ChatGPT解决这个技术问题 Extra ChatGPT

Static methods in Python?

Can I define a static method which I can call directly on the class instance? e.g.,

MyClass.the_static_method()

M
Mateen Ulhaq

Yep, using the staticmethod decorator:

class MyClass(object):
    @staticmethod
    def the_static_method(x):
        print(x)

MyClass.the_static_method(2)  # outputs 2

Note that some code might use the old method of defining a static method, using staticmethod as a function rather than a decorator. This should only be used if you have to support ancient versions of Python (2.2 and 2.3):

class MyClass(object):
    def the_static_method(x):
        print(x)
    the_static_method = staticmethod(the_static_method)

MyClass.the_static_method(2)  # outputs 2

This is entirely identical to the first example (using @staticmethod), just not using the nice decorator syntax.

Finally, use staticmethod sparingly! There are very few situations where static-methods are necessary in Python, and I've seen them used many times where a separate "top-level" function would have been clearer.

The following is verbatim from the documentation::

A static method does not receive an implicit first argument. To declare a static method, use this idiom: class C: @staticmethod def f(arg1, arg2, ...): ... The @staticmethod form is a function decorator – see the description of function definitions in Function definitions for details. It can be called either on the class (such as C.f()) or on an instance (such as C().f()). The instance is ignored except for its class. Static methods in Python are similar to those found in Java or C++. For a more advanced concept, see classmethod(). For more information on static methods, consult the documentation on the standard type hierarchy in The standard type hierarchy. New in version 2.2. Changed in version 2.4: Function decorator syntax added.


Why adding @staticmethod decoration or using a function pointer, when you can simply avoid the first self parameter? Well, for an object a, you won't be able to call a.your_static_method(), which is allowed in other languages, but it is anyway considered a bad practice and the compiler always warns about it
I am using python 2.7.12, with @staticmethod decorator I am unable to call first static method defined. Its givin error : TypeError: 'staticmethod' object is not callable
@SomethingSomething, if you don't use the variable name "self", but do not add the decorator, the first argument eg arg1 will still be the reference to the self instance. The name self is only a convention.
@GeorgeMoutsopoulos - agreed generally. But - you will anyway be able to call the method as ClassName.methodName(), as if it was a static one, and then no selfwill be provided to the method. As you said, it will still be possible to also call this method as ClassInstance.methodName(), and self will be provided as the first parameter, regardless of its name.
@SomethingSomething so many SO answers, blog posts, and references and NO ONE seems to answer this question. I completely agree with everything you said. One should neither write self.staticMethod() nor instance.staticMethod(). The only exception I found is when you use inheritance and want the staticMethod to be overridden, which seems to work in Python. Now calling self.staticMethod() acknowledges the class type of self... This might be interesting with factory methods or clones when needed within the class but is probably still rare
M
Md. Abu Nafee Ibna Zahid

I think that Steven is actually right. To answer the original question, then, in order to set up a class method, simply assume that the first argument is not going to be a calling instance, and then make sure that you only call the method from the class.

(Note that this answer refers to Python 3.x. In Python 2.x you'll get a TypeError for calling the method on the class itself.)

For example:

class Dog:
    count = 0 # this is a class variable
    dogs = [] # this is a class variable

    def __init__(self, name):
        self.name = name #self.name is an instance variable
        Dog.count += 1
        Dog.dogs.append(name)

    def bark(self, n): # this is an instance method
        print("{} says: {}".format(self.name, "woof! " * n))

    def rollCall(n): #this is implicitly a class method (see comments below)
        print("There are {} dogs.".format(Dog.count))
        if n >= len(Dog.dogs) or n < 0:
            print("They are:")
            for dog in Dog.dogs:
                print("  {}".format(dog))
        else:
            print("The dog indexed at {} is {}.".format(n, Dog.dogs[n]))

fido = Dog("Fido")
fido.bark(3)
Dog.rollCall(-1)
rex = Dog("Rex")
Dog.rollCall(0)

In this code, the "rollCall" method assumes that the first argument is not an instance (as it would be if it were called by an instance instead of a class). As long as "rollCall" is called from the class rather than an instance, the code will work fine. If we try to call "rollCall" from an instance, e.g.:

rex.rollCall(-1)

however, it would cause an exception to be raised because it would send two arguments: itself and -1, and "rollCall" is only defined to accept one argument.

