In Java you can define a new class inline using anonymous inner classes. This is useful when you need to rewrite only a single method of the class.
Suppose that you want create a subclass of OptionParser that overrides only a single method (for example exit()). In Java you can write something like this:
new OptionParser () {
public void exit() {
// body of the method
}
};
This piece of code creates a anonymous class that extends OptionParser and override only the exit() method.
There is a similar idiom in Python? Which idiom is used in these circumstances?
You can use the type(name, bases, dict) builtin function to create classes on the fly. For example:
op = type("MyOptionParser", (OptionParser,object), {"foo": lambda self: "foo" })
op().foo()
Since OptionParser isn't a new-style class, you have to explicitly include object in the list of base classes.
Java uses anonymous classes mostly to imitate closures or simply code blocks. Since in Python you can easily pass around methods there's no need for a construct as clunky as anonymous inner classes:
def printStuff():
print "hello"
def doit(what):
what()
doit(printStuff)
Edit: I'm aware that this is not what is needed in this special case. I just described the most common python solution to the problem most commonly by anonymous inner classes in Java.
You can accomplish this in three ways:
Proper subclass (of course)
a custom method that you invoke with the object as an argument
(what you probably want) -- adding a new method to an object (or replacing an existing one).
Example of option 3 (edited to remove use of "new" module -- It's deprecated, I did not know ):
import types
class someclass(object):
val = "Value"
def some_method(self):
print self.val
def some_method_upper(self):
print self.val.upper()
obj = someclass()
obj.some_method()
obj.some_method = types.MethodType(some_method_upper, obj)
obj.some_method()
Well, classes are first class objects, so you can create them in methods if you want. e.g.
from optparse import OptionParser
def make_custom_op(i):
class MyOP(OptionParser):
def exit(self):
print 'custom exit called', i
return MyOP
custom_op_class = make_custom_op(3)
custom_op = custom_op_class()
custom_op.exit() # prints 'custom exit called 3'
dir(custom_op) # shows all the regular attributes of an OptionParser
But, really, why not just define the class at the normal level? If you need to customise it, put the customisation in as arguments to __init__.
(edit: fixed typing errors in code)
Python doesn't support this directly (anonymous classes) but because of its terse syntax it isn't really necessary:
class MyOptionParser(OptionParser):
def exit(self, status=0, msg=None):
# body of method
p = MyOptionParser()
The only downside is you add MyOptionParser to your namespace, but as John Fouhy pointed out, you can hide that inside a function if you are going to do it multiple times.
Python probably has better ways to solve your problem. If you could provide more specific details of what you want to do it would help.
For example, if you need to change the method being called in a specific point in code, you can do this by passing the function as a parameter (functions are first class objects in python, you can pass them to functions, etc). You can also create anonymous lambda functions (but they're restricted to a single expression).
Also, since python is very dynamic, you can change methods of an object after it's been created object.method1 = alternative_impl1, although it's actually a bit more complicated, see gnud's answer
In python you have anonymous functions, declared using lambda statement. I do not like them very much - they are not so readable, and have limited functionality.
However, what you are talking about may be implemented in python with a completely different approach:
class a(object):
def meth_a(self):
print "a"
def meth_b(obj):
print "b"
b = a()
b.__class__.meth_a = meth_b
You can always hide class by variables:
class var(...):
pass
var = var()
instead of
var = new ...() {};
This is what you would do in Python 3.7
#!/usr/bin/env python3
class ExmapleClass:
def exit(self):
print('this should NOT print since we are going to override')
ExmapleClass= type('', (ExmapleClass,), {'exit': lambda self: print('you should see this printed only')})()
ExmapleClass.exit()
I do this in python3 usually with inner classes
class SomeSerializer():
class __Paginator(Paginator):
page_size = 10
# defining it for e.g. Rest:
pagination_class = __Paginator
# you could also be accessing it to e.g. create an instance via method:
def get_paginator(self):
return self.__Paginator()
as i used double underscore, this mixes the idea of "mangling" with inner classes, from outside you can still access the inner class with SomeSerializer._SomeSerializer__Paginator, and also subclasses, but SomeSerializer.__Paginator will not work, which might or might not be your whish if you want it a bit more "anonymous".
However I suggest to use "private" notation with a single underscore, if you do not need the mangling.
