What would pipelining look like?

Imagine a repository that is a modified version of the one in Repository API

Then:

my_repo = SubjectRepository('/some/structured/file/system')
my_designmaker = MyDesignParser() # Takes parameters from subject to create design
my_pipeline = Pipeline([
   realignerfactory('fsl'),
   slicetimerfactory('nipy', 'linear'),
   coregisterfactory('fsl', 'flirt'),
   normalizerfactory('spm'),
   filterfactory('nipy', 'smooth', 8),
   designfactory('nipy', my_designmaker),
   ])

my_analysis = SubjectAnalysis(my_repo, subject_pipeline=my_pipeline)
my_analysis.do()
my_analysis.archive()