Update

ext/NLoptMathOptInterfaceExt.jl

@@ -570,21 +570,18 @@ function MOI.set(
 end
 
 function _fill_gradient(grad, x, f::MOI.VariableIndex)
-    fill!(grad, 0.0)
     grad[f.value] = 1.0
     return
 end
 
 function _fill_gradient(grad, x, f::MOI.ScalarAffineFunction{Float64})
-    fill!(grad, 0.0)
     for term in f.terms
         grad[term.variable.value] += term.coefficient
     end
     return
 end
 
 function _fill_gradient(grad, x, f::MOI.ScalarQuadraticFunction{Float64})
-    fill!(grad, 0.0)
     for term in f.affine_terms
         grad[term.variable.value] += term.coefficient
     end
@@ -645,40 +642,36 @@ function _fill_jacobian(jac, x, offset, constraints::Vector)
 end
 
 function objective_fn(model::Optimizer, x::Vector, grad::Vector)
+    # The order of the conditions is important. NLP objectives override regular
+    # objectives.
     if length(grad) > 0
         fill!(grad, 0.0)
-    end
-    # If we don't give any objective to NLopt, it throws the error:
-    # invalid NLopt arguments: NULL args to nlopt_optimize
-    if model.sense == MOI.FEASIBILITY_SENSE
-        return 0.0
-    end
-    if length(grad) > 0
-        if model.nlp_data.has_objective
+        if model.sense == MOI.FEASIBILITY_SENSE
+            # nothing
+        elseif model.nlp_data.has_objective
             MOI.eval_objective_gradient(model.nlp_data.evaluator, grad, x)
         elseif model.objective !== nothing
             _fill_gradient(grad, x, model.objective)
         end
     end
-    # The order of the conditions is important. NLP objectives override regular
-    # objectives.
+    if model.sense == MOI.FEASIBILITY_SENSE
+        return 0.0
     if model.nlp_data.has_objective
         return MOI.eval_objective(model.nlp_data.evaluator, x)
     elseif model.objective !== nothing
         return MOI.Utilities.eval_variables(vi -> x[vi.value], model.objective)
     end
-    # No objective function set. This could happen with FEASIBILITY_SENSE.
+    # No ObjectiveFunction is set, but ObjectiveSense is?
     return 0.0
 end
 
 function _initialize_options!(model::Optimizer)
     local_optimizer = model.options["local_optimizer"]
     if local_optimizer !== nothing
-        if local_optimizer isa Symbol
-            local_optimizer = NLopt.Opt(local_optimizer, num_variables)
+        local_optimizer = if local_optimizer isa Symbol
+            NLopt.Opt(local_optimizer, num_variables)
         else
-            local_optimizer =
-                NLopt.Opt(local_optimizer.algorithm, num_variables)
+            NLopt.Opt(local_optimizer.algorithm, num_variables)
         end
         NLopt.local_optimizer!(model.inner, local_optimizer)
     end