sql - Multiple column condition check with multiple rows to be found -

 id    b   c   d   e(time) --------------------------- 1   j   1     b   1 2   j   1     s   2 3   m   1     b   1 4   m   1     b   2 5   m   2     s   3 6   m   2     s   4 7   t   1     b   1 8   t   2     s   2 9   t   1     b   3 10  k   1     b   1 11  k   1     b   2 

i need find unique values multiple column added condition. unique value combination of col a,b , d.

if col has 2 rows (like record 1 , 2) , column b same on both data , there different value in column d , s coming after b dont want see records

if col has multiple rows (like record 3 6 ) different col b , d, whereas in coulmn d s coming after b dont want see records.

if col has multiple rows (like record 7 9 ) different col b , d, whereas in coulmn d there s before b want see records.

if col has multiple rows (like record 10 11 ) different col b , same column d dont want see records.

any input , able see first , last of using partition , using unbounded in query...

seems basic logic on see if s preceds b on column d show records using partition...

desired output row 7-9: based on logic same column , had sell before buy customer on column d when order column e time.

  id    b   c   d   e(time) --------------------------------------------------- 7   t   1     b   1 8   t   2     s   2 9   t   1     b   3  

i started write query this, ran out of "spare time", criteria hard follow, if comment out "where" @ bottom of query functions doesn't yet produce desired effect.

possibly can lead in direction achive goal ...

with src (     select 1 id, 'j' a, 1 b, 'a' c, 'b' d, 1 e     union select 2, 'j', 1, 'a', 's', 2     union select 3, 'm', 1, 'a', 'b', 1     union select 4, 'm', 1, 'a', 'b', 2     union select 5, 'm', 2, 'a', 's', 3     union select 6, 'm', 2, 'a', 's', 4     union select 7, 't', 1, 'a', 'b', 1     union select 8, 't', 2, 'a', 's', 2     union select 9, 't', 1, 'a', 'b', 3     union select 10, 'k', 1, 'a', 'b', 1     union select 11, 'k', 1, 'a', 'b', 2 ), acnt (     select a, count(*) cnt     src     group ), firstd (     select a, d     src     e=1 ), firstsrow (     select a, min(e) e     src     d='s'     group ), lastbrow (     select a, max(e) e     src     d='b'     group ), mins (     select a, min(d) d, min(b) b     src     group ), maxs (     select a, max(d) d, max(b) b     src     group ) select src.* src     join acnt on acnt.a=src.a     join firstd on firstd.a=src.a     join mins on mins.a=src.a     join maxs on maxs.a=src.a     left join firstsrow on firstsrow.a=src.a     left join lastbrow on lastbrow.a=src.a      not (acnt.cnt=2 , mins.b=maxs.b , mins.d<>maxs.d , firstsrow.e < lastbrow.e)     , not (acnt.cnt>=3 , mins.b<>maxs.b , mins.d<>maxs.d , firstd.d='b')     , (acnt.cnt>=3 , mins.b<>maxs.b , mins.d<>maxs.d , firstd.d='b')