[ ] One good thing that could be added with this implementation is to check, after the file scan, if the produced attributes correspond to a valid Dataset.
For example:
Neither "label" nor "records" attributes are missing (None)
All items in the "items" attribute have all the necessary attributes.
etc.
[ ] Another thing that could be added is to define a way of stopping the loop early, this way in some cases we could skip the reading of the whole data. For example, if all attributes where defined before the "itemData" attribute we will be saving a lot of time.
[ ]
One good thing that could be added with this implementation is to check, after the file scan, if the produced attributes correspond to a validDataset.For example:Neither "label" nor "records" attributes are missing (None)All items in the "items" attribute have all the necessary attributes.etc.[ ] Another thing that could be added is to define a way of stopping the loop early, this way in some cases we could skip the reading of the whole data. For example, if all attributes where defined before the "itemData" attribute we will be saving a lot of time.