Use Optim.jl directly for Optim solvers

[sethaxen]

With LogDensityProblems, we have the ability to simultaneously compute the log density, its gradient, and even its Hessian. Optim.jl supports this kind of interface that shares work; however, Optimization.jl does not.

This PR changes the internals of Pathfinder to use Optim.jl whenever a LogDensityProblem and an Optim.jl optimizer are provided. In the best case, this halves the amount of work performed by Pathfinder. Since the optimization solution stored in PathfinderResult can now be one of two types, this change is marked as breaking.

[codecov-commenter]

Codecov Report

❌ Patch coverage is 83.33333% with 19 lines in your changes missing coverage. Please review.
✅ Project coverage is 78.26%. Comparing base (1e596a4) to head (d7f4bc3).
⚠️ Report is 107 commits behind head on main.

Files with missing lines	Patch %	Lines
src/optimize.jl	79.16%	10 Missing ⚠️
src/multipath.jl	40.00%	3 Missing ⚠️
src/trace.jl	40.00%	3 Missing ⚠️
src/callbacks.jl	95.00%	2 Missing ⚠️
src/singlepath.jl	93.75%	1 Missing ⚠️

❗ There is a different number of reports uploaded between BASE (1e596a4) and HEAD (d7f4bc3). Click for more details.

HEAD has 72 uploads less than BASE

Flag	BASE (1e596a4)	HEAD (d7f4bc3)
	77	5

Additional details and impacted files

@@             Coverage Diff             @@
##             main     #129       +/-   ##
===========================================
- Coverage   93.00%   78.26%   -14.74%     
===========================================
  Files          13       15        +2     
  Lines         572      635       +63     
===========================================
- Hits          532      497       -35     
- Misses         40      138       +98     

☔ View full report in Codecov by Harness.
📢 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.