At the moment I have a python script to insert, update or delete features from a feature layer. This works well, but I have to upload the feature class containing the new records first to AGOL and publish it as a feature layer. This is because in order to update an existing feature layer, I have to create a deepcopy from the layer containing the new records, like so:
Say my layer (with the new records) is named "updater_layer". I do a query on it, using a where clause :
updater_layer_query = updater_layer.query(where=where_clause)
# I get the features:
update_query_features = updater_layer_query.features
# and create a deepcopy which I can use to update an existing layer
new_features = deepcopy(update_query_features)
But to get those deepcopy records, I first have to upload and publish a new layer with new records to AGOL.
I am wondering if there is a way to do this on the desktop so that I end up with a deepcopy list from the feature class on the desktop, without having to upload the feature class first to AGOL as a feature layer. That deepcopy (derived from the feature class on the desktop) can then directly be used to update an existing feature layer in AGOL.
I looked but it does not seem possible to create a deepcopy from a feature class. Is there a way to do this, perhaps via a JSON representation of the feature class or perhaps by using a searchcursor on the feature class?