Enable get_face_interpolation_matrix() in 1D on FE_Q and FE_Bernstein

Currently, get_face_interpolation_matrix() on FE_Q and FE_Bernstein
throws an ExcImpossibleInDim(1) exception in 1D. This is strange
because the interpolation matrix should just be a 1-by-1 unit matrix.
Remove the dim > 1 assert in FE_Q_base and FE_Bernstein to fix this
and modify two tests to check that it works.

source/fe/fe_bernstein.cc

@@ -98,7 +98,6 @@ FE_Bernstein<dim, spacedim>::get_face_interpolation_matrix(
   FullMatrix<double> &                interpolation_matrix,
   const unsigned int                  face_no) const
 {
-  Assert(dim > 1, ExcImpossibleInDim(1));
   get_subface_interpolation_matrix(source_fe,
                                    numbers::invalid_unsigned_int,
                                    interpolation_matrix,
@@ -145,8 +144,8 @@ FE_Bernstein<dim, spacedim>::get_subface_interpolation_matrix(
       // Rule of thumb for FP accuracy, that can be expected for a given
       // polynomial degree.  This value is used to cut off values close to
       // zero.
-      double eps =
-        2e-13 * std::max(this->degree, source_fe->degree) * (dim - 1);
+      const double eps = 2e-13 * std::max(this->degree, source_fe->degree) *
+                         std::max(dim - 1, 1);
 
       // compute the interpolation matrix by simply taking the value at the
       // support points.

source/fe/fe_q_base.cc

@@ -611,7 +611,6 @@ FE_Q_Base<dim, spacedim>::get_face_interpolation_matrix(
   FullMatrix<double> &                interpolation_matrix,
   const unsigned int                  face_no) const
 {
-  Assert(dim > 1, ExcImpossibleInDim(1));
   get_subface_interpolation_matrix(source_fe,
                                    numbers::invalid_unsigned_int,
                                    interpolation_matrix,
@@ -660,7 +659,7 @@ FE_Q_Base<dim, spacedim>::get_subface_interpolation_matrix(
       // Rule of thumb for FP accuracy, that can be expected for a given
       // polynomial degree.  This value is used to cut off values close to
       // zero.
-      double eps = 2e-13 * this->q_degree * (dim - 1);
+      const double eps = 2e-13 * this->q_degree * std::max(dim - 1, 1);
 
       // compute the interpolation matrix by simply taking the value at the
       // support points.

tests/bits/face_interpolation.cc

@@ -27,9 +27,6 @@ template <int dim>
 void
 check_this(const FiniteElement<dim> &fe1, const FiniteElement<dim> &fe2)
 {
-  if (dim == 1)
-    return;
-
   // check all combinations of fe1 and fe2
   FullMatrix<double> face_constraints;
   try