【SQLServer】T-SQL查询学习总结（2）[基础知识]

PS:这是是学习微软官方教程 《技术内幕 T-SQL查询》的笔记，样例数据库可以到这里下载。因为是自己作总计笔记之用，查资料的学习的同学注意，很可能对你没有帮助。

以下全文引用的是同一个数据库，客户订单场景，创建代码如下：

SET NOCOUNT ON;
USE tempdb;

IF OBJECT_ID('dbo.Orders') IS NOT NULL DROP TABLE dbo.Orders;
IF OBJECT_ID('dbo.Customers') IS NOT NULL DROP TABLE dbo.Customers;
GO

CREATE TABLE dbo.Customers
(
  customerid  CHAR(5)     NOT NULL PRIMARY KEY,
  city        VARCHAR(10) NOT NULL
);

CREATE TABLE dbo.Orders
(
  orderid    INT     NOT NULL PRIMARY KEY,
  customerid CHAR(5)     NULL REFERENCES Customers(customerid)
);
GO

INSERT INTO dbo.Customers(customerid, city) VALUES('FISSA', 'Madrid');
INSERT INTO dbo.Customers(customerid, city) VALUES('FRNDO', 'Madrid');
INSERT INTO dbo.Customers(customerid, city) VALUES('KRLOS', 'Madrid');
INSERT INTO dbo.Customers(customerid, city) VALUES('MRPHS', 'Zion');

INSERT INTO dbo.Orders(orderid, customerid) VALUES(1, 'FRNDO');
INSERT INTO dbo.Orders(orderid, customerid) VALUES(2, 'FRNDO');
INSERT INTO dbo.Orders(orderid, customerid) VALUES(3, 'KRLOS');
INSERT INTO dbo.Orders(orderid, customerid) VALUES(4, 'KRLOS');
INSERT INTO dbo.Orders(orderid, customerid) VALUES(5, 'KRLOS');
INSERT INTO dbo.Orders(orderid, customerid) VALUES(6, 'MRPHS');
INSERT INTO dbo.Orders(orderid, customerid) VALUES(7, NULL);

SELECT * FROM dbo.Customers;
SELECT * FROM dbo.Orders;

以下直观的看一下数据：

我仅补充我之前不会的知识，学习的童鞋请绕道。

1：什么叫添加外部行

on筛选器过后，把保留表的数据行添加回来，非保留表用null填充，所谓保留表就是left outer join，那么左边的表就是保留表。跟据上一篇，添加外部行这一步骤，在on筛选器后where之前发生，亲，明白点什么了没？

2：having到底用到哪里

干四年码奴，竟然没用到过having，也许是真的对having用处不太了解。having只能和group by联用。having是分好组之后的数据进行筛选，所以必须使用聚合函数，注意如果是count那么不能count(*),只能count(key).做两个实验，暂时只能想到这么多。

场景一：订单数不是一条的客户

USE tempdb
SELECT customerid
FROM Orders
GROUP BY customerid
HAVING COUNT(orderid) <> 1

场景二：最新订单ID是6的客户

USE tempdb
SELECT customerid
FROM Orders
GROUP BY customerid
HAVING MAX(orderid) = 6