Question

I’ve got two tables in a database with the following columns:

Table1

athleteId | sportTypeId
 205      |    5
 206      |    5
 206      |    6
 207      |    4
 208      |    1
 208      |    4

Due to an error I made (whoops), the second table has some information like this:

Table2*

athleteId | sportTypeId
  205     |    4
  206     |    4
  207     |    4
  208     |    4
  NULL    |    4

I assigned the wrong athleteId to some athletes, so what I need to do is check every row in Table2 and see if it has a matching row in Table1. The rows containing athleteId 205 and 206 from Table2 would have the athleteId set to NULL (they don’t have a sportTypeId of 4 in Table1). athleteId 207 and 208 would remain (they do have a row in Table1 with sportTypeId of 4).

Notice that some athletes appear twice in Table1.

I’m using SQL Server 2008.

Thanks!

Answer 1

Just to be clear, you’re asking for a way to delete rows from Table2 where there is no matching row in Table1, correct? If so, here’s how to do it:

DELETE t2.*
FROM Table2 AS t2
LEFT JOIN Table1 AS t1 ON t2.athleteId = t1.athleteId AND t2.sportTypeId = t1.sportTypeId
WHERE t1.athleteId IS NULL;

The reason this works is because of the left join and the where clause. The left join will cause t1.athleteId to be null whenever a matching record isn’t found in Table1.

Edit: If you just want to set the athleteId to null instead of deleting the whole row, it’s a small modification:

UPDATE t2
SET t2.athleteId = null
FROM Table2 AS t2
LEFT JOIN Table1 AS t1 ON t2.athleteId = t1.athleteId AND t2.sportTypeId = t1.sportTypeId
WHERE t1.athleteId IS NULL;