Python基于Mitmproxy实现被动式浏览器请求抓取与解析

林舟

发布时间：2026-10-13| 6人浏览过

原创

本文介绍如何利用Python结合Mitmproxy库构建被动式代理，以捕获浏览器发出的HTTP/HTTPS请求。内容涵盖环境安装、目录结构配置及核心代码实现，重点解析如何过滤静态资源并提取请求关键信息，适用于安全扫描器开发场景。

项目背景与工具选择

近期开发一款基于被动模式的漏洞扫描器，其核心机制是捕获用户在浏览器浏览过程中发出的网络请求，并将这些数据交由扫描器进行后续处理。起初考虑自行编写代理程序，但由于需要支持HTTPS协议的抓取，最终选择了功能完善的Mitmproxy工具。

环境安装与目录结构

安装步骤如下：

执行命令：pip install mitmproxy

随后通过一个具体案例来演示其使用方式，项目目录结构如下所示：

sproxy

|utils

|__init__.py

|parser.py

|sproxy.py

核心代码实现

sproxy.py主要代码实现如下：


#coding=utf-8
 
from pprint import pprint
from mitmproxy import flow, proxy, controller, options
from mitmproxy.proxy.server import ProxyServer
from utils.parser import ResponseParser
 
# http static resource file extension
static_ext = ['js', 'css', 'ico', 'jpg', 'png', 'gif', 'jpeg', 'bmp']
# media resource files type
media_types = ['image', 'video', 'audio']
 
# url filter
url_filter = ['baidu','360','qq.com']
 
static_files = [
 'text/css',
 'image/jpeg',
 'image/gif',
 'image/png',
]
 
class WYProxy(flow.FlowMaster):
 
 
 def __init__(self, opts, server, state):
  super(WYProxy, self).__init__(opts, server, state)
 
 def run(self):
  try:
   pprint("proxy started successfully...")
   flow.FlowMaster.run(self)
  except KeyboardInterrupt:
   pprint("Ctrl C - stopping proxy")
   self.shutdown()
 
 def get_extension(self, flow):
  if not flow.request.path_components:
   return ''
  else:
   end_path = flow.request.path_components[-1:][0]
   split_ext = end_path.split('.')
   if not split_ext or len(split_ext) == 1:
    return ''
   else:
    return split_ext[-1:][0][:32]
 
 def capture_pass(self, flow):
 	# filter url
  url = flow.request.url
  for i in url_filter:		
			if i in url:
				return True
 
  """if content_type is media_types or static_files, then pass captrue"""
  extension = self.get_extension(flow)
  if extension in static_ext:
   return True
 
  # can't catch the content_type
  content_type = flow.response.headers.get('Content-Type', '').split(';')[:1][0]
  if not content_type:
   return False
 
  if content_type in static_files:
   return True
 
  http_mime_type = content_type.split('/')[:1]
  if http_mime_type:
   return True if http_mime_type[0] in media_types else False
  else:
   return False
 
 @controller.handler
 def request(self, f):
 	pass
 
 @controller.handler
 def response(self, f):
  try:
   if not self.capture_pass(f):
    parser = ResponseParser(f)
    result = parser.parser_data()
    if f.request.method == "GET":
     print result['url']
    elif f.request.method == "POST":
     print result['request_content'] # POST提交的参数
 
  except Exception as e:
   raise e
 
 @controller.handler
 def error(self, f):
 	pass
  # print("error", f)
 
 @controller.handler
 def log(self, l):
 	pass
  # print("log", l.msg)
 
def start_server(proxy_port, proxy_mode):
 port = int(proxy_port) if proxy_port else 8090
 mode = proxy_mode if proxy_mode else 'regular'
 
 if proxy_mode == 'http':
  mode = 'regular'
 
 opts = options.Options(
  listen_port=port,
  mode=mode,
  cadir="~/.mitmproxy/",
  )
 config = proxy.ProxyConfig(opts)
 state = flow.State()
 server = ProxyServer(config)
 m = WYProxy(opts, server, state)
 m.run()
 
 
if __name__ == '__main__':
	start_server("8090", "http")

parser.py模块代码：


# from __future__ import absolute_import
 
class ResponseParser(object):
 """docstring for ResponseParser"""
 
 def __init__(self, f):
  super(ResponseParser, self).__init__()
  self.flow = f
 
 def parser_data(self):
  result = dict()
  result['url'] = self.flow.request.url
  result['path'] = '/{}'.format('/'.join(self.flow.request.path_components))
  result['host'] = self.flow.request.host
  result['port'] = self.flow.request.port
  result['scheme'] = self.flow.request.scheme
  result['method'] = self.flow.request.method
  result['status_code'] = self.flow.response.status_code
  result['content_length'] = int(self.flow.response.headers.get('Content-Length', 0))
  result['request_header'] = self.parser_header(self.flow.request.headers)
  result['request_content'] = self.flow.request.content
  return result
 
 @staticmethod
 def parser_multipart(content):
  if isinstance(content, str):
   res = re.findall(r'name=\"(\w+)\"\r\n\r\n(\w+)', content)
   if res:
    return "&".join([k + '=' + v for k, v in res])
   else:
    return ""
  else:
   return ""
 
 @staticmethod
 def parser_header(header):
  headers = {}
  for key, value in header.items():
   headers[key] = value
  return headers
 
 @staticmethod
 def decode_response_text(content):
  for _ in ['UTF-8', 'GB2312', 'GBK', 'iso-8859-1', 'big5']:
   try:
    return content.decode(_)
   except:
    continue
  return content