SELECT列表中的标量子查询

发现了一种表连接新的写法，以前还没有这样写过或者见别人写过。跟同学聊天他们公司却很多人这样写，看来真的要学学sql了

表

CREATE TABLE `t_book` (
  `FId` int(11) NOT NULL,
  `FName` varchar(50) DEFAULT NULL,
  `FYearPublished` int(11) DEFAULT NULL,
  `FCategoryId` int(11) DEFAULT NULL
) ENGINE=InnoDB DEFAULT CHARSET=utf8;

-- ----------------------------
-- Records of t_book
-- ----------------------------
INSERT INTO `t_book` VALUES ('', 'About J2EE', '', '');
INSERT INTO `t_book` VALUES ('', 'Learning Hibernate', '', '');
INSERT INTO `t_book` VALUES ('', 'Two Cites', '', '');
INSERT INTO `t_book` VALUES ('', 'Jane Eyre', '', '');
INSERT INTO `t_book` VALUES ('', 'Oliver Twist', '', '');
INSERT INTO `t_book` VALUES ('', 'History of China', '', '');
INSERT INTO `t_book` VALUES ('', 'History of England', '', '');
INSERT INTO `t_book` VALUES ('', 'History of America', '', '');
INSERT INTO `t_book` VALUES ('', 'History of The World', '', '');
INSERT INTO `t_book` VALUES ('', 'Atom', '', '');
INSERT INTO `t_book` VALUES ('', 'RELATIVITY', '', '');
INSERT INTO `t_book` VALUES ('', 'Computer', '', '');
INSERT INTO `t_book` VALUES ('', 'Astronomy', '', '');
INSERT INTO `t_book` VALUES ('', 'How To Singing', '', '');
INSERT INTO `t_book` VALUES ('', 'DaoDeJing', '', '');
INSERT INTO `t_book` VALUES ('', 'Obedience to Authority', '', '');

CREATE TABLE `t_category` (
  `FId` int(11) NOT NULL,
  `FName` varchar(50) DEFAULT NULL
) ENGINE=InnoDB DEFAULT CHARSET=utf8;

-- ----------------------------
-- Records of t_category
-- ----------------------------
INSERT INTO `t_category` VALUES ('', 'Story');
INSERT INTO `t_category` VALUES ('', 'History');
INSERT INTO `t_category` VALUES ('', 'Theory');
INSERT INTO `t_category` VALUES ('', 'Technology');
INSERT INTO `t_category` VALUES ('', 'Art');
INSERT INTO `t_category` VALUES ('', 'Philosophy');

书表

目录表

任务：查询出每种类目的 类目名称、类目id，最晚的出版年份

一般的想法是，将book表通过类目id进行分组，求出max(出版年份)，再和类目表进行join即可：

select c.*,A.maxPub
    from t_category c
        left join (
        select b.FCategoryId cid ,max(b.fyearpublished) maxPub from t_book b group by b.FCategoryId)A
    on c.fid=A.cid

但是看见新的写法是这样的：

select c.fid,c.FName,(select max(b.FYearPublished) from t_book b where b.FCategoryId=c.FId) maxPub from t_category c

查询出来结果是一样的

这个SELECT语句首先检索FId、FName两个字段，而第三个字段不是一个列二是一个子查询。这个子查询位于主查询的内部，它返回一类图书的最新出版年份。因为聚合函数仅返回一行记录，所以这满足标量子查询的条件。通过WHERE语句，这个子查询也被连接到外部的SELECT查询语句中，因为这个连接，MAX(FYearPublished)将返回每类图书的最新出版年份。

记录一下