RandomForestRegressor - automatically parse features

I am using the sample notebook (https://github.com/awslabs/amazon-sagemaker-examples/tree/master/sagemaker-python-sdk/scikit_learn_randomforest
Which contains the following line:
parser.add_argument('--features', type=str) # in this script we ask user to explicitly name features
Is there a way not to explicitly define the features? I am using quite long feature names and plenty of them, so it is quite error-prone process to change them.
Is there a way so, it can automatically parse all the features except the target?

Start with the scikit_learn_randomforest sample notebook and the training script line containing parser.add_argument('--features', type=str). Review how the target and feature columns are currently supplied, then determine how all non-target features could be selected automatically. Done means the example supports long feature names without explicitly listing every feature and documents the expected target input.

Written by the indexing model from the issue text.