I am trying to test locally some pipelines, passing test data as inputs that I create using Python classes because it is easier in this way because data are Spark parquets. So I created a pipeline tester that takes the pipeline as argument and basically runs it giving this fake data as a catalog