I am trying to create a delete query to remove records from one table, based on whether or not one of the field exists in another master table. The situation is that I am importing new records into a database, but I want to remove the records that have already been imported, i.e. that already have an account in the master table. The field I need to join on, however is not equal: it is prefixed with a constant three letter code XYZ.
tbl_to_import.Account master_table.Account
123456 XYZ.123456
345678 XYZ.345678
To avoid using a join in the delete query I tried the following:
Delete tbl_to_import.*
From tbl_to_import
Where Exists( Select master_table.Account From master_table
Where master_table.Account = ("XYZ."& tbl_to_import.Account) ) = True;
However, the query gets hung up in Access. I’m not sure what I’m doing incorrectly. I don’t get an error message, but the query runs without producing anything and I eventually stop it. In this situation, tbl_to_import has 2,700 records and master_table has 50,000 records. Additionally, I am connecting to the master_table via ODBC.
Originally, I constructed two queries using a join to perform the delete. tbl_to_import.Account has a primary key called ID. One query, qry_find_existing_accounts, located the ID numbers in tbl_to_import for which there exists a corresponding account in master_table.Account:
SELECT DISTINCTROW tbl_to_import.ID AS DELETEID
FROM tbl_to_import LEFT JOIN master_table
ON ("XYZ."& tbl_to_import.Account) = master_table.Account
WHERE ((("XYZ." & [Account])=[master_table].[Account]));
Then I used this query to construct the delete query:
DELETE DISTINCTROW tbl_to_import.*, tbl_to_import.ID
FROM tbl_to_import RIGHT JOIN qry_find_existing_accounts
ON tbl_to_import.ID =qry_find_existing_accounts.DELETEID
WHERE (((tbl_to_import.ID)=[qry_find_existing_accounts].[DELETEID]));
The query qry_find_existing_accounts worked fine; however, when I tried to run the second query to delete, I got the error: Could not delete from specified tables. Typically, when I get this error, it is because I have not selected unique records only, however, I used DISTINCTROW in both queries.
Any ideas what I am doing wrong and how I can accomplish what I need to do?
I’d go with a simpler nested SQL statement:
This should be fairly fast, especially if your Account fields are indexed.