Open AdrienVannson opened 2 weeks ago
I don't see a problem with leaving this in
I don't see any reason to repeat the test, since everything here is fully deterministic. More importantly, this feature is not used at all: the test is always executed only one single time.
Even if, one day, repeating tests is needed, there are much simpler ways to do it (see https://stackoverflow.com/questions/21764473/how-can-i-repeat-each-test-multiple-times-in-a-py-test-run )
It is not a big change, but I am trying as much as I can to simplify the current implementation of betterproto. For now, I have to say that the code is quite messy and many simple changes can become painful because of this :)
I'm just seeing it as a connivence thing if someone wants to check that something they've written doesn't depend on execution order or something like class level state. I'm overall though fairly neutral on this however
my opinion: we can always add it back if we need it, and perhaps it will be improved the second time around
Summary
test_inputs.py
used to include some mechanism to repeat tests several times. I don't see any reasons for this since serializing / deserializing messages is deterministic. This option was not used anyway:repeat
always had a value of 1.Checklist