3rd try at scenario building

[flomnes]

• 🧹🗑️ Remove simple & legacy tests - they make it too hard to move forward with the interface
• 📜 Add basic scenarization, allowing for multiple scenarios within the same LP (💫 new !)
• 📏 Formatting according to src/.clang-format

TODO

• 💪 Adapt "standard" test-case for more than one single scenario

[pet-mit]

we need to keep testing that the new fillers are compatible with the legacy problem. instead of deleting this, you can replace the new fillers it uses with the ones use in "new-workflow/standard-test"

[flomnes]

Yes. I tried but ran into the following issue : ports are not implemented for LegacyLinearProblem/LegacyLinearProblemFillerImpl. So it's a bit hard to connect battery / thermal to legacy balance constraints. This is not related to scenarization, so I'll open another PR for this.

[pet-mit]

when using the ComponentFiller of "BALANCE" type, the used AddBalanceConstraint in it will hook the new balance terms to the legacy constraint (meaning you don't need a port, it is implicit). you only have to make sure the ComponentFiller has the same node name as the legacy node

[pet-mit]

Suggested change

	const std::map<std::string, std::vector<double>>& scalarData,
	const std::map<std::string, std::vector<std::vector<double>>>& timedData) :
	const std::map<std::string, std::vector<double>>& scalarDataPerScenario,
	const std::map<std::string, std::vector<std::vector<double>>>& timedDataPerScenario) :

[pet-mit]

I think it would be interesting to add a new test, inspired by the ones here, to test multiple scenarios

[flomnes]

objectiveValue_ is set at construction. I think it should be extracted after Solve. But then we have to manage a cache with synchronization issues, etc. So I suggest not caching the objective value for now.

[pet-mit]

MipSolution is constructed after Solve