Incidentally, rex.rollCall() would send the correct number of arguments, but would also cause an exception to be raised because now n would be representing a Dog instance (i.e., rex) when the function expects n to be numerical.

This is where the decoration comes in: If we precede the "rollCall" method with

@staticmethod

then, by explicitly stating that the method is static, we can even call it from an instance. Now,

rex.rollCall(-1)

would work. The insertion of @staticmethod before a method definition, then, stops an instance from sending itself as an argument.

You can verify this by trying the following code with and without the @staticmethod line commented out.

class Dog:
    count = 0 # this is a class variable
    dogs = [] # this is a class variable

    def __init__(self, name):
        self.name = name #self.name is an instance variable
        Dog.count += 1
        Dog.dogs.append(name)

    def bark(self, n): # this is an instance method
        print("{} says: {}".format(self.name, "woof! " * n))

    @staticmethod
    def rollCall(n):
        print("There are {} dogs.".format(Dog.count))
        if n >= len(Dog.dogs) or n < 0:
            print("They are:")
            for dog in Dog.dogs:
                print("  {}".format(dog))
        else:
            print("The dog indexed at {} is {}.".format(n, Dog.dogs[n]))


fido = Dog("Fido")
fido.bark(3)
Dog.rollCall(-1)
rex = Dog("Rex")
Dog.rollCall(0)
rex.rollCall(-1)

The first example (calling by class without @staticmethod) doesn't work for me on Python 2.7. I get "TypeError: unbound method rollCall() must be called with Dog instance as first argument (got int instance instead)"
This does not work. Both Dog.rollCall(-1) and rex.rollCall(-1) return the same TypeError: unbound method rollCall() must be called with Dog instance as first argument (got int instance instead)
@MestreLion I think this is not a huge benefit, since all it does is muddle up static and instance methods and make one look like the other. If you know a method is static, you should access it as a static method. The fact that the method does not need or use your instance should be self-evident wherever you use the method. I always use T.my_static_method() or type(my_t_instance).my_static_method(), since this is clearer and makes it immediately obvious that I'm calling a static method.
P
Paolo Bergantino

Yes, check out the staticmethod decorator:

>>> class C:
...     @staticmethod
...     def hello():
...             print "Hello World"
...
>>> C.hello()
Hello World

Many answer above were downvoted because they provide just an example. Why this is different?
h
hichris123

You don't really need to use the @staticmethod decorator. Just declaring a method (that doesn't expect the self parameter) and call it from the class. The decorator is only there in case you want to be able to call it from an instance as well (which was not what you wanted to do)

Mostly, you just use functions though...


This would not work for python 2.5.2 class Dummy: def static1(): print "hello from static1" @staticmethod def static2(): print "hello from static2" Dummy.static2() Dummy.static1() Output: hello from static2 Traceback<most recent call last>: File "ll.py", line 46, in <module> Dummy.static1() TypeError: unbound method static1() must be called with Dummy instance as first argument (got nothing instead)
This does not work inside a class. Python will pass self as the first argument unless you tell it not to. (see: decorator)
Actually this is wrong in 2.7 and legal as of 3.X (tested fine in 3.2). That is you can't call a method from context of a class without the @staticmethod decorator in 2.7 and below. In 3.2 it works and will insert a self ref appropriately depending on how its called. Test case: pastebin.com/12DDV7DB.
Technically accurate, but a terrible solution. The staticmethod decorator allows one to call the function on both the class and an instance (this solution fails when calling the function on an instance).
a method can be called like any other attribute of an object using dot notation.class C: def callme(): print('called'); C.callme()
R
Russia Must Remove Putin

Static methods in Python? Is it possible to have static methods in Python so I can call them without initializing a class, like: ClassName.StaticMethod()

Yes, static methods can be created like this (although it's a bit more Pythonic to use underscores instead of CamelCase for methods):

class ClassName(object):

    @staticmethod
    def static_method(kwarg1=None):
        '''return a value that is a function of kwarg1'''

The above uses the decorator syntax. This syntax is equivalent to

class ClassName(object):

    def static_method(kwarg1=None):
        '''return a value that is a function of kwarg1'''

    static_method = staticmethod(static_method)

This can be used just as you described:

ClassName.static_method()

A builtin example of a static method is str.maketrans() in Python 3, which was a function in the string module in Python 2.

