Question

So I screwed up and filled a table with duplicate rows… I have a table something like this (actually THOUSANDS of somethings like this 🙁 ) :

id    gname     pname    value 
1     15         13       test 
2     15         13       test 
3     15         13       test
4     18         18       test
5     18         18       test 
6     18         18       test

---

This table is only supposed to allow one row of the same gname, pname and value.

So now I need to write some sql that will group all these common rows together and delete them all except ONE! Argh!

Can anyone help me, STEP ONE, I believe is just to get a statement to return all the groups of duplicates.

** BTW – I know this is totally jacked. This is a very old system that I inherited and a big client this is using this… so I’m aware of the issues. I can fix the code that caused this ONCE I get the db back to the way its supposed to be.

Thanks!

Answer 1

If id is unique and auto-incrementing this is easy:

DELETE Table
WHERE ID NOT IN (SELECT MIN(id)
                  FROM Table
                  GROUP BY gname, pname, value)

Do a SELECT first, though.