Question

I have a data row with columns A, B, C, D, E, F…
How could I write a sql select to have the result:

Col1, Col2
A,    B
C,    D
E,    F
...

EDIT:
There will be only one row and dynamic number of columns. And I am not allowed to use information_schema and sys tables. Also would be interesting to know how to do it for entire table (multiple A,B,C,D,E.. rows), not the only row.

Answer 1

If you know the list of columns ahead of time, it should be as simple as unioning each group of two columns together:

;with Data as (
    select 'A' as A, 'B' as B, 'C' as C, 'D' as D, 'E' as E, 'F' as F
)
select A,B from Data union
select C,D from Data union
select E,F from Data

If the list of columns is dynamic, you will need Dynamic SQL to process. Here’s a fully-working example using a test table with 11 columns:

create table Test (id01 int,id02 int,id03 int,id04 int,id05 int,id06 int,id07 int,id08 int,id09 int,id10 int,id11 int);
insert Test values (1,2,3,4,5,6,7,8,9,10,11);

declare @select varchar(max)
;with ColumnInfo as (
    select
        row_number() over (order by column_name) - 1 as rownumber,
        column_name
    from information_schema.columns
    where table_name = 'Test'
)
select
    @select = coalesce(@select + ' union ' + char(13) + char(10), '') +
        'select ' + max(case when rownumber % 2 = 0 then column_name end) +
        ',' + coalesce(max(case when rownumber % 2 = 1 then column_name end), 'NULL') +
        ' from test'
from
    ColumnInfo
group by
    rownumber / 2

exec(@select)

Sample Input:

id01 id02 id03 id04 id05 id06 id07 id08 id09 id10 id11
1    2    3    4    5    6    7    8    9    10    11

Sample Output:

id01 id02
1    2
3    4
5    6
7    8
9    10
11   NULL

Live Demonstration: http://sqlfiddle.com/#!3/d0507/1