I've written lots of little shell scripts, many of which are simple enough they don't need to be tested, but some of which I've found have broken over time.
Maybe each of the scripts should have a --self-test flag that runs through a set of usage examples? That'd also act as a form of usage documentation?
It could also be useful to have a framework for testing setting up a clean instance of my dotfiles.
I've written lots of little shell scripts, many of which are simple enough they don't need to be tested, but some of which I've found have broken over time.
Maybe each of the scripts should have a
--self-testflag that runs through a set of usage examples? That'd also act as a form of usage documentation?It could also be useful to have a framework for testing setting up a clean instance of my dotfiles.