循环遍历列SQL

Question

I'm looking for a way to loop through the columns of a table to generate an output as described below.

The table looks like that:

ID  Name     OPTION1 OPTION2 OPTION3 OPTION4 OPTION5
1   MyName1  1       0       1       1       0
2   MyName2  0       0       1       0       0

And the output looks like that:

MyName1 -> OPTION1, OPTION3, OPTION4
MyName2 -> OPTION3

Any directions of doing this simply would be greatly appreciated. Otherwise, I suppose I'll have to use a cursor or a temporary table... The database engine is MSSQL. The reason I'm doing formatting at the database level is to feed its output into a limited programmable environment.

Update : the ouput can by in any form, a string or rows of strings.

Update : Would the be a way to accomplish that by building a string using @str = @str + ... ?

Update : I changed the output... this should be easier.

Thanks!

Answer 1

您可能想看一下PIVOT Tables。

Answer 2

Well, in case of a known number of columns, you can do:

SELECT  
  MyName + " ->"
  + case OPTION1 when 1 then ' OPTION1' else '' end
  + case OPTION2 when 1 then ' OPTION2' else '' end
  + ...
FROM
 Table

If columns are unknown when you create the query - I'd probably still go that way with some dynamically created SQL. The advantage is that the code probably does what you wants and is very simple.

Answer 3

You could build a dynamic statement using the system catalog:

http://msdn.microsoft.com/en-us/library/ms189082.aspx

Answer 4

Since you don't go into the specific needs of why you want to be able to do this I can't be certain, but usually when I see this kind of question there are two things that I think of:

1. You need to normalize your database. Maybe "Option1", "Option2" etc. have nothing in common, but there is also a good chance that they are a repeating group within your table.

2. Handle display issues in the display layer of your application - ie the front end, not the database.

As I said, maybe these don't apply in your case for some specific reason, but it seems like it from what I've read of your question.

Answer 5

If using pivot table, you must make sure all of your "Option" columns have the same data type and length.

I would suggest the following answer:

---

IF NOT EXISTS( SELECT TABLE_NAME FROM INFORMATION_SCHEMA.TABLES WHERE TABLE_NAME      
 = 'TABLE1' ) 
create table table1
(
   name nvarchar(50),       
   colvalue nvarchar(50)
)
else
  truncate table table1

declare @table nvarchar(50)
set @table = 'yourtable'

declare @column table
(
   ID integer identity,
   colname nvarchar(20)
)


insert into @column
SELECT c.name FROM sys.tables t 
JOIN sys.columns c ON t.Object_ID = c.Object_ID 
WHERE t.Name = @table 
and c.name in ('Option1','Option2','Option3','Option4','Option5')

declare @minID integer, @maxID integer
declare @cmd nvarchar(max)  
declare @col nvarchar(20)
declare @SQLStr nvarchar(max)

select @minID = MIN(ID), @maxID= MAX(ID)
from @column

while @minID <= @maxID
begin
    select @col = colname
    from @column
    where ID = @minID

    set @SQLStr =    
    'insert into table1 (name, colvalue)
    select name,' + @col + '
    from ' + @table + ' 
    where ' + @col + ' <> 0'    

    exec(@SQLStr)

    set @minID = @minID + 1
end

select distinct name, STUFF(
(SELECT  ',' + a.colvalue  AS [text()]
from Table1  a
where a.name = b.name
Order by a.colvalue
for xml PATH('')),1,1,''    ) AS Comments_Concatenated
from Table1 b
group by name, colvalue
ORDER BY name

---

You just have to modify the @table by putting in your table name and the list of the column you need before insret into @column.

No matter what data type you are, it will working fine.

Answer 6

Edit:

Is there a way to create a MSSQL function to “join” multiple rows from a subquery into a single delimited field?

Search for various solutions to previous questions