Allow default_counters macro to skip functions

[dpo]

Skipping certain functions is useful to avoid overwriting methods. If, e.g., neval_hprod is forwarded by

@default_counters Model inner

and the user redefines

NLPModels.neval_hprod(m::Model) = ...

then, neval_hprod is being overwritten, and that prevents precompilation. That is the case in quasi-Newton models.

[Copilot]

Pull request overview

This PR updates how NLP model wrappers can forward evaluation-counter accessors by enhancing @default_counters to support excluding selected forwarded counter functions, and adjusts module include order + tests accordingly.

Changes:

• Extend @default_counters to accept an optional exclusion list to avoid generating forwarding methods for specific counters.
• Reorder include("nlp/*.jl") so counters is loaded before utils/api.
• Expand test coverage around excluded counter forwarding behavior.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 1 comment.

File	Description
src/NLPModels.jl	Reorders include sequence so counters loads before utils/api.
src/nlp/utils.jl	Extends @default_counters with an excluded parameter and updates its docstring.
test/nlp/utils.jl	Updates tests to exercise the new excluded argument path.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[dpo]

@tmigot @MaxenceGollier The macro @default_counters forwards both the neval_* methods and the counters attribute. Because we have the definition neval_*(nlp) = nlp.counters.neval_* (in counters.jl), forwarding the counters attribute again forwards all the methods. So the part

  push!(
    ex.args,
    :(
      Base.getproperty(nlp::$(esc(Model)), s::Symbol) =
        (s == :counters ? nlp.$inner.counters : getfield(nlp, s))
    ),
  )

is redundant.

In addition, some models do not have a counters attribute (example: quasi-Newton models). So this is all a bit messy. For starters, it would surely be better to define a get_counters() accessor.

Admittedly, this is fairly old code.

[github-actions]

Package name	latest	stable
ADNLPModels
AdaptiveRegularization
AmplNLReader
BundleAdjustmentModels
CUTEst
CaNNOLeS
DCISolver
FletcherPenaltySolver
FluxNLPModels
JSOSolvers
JSOSuite
LLSModels
ManualNLPModels
NLPModelsIpopt
NLPModelsJuMP
NLPModelsKnitro
NLPModelsModifiers
NLPModelsTest
NLSProblems
PDENLPModels
PartiallySeparableNLPModels
PartiallySeparableSolvers
Percival
QuadraticModels
RegularizedOptimization
RegularizedProblems
SolverBenchmark
SolverTest
SolverTools