Hardcoded column counts and column drop

[OsakaStarbux]

Hi. I think some changes with degiro's csv output is breaking the hardcoded column counts and might be a problem for dropping unused columns too. As a workaround I just hardcoded 12 columns as per the csv format I downloaded.

[ERROR] Input data column count incorect! expected: 19

[Gaunah]

Hi,
thanks for the heads up.
Can you share the header of your "Transactions.cvs", so I can see what needs to be done.

My header still has 19 columns:
Datum,Uhrzeit,Produkt,ISIN,Referenzbörse,Ausführungsort,Anzahl,Kurs,,Wert in Lokalwährung,,Wert,,Wechselkurs,Transaktionskosten,,Gesamt,,Order-ID

[OsakaStarbux]

Perhaps it varies by country. I'm using the UK site. By the way, there is also an option to hide Money Market Fund transactions ("Hide cash movements" toggle) in the Degiro UI. Anyway, I get the following:
12 columns, 10 headers
Date (eg. 31-12-2018 )
Time (eg. 08:42 )
Value date (eg. 31-12-2018 )
Product (eg. TESLA MOTORS INC. - C)
ISIN (eg. US88160R1014)
Description (eg. Buy 10 TESLA MOTORS INC. - C@305 USD (US88160R1014))
FX (eg. 1.2621)
Change (2 columns, eg. USD | 300.00 )
Balance (2 columns, eg. USD | 300.00 )
Order Id (eg. 270b3e5e-b771-5d34-1n0t-r3a16fc60967a)

I will try to look at your code if I have time. I don't know Pandas very well but I imagine there's a neat solution because the Tradervue format is so simple and they allow a different column order if headers are present.

[Gaunah]

Okay, yes.
We just would have to define different parsing functions for the different input formats and choose these accordingly.

Would be nice to know if there are even more different formats and how to distinguish between them.

For now, it would just be parseInputDE(inputData) and pasreInputUK(inputData) or something along the lines.

[OsakaStarbux]

Or ask the user which columns to select with args? So few are used and they have the file. I emailed DEGIRO asking about exports. I’ll let you know.

[Gaunah]

I would prefer to automate it instead of having the user to think through it and fiddle with it.