Question

我让我上课，就是把自己当作一个泡菜。

我希望该课程能够从泡菜中导入数据......：

喜欢：

myFirstInstance.printV1() #variable is set to 4 by the __init__
myFirstInstance.write2Pickle(myPickleFile.pkl)

mySecondInstance.loadPickleData(myPickleFile.pkl)
mySecondInstance.printV1() # should give 4 back from the firstInstance...

Answer 1

使用pickle非常简单：

import cPickle as pickle
with open('a.pkl', 'wb') as f:
    pickle.dump(obj, f)
with open('a.pkl', 'rb') as f:
    obj = pickle.load(f)

如果您的课程需要特殊的逻辑来进行腌制/取消篡改，您应该覆盖pickle protocol使用的方法（例如__getstate__ / __setstate__）。这样，上面的用法仍然有效。

Answer 2

这有点棘手，因为您正在从现有实例中更改类实例的状态...而且您还希望从类的实例中挑选类的实例。为了简化类实例的修补，我使用dill代替pickle。

>>> import dill
>>> 
>>> class MyClass(object):
...   def __init__(self, x):
...     self.V1 = x
...   def printV1(self):
...     print(self.V1)
...   def write2Pickle(self, filename):
...     with open(filename, 'wb') as f:
...       dill.dump(self, f)
...   def loadPickleData(self, filename):
...     with open(filename, 'rb') as f:
...       obj = dill.load(f)
...     self.V1 = obj.V1
...     # similarly for other instance state variables
... 
>>> myFirstInstance = MyClass(4)
>>> myFirstInstance.printV1()
4
>>> myFirstInstance.write2Pickle('myPickleFile.pkl')
>>> del myFirstInstance
>>> mySecondInstance = MyClass(1)
>>> mySecondInstance.printV1()
1
>>> mySecondInstance.loadPickleData('myPickleFile.pkl')
>>> mySecondInstance.printV1()
4

我注意到，不要将酸洗和取消作为一种类方法进行编程练习可能更好...除非有真的它应该存在的理由

。

此外，您可以直接取消拼写并获得一个非常好的新实例，因此除非您需要覆盖现有实例的状态，否则无需将取消作为类方法。

>>> with open('myPickleFile.pkl', 'rb') as f:
...   myThirdInstance = dill.load(f)
... 
>>> myThirdInstance.printV1()
4

同样，您可以在实例外部使用dump，它也可以创建一个新的pickle实例。

如何将变量从pickle整合回对象

2 个答案: