如何基于两列值的组合来获取唯一行

Question

I have a table consisting of say following fields with values

STAFFNO NAME DESGN DEPTT SPOUSE_STAFFNO SPOUSE_NAME SPOUSE_DESGN SPOUSE_DEPTT
111111  AAA  AM    HR    999999         PPP         JM           FIN
222222  BBB  DM    MKTG  888888         QQQ         AM           HR
333333  CCC  SM    FIN   777777         RRR         DM           FIN
999999  PPP  JM    FIN   111111         AAA         AM           HR
888888  QQQ  AM    HR    222222         BBB         DM           MTKG
777777  RRR  DM    FIN   333333         CCC         SM           FIN

I need tow rite a SQL query in Oracle to fetch the list of employees along with their spouses, without any repetitions of rows. As can be seen from the values in the table, the record present in the table is for every unique employee, which repeats the entry for the spouses.

I need the output as follows, considering the list of employees so that if the spouses details is present against any employee, the same should not be repeated again in the output considering the spouse as the employee again.

STAFFNO NAME DESGN DEPTT SPOUSE_STAFFNO SPOUSE_NAME SPOUSE_DESGN SPOUSE_DEPTT
111111  AAA  AM    HR    999999         PPP         JM           FIN
222222  BBB  DM    MKTG  888888         QQQ         AM           HR
333333  CCC  SM    FIN   777777         RRR         DM           FIN

Please suggest the SQL code for the desired output. Any help would be highly appreciated.

Answer 1

use corelated subquery

select distinct t1.* from table t1
where not exists ( select 1 from table t2 where t1.STAFFNO=t2.SPOUSE_STAFFNO)

Answer 2

我没有测试它，但你可以尝试这样的事情：

select * from table t where t.STAFFNO = t.SPOUSE_STAFFNO;

Answer 3

Here is one method that works for your data:

select t.*
from t
where t.staffno < t.spouse_staffno;

However, this assumes that all pairs are in the data. Let me assume that is not the case. In that case:

select t.*
from t
where t.staffno < t.spouse_staffno
union all
select t.*
from t
where t.staffno > t.spouse_staffno and
      not exists (select 1
                  from t t2
                  where t2.staffno = t.spouse_staffno and
                        t2.spouse_staffno = t.staffno
                 );

You can also do this with a single statement and window functions:

select t.*
from (select t.*,
             row_number() over (partition by least(staffno t.spouse_staffno), greatest(staffno,  t.spouse_staffno)
                                order by staffno
                               ) as seqnum
      from t
     ) t
where seqnum = 1;

Answer 4

With NOT EXISTS:

select t.* 
from tablename t
where not exists(
  select 1 from tablename 
  where
    STAFFNO < t.STAFFNO
    and 
    STAFFNO = t.SPOUSE_STAFFNO
)

Answer 5

Slightly strange requirement... potentially indicative of a design issue.

with your_data (STAFFNO,NAME,DESGN,DEPTT,SPOUSE_STAFFNO,SPOUSE_NAME,SPOUSE_DESGN,SPOUSE_DEPTT)
as
(select 111111, 'AAA','AM',  'HR',999999,'PPP','JM', 'FIN' from dual union all
 select 222222, 'BBB','DM','MKTG',888888,'QQQ','AM',  'HR' from dual union all
 select 333333, 'CCC','SM', 'FIN',777777,'RRR','DM', 'FIN' from dual union all
 select 999999, 'PPP','JM', 'FIN',111111,'AAA','AM',  'HR' from dual union all
 select 888888, 'QQQ','AM',  'HR',222222,'BBB','DM','MTKG' from dual union all
 select 777777, 'RRR','DM', 'FIN',333333,'CCC','SM', 'FIN' from dual)
select *
from (
select t1.staffno, t1.name, t1.desgn, t1.deptt, t2.staffno spouse_staffno, t2.name spouse_name, t2.desgn spouse_degn, t2.deptt spouse_deptt
,      row_number() over (partition by (coalesce(least(t1.staffno,t2.staffno),t1.staffno)) order by t1.staffno) rn
from   your_data t1
,      your_data t2
where  t2.staffno (+) = t1.spouse_staffno)
where rn = 1;

   STAFFNO NAM DE DEPT SPOUSE_STAFFNO SPO SP SPOU         RN
---------- --- -- ---- -------------- --- -- ---- ----------
    111111 AAA AM HR           999999 PPP JM FIN           1
    222222 BBB DM MKTG         888888 QQQ AM HR            1
    333333 CCC SM FIN          777777 RRR DM FIN           1