Question

我希望能够在对象实例化期间动态加载实例方法。根据我的设计，默认行为在基类中编码。但是，如果在物体实例化期间，我会动态地满足某些条件用另一段代码改变这种行为。这是怎么回事我这样做：

默认行为以first.py编码：

class First(object):
    def __init__(self, p):
        p = str(p)
        #The decision whether or not to perform default action is done
        #in the following try/except block. In reality this block 
        #is more complicated
        #and more checks are performed in order to assure proper work
        try: 
            strImport = "__import__('module%s')"%p
            e = eval(strImport, {}, {})
            if not hasattr(e, p):
                raise ImportError()
        except ImportError:
            e = None #default behaviour
        if e is not None:
            self.act = getattr(e, p)(p).act #special behaviour
        self.p = p

    def act(self):
        print 'Default behaviour'


import cPickle as pickle



if __name__ == '__main__':
    print 'first'
    first = First('f')
    first.act()
    pickle.dump(first, open('first.dump', 'w'))

    print 'third'
    third = First('Third')
    third.act()
    pickle.dump(third, open('third.dump', 'w'))

在上面的代码中，first和third都执行默认操作。我可以改变 third通过添加文件moduleThird.py的行为，如下所示：

from temp import First
class Third(First):
    def __init__(self, p):
        p = 'Third *** %p'
        print 'third init'
        super(self.__class__, self).__init__(p)


    def act(self):
        super(self.__class__, self).act()
        print 'Third acted'

添加此文件后，third会更改其行为。不过我不是由于以下错误，无法腌制结果对象：

Traceback (most recent call last):
  File "C:\temp\temp.py", line 35, in <module>
    pickle.dump(fff, open('fff.dump', 'w'))
  File "C:\Python26\lib\copy_reg.py", line 70, in _reduce_ex
    raise TypeError, "can't pickle %s objects" % base.__name__
TypeError: can't pickle instancemethod objects

很明显，动态加载方法Third.act导致了pickle的问题。我如何才能更改我的方法以获得可选择的对象（以及更优雅的代码）？

有没有更好的方法来实现我的目标？

Answer 1

如果您按如下方式更改代码，那么它应该有效：

class First(object):
    def __init__(self, p):
        p = str(p)
        #The decision whether or not to perform default action is done
        #in the following try/except block. In reality this block 
        #is more complicated
        #and more checks are performed in order to assure proper work
        try: 
            strImport = "__import__('module%s')"%p
            print strImport
            e = eval(strImport, {}, {})
            if not hasattr(e, p):
                raise ImportError()

            self.override_obj = getattr(e, p)(p)
        except ImportError:
            e = None #default behaviour
            self.override_obj = None

        self.p = p

    def act(self):
        if self.override_obj:
            return self.override_obj.act()
        else:
            print 'Default behaviour'

在对象实例化期间动态加载实例方法

1 个答案: