How to delete duplicate records (when table doesn't have a primary key column)?

You can use the 'secret' RDB$DB_KEY feature of Firebird. RDB$DB_KEY is a 'hidden' record that identifies rows in each SQL statement. For example, if you have a table TABLE1 without primary key and have columns COL1 and COL2 which are the same, you can delete the extra rows with the following statement:

DELETE FROM TABLE1 t1
WHERE EXISTS (
SELECT 1 FROM TABLE1 t2
WHERE t1.COL1 = t2.COL1 and t1.COL2 = t2.COL2
AND t1.RDB$DB_KEY < t2.RDB$DB_KEY
);

You can control whether the record is considered duplicate by including the relevant table columns in the WHERE clause (in most cases, you will want to include all columns).

Do you find this FAQ incorrect or incomplete? Please e-mail us what needs to be changed. To ensure quality, each change is checked by our editors (and often tested on live Firebird databases), before it enters the main FAQ database. If you desire so, the changes will be credited to your name. To learn more, visit our add content page.

If you are a commercial tool maker and your tool features a great way to handle the issue written about in this FAQ, please check out our advertisement page.