Apply recipe directly on archived dataset

[heikomuller]

Is your feature request related to a problem? Please describe.
When applying the recipe operations for a sample dataset on the full dataset we currently need to checkout the previous snapshot, modify it and write it back to the archive. This is not only inefficient but can also cause problems for large datasets that do not fit into main memory.

Describe the solution you'd like
The latest version of histore supports applying modification operations directly on the archive (using the DatasetOperator). We should make use of this feature in openclean.