Completeness check fails on empty data

[cr-assafbendavid]

When the exported data set is empty, the completeness check fails.

This happens because if the query results in an empty table, adding the make_col + statsby has no effect, resulting still in an empty table. Then, when the result is read (either using pd.read_csv or pd.read_json, the resulting DataFrame is empty.

This can be addressed easily when getting the expected_number_of_lines, by replacing the [0] with a .get(0, 0), i.e. interpreting an empty result as 0 expected lines.