create_collection.py

def create_collection(rs):
  # need to do: create an integration on the Rockset console
  # Below: After, we can create a collection from the integration
  integration = rs.Integration.retrieve('s3spark-rockset')
  csv = rs.Source.csv_params(
    separator=',',
    encoding='UTF-8',
    first_line_as_column_names=True,
    column_names=None,
    column_types=None
    )

  s3 = rs.Source.s3(bucket='spark-rockset-public-nadine', prefix='OutputMovies.csv/', integration=integration, format_params=csv)
  newcoll = rs.Collection.create(name='s3sparkPythonMoviesColTrue', sources=[s3])
  # note we would want to poll the collection until it's `READY`
  return rs