Oracle选择行数为偶数/偶数的行数

问题描述 投票：0回答：3

我得到以下结果

KICAE   101000000010    110000001
KICAE   101000000010    210000001
KICAE   101000000020    110000002
KICAE   101000000020    210000002
KICAE   101000000030    110000003
KICAE   101000000030    210000003
KICAE   101000000040    110000004
KICAE   101000000040    210000004

当我运行下面的SQL时：

SELECT 'KICAE','101'||LPAD(ROW_NUMBER() OVER(ORDER BY COURSE_LIST)*10,9,0),COURSE_LIST FROM SYSADM.PS_CRSE_LST_HDR_SF where COURSE_LIST LIKE '1%' 
UNION 
SELECT 'KICAE','101'||LPAD(ROW_NUMBER() OVER(ORDER BY COURSE_LIST)*10,9,0),COURSE_LIST FROM SYSADM.PS_CRSE_LST_HDR_SF where COURSE_LIST LIKE '2%'

我想返回以1开头的课程列表的奇数倍数，以2开头的课程列表的偶数倍数

所需结果：

KICAE   SEQNUM          COURSE_LIST 
KICAE   101000000010    110000001
KICAE   101000000020    210000001
KICAE   101000000030    110000002
KICAE   101000000040    210000002
KICAE   101000000050    110000003
KICAE   101000000060    210000003
KICAE   101000000070    110000004
KICAE   101000000080    210000004
sql oracle select row-number
3个回答
0
投票

这是我的猜测（COURSES子查询在那里产生一些输入，它不是解决方案的一部分）：

with courses(course_list) as (
  select mod(level-1,2)+1 || '100000' || to_char(ceil(level/2),'fm00')
  from dual
  where mod(level-1,2)+1 = 1 or ceil(level/2) <= 39
  connect by level <= 72*2
)
, dr_added(course_list, dr) as (
  select course_list,
  dense_rank() over(partition by substr(course_list,1,1) order by course_list) dr
  from courses
)
select
  101000000000 + dr*20 - case when course_list like '1%' then 10 else 0 end seqnum,
  course_list
from dr_added
order by seqnum;

SEQNUM        COURSE_LIST
101000000010    110000001
101000000020    210000001
101000000030    110000002
101000000040    210000002
101000000050    110000003
101000000060    210000003
101000000070    110000004
101000000080    210000004
101000000090    110000005
101000000100    210000005
101000000110    110000006
101000000120    210000006
101000000130    110000007
101000000140    210000007
101000000150    110000008
101000000160    210000008
101000000170    110000009
101000000180    210000009
101000000190    110000010
101000000200    210000010
101000000210    110000011
101000000220    210000011
101000000230    110000012
101000000240    210000012
101000000250    110000013
101000000260    210000013
101000000270    110000014
101000000280    210000014
101000000290    110000015
101000000300    210000015
101000000310    110000016
101000000320    210000016
101000000330    110000017
101000000340    210000017
101000000350    110000018
101000000360    210000018
101000000370    110000019
101000000380    210000019
101000000390    110000020
101000000400    210000020
101000000410    110000021
101000000420    210000021
101000000430    110000022
101000000440    210000022
101000000450    110000023
101000000460    210000023
101000000470    110000024
101000000480    210000024
101000000490    110000025
101000000500    210000025
101000000510    110000026
101000000520    210000026
101000000530    110000027
101000000540    210000027
101000000550    110000028
101000000560    210000028
101000000570    110000029
101000000580    210000029
101000000590    110000030
101000000600    210000030
101000000610    110000031
101000000620    210000031
101000000630    110000032
101000000640    210000032
101000000650    110000033
101000000660    210000033
101000000670    110000034
101000000680    210000034
101000000690    110000035
101000000700    210000035
101000000710    110000036
101000000720    210000036
101000000730    110000037
101000000740    210000037
101000000750    110000038
101000000760    210000038
101000000770    110000039
101000000780    210000039
101000000790    110000040
101000000810    110000041
101000000830    110000042
101000000850    110000043
101000000870    110000044
101000000890    110000045
101000000910    110000046
101000000930    110000047
101000000950    110000048
101000000970    110000049
101000000990    110000050
101000001010    110000051
101000001030    110000052
101000001050    110000053
101000001070    110000054
101000001090    110000055
101000001110    110000056
101000001130    110000057
101000001150    110000058
101000001170    110000059
101000001190    110000060
101000001210    110000061
101000001230    110000062
101000001250    110000063
101000001270    110000064
101000001290    110000065
101000001310    110000066
101000001330    110000067
101000001350    110000068
101000001370    110000069
101000001390    110000070
101000001410    110000071
101000001430    110000072
0
投票

我建议使用这样的嵌套查询：

SELECT 'KICAE','101'||LPAD(RN*10,9,0) SEQNUM,COURSE_LIST
FROM (
    SELECT ROW_NUMBER() OVER(ORDER BY COURSE_LIST) RN,COURSE_LIST FROM SYSADM.PS_CRSE_LST_HDR_SF where COURSE_LIST LIKE '1%' 
    UNION 
    SELECT ROW_NUMBER() OVER(ORDER BY COURSE_LIST) RN,COURSE_LIST FROM SYSADM.PS_CRSE_LST_HDR_SF where COURSE_LIST LIKE '2%'
)
ORDER BY RN,COURSE_LIST;
0
投票

只需稍微调整一下数学：

SELECT 'KICAE','101'||LPAD(ROW_NUMBER() OVER(ORDER BY COURSE_LIST)*20-10,9,0),COURSE_LIST FROM SYSADM.PS_CRSE_LST_HDR_SF where COURSE_LIST LIKE '1%' 
UNION 
SELECT 'KICAE','101'||LPAD(ROW_NUMBER() OVER(ORDER BY COURSE_LIST)*20,9,0),COURSE_LIST FROM SYSADM.PS_CRSE_LST_HDR_SF where COURSE_LIST LIKE '2%'

我制作了第一个*20-10和第二个*20

..因为您希望rownum 1在第一个中变成10（1 * 20-10），在第二个中变成20（1 * 20），所以rownum 2变成30（2 * 20-10）和40（2 * 20 ）等

您可能还会发现这一作品：

SELECT 'KICAE','101'||LPAD(rownum*10,9,0),COURSE_LIST
FROM SYSADM.PS_CRSE_LST_HDR_SF 
ORDER BY REVERSE(course_list)

如果您的列是数字列，则最好将它们保留为数字列，而不是严格执行此操作：

SELECT 'KICAE',101000000000+rownum*10,COURSE_LIST
FROM SYSADM.PS_CRSE_LST_HDR_SF 
ORDER BY MOD(course_list, 100000000), course_list

而且我也很震惊，您也许可以完全放弃昂贵的排序操作：

SELECT 
   'KICAE',
   101000000000+
      MOD(course_list, 1000000) * 2 - 
      10*((course_list/1000000000)-1),
    COURSE_LIST
FROM SYSADM.PS_CRSE_LST_HDR_SF
最新问题
© www.soinside.com 2019 - 2024. All rights reserved.