invariant tests

[Flip-Liquid]

• Add callbacks for actors in invariant test contract
  -Add initial invariants for actors

[neodaoist]

Really nice. I see how we'll be able to track what we need to, and how we can calculate invariants using totals and other variables which may not be tracked by Engine.

We might want to refactor out these arrays into a World object, instead of the test contract itself. This will give us flexibility to use in other test harnesses and it will remove a circular dependency — we would pass the World to each Actor instead of the test suite and include it as a state variable in the test contract (kinda like Timekeeper).

LMK what you think, I'm sure there are trade-offs. Separating out the tracking / accumulating into a world object is a pattern from BDD style testing that I think fits what you're doing well and may open doors for us for some other good techniques too.

[Flip-Liquid]

agree with the idea, was mostly just working around for the sake of this PR, and I think that would be a good refactor.

but how much more test code are we going to be authoring which would be able to leverage your proposed structure? if the answer is "a lot" then we should definitely hit the refac. but, are we actually going to be writing new test harnesses at this point? should we really invest more hours into refactoring test code?

I'm skeptical that whatever additional tests we're going to write (or not write) would justify spending more time on refactors. The erc20 wrappers seem like a much higher priority.

[neodaoist]

Sounds good, let's hold on World for time being. Maybe will be helpful for Wrapper, Margin, Vaults, etc. Let's wait and see.

[0xAlcibiades]

I think the world pattern is not a bad idea, but don't feel strongly that it has to be in this PR. We do need to get this compiling, though.

[codecov-commenter]

Codecov Report

Merging #162 (9f60c1e) into audit-fixes (83a9ac8) will not change coverage.
The diff coverage is n/a.

@@             Coverage Diff              @@
##           audit-fixes     #162   +/-   ##
============================================
  Coverage        91.22%   91.22%           
============================================
  Files                2        2           
  Lines              376      376           
  Branches            57       57           
============================================
  Hits               343      343           
  Misses              27       27           
  Partials             6        6           

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[0xAlcibiades]

@Flip-Liquid a follow on PR to the litepaper with some of these invariants expressed mathematically would be useful.

[neodaoist]

Nice work @Flip-Liquid

We could do an invariant_positions_accounting but for option positions not just claim positions.

I am still confused why no logs show other than from setUp function, rn it's hard to determine if the actors are actually getting exercised.