杂计

字段名   name  
              a 
              null 
             b
              null  
              a
select * from test group by name
结果为
     a
     null
      b 
想得到的是
      a 
     null
     null
     b 
SQL语句：

select name from test where name is not null group by name
union all
select name from test where name is null;

===================

在某些情况下，or条件可以避免全表扫描的。 1 .where 语句里面如果带有or条件, myisam表能用到索引， innodb不行。1)myisam表： CREATE TABLE IF NOT EXISTS `a` (   `id` int(1) NOT NULL AUTO_INCREMENT,   `uid` int(11) NOT NULL,   `aNum` char(20) DEFAULT NULL,   PRIMARY KEY (`id`),   KEY `uid` (`uid`) ) ENGINE=MyISAM  DEFAULT CHARSET=latin1 AUTO_INCREMENT=6 ; mysql> explain select * from a where id=1 or uid =2; +----+-------------+-------+-------------+---------------+-------------+---------+------+------+---------------------------------------+ | id | select_type | table | type        | possible_keys | key         | key_len | ref  | rows | Extra                                 | +----+-------------+-------+-------------+---------------+-------------+---------+------+------+---------------------------------------+ |  1 | SIMPLE      | a     | index_merge | PRIMARY,uid   | PRIMARY,uid | 4,4     | NULL |    2 | Using union(PRIMARY,uid); Using where | +----+-------------+-------+-------------+---------------+-------------+---------+------+------+---------------------------------------+ 1 row in set (0.00 sec)   2)innodb表： CREATE TABLE IF NOT EXISTS `a` (   `id` int(1) NOT NULL AUTO_INCREMENT,   `uid` int(11) NOT NULL,   `aNum` char(20) DEFAULT NULL,   PRIMARY KEY (`id`),   KEY `uid` (`uid`) ) ENGINE=InnoDB  DEFAULT CHARSET=latin1 AUTO_INCREMENT=6 ; mysql>  explain select * from a where id=1 or uid =2; +----+-------------+-------+------+---------------+------+---------+------+------+-------------+ | id | select_type | table | type | possible_keys | key  | key_len | ref  | rows | Extra       | +----+-------------+-------+------+---------------+------+---------+------+------+-------------+ |  1 | SIMPLE      | a     | ALL  | PRIMARY,uid   | NULL | NULL    | NULL |    5 | Using where | +----+-------------+-------+------+---------------+------+---------+------+------+-------------+ 1 row in set (0.00 sec) 2 .必须所有的or条件都必须是独立索引：+-------+---------------------------------------------------------------------------------------------------------------------- | Table | Create Table +-------+---------------------------------------------------------------------------------------------------------------------- | a     | CREATE TABLE `a` (   `id` int(1) NOT NULL AUTO_INCREMENT,   `uid` int(11) NOT NULL,   `aNum` char(20) DEFAULT NULL,   PRIMARY KEY (`id`) ) ENGINE=MyISAM AUTO_INCREMENT=6 DEFAULT CHARSET=latin1 | +-------+---------------------------------------------------------------------------------------------------------------------- 1 row in set (0.00 sec) explain查看： mysql> explain select * from a where id=1 or uid =2; +----+-------------+-------+------+---------------+------+---------+------+------+-------------+ | id | select_type | table | type | possible_keys | key  | key_len | ref  | rows | Extra       | +----+-------------+-------+------+---------------+------+---------+------+------+-------------+ |  1 | SIMPLE      | a     | ALL  | PRIMARY       | NULL | NULL    | NULL |    5 | Using where | +----+-------------+-------+------+---------------+------+---------+------+------+-------------+ 1 row in set (0.00 sec) 全表扫描了。 3. 用UNION替换OR (适用于索引列)       通常情况下, 用UNION替换WHERE子句中的OR将会起到较好的效果. 对索引列使用OR将造成全表扫描.         注意, 以上规则只针对多个索引列有效. 如果有column没有被索引, 查询效率可能会因为你没有选择OR而降低.        在下面的例子中, LOC_ID 和REGION上都建有索引.        高效:   select loc_id , loc_desc , region from location where loc_id = 10    union    select loc_id , loc_desc , region  from location where region = "melbourne"         低效:  select loc_id , loc desc , region from location where loc_id = 10 or region = "melbourne"     如果你坚持要用OR, 那就需要返回记录最少的索引列写在最前面. 4. 用in来替换or       这是一条简单易记的规则，但是实际的执行效果还须检验，在oracle8i下，两者的执行路径似乎是相同的．　 低效:  select…. from location where loc_id = 10 or loc_id = 20 or loc_id = 30  高效  select… from location where loc_in  in (10,20,30);

转载于:https://my.oschina.net/u/3255899/blog/1499268