https://bugs.kde.org/show_bug.cgi?id=362139
--- Comment #10 from allan <agande...@gmail.com> --- (In reply to NSLW from comment #9) > (In reply to allan from comment #7) > > I think I've traced it, possibly, to the recent fee patch. > > In investprocessing.cpp(), at circa line 1723 - > > tr.m_amount = tr.m_amount.abs() - m_trInvestData.fee.abs(); > > appears to drop the sign. > > I'm glad you did it and gave feedback. I apologize if something is broken > for you lately with CSV imports. I don't have any problem, as I was using a test file, not live data. > I would like to help fixing that. Could you > send your problematic, anonymized CSV test file and explain me what result > you expect after importing? > As I understand, you've got two CashDividends: one negative and one > positive. Frankly I don't get it, shouldn't CashDividends be always > positive, as it its you who gets the cash? The test file is one I've used for several years, and originally I obtained it from another user. At that time, I was developing the CSV investment handling and I found the file quite useful, as the data was not the usual straight-forward simple investment transactions. I cannot now remember if the CashDividend with the negative amount was an original entry, or whether I modified it for the purpose. My "justification/rationale" for the entry was that it was documenting a refund of an erroneous earlier transaction. I suppose a "miscexp" would be similar (or perhaps not). The file was from a US broker, who did produce some odd methods in his files. As it's quite small, here it is below. "Trade Date","Settlement Date","Type","Description 1 ","Description 2","Symbol/CUSIP","Quantity","Price ($)","Amount ($)" "","2/24/2010","DividendAndInterest","Div","description","NECZX","","","504.72" "","3/28/2010","Other","Div","description",""NECZX"","","","-504.72" -- You are receiving this mail because: You are watching all bug changes.