On Mittwoch, 12. Oktober 2022 22:22:52 CEST Jack via KMyMoney-devel wrote:
> On 2022.10.12 16:10, Thomas Baumgart via KMyMoney-devel wrote: > > On Dienstag, 11. Oktober 2022 21:53:01 CEST Jack via KMyMoney-devel > > wrote: > > > > > I've been working with my contact at Merrill Lynch to try to at > > least > > > be able to talk with someone with some technical knowledge, and I > > just > > > forwarded her a (redacted) copy of the ofxlog.txt from my latest > > > attempt to download data - which "succeeds" but does not contain any > > > transaction data, just stock position data. However, in looking at > > the > > > file, I see a problem in the line > > > > > > <DTSTART>546361101 > > > > > > Running a 5.1.3 appimage, that line is > > > > > > <DTSTART>20220925 > > > > > > and that version actually does still download transactions from my > > > Merril Lynch accounts. > > > > The ofxlog.txt file contains the data as received from the bank. If > > it shows these weird values, then it is a problem introduced by the > > sender. Does the request part show a date range and is that valid? > > Is there a difference in the logs (esp. the request part) between > > 5.1.3 and the version you use to get those weird results? > That date is in the send part, not what comes back from the bank. I've > cleared out my old logs, but will generate another one and send it to > you privately. > > > > > 546361101 seems to translate to 1987-04-25 although I have no idea > > > where that date would have come from. I also don't see any code > > that > > > would use that form of the date without converting to ISODate. I've > > > looked at the git log for > > > kmymoney/plugins/ofx/import/dialogs/mymoneyofxconnector.cpp and > > don't > > > see anything related to any lines dealing with dtstart, although I > > may > > > well have misread that log and/or be looking at the wrong file. > > > > That data is parsed by libofx so you cannot find anything in KMyMoney > > other than the code you find in OFXImporter::ofxStatementCallback but > > that uses the data extracted by libofx. > Is that also true of the request part at the top of the log? > > > > > Any thoughts? With the recent discussions about the Open Banking > > > related changes, I don't want to lose access to accounts that still > > > work with direct connect. > > > > I do understand that, but we can't do anything when the bank sends > > that kind of data. Ooops, that is in the request. Thought it to show up in the response. My bad. Anyway, fixed with https://invent.kde.org/office/kmymoney/commit/af8d7fc7a6a334e0e2cae85616b1d7508e3d1c5c which also shows what the problem was which was created around the end of August in https://invent.kde.org/office/kmymoney/commit/bc9f4d95f1c40cd51cfff36076640048b5d4273f -- Regards Thomas Baumgart ------------------------------------------------------------- Windows: It's not pretty. It's not ugly. But it's pretty ugly. -------------------------------------------------------------
signature.asc
Description: This is a digitally signed message part.
