Be smart about column names in CSV import

[efstajas]

If there's a header line with column names, we should look for column names matching the example ("receiver", "percentage"), and discard any other columns present. If no header, we should apply the current logic (assume receiver is column 1, percentage is column 2).