csvlink with different columns

[joernhees]

I have a use-case for csvlink with two files containing different amounts of columns. There is no clear 1:1 mapping of colX from file1 to colY from file2.

Looking at the code i don't understand what the intention of this is:

csvdedupe/csvdedupe/csvlink.py

Lines 91 to 97 in e4ed504

	if self.field_names_1 != self.field_names_2:
	for record_id, record in data_2.items():
	remapped_record = {}
	for new_field, old_field in zip(self.field_names_1,
	self.field_names_2):
	remapped_record[new_field] = record[old_field]
	data_2[record_id] = remapped_record

It looks as if data_2 records are re-written pairwise into the field names of file 1... if the field names have different lengths things are dropped. Is that meant to happen?