今天在写一个sql 的时候可能是写的少的原因 花了好长的时间
现在吧代码贴出来 大家一起看看
select team.name tname,sum(c.none) as none ,sum(c.one) as one,sum(c.two) as two,sum(c.three) as three,sum(c.four) as four,sum(c.five) as five from (
select * from
(select hometeamid tid,
sum(case when FirstHomeScores + SecondHomeScores = 0 then 1 else 0 end) none,
sum(case when FirstHomeScores + SecondHomeScores = 1 then 1 else 0 end) one,
sum(case when FirstHomeScores + SecondHomeScores = 2 then 1 else 0 end) two,
sum(case when FirstHomeScores + SecondHomeScores = 3 then 1 else 0 end) three,
sum(case when FirstHomeScores + SecondHomeScores = 4 then 1 else 0 end) four,
sum(case when FirstHomeScores + SecondHomeScores >= 5 then 1 else 0 end) five
from match where liansaiid=3 group by hometeamid )as a
union
select * from
(select awayteamid as tid,
sum(case when FirstAwayScores + SecondAwayScores = 0 then 1 else 0 end) none,
sum(case when FirstAwayScores + SecondAwayScores = 1 then 1 else 0 end) one,
sum(case when FirstAwayScores + SecondAwayScores = 2 then 1 else 0 end) two,
sum(case when FirstAwayScores + SecondAwayScores = 3 then 1 else 0 end) three,
sum(case when FirstAwayScores + SecondAwayScores = 4 then 1 else 0 end) four,
sum(case when FirstAwayScores + SecondAwayScores >= 5 then 1 else 0 end) five
from match where liansaiid=3 group by awayteamid ) as b
)as c,team
where c.tid=team.teamid
group by team.name
首先这里使用了case when 条件 then 真值 else 假值 语法 。
然后这里采用了一个union 连接两张结构一样的表,问题就处在这里。
原以为两张表经过union之后变成了一张表,可以对这样表进行操作了。可是死活就是编译不过去。后来实在没办法 就想到了 ,把两张表合并在一起以后进行一次查询。这样就获得了一张可以操作的表,并且
这张表可以使用 前面的别名。group 不要中是否能使用 还没试过。