imagej
ij.py.run_script returns the wrong type
ij.py.run_script promises to return a java map, but it returns a ScriptModule instead. Should be a simple change to simply call getOutputs on the return of .get().
@elevans was thinking about just fixing this, but I'd want to write a test - where do you think it should live?
Great catch and definitely something we should fix. ij.py.run_macro will also need the same treatment. I have two commits ready that applies the fixes, clarifies the doc string and also adds test_macro_script.py file to our tests.
My main question is do we need to test all the supported languages?
Yeah, I think all you need to test is the return type from ij.py.run_script from one language - it's pyimagej's job to get the correct outputs from the ScriptService, and it's each scijava/scripting-* module to process a script in a uniform way. I'd think just testing each of the pyimagej code paths is probably sufficient