Update reference wrapping

cpp/demo/README.md

@@ -50,9 +50,9 @@ The C++ and Python documentation reST source files should
   and explaining code snippets from the demo source code.
 
 * Include links to the API documentation using the :cpp:class: and
-  :py:class: directives. Note that for the Python classes, the
+  {py:class} directives. Note that for the Python classes, the
   full module path is required (for instance
-  py:class:dolfinx.cpp.NewtonSolver)
+  {py:class}`dolfinx.cpp.NewtonSolver`)
 
 * Include the complete set of files needed to run the demo using the
   include directive.

cpp/demo/biharmonic/biharmonic.py

@@ -23,7 +23,7 @@
 
 e = element("Lagrange", "triangle", 2)
 
-# The first argument to :py:class:`FiniteElement` is the finite element
+# The first argument to {py:class}`FiniteElement` is the finite element
 # family, the second argument specifies the domain, while the third
 # argument specifies the polynomial degree. Thus, in this case, our
 # element `element` consists of second-order, continuous Lagrange basis

cpp/demo/poisson/poisson.py

@@ -19,7 +19,7 @@
 
 e = element("Lagrange", "triangle", 1)
 
-# The first argument to :py:class:`FiniteElement` is the finite element
+# The first argument to {py:class}`FiniteElement` is the finite element
 # family, the second argument specifies the domain, while the third
 # argument specifies the polynomial degree. Thus, in this case, our
 # element `element` consists of first-order, continuous Lagrange basis

python/demo/demo_scipy-solvers.py

@@ -116,9 +116,9 @@ def compute_residual(x) -> npt.NDArray[dolfinx.default_scalar_type]:
 jacobian = ufl.derivative(residual, uh, dv)
 
 # We assemble the Jacobian matrix and its inverse.
-# Note that we assemble the Jacobian into a :class:`dolfinx.la.MatrixCSR`.
+# Note that we assemble the Jacobian into a {py:class}`dolfinx.la.MatrixCSR`.
 # Then, we get a view of the data in this matrix by using
-# :func:`dolfinx.la.MatrixCSR.to_scipy`.
+# {py:meth}`dolfinx.la.MatrixCSR.to_scipy`.
 
 J_compiled = dolfinx.fem.form(jacobian)
 A = dolfinx.fem.assemble_matrix(J_compiled)