While finding the diff, most often the first few columns should be ignored because they contain timestamps which certainly will differ across different runs.

This makes the diffs meaningless.
Add an option that takes either number_of_columns_to_skip or if its too tedious to count manually, takes a string and uses its length to skip from left. Or maybe think of a better UX