In my case, all I need is a fast subclass to set some class attributes, followed up by assigning it to the class attribute of my RestSerializer class, so the double underscore would denote to "not use it at all further" and might change to no underscores, if I start reusing it elsewhere.
Being perverse, you could use the throwaway name _ for the derived class name:
class _(OptionParser):
def exit(self):
pass # your override impl
Here is a more fancy way of doing Maciej's method.
I defined the following decorator:
def newinstance(*args, **kwargs):
def decorator(cls):
return cls(*args, **kwargs)
return decorator
The following codes are roughly equivalent (also works with args!)
// java
MyClass obj = new MyClass(arg) {
public void method() {
// body of the method
}
};
# python
#newinstance(arg)
class obj(MyClass):
def method(self):
pass # body of the method
You can use this code from within a class/method/function if you want to define an "inner" class instance.
Related
Assuming I have a class which requires a function (or should I say method) which is:
independent from my class instance - doesn't need self argument;
is called only inside my class object
I won't need access to it at any point (to override it for example);
should I (A) place it inside the class and mark it as a #staticmethod or should I (B) define it outside my class object (but in the same namespace)? Why?
Example:
class A:
def __init__(self, my_int):
self.my_int = my_int
def my_int_and_4(self):
print(self.adder(self.my_int,4))
#staticmethod
def adder(a,b):
return a+b
or
def adder(a,b):
return a+b
class B:
def __init__(self, my_int):
self.my_int = my_int
def my_int_and_4(self):
print(adder(self.my_int,4))
EDIT: maybe the example is a bit oversimplified. I should have added that my version of "adder" is specificly used with my class and in no other case.
This is a textbook use case for a private static method.
They key point here is that you should make it a private method of that class. That way you're certain nothing else will use it and depend on its implementation. You'll be free to change it in the future, or even delete it, without breaking anything outside that class.
And yeah, make it static, because you can.
In Python, there is no way to make a method truly private, but by convention, prefixing the method name by a _ means it should be treated as private.
#staticmethod
def _adder(a,b): ## <-- note the _
return a+b
If at some point you suddenly need to use it outside the class, then exposing it will be no trouble at all, e.g. using a public wrapper method.
The reverse, however, isn't true; once exposed, it's difficult to retract that exposure.
I would definitely use a private static method in this case, for the reasons described by Jean-Francois Corbett. There are two types of methods in Python that belong to the class itself, rather than an instance: class methods and static methods.
The first parameter of a class method (created with #classmethod) references the class in exactly the same manner that the first parameter of an instance method (self) references an instance. It is the equivalent of static methods in most other languages. If your method requires access to other class members, use a class method.
A static method (created with #staticmethod) does not contain a reference to the class, and therefore cannot reference other class members. It's generally used for private helper methods and the like.
For your adder method, I would definitely use a static method. However, in this modified (and rather useless) version, a class method is necessary:
class A:
x = 1
def __init__(self, my_int):
self.my_int = my_int
def my_int_and_4(self):
print(self._adder(self.my_int,4))
#staticmethod
def _adder(a,b):
return a+b
#classmethod
def _increment(cls, n):
return n + cls.x
Both approaches will work, so it's the matter of readability and following conventions.
Does the method need to look at the instance's private attributes? If yes, it's a good reason to keep it in the class.
Is the method only used as a helper for one of different methods? If yes, it's a good reason to put it right after the calling method so that the code can be read top-down.
Does the method seem to make sense outside of the context of your class? If yes, it's a good reason to make it a free function or even move it to a different file, like utils.
This is about multiple inheritance. Parent class A provides a few methods and B parent class B a few additional ones. By creating a class inheriting from A and B I could instantiate an object having both method sets.
Now my problem is, that I detect only after having instantiated A, that the methods from B would be helpful too (or more strictly stated, that my object is also of class B).
While
aInstance.bMethod = types.MethodType(localFunction, aInstance)
works in principle, it has to be repeated for any bMethod, and looks unnecessary complicated. It also requires stand-alone (local) functions instead of a conceptually cleaner class B. Is there a more streamlined approach?
Update:
I tried abstract base class with some success, but there only the methods of one additional class could be added.
