Execute benchmarks on aws

[greole]

This PR refactores our benchmark setup and aims to be a first step towards a continuos benchmarking approach.

It executes the benchmarks on aws, collects the results and pushes it as json to https://github.com/exasim-project/NeoFOAM-BenchmarkData for storage. All advanced post-processing etc. can be done on the NeoFOAM-BenchmarkData repo.

In order for the current approach to work size and then executor need to be generated in a particular order see:

    auto size = GENERATE(1 << 16, 1 << 17, 1 << 18, 1 << 19, 1 << 20);
    NeoFOAM::Executor exec = GENERATE(
        NeoFOAM::Executor(NeoFOAM::SerialExecutor {}),
        NeoFOAM::Executor(NeoFOAM::CPUExecutor {}),
        NeoFOAM::Executor(NeoFOAM::GPUExecutor {})
    );
    std::string execName = std::visit([](auto e) { return e.name(); }, exec);

Additional Changes:

• The presets are modified slightly, such that benchmarks are performed in the profiling preset but not in the debug preset.
• I removed the plotting script and just added a converter between catch2 xml and json

Limitations:
The benchmark output of Catch2 is very messy IMO and we need a single BENCHMARK statement per TEST_CASE.

Future work:
Once we merge this PR into main we can produce benchmark data on main to. Then we can execute the benchmark on aws twice, once for main once for the PR branch and compare the results. This allows to show the impact of a particular PR.

[github-actions]

Deployed test documentation to https://exasim-project.com/NeoFOAM/Build_PR_241

[MarcelKoch]

maybe add a comment why this is needed for windows? (especially the generator stuff).

[greole]

I removed the generator expressions, I think they actually don't do anything atm.