Change perp to being a named operator

test/test_tensoralgebra.py

@@ -105,6 +105,13 @@ def test_cross(self):
     self.assertEqualValues(C, D)
 
 
+def test_perp(self):
+    u = as_vector([3, 1])
+    v = perp(u)
+    w = as_vector([-1, 3])
+    self.assertEqualValues(v, w)
+
+
 def xtest_dev(self, A):
     C = dev(A)
     D = 0*C  # FIXME: Add expected value here

ufl/algorithms/apply_algebra_lowering.py

@@ -55,6 +55,9 @@ def c(i, j):
             return Product(a[i], b[j]) - Product(a[j], b[i])
         return as_vector((c(1, 2), c(2, 0), c(0, 1)))
 
+    def perp(self, o, a):
+        return as_vector([-a[1], a[0]])
+
     def dot(self, o, a, b):
         ai = indices(len(a.ufl_shape) - 1)
         bi = indices(len(b.ufl_shape) - 1)

ufl/operators.py

@@ -19,7 +19,7 @@
 from ufl.form import Form
 from ufl.constantvalue import Zero, RealValue, ComplexValue, as_ufl
 from ufl.differentiation import VariableDerivative, Grad, Div, Curl, NablaGrad, NablaDiv
-from ufl.tensoralgebra import Transposed, Inner, Outer, Dot, Cross, \
+from ufl.tensoralgebra import Transposed, Inner, Outer, Dot, Cross, Perp, \
     Determinant, Inverse, Cofactor, Trace, Deviatoric, Skew, Sym
 from ufl.coefficient import Coefficient
 from ufl.variable import Variable
@@ -174,7 +174,7 @@ def perp(v):
     v = as_ufl(v)
     if v.ufl_shape != (2,):
         raise ValueError("Expecting a 2D vector expression.")
-    return as_vector((-v[1], v[0]))
+    return Perp(v)
 
 
 def cross(a, b):

ufl/tensoralgebra.py

@@ -217,6 +217,34 @@ def __str__(self):
                             parstr(self.ufl_operands[1], self))
 
 
+@ufl_type(is_index_free=True, num_ops=1)
+class Perp(CompoundTensorOperator):
+    __slots__ = ()
+
+    def __new__(cls, A):
+        sh = A.ufl_shape
+
+        # Checks
+        if not len(sh) == 1:
+            raise ValueError(f"Perp requires arguments of rank 1, got {ufl_err_str(A)}")
+        if not sh[0] == 2:
+            raise ValueError(f"Perp can only work on 2D vectors, got {ufl_err_str(A)}")
+
+        # Simplification
+        if isinstance(A, Zero):
+            return Zero((), A.ufl_free_indices, A.ufl_index_dimensions)
+
+        return CompoundTensorOperator.__new__(cls)
+
+    def __init__(self, A):
+        CompoundTensorOperator.__init__(self, (A,))
+
+    ufl_shape = (2,)
+
+    def __str__(self):
+        return "perp(%s)" % self.ufl_operands[0]
+
+
 @ufl_type(num_ops=2)
 class Cross(CompoundTensorOperator):
     __slots__ = ("ufl_free_indices", "ufl_index_dimensions")