Am Freitag, 26. Januar 2007 18:50 schrieb Ethy H. Brito: > > I would strongly suggest to simply add a numerical comparison (in > > additiona to the string comparison) in import-backend.c, as this > > achieves your original goal of "good scores to imported transactions" > > obviously much much easier. > > I agree and already did that. It is working. > > But somebody (maybe yourself) stated that "perhaps the importer could try > to be more intelligent" and I simply offered to give my 2 cents.
Ok. My personal feeling is that the importer has quite a number of fields to choose from, so for one particular field like e.g. the check number, you shouldn't introduce too much additional complexity *only for this field*. With the current rule-based import matching algorithm, your original issue is solved by a numerical comparison just fine. > If you think this is not necessary (maybe this is a "problem" complained by > myself only) I am satisfied with my (unsent) small patch. My sole intention > was to contribute. ;-) Feel free to submit the patch, even if it seems small to you. More complexity might be a good thing in places where it is really adequate. In this particular case, I think it would divert from the original problem and its easy solution. Thanks for any contribution, Christian _______________________________________________ gnucash-devel mailing list [email protected] https://lists.gnucash.org/mailman/listinfo/gnucash-devel
