Python is used in combination with multiple decorators to implement Aspect-Oriented AOP programming.

First, rewrite the two callhandler (performancecountcallhandler and cachecallhandler) described in the previous articles.

The Code is as follows:

#-*-Coding: UTF-8-*-# features # Name: module 2 # purpose: # Author: ankier # created: 22-12-2012 # copyright: (c) ankier 2012 # licence: <2012 ~ 2020> # Export Import timefrom functools import wraps_cache ={} def performancecountcallhandler (): def _ performancecountcallhandler (originalfunc): def _ performancecountcallhandler (fun): def wrap (ARGs, kW ): glos = originalfunc. func_globals package = Glos ['_ package _'] model = Glos ['_ name _'] methodname = originalfunc. func_name timestart = time. time () Result = fun (ARGs, kW) timeend = time. time () print 'package: ', package,', model: ', model,', methodname: ', methodname ,'. ', timeend-timestart,'s 'Return result return wrap return _ performancecountcallhandler

 

#-*-Coding: UTF-8-*-# features # Name: module 2 # purpose: # Author: ankier # created: 22-12-2012 # copyright: (c) ankier 2012 # licence: <2012 ~ 2020> # define import timeimport hashlibimport pickleimport sysfrom functools import wraps_cache ={} def _ hashparamskey (function, argS, kW): Glos = function. func_globals package = Glos ['_ package _'] model = Glos ['_ name _'] methodname = function. func_name key = pickle. dumps (package, model, methodname, argS, kW) return hashlib. sha1 (key ). hexdigest () def _ isobsolete (entry, duration): return time. time ()-entry ['time']> durationdef cachecallhandler (duration = sys. maxint): def _ cachecallhandler (originalfunc): def _ cachecallhandler (fun): def wrap (ARGs, kW): Key = _ hashparamskey (originalfunc, argS, kW) if key not in _ cache or _ isobsolete (_ cache [Key], duration): # Save result = fun (ARGs, kW) _ cache [Key] = {'value': result, 'time': time. time ()} return _ cache [Key] ['value'] Return wrap return _ cachecallhandler

You need a register decorator to manage and register multiple decorator.

#-*- coding: UTF-8 -*-#-------------------------------------------------------------------------------# Name:        registerDecorators# Purpose:## Author:      ankier## Created:     23-12-2012# Copyright:   (c) Ankier 2012# Licence:     <2012~2020>#-------------------------------------------------------------------------------def RegisterDecorators(*_decorators):    def _RegisterDecorators(func):        wrap = func        for _deco in _decorators:            __deco = _deco(func)            wrap = __deco(wrap)        return wrap    return _RegisterDecorators

The following describes how to use these classes. cache is at the beginning and performance handler is at the end.

The usage is as follows:

#-*- coding: UTF-8 -*-import timefrom performanceCountCallHandler import PerformanceCountCallHandlerfrom cacheCallHandler import CacheCallHandlerfrom registerDecorators import RegisterDecorators@RegisterDecorators(CacheCallHandler(10000), PerformanceCountCallHandler())def Sum(xx , yy ):    sum = xx + yy    print '------Execute Sum----- '    time.sleep(5)                return sum@RegisterDecorators(CacheCallHandler(1), PerformanceCountCallHandler())def Count(xx , yy ):    sum = xx + yy    print '------Execute Count----- '    time.sleep(5)                return sum

Running effect:

------Execute Sum----- package: None  , model: mainFrame  , methodName: Sum .  5.0 s9package: None  , model: mainFrame  , methodName: Sum .  0.0 s9------Execute Count----- package: None  , model: mainFrame  , methodName: Count .  5.0 s2------Execute Count----- package: None  , model: mainFrame  , methodName: Count .  5.0 s3------Execute Count----- package: None  , model: mainFrame  , methodName: Count .  5.01600003242 s2