

ID，Name，Price

ID	Name	Price
1	ABC	10
2	DEF	16
3	GHI	18
4	JKL	9
5	MNO	100

SELECT 
       C.Name AS CustomerName, 
       O.OrderDate AS OrderDate, 
       O.Qua,
       P.Name AS ProductName,
       P.Price AS Price
  FROM Customer C
       INNER JOIN OrderInfo O
            ON C.ID = O.Cus_ID
      INNER JOIN Product P
          ON O.Product_ID = P.ID

现在我们根据上一篇文章中的逻辑执行顺序分析一下查询的结果。

1) FROM Customer 表作为主表 T1

2) JOIN OrderInfo 表作为表 T2，这时将 T1和T2做笛卡尔乘积。得到的虚拟表如下：

ID	Name	Adress	PhoneNumber	ID	Cus_ID	Product_ID	OrderDate	Qua
1	CompanyA	No.1 Street	123456	1	2	2	1/1/2012	1000
2	CompanyB	No.2 Street	23453	1	2	2	1/1/2012	1000
3	CompanyC	No.3 Street	45321	1	2	2	1/1/2012	1000
4	CompanyD	No.4 Street	4567890	1	2	2	1/1/2012	1000
5	CompanyE	No.5 Street	123890	1	2	2	1/1/2012	1000
1	CompanyA	No.1 Street	123456	2	4	1	5/2/2012	500
2	CompanyB	No.2 Street	23453	2	4	1	5/2/2012	500
3	CompanyC	No.3 Street	45321	2	4	1	5/2/2012	500
4	CompanyD	No.4 Street	4567890	2	4	1	5/2/2012	500
5	CompanyE	No.5 Street	123890	2	4	1	5/2/2012	500
1	CompanyA	No.1 Street	123456	3	1	5	9/10/2013	3069
2	CompanyB	No.2 Street	23453	3	1	5	9/10/2013	3069
3	CompanyC	No.3 Street	45321	3	1	5	9/10/2013	3069
4	CompanyD	No.4 Street	4567890	3	1	5	9/10/2013	3069
5	CompanyE	No.5 Street	123890	3	1	5	9/10/2013	3069

 执行 ON 后边的条件：C.ID = O.Cus_ID，保留结果为 TRUE 的记录，结果如下：

ID	Name	Adress	PhoneNumber	ID	Cus_ID	Product_ID	OrderDate	Qua	Result
1	CompanyA	No.1 Street	123456	1	2	2	1/1/2012	1000	FALSE
2	CompanyB	No.2 Street	23453	1	2	2	1/1/2012	1000	TRUE
3	CompanyC	No.3 Street	45321	1	2	2	1/1/2012	1000	FALSE
4	CompanyD	No.4 Street	4567890	1	2	2	1/1/2012	1000	FALSE
5	CompanyE	No.5 Street	123890	1	2	2	1/1/2012	1000	FALSE
1	CompanyA	No.1 Street	123456	2	4	1	5/2/2012	500	FALSE
2	CompanyB	No.2 Street	23453	2	4	1	5/2/2012	500	FALSE
3	CompanyC	No.3 Street	45321	2	4	1	5/2/2012	500	FALSE
4	CompanyD	No.4 Street	4567890	2	4	1	5/2/2012	500	TRUE
5	CompanyE	No.5 Street	123890	2	4	1	5/2/2012	500	FALSE
1	CompanyA	No.1 Street	123456	3	1	5	9/10/2013	3069	TRUE
2	CompanyB	No.2 Street	23453	3	1	5	9/10/2013	3069	FALSE
3	CompanyC	No.3 Street	45321	3	1	5	9/10/2013	3069	FALSE
4	CompanyD	No.4 Street	4567890	3	1	5	9/10/2013	3069	FALSE
5	CompanyE	No.5 Street	123890	3	1	5	9/10/2013	3069	FALSE

由于用的是 INNER JOIN,所以不考虑 T1(T2)中存在而T2(T1)中不存在的情况，继续执行笛卡尔乘积和 ON 条件筛选，得到结果如下表：

ID	Name	Adress	PhoneNumber	ID	Cus_ID	Product_ID	OrderDate	Qua	ID	Name	Price
4	CompanyD	No.4 Street	4567890	2	4	1	5/2/2012	500	1	ABC	10
2	CompanyB	No.2 Street	23453	1	2	2	1/1/2012	1000	2	DEF	16
1	CompanyA	No.1 Street	123456	3	1	5	9/10/2013	3069	5	MNO	100

由于没有用到 WHERE 条件，没有 GROUP BY，没有 HAVING 之类的操作。直接执行 SELECT 中需要选中的列，返回数据查询结果。

CustomerName	OrderDate	Qua	ProductName	Price
CompanyD	5/2/2012	500	ABC	10
CompanyB	1/1/2012	1000	DEF	16
CompanyA	9/10/2013	3069	MNO	100

至此，SQL 的逻辑执行过程处理完毕。