Open joosthooz opened 2 years ago
We should create an intersection of the sets of datatypes that both parquet and csv support. Right now the datageneration function in the unit tests outputs data for one of them. If we have data for both, we can create a matrix of the conversion parameters (format, partitioning, compression), and automatically test all conversions between them.
Since due to #40, users' primary integration with the package is now stdout, we should add tests that capture stdout and ensure it's parseable
Right now there's only a very small amount of unit tests, this needs to be expanded