output_csv_include_input_csv and ignore_csv_columns don't play nicely together

[mjordan]

When output_csv_include_input_csv and ignore_csv_columns are both present in the config file, the field names that are in the ignore_csv_columns list are included in the output CSV but their values are empty, regardless of whether they are empty in the input CSV. Also, only valid Drupal field names appear in the output CSV. In order to support integration with other systems, non-Drupal field names present in the ignore_csv_columns list should also appear in the output CSV.

[mjordan]

The data in the input_csv_row argument to write_to_output_csv() already has the columns listed in ignore_csv_columns removed so we'll need to figure out a way to access the raw (i.e. prior to those columns being removed) input CSV in order to put them back into the output CSV.

[mjordan]

This feature needs a little attention. For example, should the "base fields" be included in the output CSV? Should the raw values from the input CSV, or the serialized values from the node fields (as is currently the case) be included in the output CSV?