Another option that can be used as you describe is the classmethod, the difference is that the classmethod gets the class as an implicit first argument, and if subclassed, then it gets the subclass as the implicit first argument.

class ClassName(object):

    @classmethod
    def class_method(cls, kwarg1=None):
        '''return a value that is a function of the class and kwarg1'''

Note that cls is not a required name for the first argument, but most experienced Python coders will consider it badly done if you use anything else.

These are typically used as alternative constructors.

new_instance = ClassName.class_method()

A builtin example is dict.fromkeys():

new_dict = dict.fromkeys(['key1', 'key2'])

M
Matt Woelk

Perhaps the simplest option is just to put those functions outside of the class:

class Dog(object):
    def __init__(self, name):
        self.name = name

    def bark(self):
        if self.name == "Doggy":
            return barking_sound()
        else:
            return "yip yip"

def barking_sound():
    return "woof woof"

Using this method, functions which modify or use internal object state (have side effects) can be kept in the class, and the reusable utility functions can be moved outside.

Let's say this file is called dogs.py. To use these, you'd call dogs.barking_sound() instead of dogs.Dog.barking_sound.

If you really need a static method to be part of the class, you can use the staticmethod decorator.


what's considered to be the best practice? because I prefer keeping functions outside of the class, but it might be more maintainable when you define static methods within classes?
J
JJD

Aside from the particularities of how static method objects behave, there is a certain kind of beauty you can strike with them when it comes to organizing your module-level code.

# garden.py
def trim(a):
    pass

def strip(a):
    pass

def bunch(a, b):
    pass

def _foo(foo):
    pass

class powertools(object):
    """
    Provides much regarded gardening power tools.
    """
    @staticmethod
    def answer_to_the_ultimate_question_of_life_the_universe_and_everything():
        return 42

    @staticmethod
    def random():
        return 13

    @staticmethod
    def promise():
        return True

def _bar(baz, quux):
    pass

class _Dice(object):
    pass

class _6d(_Dice):
    pass

class _12d(_Dice):
    pass

class _Smarter:
    pass

class _MagicalPonies:
    pass

class _Samurai:
    pass

class Foo(_6d, _Samurai):
    pass

class Bar(_12d, _Smarter, _MagicalPonies):
    pass

...

# tests.py
import unittest
import garden

class GardenTests(unittest.TestCase):
    pass

class PowertoolsTests(unittest.TestCase):
    pass

class FooTests(unittest.TestCase):
    pass

class BarTests(unittest.TestCase):
    pass

...

# interactive.py
from garden import trim, bunch, Foo

f = trim(Foo())
bunch(f, Foo())

...

# my_garden.py
import garden
from garden import powertools

class _Cowboy(garden._Samurai):
    def hit():
        return powertools.promise() and powertools.random() or 0

class Foo(_Cowboy, garden.Foo):
    pass

It now becomes a bit more intuitive and self-documenting in which context certain components are meant to be used and it pans out ideally for naming distinct test cases as well as having a straightforward approach to how test modules map to actual modules under tests for purists.

I frequently find it viable to apply this approach to organizing a project's utility code. Quite often, people immediately rush and create a utils package and end up with 9 modules of which one has 120 LOC and the rest are two dozen LOC at best. I prefer to start with this and convert it to a package and create modules only for the beasts that truly deserve them:

# utils.py
class socket(object):
    @staticmethod
    def check_if_port_available(port):
        pass

    @staticmethod
    def get_free_port(port)
        pass

class image(object):
    @staticmethod
    def to_rgb(image):
        pass

    @staticmethod
    def to_cmyk(image):
        pass

A
Aman Raheja

So, static methods are the methods which can be called without creating the object of a class. For Example :-

    @staticmethod
    def add(a, b):
        return a + b

b = A.add(12,12)
print b

In the above example method add is called by the class name A not the object name.


H
HIMANSHU PANDEY

Summarizing others' answers and adding, there are many ways to declare Static Methods or Variables in python.

