日期:2014-05-16 浏览次数:20898 次
mysql中最近用到的函数,记录下
1、IF(expr1,expr2,expr3)与我们常用的三目运算类似。expr1是一个表达式,如果TRUE,返回expr2否则为expr3
如下数据:
INSERT INTO a(id,a,b) VALUES ('1', '1', '1'); INSERT INTO a(id,a,b) VALUES ('2', '1', '0'); INSERT INTO a(id,a,b) VALUES ('3', '1', '0'); INSERT INTO a(id,a,b) VALUES ('4', '1', '0'); INSERT INTO a(id,a,b) VALUES ('5', '0', '0'); INSERT INTO a(id,a,b) VALUES ('6', '0', '1');
比如要查询a的返回状态,1代表是,0代表否有:
SELECT IF(a=1,'是','否') as flag FROM a
有时需要比较两列数据,如同时比较a、b其取值通过(1,1),(1,0),(0,1),(0,0)来统计:
SELECT SUM(IF (a=1 AND b= 1, 1, 0)) as flag1, SUM(IF (a=1 AND b= 0, 1, 0)) as flag2, SUM(IF (a=0 AND b= 1, 1, 0)) as flag3, SUM(IF (a=0 AND b= 0, 1, 0)) as flag4 FROM a
1 3 1 1
INSERT INTO `table2(id, a)` VALUES ('1', '0'); INSERT INTO `table2(id, a)` VALUES ('1', '1'); INSERT INTO `table2(id, a)` VALUES ('2', '0'); INSERT INTO `table2(id, a)` VALUES ('2', '3'); INSERT INTO `table2(id, a)` VALUES ('1', '4');
id a ----------- 1 |0,1,4 2 |0,3
SELECT id, GROUP_CONCAT(a) FROM table2 GROUP BY id;
GROUP_CONCAT([DISTINCT] expr [,expr ...] [ORDER BY {unsigned_integer | col_name | expr} [ASC | DESC] [,col_name ...]] [SEPARATOR str_val])
参考该函数的doc :可以DISTINCT去重, ORDER BY排序,SEPARATOR 来指定分隔符(默认为“,”)如有下面数据
INSERT INTO `table2(id, a)` VALUES ('1', '0'); INSERT INTO `table2(id, a)` VALUES ('1', '1'); INSERT INTO `table2(id, a)` VALUES ('2', '0'); INSERT INTO `table2(id, a)` VALUES ('2', '3'); INSERT INTO `table2(id, a)` VALUES ('1', '4'); INSERT INTO `table2(id, a)` VALUES ('2', '3');
我们需要显示出来的按照a降序、不能重复:
SELECT id, GROUP_CONCAT(DISTINCT a ORDER BY a DESC SEPARATOR '-') FROM table2 GROUP BY id;
这样输出结果:
id a ----------------------- 1 4-1-0 2 3-0
?有了这个函数我们就可以处理一些业务上的事情了,比如现在有两张表其中一张table3的id一对多与另一张表table4的rid关联,现在要统计ipad和mac的具体版本,那么我们就可以直接用sql实现了
-- table3(id, name) INSERT INTO `table3(id, name)` VALUES ('1', 'ipad'); INSERT INTO `table3(id, name)` VALUES ('2', 'mac'); -- table4(id, rid, name) INSERT INTO `table4(id, rid, name)` VALUES ('1', '1', 'ipad1'); INSERT INTO `table4(id, rid, name)` VALUES ('2', '1', 'ipad2'); INSERT INTO `table4(id, rid, name)` VALUES ('3', '1', 'ipad3'); INSERT INTO `table4(id, rid, name)` VALUES ('4', '2', 'pro'); INSERT INTO `table4(id, rid, name)` VALUES ('5', '2', 'air'); INSERT INTO `table4(id, rid, name)` VALUES ('6', '2', 'mini');
SELECT a.id, a.name, GROUP_CONCAT(b.name) as version FROM table3 a JOIN table4 b ON a.id = b.rid GROUP BY a.id; --- id name version 1 ipad ipad1,ipad2,ipad3 2 mac pro,air,mini
?
注意事项:
1、连接的长度受group_concat_max_len参数限制,也就是说这个返回这个长度不是所有都会返回,但是默认为1024也很长了,当然具体可能会到当前concat字段的类型限制同时和max_allowed_packet的限制
2、连接返回二进制和非二进制string,依赖当前连接的类型。有可能超过512个后就返回TEXT或BLOB。如果连接的是int或其他最好先转成Char,如使用函数CAST(expr AS type), CONVERT(expr,type),见CAST文档
-- CAST SELECT CAST(id as CHAR) FROM table4; -- Convert SELECT Convert(id, CHAR) FROM table4;?
?
参考了以下blog
http://blog.sina.com.cn/s/blog_4e808acf01009qna.html