Hi, could you please consider the following improvements in CSV import ?
- handling of negative values - E.g. if a row is of type “Interest” positive values should be handled as “Interest” but negative should be handled as “Interest Charge” and similarly for all other types. Many CSVs don’t have a way to set correct type from column that would hint transaction direction but only positive/negative value and it’s never a good thing to just ignore that and use absolute value.
- select currency - If there is no currency column in the CSV then default currency is forced and it’s impossible to import it to account in different currency. It should be possible to select currency for each import template.
- no target account pre-selecting - It would be much safer not to have target account pre-selected from previous import. I have imported to incorrect account so many times and it may be pretty hard to recover from such mistake.
- datetime handling - Currently there are date and time columns, but often CSV contains single column with datetime so it would be nice to be able to map that too (maybe just adding combined pattern to date).
- resizable type mapping window - The window “Mapped to Field” used to map column values to type of the row is too narrow and non-resizable which makes it unnecessarily dificult to use (especially with more complex patterns). I suggest to make the windows resizable, by default much wider and ideally with autoresized columns to make all content visible.
- use the same column for type and note - Quite often information about row type is within the note column so I need to select that column as type and cannot import it as note. It would be better if we can select which column(s) contain type information without blocking of using them for actual import.
- templates showing all columns - When I load template, only columns specified in in are visible and rest of columns are hidden. It would be better if we can always see all columns so that we can reuse templates for CSV with more columns.
- templates using header for mapping - Currently it seems templates store all columns mappings by column index position so adding a new column breaks it. It would be better if it can be (optionally?) mapped according to header name (if there is one).
- multiple note columns - It would be nice to be able to select multiple note columns that would be concatenanted by (customizable) separator.