Refactor SemOptimizer API

[alyst]

Extracted from #245 -- moving all engine-specific code (types) into extensions for modularity and leaner API, updating docstrings to make all optimizer help available for the user.

Requires #245 to be merged first.

[codecov]

Codecov Report

❌ Patch coverage is 46.75325% with 41 lines in your changes missing coverage. Please review.
✅ Project coverage is 71.83%. Comparing base (8b0f880) to head (1f8d2a9).
⚠️ Report is 87 commits behind head on devel.

Files with missing lines	Patch %	Lines
src/optimizer/abstract.jl	17.64%	14 Missing ⚠️
ext/SEMProximalOptExt/ProximalAlgorithms.jl	36.84%	12 Missing ⚠️
src/frontend/fit/SemFit.jl	0.00%	6 Missing ⚠️
ext/SEMNLOptExt/NLopt.jl	88.00%	3 Missing ⚠️
src/optimizer/optim.jl	57.14%	3 Missing ⚠️
src/frontend/fit/summary.jl	0.00%	2 Missing ⚠️
src/optimizer/Empty.jl	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##            devel     #299      +/-   ##
==========================================
- Coverage   72.94%   71.83%   -1.11%     
==========================================
  Files          50       51       +1     
  Lines        2218     2223       +5     
==========================================
- Hits         1618     1597      -21     
- Misses        600      626      +26     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[alyst]

Rebased to the current devel

[Maximilian-Stefan-Ernst]

I opened alyst#4 to try to streamline the access of the docs a bit further - let me know what you think.

[alyst]

@Maximilian-Stefan-Ernst I've merged your PR into my branch, thank you!
I think I've fixed the remaining parts:

• added SemOptimizer_impltype(engine) which just returns the type for the engine, so the optimizer_engine_doc() can be defined just once.
• added the SemOptimizerResult abstract basic optimization wrapper type for all engines. The engines have to implement it. It expects the optimizer to be always present.
  It does not add much overhead for Optim.jl engine and allows nice cleanups for NLoptl, Proximal
• fixed the generation of the docs for the extensions
• renamed SemOptimizerName to SemOptimizerMyopt in the tutorial

[alyst]

@Maximilian-Stefan-Ernst Thank you for another review round! Is there anything else except optimizer_engine_type to tweak?

[Maximilian-Stefan-Ernst]

Thanks a lot for this great set of changes! Once the last remaining comments are addressed, I would merge it.

[alyst]

I've addressed the last changes + a few other things:

• Since we have sem_optimizer_subtype(), we actually don't need to define SemOptimizer(Val{E}, ...) ctor for each engine, the generic SemOptimizer(val::Val{E}, ...) = sem_optimizer_subtype(val)(...) works just fine
• Some small tweaks to docs
• Remaining formatting fixes
• Some definitions moved around to maintain a natural order

I have cleaned up the commit history and force-pushed branch, but here's the patch of what's has changed since your last code review: compare.patch

[alyst]

@Maximilian-Stefan-Ernst I've tweaked the FormatCheck to run in the context of the PR, rather then its target, because it looks like it was checking out incorrect commit (that's why false alarms and incorrect suggestions).
I think there are no security concerns for running the format checker in the PR context.

[Maximilian-Stefan-Ernst]

Thanks a lot! Loks great, also the changes to the docs. I already merged the FormatCheck changes into main, so they take effect now.