I am trying to import from multiple csv files, but KMyMoney does not recognize the duplicates from the second one. Does KMyMoney require the Number / ID field to be present for duplicate detection? CSV import settings: Number: - Date: 1 Description: 6 Note: 4 Amount: 7 Field delimiter: ; Decimal sign: , The CSV (shortened to a single line): "08.03.2013";"08.03.2013";"Lastschrift";"070312478146424161234567890 ELV12345678 07.03 12.47 ME8 ";"Daniel Hahler";"KARSTADT PERFETTO SAGT DANK";"-19,97 ";"1.262,99 " The file is encoded as latin1, and therefore the Euro sign is not displayed properly. Couldn't the other fields (especially "Note") be used to identify duplicates? btw: when I've then tried to re-import the same (second) file again (without loading a new file), it even crashed: Start matching payee Found match with 'VEGANE GESELLS.DEUTSCHL. EV' on 'VEGANE GESELLS.DEUTSCHL. EV' Single matches Adding second split to Spenden(A000079) Looking for a match with transaction: "2013-03-04" , "VEGANE GESELLS.DEUTSCHL. EV" , "(5.00)" (referenced account: "Postbank" ) Considering 5 existing transaction(s) for matching Looking for a match with transaction: "2013-03-04" , "VEGANE GESELLS.DEUTSCHL. EV" , "(5.00)" (referenced account: "Postbank" ) Considering 0 schedule(s) for matching the transaction Processing transactions done (Postbank) KCrash: Application 'kmymoney' crashing... KCrash: Attempting to start /usr/lib/kde4/libexec/drkonqi from kdeinit [1] 8786 alarm ./kmymoney/kmymoney But Dr. Konqi did not came up. Does it need a special setup when running KMyMoney from the build dir of a Git checkout? Reproducible: Always This happened with 4.6.6-1, and then I've tried it on Git master - since 4.7 is supposed to have a revamped CSV importer.
I'm afraid your heading is misleading as the CSV importer does no matching. All it does is to attend to the importing, and after that control is passed to the KMyMoney main app, where all the matching takes place. I've altered the Component from csvimporter to General, but could I ask you alter the heading, perhaps to 'Duplicates not matched', or similar. The mention of csv importer there could result in the correct developer missing the item. Thanks.
Thanks for your feedback! I've changed the title. I am usually using the Aqbanking importer (CSV) and mostly the Aqbanking online module, which has no problems finding duplicates.
It appears to be fixed in Git, despite what I said before - I must have used the 4.6.x version then. I have just tried it again, and hundreds of duplicates where recognized! :) Thanks!