Detailed explanation of python decorators

The syntax of the decorator starts with @, followed by the name of the decorator function and optional parameters.

Following the decorator declaration are the decorated function and the optional parameters of the decorated function, as follows:

@decorator(dec_opt_args)

def func(func_args):

....

In fact, overall Speaking of which, a decorator is actually a function, a function used to wrap a function. The decorator is called when the function declaration is completed, and the function declared after the call is replaced by a function decorated by the decorator.

For example:

def deco(func):
   ...
   return func
@deco
def foo():
      print 'foo'
#-----------------------------------
#等价如下：
def deco(func):
   ...
   return func
def foo():
      print 'foo'
foo = deco(foo)
如下例子：
def deco1(func):
    print 'ok'
    return func
@deco1
def foo():
    print 'foo'
foo()
#输出--------------
#ok
#foo
#------------------

If you don’t use decorators, you can do the following:

def deco1(func):
    print 'ok'
    return func
def foo():
    print 'foo'
print foo           #<function foo at 0x00AFE6F0>
foo = deco1(foo)   
foo()
#输出--------------
#ok
#foo
#------------------

Comparing the two, you can see that using decorators is so simple and flexible. Especially in enterprise-level development.

You can also use multiple decorators overlappingly:

def deco1(func):
    print 'deco1'
    return func
def deco2(func):
    print 'deco2'
    return func   
@deco1
@deco2
def foo():
    print 'foo'
foo()
#输出如下：-----------
#deco2
#deco1
#foo
#---------------------

Equivalent to:

@deco1
@deco2
def foo(arg):pass
-----------与下等效----------
foo = deco1(deco2(foo()))

2. Decorators with and without parameters

The above examples basically have parameters, and no parameters are simpler .

1. No ginseng

@deco1

@deco2

def foo(arg):pass

------------------------

foo = deco1(deco2(foo()))

2, there are ginseng

@deco1(deco_arg)

@deco2

def foo(arg):pass

---------- ----------

foo = deco1(deco_arg)(deco2(foo()))

Returns a decorator that takes a function as a parameter

3. Uses

1. Reference log

2 , Add timing logic to test performance

3. The ability to add transactions to functions

4. Example

from time import ctime,sleep
def deco(func):
    def decoIn():
        print '[%s]:%s called' %(ctime(),func.__name__)
        return func
    return decoIn
@deco
def foo():
    pass
foo()
sleep(4)
for i in range(2):
    sleep(1)
    foo()
      
#输出如下：--------
#[Fri Jul 05 10:45:04 2013]:foo called
#[Fri Jul 05 10:45:09 2013]:foo called
#[Fri Jul 05 10:45:10 2013]:foo called
#------------------