Permit creation of NumPy arrays with a "base" object that owns the data

This patch adds an extra base handle parameter to most ``py::array`` and
``py::array_t<>`` constructors. If specified along with a pointer to
data, the base object will be registered within NumPy, which increases
the base's reference count. This feature is useful to create shallow
copies of C++ or Python arrays while ensuring that the owners of the
underlying can't be garbage collected while referenced by NumPy.

The commit also adds a simple test function involving a ``wrap()``
function that creates shallow copies of various N-D arrays.

tests/test_numpy_array.cpp

@@ -99,4 +99,14 @@ test_initializer numpy_array([](py::module &m) {
     sm.def("make_c_array", [] {
         return py::array_t<float>({ 2, 2 }, { 8, 4 });
     });
+
+    sm.def("wrap", [](py::array a) {
+        return py::array(
+            a.dtype(),
+            std::vector<size_t>(a.shape(), a.shape() + a.ndim()),
+            std::vector<size_t>(a.strides(), a.strides() + a.ndim()),
+            a.data(),
+            a
+        );
+    });
 });