Using staticmethod() as a decorator: One can simply put a decorator above a method(function) declared to make it a static method. For eg.

class Calculator:
    @staticmethod
    def multiply(n1, n2, *args):
        Res = 1
        for num in args: Res *= num
        return n1 * n2 * Res

print(Calculator.multiply(1, 2, 3, 4))              # 24

Using staticmethod() as a parameter function: This method can receive an argument which is of function type, and it returns a static version of the function passed. For eg.

class Calculator:
    def add(n1, n2, *args):
        return n1 + n2 + sum(args)

Calculator.add = staticmethod(Calculator.add)
print(Calculator.add(1, 2, 3, 4))                   # 10

Using classmethod() as a decorator: @classmethod has similar effect on a function as @staticmethod has, but this time, an additional argument is needed to be accepted in the function (similar to self parameter for instance variables). For eg.

class Calculator:
    num = 0
    def __init__(self, digits) -> None:
        Calculator.num = int(''.join(digits))

    @classmethod
    def get_digits(cls, num):
        digits = list(str(num))
        calc = cls(digits)
        return calc.num

print(Calculator.get_digits(314159))                # 314159

Using classmethod() as a parameter function: @classmethod can also be used as a parameter function, in case one doesn't want to modify class definition. For eg.

class Calculator:
    def divide(cls, n1, n2, *args):
        Res = 1
        for num in args: Res *= num
        return n1 / n2 / Res

Calculator.divide = classmethod(Calculator.divide)

print(Calculator.divide(15, 3, 5))                  # 1.0

Direct declaration A method/variable declared outside all other methods, but inside a class is automatically static.

class Calculator:   
    def subtract(n1, n2, *args):
        return n1 - n2 - sum(args)

print(Calculator.subtract(10, 2, 3, 4))             # 1

The whole program

class Calculator:
    num = 0
    def __init__(self, digits) -> None:
        Calculator.num = int(''.join(digits))
    
    
    @staticmethod
    def multiply(n1, n2, *args):
        Res = 1
        for num in args: Res *= num
        return n1 * n2 * Res


    def add(n1, n2, *args):
        return n1 + n2 + sum(args)
    

    @classmethod
    def get_digits(cls, num):
        digits = list(str(num))
        calc = cls(digits)
        return calc.num


    def divide(cls, n1, n2, *args):
        Res = 1
        for num in args: Res *= num
        return n1 / n2 / Res


    def subtract(n1, n2, *args):
        return n1 - n2 - sum(args)
    



Calculator.add = staticmethod(Calculator.add)
Calculator.divide = classmethod(Calculator.divide)

print(Calculator.multiply(1, 2, 3, 4))              # 24
print(Calculator.add(1, 2, 3, 4))                   # 10
print(Calculator.get_digits(314159))                # 314159
print(Calculator.divide(15, 3, 5))                  # 1.0
print(Calculator.subtract(10, 2, 3, 4))             # 1

Refer to Python Documentation for mastering OOP in python.


C
Codemaker

Python Static methods can be created in two ways.

Using staticmethod() class Arithmetic: def add(x, y): return x + y # create add static method Arithmetic.add = staticmethod(Arithmetic.add) print('Result:', Arithmetic.add(15, 10))

Output:

Result: 25

Using @staticmethod class Arithmetic: # create add static method @staticmethod def add(x, y): return x + y print('Result:', Arithmetic.add(15, 10))

Output:

Result: 25


You just provided examples and shared the ways of how you can do it. You didn't explain what the methods mean.
E
Eric Aya

I encounter this question from time to time. The use case and example that I am fond of is:

jeffs@jeffs-desktop:/home/jeffs  $ python36
Python 3.6.1 (default, Sep  7 2017, 16:36:03) 
[GCC 6.3.0 20170406] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import cmath
>>> print(cmath.sqrt(-4))
2j
>>>
>>> dir(cmath)
['__doc__', '__file__', '__loader__', '__name__', '__package__', '__spec__', 'acos', 'acosh', 'asin', 'asinh', 'atan', 'atanh', 'cos', 'cosh', 'e', 'exp', 'inf', 'infj', 'isclose', 'isfinite', 'isinf', 'isnan', 'log', 'log10', 'nan', 'nanj', 'phase', 'pi', 'polar', 'rect', 'sin', 'sinh', 'sqrt', 'tan', 'tanh', 'tau']
>>> 

It does not make sense to create an object of class cmath, because there is no state in a cmath object. However, cmath is a collection of methods that are all related in some way. In my example above, all of the functions in cmath act on complex numbers in some way.


You didn't answer the question but you provided a example instead.