Skip the inner (inner) cache if bridges are not used

[odow]

One option to improve performance is to skip the inner-most cache if no bridges are used.

julia> model = Model(Clp.Optimizer)
A JuMP Model
Feasibility problem with:
Variables: 0
Model mode: AUTOMATIC
CachingOptimizer state: EMPTY_OPTIMIZER
Solver name: Clp

julia> @variable(model, x)
x

julia> backend(model)
MOIU.CachingOptimizer{MOI.AbstractOptimizer,MOIU.UniversalFallback{MOIU.Model{Float64}}}
in state EMPTY_OPTIMIZER
in mode AUTOMATIC
with model cache MOIU.UniversalFallback{MOIU.Model{Float64}}
  fallback for MOIU.Model{Float64}
with optimizer MOIB.LazyBridgeOptimizer{MOIU.CachingOptimizer{Clp.Optimizer,MOIU.UniversalFallback{MOIU.Model{Float64}}}}
  with 0 variable bridges
  with 0 constraint bridges
  with 0 objective bridges
  with inner model MOIU.CachingOptimizer{Clp.Optimizer,MOIU.UniversalFallback{MOIU.Model{Float64}}}
    in state ATTACHED_OPTIMIZER
    in mode AUTOMATIC
    with model cache MOIU.UniversalFallback{MOIU.Model{Float64}}
      fallback for MOIU.Model{Float64}
    with optimizer Clp.Optimizer

[odow]

The idea is to start off like

julia> backend(model)
MOIU.CachingOptimizer{MOI.AbstractOptimizer,MOIU.UniversalFallback{MOIU.Model{Float64}}}
in state EMPTY_OPTIMIZER
in mode AUTOMATIC
with model cache MOIU.UniversalFallback{MOIU.Model{Float64}}
  fallback for MOIU.Model{Float64}
with optimizer A GLPK model

and if the copy_to fails, then re-try with

julia> backend(model)
MOIU.CachingOptimizer{MOI.AbstractOptimizer,MOIU.UniversalFallback{MOIU.Model{Float64}}}
in state EMPTY_OPTIMIZER
in mode AUTOMATIC
with model cache MOIU.UniversalFallback{MOIU.Model{Float64}}
  fallback for MOIU.Model{Float64}
with optimizer MOIB.LazyBridgeOptimizer{MOIU.CachingOptimizer{GLPK.Optimizer,MOIU.UniversalFallback{MOIU.Model{Float64}}}}
  with 0 variable bridges
  with 0 constraint bridges
  with 0 objective bridges
  with inner model MOIU.CachingOptimizer{GLPK.Optimizer,MOIU.UniversalFallback{MOIU.Model{Float64}}}
    in state ATTACHED_OPTIMIZER
    in mode AUTOMATIC
    with model cache MOIU.UniversalFallback{MOIU.Model{Float64}}
      fallback for MOIU.Model{Float64}
    with optimizer GLPK.Optimizer

[odow]

I have a nice way to do this.

[odow]

Closing because we don't be doing this in MOI. Here's the proposed approach on the JuMP front: jump-dev/JuMP.jl#2610.