Hackathon ideas for zppy

[forsyth2]

Ideas that would be useful but aren't on the immediate horizon (i.e., ideas that could benefit from a dedicated development time like in a hackathon).

Infrastructure & software architecture:

• Learn about Mocks in unit testing and implement them into zppy. One of the things that makes zppy challenging to test is that it relies on a large number of external dependencies (e.g., "does that directory exist?", "does it have data for the correct component for the correct years"? "does the input actually have the data to run all the sets we want?", "how do we store so much simulation data for testing without having it wiped out in ocassional file system purges?") which must then be tested under a variety of parameter combinations. Add minimal case cfgs #608 added "minimum case" cfg files to try to test a number of parameter combinations, but these are not run automatically (because it would take an unreasonable amount of time to run them all). Improve input validation and testing #628 made zppy code more modular, allowing for true unit testing (rather than integration testing) on functions that don't interact with the outside world (e.g., file systems, these are functions that would be wrapped in the IO monad were the code written in Haskell). Nevertheless, integration testing remains a tedious and lengthy process, hence the need for simulating external dependencies with Mocks.
• Make an independent package for Global Time Series #398 -- The global_time_series task has grown to be outside zppy's true scope of coordinating post-processing tools. It is an analysis package similar to e3sm_diags or mpas_analysis. Indeed Create global time series Viewers #616 will be adding a viewer as can be seen in e3sm_diags. Furthermore, global_time_series is more-or-less a Python application. That means the call structure is python global_time_series.py that calls bash global_time_series.bash that calls Python again: coupled_global.py. This level of language layering increases debugging complexity.

Features:

• Add task for zstash #202 -- It would be nice for users to include long-term archiving (i.e., through zstash) as part of their post-processing workflow