A collection of small improvements

[kellertuer]

While writing some answers in #264 I noticed that

• get_stoping_criterion might be nice
• indicated_convergence was not documented
• get_state would be nicer when it also had the recursive parameter (like get_objective)
• get_state replaces a few access functions

Internally

• maybe it would be reasonable to switch the state also to a AbstractDecoratedManoptState pattern like the objective, since there seems not much sense to keep the dispatch-with-THTT approach, since with an abstract decorator type it would be nicer to write and unified with the objective. -> this is a bit more work for later
• one could also implement a more. precise way for convergence than just the string returned by get_reason. – maybe after we have MOI states internally (with their dependency) but not before.

I will collect a few small internal improvements here for a while.

[codecov[bot]]

Codecov Report

Merging #266 (5b00f51) into master (3fe3acb) will increase coverage by 0.00%. The diff coverage is 100.00%.

@@           Coverage Diff           @@
##           master     #266   +/-   ##
=======================================
  Coverage   99.70%   99.70%           
=======================================
  Files          73       73           
  Lines        6452     6465   +13     
=======================================
+ Hits         6433     6446   +13     
  Misses         19       19           

Impacted Files	Coverage Δ
src/Manopt.jl	87.50% <ø> (ø)
src/plans/stopping_criterion.jl	100.00% <ø> (ø)
src/plans/objective.jl	100.00% <100.00%> (ø)
src/plans/solver_state.jl	100.00% <100.00%> (ø)
src/solvers/difference-of-convex-proximal-point.jl	100.00% <100.00%> (ø)
src/solvers/difference_of_convex_algorithm.jl	100.00% <100.00%> (ø)

:mega: We’re building smart automated test selection to slash your CI/CD build times. Learn more

[kellertuer]

Hm, I am a bit confused here. I could need this PR a lot on ManoptExamples just now but it seems my most recent commit made that every first run of the CI fails but every second one passes? Strange.

edit: I think I found it.