前幾天發生在羣裏的討論
下面有如下需求
c1 球隊ID
c2 球隊名稱
SQL> with dao as
2 (
3 select 1 c1,’a’ c2 from dual
4 union all
5 select 2 c1,’b’ c2 from dual
6 union all
7 select 3 c1,’c’ c2 from dual
8 )
9 select *
10 from dao
11 /C1 C
———- -
1 a
2 b
3 c
現在需要生成一張對戰表
每個球隊都要與其他球隊對戰一次。
解法1:簡單不等連接
SQL> with dao as
2 (
3 select 1 c1,’a’ c2 from dual
4 union all
5 select 2 c1,’b’ c2 from dual
6 union all
7 select 3 c1,’c’ c2 from dual
8 )
9 select t1.c2||’ PK ‘||t2.c2
10 from dao t1,dao t2
11 where t1.c1 <t2.c1
12 /
T1.C2|
——
a PK b
a PK c
b PK c
解法2:分析函數
SQL> with dao as
2 (
3 select 1 c1,’a’ c2 from dual
4 union all
5 select 2 c1,’b’ c2 from dual
6 union all
7 select 3 c1,’c’ c2 from dual
8 )
9 select res
10 from
11 (
12 select row_number() over ( order by t1.c1+t1.c1) rn ,t1.c2||’ PK ‘||t2.c2 res
13 from dao t1,dao t2
14 where t1.c2 !=t2.c2 ) inner
15 where mod(inner.rn,2) =1
16 /RES
——
a PK b
b PK a
c PK a
解法3:遞歸查詢
SQL> with dao as
2 (
select 1 c1,’a’ c2 from dual
3 4 union all
5 select 2 c1,’b’ c2 from dual
6 union all
7 select 3 c1,’c’ c2 from dual
8 )
9 select cola||’ PK ‘||colb
10 from (
11 select prior c2 cola,c2 colb
12 from dao
13 connect by prior c1=c1+1)
14 where cola is not null
15 ;COLA||
——
b PK a
c PK b
b PK a
從實現方法上來看第一種是最好的方式,簡單,高效。
我們寫程序不一定非得要用某些複雜的特性,
只要能滿足需求,越簡單的,往往意味越高效。
用SQL生成對戰表
發表評論
所有評論
還沒有人評論,想成為第一個評論的人麼? 請在上方評論欄輸入並且點擊發布.