Call a C++ function from Python that uses a complex array pointer as an argument

I want to call a C++ function from Python that uses a complex array pointer as an argument.

My function looks like this.

extern "C" void myfunction(complex<float> *arr)

And I have made a library called mylib.so.

In Python, I have the following code to call the library (following: How to interface a NumPy complex array with C function using ctypes?):

complex_array = np.array([1+2j, 2+3j, 3+4j], dtype=np.complex64)

mylibrary_dll = CDLL('./mylib.so')
mylibrary_c = mylibrary_dll.myfunction
mylibrary_c.argtype = [np.ctypeslib.ndpointer(np.complex64,ndim=1,flags='c')]

And I am using:

mylibrary_c(complex_array.ctypes.data_as(POINTER(c_float)))

Reading Is it legal to cast float* to std::complex<float>*, I am not sure it is safe.

Should I use something like the following?

complex_array = np.array([1+2j, 2+3j, 3+4j], dtype=np.complex64)

# I define a new structure for complex
class Complex(ctypes.Structure):
    _fields_ = [("real", ctypes.c_float),
                ("imag", ctypes.c_float)]

mylibrary_dll = CDLL('./mylib.so')
mylibrary_c = mylibrary_dll.myfunction
mylibrary_c.argtype = [POINTER(Complex)] # This is different from above

mylibrary_c(complex_array.ctypes.data_as(POINTER(Complex))) # This is different from above

as, in C++:

  /// 26.2.3  complex specializations
  /// complex<float> specialization
  template<>
    struct complex<float>
    {
      typedef float value_type;
      typedef __complex__ float _ComplexT;

      _GLIBCXX_CONSTEXPR complex(_ComplexT __z) : _M_value(__z) { }

      _GLIBCXX_CONSTEXPR complex(float __r = 0.0f, float __i = 0.0f)
#if __cplusplus >= 201103L
      : _M_value{ __r, __i } { }
#else
      {
    __real__ _M_value = __r;
    __imag__ _M_value = __i;
      }
#endif