Importing a CSV file can fail due to unexpected thousands separator
Bug #370571 reported by
Steve Dodier-Lazaro
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
wxBanker |
Fix Released
|
Medium
|
Unassigned |
Bug Description
Hi,
A CVS file can fail to be parsed if for instance there is an occurence of "+1 000" instead of "+1000". The spaces between numbers should be parsed in order to avoid this.
Also, I didn't check but it might be good to check if the currencies are parsed out of the amount's value aswell (for instance, € $ DOL EUR $ , etc).
Finally, i suggest you change the layout of the import form, and use something similar to :
1 | 2 | 3 | 4 | ...
For each column, a list of options (Date/Desc/Amount).
This should be much easier to understand than the current approach.
Cordially, SD.
Related branches
lp:~kolmis/wxbanker/csvimport
(Merged)
Changed in wxbanker: | |
status: | Fix Committed → Fix Released |
To post a comment you must log in.
I guess the right way to solve the first issue is to add a "thousand separator" setting next to the decimal one in the csv import dialog.
As for the second one, removal of non-digit characters on both ends of the column value might be an option, i'll think about that.
About the csv import dialog design: I actually prefer the import dialog mapping wxBanker keys to csv file columns instead of the other way around as you're suggesting. Bank exported csv files will usuallly have 10+ columns, you're mapping to 3 wxBanker values - date, amount, description, and the way it is is easier. The chaining of many csv columns into the description one seems intuitive too. Moneydance (or was that some other java personal banking app ?) has a csv import tool that works the way you would like, you map column 1 to something, column 2 to something, ... etc. I absolutely hated it when trying it out....