ability to prevent force casts in numpy arguments

docs/advanced.rst

@@ -1100,10 +1100,12 @@ or ``py::array::f_style``.
 
 .. code-block:: cpp
 
-    void f(py::array_t<double, py::array::c_style> array);
+    void f(py::array_t<double, py::array::c_style | py::array::forcecast> array);
 
-As before, the implementation will attempt to convert non-conforming arguments
-into an array satisfying the specified requirements.
+The ``py::array::forcecast`` argument is the default value of the second
+template paramenter, and it ensures that non-conforming arguments are converted
+into an array satisfying the specified requirements instead of trying the next
+function overload.
 
 Vectorizing functions
 =====================

example/example10.cpp

@@ -33,4 +33,9 @@ void init_ex10(py::module &m) {
 
     // Vectorize a complex-valued function
     m.def("vectorized_func3", py::vectorize(my_func3));
+
+    /// Numpy function which only accepts specific data types
+    m.def("selective_func", [](py::array_t<int, py::array::c_style>) { std::cout << "Int branch taken. "<< std::endl; });
+    m.def("selective_func", [](py::array_t<float, py::array::c_style>) { std::cout << "Float branch taken. "<< std::endl; });
+    m.def("selective_func", [](py::array_t<std::complex<float>, py::array::c_style>) { std::cout << "Complex float branch taken. "<< std::endl; });
 }

example/example10.py

@@ -27,3 +27,8 @@ for f in [vectorized_func, vectorized_func2]:
     print(f(np.array([[1, 2, 3], [4, 5, 6]]), np.array([[2], [3]]), 2))
     print(np.array([[1, 2, 3], [4, 5, 6]])* np.array([[2], [3]])* 2)
 
+from example import selective_func
+
+selective_func(np.array([1], dtype=np.int32))
+selective_func(np.array([1.0], dtype=np.float32))
+selective_func(np.array([1.0j], dtype=np.complex64))

example/example10.ref

@@ -73,3 +73,6 @@ my_func(x:int=6, y:float=3, z:float=2)
  [ 24.  30.  36.]]
 [[ 4  8 12]
  [24 30 36]]
+Int branch taken. 
+Float branch taken. 
+Complex float branch taken. 

include/pybind11/numpy.h

@@ -78,7 +78,8 @@ public:
 
     enum {
         c_style = API::NPY_C_CONTIGUOUS_,
-        f_style = API::NPY_F_CONTIGUOUS_
+        f_style = API::NPY_F_CONTIGUOUS_,
+        forcecast = API::NPY_ARRAY_FORCECAST_
     };
 
     template <typename Type> array(size_t size, const Type *ptr) {
@@ -124,7 +125,7 @@ protected:
     }
 };
 
-template <typename T, int ExtraFlags = 0> class array_t : public array {
+template <typename T, int ExtraFlags = array::forcecast> class array_t : public array {
 public:
     PYBIND11_OBJECT_CVT(array_t, array, is_non_null, m_ptr = ensure(m_ptr));
     array_t() : array() { }
@@ -135,10 +136,9 @@ public:
             return nullptr;
         API &api = lookup_api();
         PyObject *descr = api.PyArray_DescrFromType_(detail::npy_format_descriptor<T>::value);
-        PyObject *result = api.PyArray_FromAny_(
-            ptr, descr, 0, 0,
-            API::NPY_ENSURE_ARRAY_ | API::NPY_ARRAY_FORCECAST_ | ExtraFlags,
-            nullptr);
+        PyObject *result = api.PyArray_FromAny_(ptr, descr, 0, 0, API::NPY_ENSURE_ARRAY_ | ExtraFlags, nullptr);
+        if (!result)
+            PyErr_Clear();
         Py_DECREF(ptr);
         return result;
     }
@@ -318,11 +318,11 @@ struct vectorize_helper {
     template <typename T>
     vectorize_helper(T&&f) : f(std::forward<T>(f)) { }
 
-    object operator()(array_t<Args>... args) {
+    object operator()(array_t<Args, array::c_style | array::forcecast>... args) {
         return run(args..., typename make_index_sequence<sizeof...(Args)>::type());
     }
 
-    template <size_t ... Index> object run(array_t<Args>&... args, index_sequence<Index...> index) {
+    template <size_t ... Index> object run(array_t<Args, array::c_style | array::forcecast>&... args, index_sequence<Index...> index) {
         /* Request buffers from all parameters */
         const size_t N = sizeof...(Args);
 
@@ -332,7 +332,7 @@ struct vectorize_helper {
         int ndim = 0;
         std::vector<size_t> shape(0);
         bool trivial_broadcast = broadcast(buffers, ndim, shape);
-                
+
         size_t size = 1;
         std::vector<size_t> strides(ndim);
         if (ndim > 0) {
@@ -384,7 +384,7 @@ struct vectorize_helper {
     }
 };
 
-template <typename T> struct handle_type_name<array_t<T>> {
+template <typename T, int Flags> struct handle_type_name<array_t<T, Flags>> {
     static PYBIND11_DESCR name() { return _("numpy.ndarray[dtype=") + type_caster<T>::name() + _("]"); }
 };