Current behaviour: the CSV file represents null as the 8-character string “<NULL>” . e.g. If I have a table with an integer primary key column id, a varchar column a, and a varchar column b, and the table has one newly-inserted row (1, NULL, “asdf”), the CSV file would look like this:

o “Type”,”idk”,”as”,”at”,”bs”,”b_t”
o “In1”,”1”,”<NULL>”,””,”asdf”,””

Desired behaviour: allow the CSV file to represent null as a completely empty entry. E.g. the CSV file would look like this:

o “Type”,”idk”,”as”,”at”,”bs”,”b_t”
o “In1”,”1” ,, ””,”asdf”,””

Suggested options:

1) Command line switch toggling between “<NULL>” and empty string
2) Allow users to define how to represent null in the UI