Test symbolic stoichiometry through public API

test/reactionsystem_core/symbolic_stoichiometry.jl

@@ -2,7 +2,6 @@
 
 # Fetch packages.
 using Catalyst, JumpProcesses, OrdinaryDiffEq, StochasticDiffEq, Statistics, Test
-using Symbolics: unwrap
 import SymbolicIndexingInterface as SII
 
 # Mock integrator type for testing generated VRJ affect functions.
@@ -207,35 +206,26 @@ let
     end
 end
 
-# Tests that non-species unknowns (regular variables) in stoichiometry are properly wrapped in Pre().
-# This ensures MTKBase generates explicit affects rather than implicit ones.
+# Tests jump simulation with a non-species unknown in symbolic stoichiometry.
 let
     @parameters k
     @variables V(t)
     @species X(t) Y(t)
 
-    # Reaction where V (a non-species variable) appears in the stoichiometry
     rs = @reaction_network begin
         @parameters k
         @variables V(t)
-        k, X --> V*Y
+        k, X --> V * Y
     end
 
-    # The jump System should have explicit affects (no equations after mtkcompile)
-    jsys = jump_model(rs)
-    js = ModelingToolkitBase.jumps(jsys)
-    @test length(js) == 1
-
-    # Check that V appears as Pre(V(t)) in the affect equations
-    affect_str = string(js[1].affect!)
-    @test occursin("Pre(V(t))", affect_str)
-
-    # Verify it creates an explicit affect (empty equations after mtkcompile)
-    iv = unwrap(t)
-    aff = ModelingToolkitBase.AffectSystem(js[1].affect!; iv)
-    affsys = ModelingToolkitBase.system(aff)
-    eqs = ModelingToolkitBase.equations(ModelingToolkitBase.unhack_system(affsys))
-    @test isempty(eqs)  # Should be explicit, not implicit
+    jprob = JumpProblem(
+        rs, [X => 1, Y => 0, V => 2], (0.0, 1.0), [k => 100.0];
+        rng = StableRNG(12345)
+    )
+    sol = solve(jprob, SSAStepper())
+    @test sol[X][end] == 0
+    @test sol[Y][end] == 2
+    @test sol[V][end] == 2
 end
 
 # Tests symbolic stoichiometries in simulations.