I notice that it seems possible, both in mymoneystatementreader.cpp and in mymoneyqifreader.cpp, to create an account with the same name and type as an existing account. > It seems wrong. Are you sure it's not prevented by the engine? > OTOH, perhaps it's allowed only for imports, in order to allow an import > from apps with looser rules. > > Regards, > Alvaro > No, and mymoneystatementreader.cpp is used not just for imports. I can create a duplicate from the UI, too. I can then enter a transfer transaction between them, so they both seem functional. Also, what seems strange is that the new account has as default opening date of 6 April 2014, the start of the next fiscal year here. In addition, the transfer I entered to the new account, was accepted with today's date, before the opening date, although it couldn't then be edited. Reproducible: Always Steps to Reproduce: 1.Enter a new account with the same name and type as an existing account. Actual Results: The new account seems fully functional, once the default opening date is changed to something more sensible. Expected Results: A duplicate account name should be detected.
Yes, a duplicate account can be created by design. The false opening date is a problem. I agree, that if the account is created using the UI the user should be warned that an account exists somehow.
A duplicate (Brokerage) account is created if the QIF import contains an account by the same name. In this case both accounts are created during the QIF processing. I think the import code should detect that the QIF is already doing the security/brokerage split thing and not create a new (Brokerage) account.