如何将C方法附加到现有的Python类?
此方法显示如何从C扩展模块定义新的Python类。该类的方法在C中实现,但是仍可以从Python实例化,子类化和扩展该类。带有继承的相同技术也可以用于扩展现有的Python类,并使用C语言编写的方法。在该技术中,PyClass_New的第一个参数作为NULL传递,表明新类没有基类。然后,我们在该位置传递基类的元组,即使我们的新类是在C扩展而不是Python源代码中构建的,我们也将获得正常的Python继承行为。
示例
#include <Python.h> static PyObject* Foo_init(PyObject *self, PyObject *args) { printf("Foo._ _init_ _ called\n"); Py_INCREF(Py_None); return Py_None; } static PyObject* Foo_doSomething(PyObject *self, PyObject *args) { printf("Foo.doSomething called\n"); Py_INCREF(Py_None); return Py_None; } static PyMethodDef FooMethods[] = { {"_ _init_ _", Foo_init, METH_VARARGS, "doc string"}, {"doSomething", Foo_doSomething, METH_VARARGS, "doc string"}, {0, 0}, }; static PyMethodDef ModuleMethods[] = { {0, 0} }; #ifdef _ _cplusplus extern "C" #endif void initFoo( ) { PyMethodDef *def; /* create new module and class objects */ PyObject *module = Py_InitModule("Foo", ModuleMethods); PyObject *moduleDict = PyModule_GetDict(module); PyObject *classDict = PyDict_New( ); PyObject *className = PyString_FromString("Foo"); PyObject *fooClass = PyClass_New(NULL, classDict, className); PyDict_SetItemString(moduleDict, "Foo", fooClass); Py_DECREF(classDict); Py_DECREF(className); Py_DECREF(fooClass); /* add methods to class */ for (def = FooMethods; def->ml_name != NULL; def++) { PyObject *func = PyCFunction_New(def, NULL); PyObject *method = PyMethod_New(func, NULL, fooClass); PyDict_SetItemString(classDict, def->ml_name, method); Py_DECREF(func); Py_DECREF(method); } }