SQL Server：从链接到另一个表的ID列动态创建列

Question

I'm trying to dynamically create columns based on an ID that links to another table for column names. (I have no idea if this is even possible it will be used for ssrs reports and gridviews on a webpage.)

Take a look at the example below - the "Current table" is how it looks now. I have a column for every name. I already have the definition table created. I want to make the current table look like the proposed table and "CREATE" the columns in a select or something using the definition table for names to look like it does now.

Answer 1

Here is one approach that will "dynamically" unpivot your data without actually using dynamic SQL.

Clearly Gordon's would be more performant, but here, you don't have to declare the columns.

Example

Select  D.id
       ,[count] = C.Value
       ,A.ts
 From  CurrentTable A
 Cross Apply ( values (cast((Select A.* for XML RAW) as xml))) B(XMLData)
 Cross Apply (
                Select Item = a.value('local-name(.)','varchar(100)')
                      ,Value = a.value('.','int')    -- Use appropriate type
                 From  B.XMLData.nodes('/row')  as C1(n)
                 Cross Apply C1.n.nodes('./@*') as C2(a)
                 Where a.value('local-name(.)','varchar(100)') not in ('ts','OtherColToExclude')
             ) C
 Join   definition D on C.item=D.Name

Returns

id  count   ts
1   5       2018-12-06
2   88      2018-11-01
2   18      2018-10-18
3   32      2018-10-27
4   6       2018-10-27

Answer 2

This is a very strange data structure. Here is one method:

select d.id as definition,
       coalesce(c.p1, c.p2, c.p3, c.p4) as "count",
       c.ts
from current_table c join
     definition d
     on (d.name = 'p1' and p1 is not null) or
        (d.name = 'p2' and p2 is not null) or
        (d.name = 'p3' and p3 is not null) or
        (d.name = 'p4' and p4 is not null) ;

Answer 3

Also you can try to use unpivot to achieve your requirement. For more details, you can refer to this article: https://docs.microsoft.com/en-us/sql/t-sql/queries/from-using-pivot-and-unpivot?view=sql-server-2017

----drop table [current_table] 
----drop table [definition] 
create table [current_table] 
(p1 int,
p2 int,
p3 int,
p4 int,
ts date )
insert into [current_table] values
(5,null,null,null,'20181206'),
(null,88,null,null,'20181101'),
(null,18,null,null,'20181018'),
(null,null,32,null,'20181027'),
(null,null,null,6,'20181014')

create table [definition] 
(id int,
name varchar(5) )
insert into [definition] values 
(1,'p1'),(2,'p2'),
(3,'p3'),(4,'p4')

;with cte as (
SELECT [name] ,[count],ts 
FROM [current_table]  t
UNPIVOT ([count] FOR [name] IN ([p1],[p2],[p3],[p4])) u)
select b.id,a.count,a.ts from  cte a  join [definition] b on a.name=b.name
/*
id          count       ts
----------- ----------- ----------
1           5           2018-12-06
2           88          2018-11-01
2           18          2018-10-18
3           32          2018-10-27
4           6           2018-10-14
*/


------dynamic-------------
DECLARE @ColNames varchar(1000);
SET @ColNames = '';
SELECT @ColNames =stuff((
    SELECT DISTINCT ',' + QUOTENAME(COLUMN_NAME)
    FROM INFORMATION_SCHEMA.COLUMNS p2
    WHERE TABLE_NAME = 'current_table'
      AND COLUMN_NAME like 'p%'
    FOR XML PATH(''), TYPE).value('.', 'varchar(max)')
            ,1,1,'')
DECLARE @CMD nvarchar(2000);
SET @CMD = ';with cte as (
SELECT [name] ,[count],ts 
FROM [current_table]  t
UNPIVOT ([count] FOR [name] IN ('+@ColNames+')) u)
select b.id,a.count,a.ts from  cte a  join [definition] b on a.name=b.name'
PRINT @CMD
execute sp_executesql @CMD
/*
;with cte as (
SELECT [name] ,[count],ts 
FROM [current_table]  t
UNPIVOT ([count] FOR [name] IN ([p1],[p2],[p3],[p4])) u)
select b.id,a.count,a.ts from  cte a  join [definition] b on a.name=b.name
id          count       ts
----------- ----------- ----------
1           5           2018-12-06
2           88          2018-11-01
2           18          2018-10-18
3           32          2018-10-27
4           6           2018-10-14
*/

Hope it can help you.

Best Regards,

Rachel