Problem solve Get help with specific problems with your technologies, process and projects.

Stringing together columns with UPDATE SQL

Is it possible in T-SQL or PL/SQL to string together the values in a column grouped by another column?

I have a Budgets table with two columns:

BudgetID     BudgetPath
   1           HR
   1           HRIT           
   1           Americas
   2           HR
   2           Staffing
   3           IT
   3           Finance
   3           HR
   3           Asia Pacific

Is it possible in T-SQL or PL/SQL to string together the values in BudgetPath, grouped by BudgetID, so the output would be:

BudgetID     BudgetPath
   1           HR>HRIT>Americas
   2           HR>Staffing
   3           IT>Finance>HR>Asia Pacific

Sadly, there is nothing quite as elegant for T-SQL or PL/SQL as MySQL's wonderful GROUP_CONCAT function. The drop-dead simple solution using GROUP_CONCAT is:

select BudgetID 
     , group_concat(BudgetPath separator '>')
              as BudgetPaths
  from Budgets 
    by BudgetID

For Oracle, have a look at: Displaying multiple records in one row.

For T-SQL, here's a method that will work. First, create a temporary table like this:

create table TempBudgetPaths
( BudgetID    integer      not null primary key 
, BudgetPaths varchar(max) null

Next, populate this table with keys:

  into TempBudgetPaths
     ( BudgetID )
select distinct 
  from Budgets

Finally, update the temporary table like this:

SELECT @row=1, @maxrows=count(*) FROM Budgets
WHILE @row <= @maxrows
  UPDATE TempBudgetPaths
     SET BudgetPaths 
       = COALESCE(t1.BudgetPaths+'>','')
                + t2.BudgetPath
    FROM TempBudgetPaths t1
    JOIN (
         SELECT BudgetID
              , BudgetPath 
              , ROW_NUMBER() OVER 
                  ( ORDER BY BudgetID
                           , BudgetPath ) 
                         as rownum
           FROM Budgets 
         ) AS t2
      ON t2.BudgetID = t1.BudgetID
   WHERE t2.rownum = @row
  SELECT @row = @row + 1
SELECT * FROM TempBudgetPaths

Not nearly as pretty as GROUP_CONCAT, is it. Nevertheless, it works, and produces these results:

BudgetID     BudgetPaths
   1          Americas>HR>HRIT
   2          HR>Staffing
   3          Asia Pacific>Finance>HR>IT

Notice the sequence in which the BudgetPaths were appended. This is a result of the OVER (ORDER BY ...) clause. This is not the same as your sequence, which appears to be some kind of hierarchy. To achieve your sequence will require an additional data column, to indicate sequence of BudgetPath within BudgetID.

This was last published in August 2007

Dig Deeper on Oracle and SQL

Have a question for an expert?

Please add a title for your question

Get answers from a TechTarget expert on whatever's puzzling you.

You will be able to add details on the next page.

Start the conversation

Send me notifications when other members comment.

By submitting you agree to receive email from TechTarget and its partners. If you reside outside of the United States, you consent to having your personal data transferred to and processed in the United States. Privacy

Please create a username to comment.