1 문제

2 테스트 데이터

3 솔루션

       1열     2열    3열     4열
1행     A       B       C       D
2행     E       F       G       H
3행     I       J       K

        1열     2열    3열     4열
1행                          
2행             A       B       C
3행     D       E       F       G
4행     H       I       J       K

--drop table #temp
create table #temp (
	seq int 
,	열1 varchar(20)
,	열2 varchar(20)
,	열3 varchar(20)
,	열4 varchar(20)
)

insert #temp values(1, 'a', 'b', 'c', 'd')
insert #temp values(2, 'e', 'f', 'g', 'h')
insert #temp values(3, 'i', 'j', 'k', null)

create table #dumy(seq int)
insert #dumy values(1)
insert #dumy values(2)
insert #dumy values(3)
insert #dumy values(4)
insert #dumy values(5)
insert #dumy values(6)
insert #dumy values(7)
insert #dumy values(8)
insert #dumy values(9)
insert #dumy values(10)
insert #dumy values(11)
insert #dumy values(12)
insert #dumy values(13)
insert #dumy values(14)
insert #dumy values(15)
insert #dumy values(16)
insert #dumy values(17)
insert #dumy values(18)
insert #dumy values(19)
insert #dumy values(20)

declare 
	@col_cnt int
,	@set_num int
set @col_cnt = 4
set @set_num = 9 --요기에 시작할 곳 입력

select 
	isnull(min(case when b.cnt1 = a.seq then col end), '') 열1
,	isnull(min(case when b.cnt2 = a.seq then col end), '') 열2
,	isnull(min(case when b.cnt3 = a.seq then col end), '') 열3
,	isnull(min(case when b.cnt4 = a.seq then col end), '') 열4
from (
	select
		b.seq + (@col_cnt * (a.seq - 1)) +  @set_num - 1 seq
	,	case 
			when b.seq = 1 then 열1
			when b.seq = 2 then 열2
			when b.seq = 3 then 열3
			when b.seq = 4 then 열4
		end col
	from #temp a cross join #dumy b
	where b.seq <= @col_cnt) a right join (
					select 
						(b.seq * (@col_cnt)) - @col_cnt + 1 cnt1
					,	(b.seq * (@col_cnt)) - @col_cnt + 2 cnt2
					,	(b.seq * (@col_cnt)) - @col_cnt + 3 cnt3
					,	(b.seq * (@col_cnt)) - @col_cnt + 4 cnt4
					,	a.seq
					from #temp a right join #dumy b
					on a.seq + @set_num/@col_cnt = b.seq
					where (select count(*) from #temp) + @set_num/@col_cnt   >= b.seq) b	
on a.seq in (b.cnt1, b.cnt2, b.cnt3, b.cnt4)
group by b.cnt1

--결과셋을 적절히 Top으로 끊으면 된다.
/*
열1                   열2                   열3                   열4
-------------------- -------------------- -------------------- --------------------
                                                               
                                                               
a                    b                    c                    d
e                    f                    g                    h
i                    j                    k                    
*/