我怎样才能简单地从现有实例继承方法？

Below I have a very simple example of what I'm trying to do. I want to be able to use HTMLDecorator with any other class. Ignore the fact it's called decorator, it's just a name.

下面我有一个非常简单的例子,说明我正在尝试做什么。我希望能够将HTMLDecorator与任何其他类一起使用。忽略它被称为装饰器的事实,它只是一个名字。

import cgi

class ClassX(object):
  pass # ... with own __repr__

class ClassY(object):
  pass # ... with own __repr__

inst_x=ClassX()

inst_y=ClassY()

inst_z=[ i*i for i in range(25) ]

inst_b=True

class HTMLDecorator(object):
   def html(self): # an "enhanced" version of __repr__
       return cgi.escape(self.__repr__()).join(("<H1>","</H1>"))

print HTMLDecorator(inst_x).html()
print HTMLDecorator(inst_y).html()
wrapped_z = HTMLDecorator(inst_z)
inst_z[0] += 70
wrapped_z[0] += 71
print wrapped_z.html()
print HTMLDecorator(inst_b).html()

Output:

Traceback (most recent call last):
  File "html.py", line 21, in 
    print HTMLDecorator(inst_x).html()
TypeError: default __new__ takes no parameters

Is what I'm trying to do possible? If so, what am I doing wrong?

我正在尝试做什么?如果是这样,我做错了什么?

6 个解决方案

#1

Very close, but then I lose everything from ClassX. Below is something a collegue gave me that does do the trick, but it's hideous. There has to be a better way.

非常接近,但后来我失去了ClassX的所有东西。下面是一个同事给我的东西,它确实可以做到这一点,但它很可怕。一定有更好的方法。

Looks like you're trying to set up some sort of proxy object scheme. That's doable, and there are better solutions than your colleague's, but first consider whether it would be easier to just patch in some extra methods. This won't work for built-in classes like bool, but it will for your user-defined classes:

看起来你正在尝试设置某种代理对象方案。这是可行的,并且有比你的同事更好的解决方案,但首先考虑是否更容易修补一些额外的方法。这不适用于像bool这样的内置类,但它适用于您的用户定义的类:

def HTMLDecorator (obj):
    def html ():
        sep = cgi.escape (repr (obj))
        return sep.join (("<H1>", "</H1>"))
    obj.html = html
    return obj

And here is the proxy version:

这是代理版本:

class HTMLDecorator(object):
    def __init__ (self, wrapped):
        self.__wrapped = wrapped

    def html (self):
        sep = cgi.escape (repr (self.__wrapped))
        return sep.join (("<H1>", "</H1>"))

    def __getattr__ (self, name):
        return getattr (self.__wrapped, name)

    def __setattr__ (self, name, value):
        if not name.startswith ('_HTMLDecorator__'):
            setattr (self.__wrapped, name, value)
            return
        super (HTMLDecorator, self).__setattr__ (name, value)

    def __delattr__ (self, name):
        delattr (self.__wraped, name)

#2

Both of John's solutions would work. Another option that allows HTMLDecorator to remain very simple and clean is to monkey-patch it in as a base class. This also works only for user-defined classes, not builtin types:

约翰的两个解决方案都可行。允许HTMLDecorator保持非常简单和干净的另一个选项是将其作为基类进行修补。这也适用于用户定义的类,而不是内置类型:

import cgi

class ClassX(object):
    pass # ... with own __repr__

class ClassY(object):
    pass # ... with own __repr__

inst_x=ClassX()
inst_y=ClassY()

class HTMLDecorator:
    def html(self): # an "enhanced" version of __repr__
        return cgi.escape(self.__repr__()).join(("<H1>","</H1>"))

ClassX.__bases__ += (HTMLDecorator,)
ClassY.__bases__ += (HTMLDecorator,)

print inst_x.html()
print inst_y.html()

Be warned, though -- monkey-patching like this comes with a high price in readability and maintainability of your code. When you go back to this code a year later, it can become very difficult to figure out how your ClassX got that html() method, especially if ClassX is defined in some other library.

