Context
I built a quite long pipeline with lots of transformer and too many lines of code for it to be practical in a single file.
Following the documentation, I managed to have the Database in a separated DB.py file, but the rest of the pipeline: Transformers and BioAlgorithm are in a single file. I currently run my pipeline with bob bio pipeline simple DB.py pipeline.py.
Goal
I would like to be able to run the whole pipeline, with multiple files (one DB.py for the Database, one per Transformer and one for the BioAlgorithm bioAlgo.py), with one file pipeline.py instantiating everything, creating the pipeline and defining the configuration (log levels,..). I would imagine that such pipeline could be run with bob bio pipeline simple pipeline.py db.py transformer{1,2,..} bioAlgo.py.
Failed attempt
I tried to isolate my custom BioAlgorithm class (also tried with one transformer), import it and instantiate it in pipeline.py but it does not seems to work the way python import usually do..
Could you please provide some insight or example on how to achieve it (that could maybe added to the documentation later on)?
