SQL执行顺序

1.from子句用来组装不同数据源的数据
2,where子句用于基于指定条件的筛选
3.group by 子句用来将数据划分为多个分组

3.1：Distinct 对数据进行去重：注意Distinct 会改变结果集的的名称：所以要根椐

如：
错：select  DISTINCT convert(datetime,Convert(char,modle.lotteryTime,111),111)  from game_draw_date as modle group by  modle.lotteryTime  order by   modle.lotteryTime me  desc

正确：select  DISTINCT (convert(datetime,Convert(char,modle.lotteryTime,111),111)) as tiemKey  from game_draw_date as modle group by  modle.lotteryTime  order by   tiemKey   desc

4,使用聚集函数进行计算（）：聚合函数对一组值执行计算并返回单一的值：如果有分组则对每组进行统计，如果没有分组则全个结果集进行统计（可以理解成：分组后每一个组相当于一个结果集）

count():满足条件的记录数

max():最大值。如果是字符串则按拼音排序
min():
sum():只能对数值类型做计算
avg():

5.使用having子句对分组进行筛选

7.使用order by 对所有结果集进行排序

而你的top 和like 都是属于where语句里面的,所以何时执行,一目了然

 

 

select * from game_draw_date

select * from game_draw_date as model where model.lotteryTime<'2014-12-26 15:00:00.000'

select model.lotteryTime from game_draw_date as model where model.lotteryTime<'2014-12-26 15:00:00.000' group by model.lotteryTime

select DISTINCT convert(datetime,Convert(char,modle.lotteryTime,111),111) as timeKey from game_draw_date as modle
group by modle.lotteryTime order by timeKey desc

select DISTINCT convert(datetime,Convert(char,modle.lotteryTime,111),111) as timeKey , SUM(modle.id) from game_draw_date as modle
group by modle.lotteryTime

select DISTINCT convert(datetime,Convert(char,modle.lotteryTime,111),111) as timeKey , SUM(modle.id) as sumId from game_draw_date as modle
group by modle.lotteryTime  HAVING SUM(modle.id)>10000

select DISTINCT convert(datetime,Convert(char,modle.lotteryTime,111),111) as timeKey , SUM(modle.id) as sumId from game_draw_date as modle
group by modle.lotteryTime HAVING SUM(modle.id)>10000 order by sumId desc