但是要注意 - 像这样的猴子修补程序会带来代码的可读性和可维护性的高昂代价。当你在一年后回到这段代码时,很难弄清楚你的ClassX如何得到html()方法,特别是如果ClassX在其他库中定义的话。

#3

Is what I'm trying to do possible? If so, what am I doing wrong?

我正在尝试做什么?如果是这样,我做错了什么?

It's certainly possible. What's wrong is that HTMLDecorator.__init__() doesn't accept parameters.

这当然是可能的。错误的是HTMLDecorator .__ init __()不接受参数。

Here's a simple example:

这是一个简单的例子:

def decorator (func):
    def new_func ():
        return "new_func %s" % func ()
    return new_func

@decorator
def a ():
    return "a"

def b ():
    return "b"

print a() # new_func a
print decorator (b)() # new_func b

#4

@John (37448):

Sorry, I might have misled you with the name (bad choice). I'm not really looking for a decorator function, or anything to do with decorators at all. What I'm after is for the html(self) def to use ClassX or ClassY's __repr__. I want this to work without modifying ClassX or ClassY.

对不起,我可能误导了你的名字(糟糕的选择)。我不是在寻找装饰器功能,或者根本不需要装饰器。我所追求的是html(self)def使用ClassX或ClassY的__repr__。我希望这可以在不修改ClassX或ClassY的情况下工作。

#5

Ah, in that case, perhaps code like this will be useful? It doesn't really have anything to do with decorators, but demonstrates how to pass arguments to a class's initialization function and to retrieve those arguments for later.

啊,在这种情况下,也许这样的代码会有用吗?它与装饰器没有任何关系,但演示了如何将参数传递给类的初始化函数并检索这些参数以供日后使用。

import cgi

class ClassX(object):
    def __repr__ (self):
        return "<class X>"

class HTMLDecorator(object):
    def __init__ (self, wrapped):
        self.__wrapped = wrapped

    def html (self):
        sep = cgi.escape (repr (self.__wrapped))
        return sep.join (("<H1>", "</H1>"))

inst_x=ClassX()
inst_b=True

print HTMLDecorator(inst_x).html()
print HTMLDecorator(inst_b).html()

#6

@John (37479):

Very close, but then I lose everything from ClassX. Below is something a collegue gave me that does do the trick, but it's hideous. There has to be a better way.

非常接近,但后来我失去了ClassX的所有东西。下面是一个同事给我的东西,它确实可以做到这一点,但它很可怕。一定有更好的方法。

import cgi
from math import sqrt

class ClassX(object): 
  def __repr__(self): 
    return "Best Guess"

class ClassY(object):
  pass # ... with own __repr__

inst_x=ClassX()

inst_y=ClassY()

inst_z=[ i*i for i in range(25) ]

inst_b=True

avoid="__class__ __init__ __dict__ __weakref__"

class HTMLDecorator(object):
    def __init__(self,master):
        self.master = master
        for attr in dir(self.master):
            if ( not attr.startswith("__") or 
                attr not in avoid.split() and "attr" not in attr):
                self.__setattr__(attr, self.master.__getattribute__(attr))

    def html(self): # an "enhanced" version of __repr__
        return cgi.escape(self.__repr__()).join(("<H1>","</H1>"))

    def length(self):
        return sqrt(sum(self.__iter__()))

print HTMLDecorator(inst_x).html()
print HTMLDecorator(inst_y).html()
wrapped_z = HTMLDecorator(inst_z)
print wrapped_z.length()
inst_z[0] += 70
#wrapped_z[0] += 71
wrapped_z.__setitem__(0,wrapped_z.__getitem__(0)+ 71)
print wrapped_z.html()
print HTMLDecorator(inst_b).html()

Output:

<H1>Best Guess</H1>
<H1><__main__.ClassY object at 0x891df0c></H1>
70.0
<H1>[141, 1, 4, 9, 16, 25, 36, 49, 64, 81, 100, 121, 144, 169, 196, 225, 256, 289, 324, 361, 400, 441, 484, 529, 576]</H1>
<H1>True</H1>