Fix maxiters warning by implementing solver-specific parameter mapping

- Add _set_maxiters! function to map common maxiters to solver-specific parameters
- Support major MOI solvers: Ipopt, Gurobi, CPLEX, SCIP, Mosek, OSQP, ECOS, SCS, COSMO
- Implement generic fallback that tries common parameter names
- Add comprehensive tests for the new functionality
- Resolves #844

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <[email protected]>

Author: ChrisRackauckas

lib/OptimizationMOI/src/OptimizationMOI.jl

@@ -60,6 +60,56 @@ function _create_new_optimizer(opt::MOI.AbstractOptimizer)
     return opt_setup
 end
 
+"""
+    _set_maxiters!(optimizer, maxiters)
+
+Sets the maximum number of iterations for the optimizer using solver-specific parameter names.
+Supports common MOI solvers including Ipopt, Gurobi, CPLEX, and SCIP.
+"""
+function _set_maxiters!(optimizer, maxiters::Number)
+    optimizer_name = string(typeof(optimizer))
+    
+    # Try to set maxiters based on common solver patterns
+    try
+        if contains(optimizer_name, "Ipopt")
+            MOI.set(optimizer, MOI.RawOptimizerAttribute("max_iter"), Int(maxiters))
+        elseif contains(optimizer_name, "Gurobi")
+            MOI.set(optimizer, MOI.RawOptimizerAttribute("IterationLimit"), Int(maxiters))
+        elseif contains(optimizer_name, "CPLEX") || contains(optimizer_name, "Cplex")
+            MOI.set(optimizer, MOI.RawOptimizerAttribute("CPX_PARAM_ITLIM"), Int(maxiters))
+        elseif contains(optimizer_name, "SCIP") || contains(optimizer_name, "Scip")
+            MOI.set(optimizer, MOI.RawOptimizerAttribute("limits/iterations"), Int(maxiters))
+        elseif contains(optimizer_name, "Mosek") || contains(optimizer_name, "MOSEK")
+            MOI.set(optimizer, MOI.RawOptimizerAttribute("MSK_IPAR_INTPNT_MAX_ITERATIONS"), Int(maxiters))
+        elseif contains(optimizer_name, "OSQP")
+            MOI.set(optimizer, MOI.RawOptimizerAttribute("max_iter"), Int(maxiters))
+        elseif contains(optimizer_name, "ECOS")
+            MOI.set(optimizer, MOI.RawOptimizerAttribute("maxit"), Int(maxiters))
+        elseif contains(optimizer_name, "SCS")
+            MOI.set(optimizer, MOI.RawOptimizerAttribute("max_iters"), Int(maxiters))
+        elseif contains(optimizer_name, "COSMO")
+            MOI.set(optimizer, MOI.RawOptimizerAttribute("max_iter"), Int(maxiters))
+        else
+            # Generic fallback - try common parameter names
+            for param_name in ["max_iter", "maxiter", "IterationLimit", "max_iterations"]
+                try
+                    MOI.set(optimizer, MOI.RawOptimizerAttribute(param_name), Int(maxiters))
+                    return # Success, exit early
+                catch
+                    continue # Try next parameter name
+                end
+            end
+            # If all attempts fail, show warning with guidance
+            @warn "common maxiters argument could not be mapped for $(typeof(optimizer)). " *
+                  "Set number of iterations via optimizer specific keyword arguments."
+        end
+    catch e
+        # Catch any errors during parameter setting and show informative warning
+        @warn "Failed to set maxiters parameter for $(typeof(optimizer)): $(e). " *
+              "Set number of iterations via optimizer specific keyword arguments."
+    end
+end
+
 function __map_optimizer_args(cache,
         opt::Union{MOI.AbstractOptimizer, MOI.OptimizerWithAttributes
         };
@@ -82,7 +132,7 @@ function __map_optimizer_args(cache,
         @warn "common abstol argument is currently not used by $(optimizer). Set tolerances via optimizer specific keyword arguments."
     end
     if !isnothing(maxiters)
-        @warn "common maxiters argument is currently not used by $(optimizer). Set number of iterations via optimizer specific keyword arguments."
+        _set_maxiters!(optimizer, maxiters)
     end
     return optimizer
 end

lib/OptimizationMOI/test/runtests.jl

@@ -276,3 +276,40 @@ end
     prob = OptimizationProblem(optprob, x0, _p, lcons = [1.0, 0.5], ucons = [1.0, 0.5])
     sol = solve(prob, Ipopt.Optimizer())
 end
+
+@testset "common maxiters interface" begin
+    # Test that the common maxiters interface works without warnings
+    rosenbrock(x, p) = (p[1] - x[1])^2 + p[2] * (x[2] - x[1]^2)^2
+    x0 = zeros(2)
+    _p = [1.0, 100.0]
+    
+    optprob = OptimizationFunction(rosenbrock, Optimization.AutoZygote())
+    prob = OptimizationProblem(optprob, x0, _p)
+    
+    # Test with Ipopt using maxiters parameter
+    @testset "Ipopt maxiters" begin
+        # This should not produce a warning and should respect the iteration limit
+        sol = solve(prob, Ipopt.Optimizer(); maxiters = 5, print_level = 0)
+        # Should terminate due to iteration limit
+        @test sol.stats.iterations <= 5
+    end
+    
+    # Test with cache interface
+    @testset "Cache interface maxiters" begin
+        cache = init(prob, Ipopt.Optimizer(); maxiters = 3, print_level = 0)
+        sol = solve!(cache)
+        @test sol.stats.iterations <= 3
+    end
+    
+    # Test that unknown solver fallback works gracefully
+    @testset "Generic fallback" begin
+        # Mock optimizer that doesn't match any known pattern
+        struct MockOptimizer <: MathOptInterface.AbstractOptimizer end
+        
+        # This should not error, but may show a warning for unknown solver
+        mock_opt = MockOptimizer()
+        # We can't actually solve with this mock optimizer, but we can test
+        # that the parameter setting doesn't crash
+        @test_nowarn OptimizationMOI._set_maxiters!(mock_opt, 10)
+    end
+end