Improve testing doc

Open gpsaggese opened this issue 6 months ago • 1 comments

My approach is to start testing end-to-end behaviors and then go towards the inner working. We don't want to test implementation details, otherwise the maintenance cost is insane.

This looks like testing too details.

Also we don't want to test that OpenAI specific stuff. E.g., we don't have to test that OpenAI follows its own API.

Can you start adding some of these suggestions to the unit test docs? ./docs/tools/unit_test/all.run_unit_tests.how_to_guide.md ./docs/tools/unit_test/all.write_unit_tests.how_to_guide.md

I'll file a bug

Originally posted by @gpsaggese in https://github.com/causify-ai/helpers/pull/777#discussion_r2115982890

May 30 '25 14:05 gpsaggese

We should also document

[ ] the hunit_test framework
[ ] add it to our working paper
[ ] make a blog entry on this

May 30 '25 14:05 gpsaggese