Hi all,
I need some clarification on the scope and scale of the proposed project hardening astropy’s core stability about testing low-level layer extensions.
Does building a more fundamental test suite for this low-level layer mean testing the functions exposed by the compiled `.so` modules, OR does it mean more in-depth testing of internal helper functions in `.c` files as well?
TIA
Hi,
It could be both. Though I think right now astropy as zero C testing going on and as far as I understand that would require an entire new layer of architecture and these helper functions *might* proove too volatile to be worth the effort, so we’d need to first justify that it’s worth doing, and I wouldn’t advise attempting it now.
So would it be reasonable to assume (for now) that if python-callable extension functions have sufficient test coverage, then the helper functions are indirectly validated through those tests?
Sorry for the delay in my answer. I’m not checking on this forum daily. The answer is yes