What I finally achieved is a little routine, which adds all top-level procedures of a given module:
from types import MethodType
from inspect import ismodule, isfunction, getmembers
# adds all functions found in module as methods to given obj
def classMagic(obj, module):
assert(ismodule(module))
for name, fn in getmembers(module, isfunction):
if not name.startswith("__"):
setattr(obj, name, MethodType(fn, obj))
Functionally this is sufficient, and I'm also pleased with the automatism, that all functions are processed and I don't have separate places of function definition and adding it as method, so maintenace is easy. The only remaining issue is reflected by the startswith line, as an example for a neccessary naming convention, if selected functions shall not be added.
If I understand correctly, you want to add mixins to your class at run time. A very common way of adding mixins in Python is through decorators (rather than inheritance), so we can borrow this idea to do something runtime to the object (instead to the class).
I used functools.partial to freeze the self parameter, to emulate the process of binding a function to an object (i.e. turn a function into a method).
from functools import partial
class SimpleObject():
pass
def MixinA(obj):
def funcA1(self):
print('A1 - propertyA is equal to %s' % self.propertyA)
def funcA2(self):
print('A2 - propertyA is equal to %s' % self.propertyA)
obj.propertyA = 0
obj.funcA1 = partial(funcA1, self=obj)
obj.funcA2 = partial(funcA2, self=obj)
return obj
def MixinB(obj):
def funcB1(self):
print('B1')
obj.funcB1 = partial(funcB1, self=obj)
return obj
o = SimpleObject()
# need A characteristics?
o = MixinA(o)
# need B characteristics?
o = MixinB(o)
Instead of functools.partial, you can also use types.MethodType as you did in your question; I think that is a better/cleaner solution.
I am writing a python API/server to allow an external device (microcontroller) to remotely call methods of an object by sending a string with the name of the method. These methods would be stored in a dictionary. e.g. :
class Server:
...
functions = {}
def register(self, func):
self.functions[func.__name__] = func
def call(self, func_name, args):
self.functions[func_name](*args)
...
I know that I could define functions externally to the class definition and register them manually, but I would really like that the registering step would be done automatically. Consider the following class:
class MyServer(Server):
...
def add(self,a,b):
print a+b
def sub(self,a,b):
print a-b
...
It would work by subclassing a server class and by defining methods to be called. How could I get the methods to be automatically registered in the functions dictionary?
One way that I thought it could be done is with a metaclass that look at a pattern in the methods name add if a match is found, add that methods to the functions dictionary. It seems overkill...
Would it be possible to decorate the methods to be registered? Can someone give me a hint to the simplest solution to this problem?
There is no need to construct a dictionary, just use the getattr() built-in function:
def call(self, func_name, args):
getattr(self, func_name)(*args)
Python actually uses a dictionary to access attributes on objects anyway (it's called __dict__, - but using getattr() is better than accessing it directly).
If you really want to construct that dict for some reason, then look at the inspect module:
def __init__(self, ...):
self.functions = dict(inspect.getmembers(self, inspect.ismethod))
If you want to pick specific methods, you could use a decorator to do that, but as BrenBarn points out, the instance doesn't exist at the time the methods are decorated, so you need to use the mark and recapture technique to do what you want.
I have inherited code in which there are standalone functions, one per country code. E.g.
def validate_fr(param):
pass
def validate_uk(param):
pass
My idea is to create a class to group them together and consolidate the code into one method. Unfortunately that breaks cohesion. Another option is to dispatch to instance methods ?
class Validator(object):
def validate(param, country_code):
# dispatch
Alas, python does not have a switch statement.
UPDATE: I am still not convinced why I should leave them as global functions in my module. Lumping them as class methods seems cleaner.
I would keep the functions at module level -- no need for a class if you don't want to instantiate it anyway. The switch statement can easily be simulated using a dicitonary:
def validate_fr(param):
pass
def validate_uk(param)
pass
validators = {"fr": validate_fr,
"uk": validate_uk}
def validate(country_code, param):
return validators[country_code](param)
Given the naming scheme, you could also do it without the dictionary:
def validate(country_code, param):
return gloabls()["validate_" + country_code](param)
You do not need a switch statement for this.
validators = {
'fr': Validator(...),
'uk': Validator(...),
...
