SQL 服务器 - 根据其他列的值更新一列

Question

Say there are five columns in the table and I'd like to update the column E based on the value of column A,B,C,D. If the corresponding value is Y, then write the column name as part of the value in column E. The expected result is shown in the second image.

Originally, the data in the table looks like this

Expected result:

Answer 1

If you number of columns is fixed, then you can use a CASE statement.

Sample data

create table data
(
  A nvarchar(1),
  B nvarchar(1),
  C nvarchar(1),
  D nvarchar(1),
  E nvarchar(10)
);

insert into data (A, B, C, D) values
('Y', 'N', 'Y', 'Y'),
('N', 'N', 'N', 'Y'),
('N', 'Y', 'Y', 'N'),
('Y', 'Y', 'Y', 'N');

Solution

update d
set d.E = substring(
          case d.A when 'Y' then ',A' else '' end
        + case d.B when 'Y' then ',B' else '' end
        + case d.C when 'Y' then ',C' else '' end
        + case d.D when 'Y' then ',D' else '' end,
         2, 100)
from data d;

Result

select * from data;

A   B   C   D   E
--- --- --- --- -------
Y   N   Y   Y   A,C,D
N   N   N   Y   D
N   Y   Y   N   B,C
Y   Y   Y   N   A,B,C

SQL Fiddle

Answer 2

SQL Server now supports CONCAT_WS() , which I would recommend:

update t
    set E = concat_ws(',',
                      case d.A when 'Y' then 'A' end,
                      case d.B when 'Y' then 'B' end,
                      case d.C when 'Y' then 'C' end,
                      case d.D when 'Y' then 'D' end
                     );

However, it seems silly to store this information when it can be calculated -- and always up-to-date:

alter table t add E as (concat_ws(',',
                                  case d.A when 'Y' then 'A' end,
                                  case d.B when 'Y' then 'B' end,
                                  case d.C when 'Y' then 'C' end,
                                  case d.D when 'Y' then 'D' end
                                 )
                        );

Computed columns are calculated when they are queried. They are always up-to-date and never have to be updated.

Answer 3

update table_name set E='A,C,D' where A='Y' and B='N' and C='Y' and D='Y'

If you set id column for your table it is easy to update