Hydra includes a very simple "tabular" data model which has proven a little too simple for some applications. It was mainly intended for CSVs and TSVs, and conceives of a table as an optional header row plus a list of data rows, each of which is just a list of string-valued cells. Without making the model too much more complex, we can parameterize the type of the cells, and add explicit support for null values. This will serve a much larger set of use cases.
Hydra includes a very simple "tabular" data model which has proven a little too simple for some applications. It was mainly intended for CSVs and TSVs, and conceives of a table as an optional header row plus a list of data rows, each of which is just a list of string-valued cells. Without making the model too much more complex, we can parameterize the type of the cells, and add explicit support for null values. This will serve a much larger set of use cases.