SQL 表的自连接 不等值连接

CREATE TABLE T_20161004(
ID INT IDENTITY(1,1) NOT NULL,
uPrice INT,
Name VARCHAR(10)
)
INSERT INTO T_20161004 VALUES(60,‘哈尔滨‘)
INSERT INTO T_20161004 VALUES(70,‘长春‘)
INSERT INTO T_20161004 VALUES(80,‘沈阳‘)
INSERT INTO T_20161004 VALUES(50,‘北京‘)
INSERT INTO T_20161004 VALUES(90,‘郑州‘)
INSERT INTO T_20161004 VALUES(75,‘武汉‘)
INSERT INTO T_20161004 VALUES(80,‘长沙‘)
INSERT INTO T_20161004 VALUES(90,‘广州‘)


-- 方法一 交叉连接  不等值连接 不容易理解
SELECT TOP 1 b.id, b.NAME, SUM(a.uprice) s_sum
FROM T_20161004 a , T_20161004 b 
WHERE a.id<=b.id
GROUP BY b.id,b.name
HAVING SUM(a.uprice)<=500
ORDER BY b.id DESC

-- 方法二 自连接
SELECT TOP 1 b.id,b.NAME,SUM(a.uprice) s_sum
FROM  T_20161004 a 
INNER JOIN T_20161004 b ON a.id<=b.id  --不等值连接
GROUP  BY  b.id,b.name
HAVING SUM(a.uprice)<=500
ORDER BY b.id DESC

-- 方法三 子查询 比较容易理解

SELECT TOP 1 ID, NAME, M_SUM
FROM (SELECT *, (SELECT SUM(uprice) from T_20161004 
		WHERE  ID <= A.ID) AS M_SUM
		FROM T_20161004 A ) B 
WHERE M_SUM<=500
ORDER BY ID DESC 

/* 子查询步骤
1. 外部查询获得一条记录并将其传入内部查询;
2. 基于传入的值进行内部查询
3. 内部查询将自己返回的结果值传给外部查询，外部查询利用这些值完成自己的处理。
*/
SELECT *, (SELECT SUM(uprice) from T_20161004 
		WHERE  ID <= A.ID) AS M_SUM
		FROM T_20161004 A