Swig-结构的C ++向量以及如何连接它们

Question

I am writing a c++ library and I would love to be able to call it in python. I would love to use swig I am successfully able to create and compile the python module, but I struggle a bit on understanding how to interface python and c++.

struct twod {
        double x; ///< x value
        double y; ///< y value
    };

    double distance_calculation(std::vector <twod>  A, std::vector <twod> B);

This is a snap of my header file. Following my .i file:

%module hausdorff
%{
#include "Hausdorff.h"
using namespace hausdorff;
%}


%include "std_vector.i"
%include "Hausdorff.h"
namespace std {
    %template(vector2d) vector<twod>;
}

In python I am able to create the object:

In [13]: vector = hausdorff.vector2d

In [14]: vector = ([1,2], [3,4])

In [15]: result = hausdorff.distance_calculation(vector, vector)

And I get as error:

TypeError: in method 'distance_calculation', argument 1 of type 'std::vector< hausdorff::twod,std::allocator< hausdorff::twod > >'

How can I pass the right object to the function?

Answer 1

It's a little more complicated than that, at least without more work:

>>> import hausdorff
>>> v = hausdorff.vector2d()
>>> a = hausdorff.twod()
>>> a.x = 1.2
>>> a.y = 2.3
>>> v.push_back(a)
>>> a.x = 3.4
>>> a.y = 4.5
>>> v.push_back(a)
>>> test.distance_calculation(v,v)

If you provide constructors for your structure, you can simplify:

>>> test.distance_calculation([test.twod(1.2,3.4),test.twod(1.2,3.4)],
                              [test.twod(4.5,6.7),test.twod(1.2,3.4)])

If you provide typemap conversions, which I leave as a exercise (or another SO question :^) this can be done:

>>> test.distance_calculation([(1.1,2.2),(3.3,4.4)],[(5.5,6.6),(7.7,8.8)])