[Nonlinear] improve code coverage of operators.jl (#2671)

Author: odow

src/Nonlinear/operators.jl

@@ -21,22 +21,26 @@ function _create_binary_switch(ids, exprs)
     )
 end
 
-# We use a let block here for `expr` to create a local variable that does not
-# persist in the scope of the module. All we care about is the _eval_univariate
-# function that is eval'd as a result.
-let
+function _generate_eval_univariate()
     exprs = map(SYMBOLIC_UNIVARIATE_EXPRESSIONS) do arg
         return :(return $(arg[1])(x), $(arg[2]))
     end
-    @eval @inline function _eval_univariate(id, x::T) where {T}
-        $(_create_binary_switch(1:length(exprs), exprs))
-        return error("Invalid id for univariate operator: $id")
-    end
-    ∇²f_exprs = map(arg -> :(return $(arg[3])), SYMBOLIC_UNIVARIATE_EXPRESSIONS)
-    @eval @inline function _eval_univariate_2nd_deriv(id, x::T) where {T}
-        $(_create_binary_switch(1:length(∇²f_exprs), ∇²f_exprs))
-        return error("Invalid id for univariate operator: $id")
-    end
+    return _create_binary_switch(1:length(exprs), exprs)
+end
+
+@eval @inline function _eval_univariate(id, x::T) where {T}
+    $(_generate_eval_univariate())
+    return error("Invalid id for univariate operator: $id")
+end
+
+function _generate_eval_univariate_2nd_deriv()
+    exprs = map(arg -> :(return $(arg[3])), SYMBOLIC_UNIVARIATE_EXPRESSIONS)
+    return _create_binary_switch(1:length(exprs), exprs)
+end
+
+@eval @inline function _eval_univariate_2nd_deriv(id, x::T) where {T}
+    $(_generate_eval_univariate_2nd_deriv())
+    return error("Invalid id for univariate operator: $id")
 end
 
 struct _UnivariateOperator{F,F′,F′′}

test/Nonlinear/Nonlinear.jl

@@ -1401,6 +1401,22 @@ function test_create_binary_switch()
         ),
     )
     @test MOI.Nonlinear._create_binary_switch(1:4, [:a, :b, :c, :d]) == target
+    # Just test that these functions don't error. We'll test their contents by
+    # evaluating the actual fuctions that are `@eval`ed.
+    MOI.Nonlinear._generate_eval_univariate()
+    MOI.Nonlinear._generate_eval_univariate_2nd_deriv()
+    return
+end
+
+function test_intercept_ForwardDiff_MethodError()
+    r = Nonlinear.OperatorRegistry()
+    f(x::Float64) = sin(x)^2
+    g(x) = x > 1 ? f(x) : zero(x)
+    Nonlinear.register_operator(r, :g, 1, g)
+    @test Nonlinear.eval_univariate_function(r, :g, 0.0) == 0.0
+    @test Nonlinear.eval_univariate_function(r, :g, 2.0) ≈ sin(2.0)^2
+    @test Nonlinear.eval_univariate_gradient(r, :g, 0.0) == 0.0
+    @test_throws ErrorException Nonlinear.eval_univariate_gradient(r, :g, 2.0)
     return
 end