QIF import duplicate question

[James A. de Haseth]

My experience has been similar, but what I am doing is taking the QIF
files for each account individually and running them through a sed
script.  As I have many old account categories (from InCharge) the sed
script has become quite large.  Nonetheless I find that this is a much
more efficient way to assure that the accounts are put in the correct
places.  I also rm ~/.gnucash/qif-accounts-map as I find errors and
bogus accounts created if accounts were ever incorrectly assigned.

I actually make a copy of the "master" empty account tree just by saving
it as "testcase."  I try all my imports into testcase, note all the
errors, fix them, then import the accounts into the active account tree.

It's a pain, but every account I tackle has fewer errors as I have
already seen most of them in previous imports.  I just expand the sed
script as necessary.

The first time I saw the duplicates window was when I imported files
sequentially instead of simultaneously.  The QIF files were identical,
and hence the set-ups were identical, so it's a bit of a mystery.  As
Derek said, the QIF importer is a bit "magic."

Jim

On Thu, 2007-01-25 at 16:35 -0600, Tommy Trussell wrote:

> in my ongoing attempts to get a massive qif to import, I have noticed
> that I DON'T see the duplicates window if I get the setup done right.
> Maybe I've just been lucky, and I've had a lot of crashes so maybe my
> experience isn't typical.
> 
> For example, if I just let the Quicken categories create new accounts
> in GnuCash, not only are they not in the right places in the Assets /
> Equity / Expenses/ Income / Liabilities tree, but the qif importer has
> much less of a clue about matching duplicates.
> 
> One thing I did was spend a lot of time creating the proper hierarchy
> based on my data. At one point I wiped everything out and started
> over, but I copied my qif-accounts-map from the old to the new
> ~/.gnucash directory and it was much faster going through the import
> the next time.
>