We help IT Professionals succeed at work.

Access Query Delete with Inner Join

kg6lfz
kg6lfz asked
on
109 Views
Last Modified: 2018-11-08
ACCESS 2013 Query
I want to delete records from table 1 with inner join.  My query is

Delete Table1.* from Table1
Inner Join Table2 on Table1.invoice=Table2.Invoice

The above query works but it deletes records from Table2.  I switched Table1 with Table2 but it still deletes records from Table2.  How can I delete records from Table1?

Thank you very much in advance.
Comment
Watch Question

Senior Application Developer
CERTIFIED EXPERT
Commented:
This problem has been solved!
(Unlock this solution with a 7-day Free Trial)
UNLOCK SOLUTION
CERTIFIED EXPERT
Distinguished Expert 2017

Commented:
Paul's suggestion of a subquery should solve the problem so give him the points.

When you join multiple tables in a query and the tables are related hierarchically and then delete, you ALWAYS ONLY delete rows from the lowest level child table.  That means if your join is tblCustomer==> tblOrders==>tblOrderDetails  -- the delete would delete only from tblOrderDetails.   However, if the tables are joined PK to PK, you will delete from both tables.  But if you join non-pk to non-pk, you can't delete at all.    Deletes in queries with joins are very confusing and the rules may vary from RDBMS to RDBMS.  Experiment on tables with junk data that allows you to easily identify what will happen.

I disagree on Paul's advice regarding naming.  What are you going to call Invoice in the third table?  If you prefer to specifically identify foreign keys, that's fine but in that case Invoice would be FKInvoice in all tables where it appears.
Paul Cook-GilesSenior Application Developer
CERTIFIED EXPERT

Commented:
Pat, thanks for the info about hierarchic priority in deletion...  I didn't know that.  :)

And Pat is correct;  if the same data is a foreign key in multiple tables, I would call it FKInvoice in all of the tables.
CERTIFIED EXPERT
Distinguished Expert 2017

Commented:
I use "ID" as the suffix for ALL autonumbers and "CD" for user assigned unique IDs.  Some primary keys that are common may have no suffix.  State is one example since the common use of State is the two digit code rather than the long name.  The PK names relate to the table they control and are NEVER named simply ID.  The foreign keys always use the name of the PK unless there are multiple relationships or a self referencing situation such as SupervisorID in the Employee table.  EmployeeID has to be the PK of the individual record but SupervisorID points to the EmployeeID in a different row.  in other cases, we may end up two references to state from the same record as in BillingST and ShippingST.  These are the exception though rather than the rule since self-referencing and multiple references just don't occur very often.

The bottom line is - pick a standard and stick with it.
Paul Cook-GilesSenior Application Developer
CERTIFIED EXPERT

Commented:
Glad we could help!  :)