Add ReturnCodes for NonlinearSolve.jl

[utkarsh530]

Meant for SciML/NonlinearSolve.jl#92

[utkarsh530]

I did not remove these conversion rules, because I am not sure if they will not break things (Although it might be fine since some of these are only used by NonlinearSolve.jl) https://github.com/SciML/SciMLBase.jl/pull/287/files#diff-4c19a4b567174fa747f0e1554877954e98ae7b8799729961e524ce48fbd0b4c3L15

[codecov]

Codecov Report

Merging #287 (dd57d23) into master (c4a29dd) will decrease coverage by 1.44%.
The diff coverage is 0.00%.

@@            Coverage Diff             @@
##           master     #287      +/-   ##
==========================================
- Coverage   55.51%   54.06%   -1.45%     
==========================================
  Files          42       42              
  Lines        3035     3035              
==========================================
- Hits         1685     1641      -44     
- Misses       1350     1394      +44     

Impacted Files	Coverage Δ
src/retcodes.jl	77.77% <0.00%> (-3.00%)	⬇️
src/operators/common_defaults.jl	52.94% <0.00%> (-20.59%)	⬇️
src/operators/basic_operators.jl	50.58% <0.00%> (-5.89%)	⬇️
src/integrator_interface.jl	41.13% <0.00%> (-5.36%)	⬇️
src/operators/operators.jl	7.14% <0.00%> (-3.58%)	⬇️
src/solutions/ode_solutions.jl	94.94% <0.00%> (-2.03%)	⬇️
src/scimlfunctions.jl	57.83% <0.00%> (-1.93%)	⬇️
src/problems/ode_problems.jl	55.84% <0.00%> (-1.30%)	⬇️
src/solutions/solution_interface.jl	34.02% <0.00%> (-1.04%)	⬇️
src/operators/diffeq_operator.jl	26.02% <0.00%> (+0.35%)	⬆️

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