}
...
validators['uk'](foo)
Classes are not meant to group functions together, modules are. Functions in a class should be either methods that operate on the object itself (changing it's state, emitting information about the state, etc.) or class methods that do the same, but for the class itself (classes in Python are also objects). There's not even a need for static methods in Python, since you can always have functions at module level. As they say: Flat is better than nested.
If you want to have a set of functions place them in separate module.
Trying to change the __unicode__ method on an instance after it's created produces different results on Python 2.5 and 2.6.
Here's a test script:
class Dummy(object):
def __unicode__(self):
return u'one'
def two(self):
return u'two'
d = Dummy()
print unicode(d)
d.__unicode__ = d.two
print unicode(d)
print d.__unicode__()
On Python 2.5, this produces
one
two
two
That is, changing the instance's __unicode__ also changes unicode(instance)
On Python 2.6, this produces
one
one
two
So, after a change, unicode(instance) and instance.__unicode__() return different results.
Why? How can I get this working on Python 2.6?
(For what it's worth, the use case here is that I want to append something to the output of __unicode__ for all subclasses of a given class, without having to modify the code of the subclasses.)
Edit to make the use case a little clearer
I have Class A, which has many subclasses. Those subclasses define simple __unicode__ methods. I want to add logic so that, for instances of a Class A subclass, unicode(instance) gets something tacked on to the end. To keep the code simple, and because there are many subclasses I don't want to change, I'd prefer to avoid editing subclass code.
This is actually existing code that works in Python 2.5. It's something like this:
class A(object):
def __init__(self):
self._original_unicode = self.__unicode__
self.__unicode__ = self.augmented_unicode
def augmented_unicode(self):
return self._original_unicode() + u' EXTRA'
It's this code that no longer works on 2.6. Any suggestions on how to achieve this without modifying subclass code? (If the answer involves metaclasses, note that class A is itself a subclass of another class -- django.db.models.Model -- with a pretty elaborate metaclass.)
It appears that you are not allowed to monkey-patch protocol methods (those that begin and end with double underscores) :
Note
In practise there is another exception
that we haven't handled here. Although
you can override methods with instance
attributes (very useful for monkey
patching methods for test purposes)
you can't do this with the Python
protocol methods. These are the 'magic
methods' whose names begin and end
with double underscores. When invoked
by the Python interpreter they are
looked up directly on the class and
not on the instance (however if you
look them up directly - e.g.
x.repr - normal attribute lookup
rules apply).
That being the case, you may be stuck unless you can go with ~unutbu's answer.
EDIT: Or, you can have the base class __unicode__ method search the instance object's dict for a __unicode__ attribute. If it's present, then __unicode__ is defined on the instance object, and the class method calls the instance method. Otherwise, we fall back to the class definition of __unicode__.
I think that this could allow your existing subclass code to work without any changes. However, it gets ugly if the derived class wants to invoke the class implementation -- you need to be careful to avoid infinite loops. I haven't implemented such hacks in this example; merely commented about them.
import types
class Dummy(object):
def __unicode__(self):
func = self.__dict__.get("__unicode__", None)
if func:
// WARNING: if func() invokes this __unicode__ method directly,
// an infinite loop could result. You may need an ugly hack to guard
// against this. (E.g., set a flag on entry / unset the flag on exit,
// using a try/finally to protect against exceptions.)
return func()
return u'one'
def two(self):
return u'two'
d = Dummy()
print unicode(d)
funcType = type(Dummy.__unicode__)
d.__unicode__ = types.MethodType(Dummy.two, d)
print unicode(d)
print d.__unicode__()
Testing with Python 2.6 produces the following output:
> python dummy.py
one
two
two
Edit: In response to the OP's comment: Adding a layer of indirection can allow you to change the behavior of unicode on a per-instance basis:
class Dummy(object):
def __unicode__(self):
return self._unicode()
def _unicode(self):
return u'one'
def two(self):
return u'two'
d = Dummy()
print unicode(d)
# one
d._unicode = d.two
print unicode(d)
# two
print d.__unicode__()
# two
Looks like Dan is correct about monkey-patching protocol methods, and that this was a change between Python 2.5 and Python 2.6.
My fix ended up being making the change on the classes rather the instances:
class A(object):
def __init__(self):
self.__class__.__unicode__ = self.__class__.augmented_unicode