From ee6ddb55c4773c16b255d7daffac4d61f0020b28 Mon Sep 17 00:00:00 2001
From: odow <o.dowson@gmail.com>
Date: Mon, 29 Jan 2024 13:50:54 +1300
Subject: [PATCH] Fix feasibility pump with ScalarNonlinearFunction constraints

---
 src/filter.jl | 11 ++++++-----
 test/fpump.jl | 32 ++++++++++++++++++++++++++++++++
 2 files changed, 38 insertions(+), 5 deletions(-)

diff --git a/src/filter.jl b/src/filter.jl
index 6beb507..ad51ed1 100644
--- a/src/filter.jl
+++ b/src/filter.jl
@@ -42,11 +42,12 @@ function MOI.get(f::LinearFilter, attr::MOI.ListOfModelAttributesSet)
     end
 end
 function MOI.get(f::LinearFilter, attr::MOI.ListOfConstraintTypesPresent)
-    return filter(MOI.get(f.inner, attr)) do FS
-        F = FS[1]
-        return !(
-            F <: MOI.ScalarQuadraticFunction ||
-            F <: MOI.VectorQuadraticFunction
+    return filter(MOI.get(f.inner, attr)) do (F, _)
+        return (
+            F <: MOI.VariableIndex ||
+            F <: MOI.ScalarAffineFunction ||
+            F <: MOI.VectorOfVariables ||
+            F <: MOI.VectorAffineFunction
         )
     end
 end
diff --git a/test/fpump.jl b/test/fpump.jl
index 3451507..439a3fc 100644
--- a/test/fpump.jl
+++ b/test/fpump.jl
@@ -292,4 +292,36 @@ include("basic/gamsworld.jl")
         end
         @test JuMP.objective_value(m) ≈ 0.0
     end
+
+    @testset "FP: Issue 263: FPump with ScalarNonlinearFunction" begin
+        println("==================================")
+        println("FP: Issue 263: FPump with ScalarNonlinearFunction")
+        println("==================================")
+        ipopt_solver = JuMP.optimizer_with_attributes(
+            Ipopt.Optimizer,
+            "print_level" => 0,
+            "sb" => "yes",
+            "max_iter" => 50000,
+        )
+        highs_solver = JuMP.optimizer_with_attributes(
+            HiGHS.Optimizer,
+            "output_flag" => false,
+        )
+        juniper_solver = JuMP.optimizer_with_attributes(
+            Juniper.Optimizer,
+            "nl_solver" => ipopt_solver,
+            "mip_solver" => highs_solver,
+            "log_levels" => [],
+        )
+        m = Model(juniper_solver)
+        @variable(m, x[1:3], Int)
+        @variable(m, y)
+        @constraint(m, x[1] * x[2] * x[3] * y >= 5)
+        optimize!(m)
+        for i in 1:3
+            xval = JuMP.value(x[i])
+            @test isapprox(round(xval) - xval, 0; atol = sol_atol)
+        end
+        @test JuMP.objective_value(m) ≈ 0.0
+    end
 end