CODEWITHSUNDEEP
sql-servert-sqlsql-server-2008
prav
prav

Reputation: 450

how to convert column values into comma seperated row vlaues

I have a table with values as 

FKTABLE_NAME               FKCOLUMN_NAME               PKCOLUMN_NAME
table1                     column1                        column1
table1                     column2                        column2
table2                     column1                        column1
table2                     column2                        column2

How I need to convert this into

FKTABLE_NAME               FKCOLUMN_NAME               PKCOLUMN_NAME

tablel1                    column1,column2                column1,column2
table12                    column1,column2                column1,column2

Basically, I am trying to get the comma seperated columns group by the table name.

thanks

Upvotes: 0

Views: 2722

Answers (4)

Vikash Sinha
Vikash Sinha

Reputation: 193

Below SQL statement is used for converting single column data in a single row:

declare @testdata varchar(8000)
set @testdata=''

select @testdata=@testdata+NAME+', ' from User where ID in('1','2','3','4','5')
print @testdata

select substring(@testdata,0,len(@testdata))

In above the above example:

  • User is a table
  • NAME is column of user which data will be manuplate in row
  • ID is column of user

For reference see below link:

http://www.devdescent.com/2012/02/column-data-in-to-single-row-with-comma.html

Upvotes: 0

luckyluke
luckyluke

Reputation: 1563

As with all string aggregation problems I advise writing a custom aggregate (and I STRONGLY advise using a clr aggregate for best performance) and write a query like this

select fktablename, stringAggregate(fkcolumnname), stringAggregate(pkcolumnname)
group by 
fktablename

some tips here

other nice reading

Upvotes: 0

RichardTheKiwi
RichardTheKiwi

Reputation: 107736

Here's a working query on any db

select distinct table_name,
  stuff((select ','+data_type
   from information_schema.columns b
   where b.table_name=a.table_name
   for xml path(''),type).value('.[1]','nvarchar(max)'),1,1,'') AS data_types,
  stuff((select ','+column_name
   from information_schema.columns b
   where b.table_name=a.table_name
   for xml path(''),type).value('.[1]','nvarchar(max)'),1,1,'') AS column_names
from information_schema.columns a

And here is your query

select distinct FKTABLE_NAME,
  stuff((select ','+FKCOLUMN_NAME
   from tbl b
   where b.FKTABLE_NAME=a.FKTABLE_NAME
   for xml path(''),type).value('.[1]','nvarchar(max)'),1,1,'') AS FKCOLUMN_NAMES,
  stuff((select ','+PKCOLUMN_NAME
   from tbl b
   where b.FKTABLE_NAME=a.FKTABLE_NAME
   for xml path(''),type).value('.[1]','nvarchar(max)'),1,1,'') AS PKCOLUMN_NAMES
from tbl a

Upvotes: 3

HuBeZa
HuBeZa

Reputation: 4761

SELECT [Col1Name] + ',' + [Col2Name] + ',' ... + [ColNName]
FROM [tableName]

Upvotes: 0

Related Questions