Pipeline.run() argument

[jacob-rosenthal]

The argument of Pipeline.run()should be an object inheriting from BaseSlide, rather than a file path.
This means that whenever we run a pipeline, we can trust that it implements everything from BaseSlide.

If we just pass a path, it may be ambiguous how to read it (is it a H&E slide, or a multiplex slide, or...?). All the work in reading the file, etc. should happen when creating the slide object, not in the pipeline object.

[jacob-rosenthal]

This is obsolete with refactor in #92
run() is now a method of SlideData