Background: We're working on a spark pipeline that generates models from data. Standard datasciency stuff. We're going to be using these models in a variety of places, and doing A/B testing to ensure parity with older models. I'd like to add data to all of our outputted models that tells us which version of the code generated the model, but rather than just having a version, I want to include unique information about the .jar. Compilation time, computer it was compiled on, username, etc etc. Lets us do A/B testing on stuff that was compiled by devs outside of our release process as well.
Is there a good way in SBT to compute that stuff at compile time and embed it into the program?