compre data row by row

• Scott Rankin

  Ten Centuries

  Points: 1217

  More actions

  June 8, 2004 at 9:49 am

  #509512

  Try this.

  1)  Union the two tables together (If they have exactly the same formats... see NOTE: below)

  SELECT * INTO ##Compare FROM Task UNION SELECT * FROM Tracker..Task

  2)  See if there are any Task IDs with different values along the row

  SELECT TaskID, COUNT(*) FROM ##Compare GROUP BY TaskID HAVING COUNT(*) > 1

  Now you can actually pull up any TaskID and see the rows right next to one another for easy identification, if necessary.

  SELECT * FROM ##Compare WHERE TaskID IN (SELECT TASKID FROM ##Compare GROUP BY TaskID HAVING COUNT(*) > 1) ORDER BY TaskID

  3) Now see if any of the task ID in the ##Compare Table don't exist in the original tables.

  NOTE:

  Make any adjustments to the code as necessary to handle inconsistencies in data formats.

• rockmoose

  SSCarpal Tunnel

  Points: 4601

  More actions

  June 8, 2004 at 10:06 am

  #509519

  That was a really nice technique Scott

  Unite And Conquer !

  /rockmoose

  
  You must unlearn what You have learnt

• Scott Rankin

  Ten Centuries

  Points: 1217

  More actions

  June 8, 2004 at 12:50 pm

  #509551

  Sometimes it's more important to get the job done that to find the snazziest way to do it.  Brute force works... most of the time... and in a lot of cases it's faster than learning something new.  Then again we all know if we don't continually learn new things we'll be out of jobs in 3-5 years.

  Thanks for the compliment.

  Scott M. Rankin

  Database Administrator

   

  Coolsavings, Inc

  360 North Michigan Ave.

  Suite 1900

  Chicago, IL <st1ostalCode>60601</st1ostalCode>

  http://www.coolsavings.com

• sqlpro

  Hall of Fame

  Points: 3859

  More actions

  June 8, 2004 at 2:47 pm

  #509567

  Thank u very much all of u guys.

  U r very helpful

  I got fair idea what to do now.