WEB 自动化测试工具 Selenium 简介及其应用

jopen 10年前

最近我们要对站点的指定 url 测试其页面请求完整性,并分析每个请求参数是否正常。如果不使用自动化测试工具,你当然可以人肉点击、查看是否页面请求遗漏和每个请求参数是否正常。当然也有如 Fiddler 之类的抓包工具,但这个工具貌似没开源代码,也没有提供接口供第三方调用和扩展。找来找去发现 Selenium 可以满足当前的业务需求。

Selenium是一个用于Web应用程序测试的工具。Selenium测试直接运行在浏览器中,就像真正的用户在操作一样。支持的浏览器包括IE(7、8、9)、Mozilla Firefox、Mozilla Suite等。这个工具的主要功能包括:测试与浏览器的兼容性——测试你的应用程序看是否能够很好得工作在不同浏览器和操作系统之上。测试系统功能——创建衰退测试检验软件功能和用户需求。支持自动录制动作和自动生成 .Net、Java、Perl等不同语言的测试脚本。Selenium 是ThoughtWorks专门为Web应用程序编写的一个验收测试工具。

1、Selenium 抓取请求的原理

其实 Selenium 抓取请求的原理和 fiddler 之类的工具是类似的:

WEB 自动化测试工具 Selenium 简介及其应用

只是这里 Selenium 代理取代了 Fiddler 代理,而自动化测试脚本则通过 Selenium 去操纵浏览器或者其里面的页面元素去模拟人的行为,进而完成预设的测试任务。

2、Selenium 自动化测试实例:获取指定 url 的所有请求并简单统计分析

废话就不说了,直接上代码了。需要提示下的是,如果你想测试下面的代码,请按如下步骤来:

(1)Install Python
(2)Install Java
(3)Install Selenium bindings for Python
(4)Run Selenium server: 'java -jar selenium-server.jar'
(5)Run web_profiler.py

# -*- coding: utf-8 -*-  #!/usr/bin/env python  #  #  Selenium Web/HTTP Profiler  #  Copyright (c) 2009-2011 Corey Goldberg (corey@goldb.org)  #  License: GNU GPLv3    #  run:  #  python web_profiler.py http://baidu.com/    import json  import socket  import sys  import time  import urlparse  import xml.etree.ElementTree as etree  from datetime import datetime    from selenium import selenium        def main():      if len(sys.argv) < 2:          print 'usage:'          print '  %s <url> [browser_launcher]' % __file__          print 'examples:'          print '  $ python %s www.google.com' % __file__          print '  $ python %s http://www.google.com/ *firefox\n' % __file__          sys.exit(1)      else:          url = sys.argv[1]          if not url.startswith('http'):              url = 'http://' + url          parsed_url = urlparse.urlparse(url)          site = parsed_url.scheme + '://' + parsed_url.netloc          path = parsed_url.path          if path == '':              path = '/'          browser = '*firefox'        if len(sys.argv) == 3:          browser = sys.argv[2]        run(site, path, browser)        def run(site, path, browser):      sel = selenium('127.0.0.1', 4444, browser, site)        try:          sel.start('captureNetworkTraffic=true')      except socket.error:          print 'ERROR - can not start the selenium-rc driver. is your selenium server running?'          sys.exit(1)        sel.open(path)      sel.wait_for_page_to_load(90000)      end_loading = datetime.now()        raw_xml = sel.captureNetworkTraffic('xml')        sel.stop()        traffic_xml = raw_xml.replace('&', '&amp;').replace('=""GET""', '="GET"').replace('=""POST""', '="POST"') # workaround selenium bugs        nc = NetworkCapture(traffic_xml)        #json_results = nc.get_json()        num_requests = nc.get_num_requests()      total_size = nc.get_content_size()      status_map = nc.get_http_status_codes()      file_extension_map = nc.get_file_extension_stats()      http_details = nc.get_http_details()      start_first_request, end_first_request, end_last_request = nc.get_network_times()        end_load_elapsed = get_elapsed_secs(start_first_request, end_loading)      end_last_request_elapsed = get_elapsed_secs(start_first_request, end_last_request)      end_first_request_elapsed = get_elapsed_secs(start_first_request, end_first_request)        print '--------------------------------'      print 'results for %s' % site        print '\ncontent size: %s kb' % total_size        print '\nhttp requests: %s' % num_requests      for k,v in sorted(status_map.items()):          print 'status %s: %s' % (k, v)        print '\nprofiler timing:'      print '%.3f secs (page load)' % end_load_elapsed      print '%.3f secs (network: end last request)' % end_last_request_elapsed      print '%.3f secs (network: end first request)' % end_first_request_elapsed        print '\nfile extensions: (count, size)'      for k,v in sorted(file_extension_map.items()):          print '%s: %i, %.3f kb' % (k, v[0], v[1])        print '\nhttp timing detail: (status, method, doc, size, time)'      for details in http_details:          print '%i, %s, %s, %i, %i ms' % (details[0], details[1], details[2], details[3], details[4])        def get_elapsed_secs(dt_start, dt_end):      return float('%.3f' % ((dt_end - dt_start).seconds +          ((dt_end - dt_start).microseconds / 1000000.0)))        class NetworkCapture(object):      def __init__(self, xml_blob):          self.xml_blob = xml_blob          self.dom = etree.ElementTree(etree.fromstring(xml_blob))          def get_json(self):          results = []          for child in self.dom.getiterator():              if child.tag == 'entry':                  url = child.attrib.get('url')                  start_time = child.attrib.get('start')                  time_in_millis = child.attrib.get('timeInMillis')                  results.append((url, start_time, time_in_millis))          return json.dumps(results)          def get_content_size(self):  # total kb passed through the proxy          byte_sizes = []          for child in self.dom.getiterator():              if child.tag == 'entry':                  byte_sizes.append(child.attrib.get('bytes'))          total_size = sum([int(bytes) for bytes in byte_sizes]) / 1000.0          return total_size          def get_num_requests(self):          num_requests = 0          for child in self.dom.getiterator():              if child.tag == 'entry':                  num_requests += 1          return num_requests          def get_http_status_codes(self):          status_map = {}          for child in self.dom.getiterator():              if child.tag == 'entry':                  try:                      status_map[child.attrib.get('statusCode')] += 1                  except KeyError:                      status_map[child.attrib.get('statusCode')] = 1          return status_map          def get_http_details(self):          http_details = []          for child in self.dom.getiterator():              if child.tag == 'entry':                  url = child.attrib.get('url') + '?'                  # url = child.attrib.get('url')                  print "---->> " + url                  url_stem = url.split('?')[0]                  doc = '/' + url_stem.split('/')[-1]                  status = int(child.attrib.get('statusCode'))                  method = child.attrib.get('method').replace("'", '')                  size = int(child.attrib.get('bytes'))                  time = int(child.attrib.get('timeInMillis'))                  http_details.append((status, method, doc, size, time))          http_details.sort(cmp=lambda x,y: cmp(x[3], y[3])) # sort by size          return http_details          def get_file_extension_stats(self):          file_extension_map = {}  # k=extension v=(count,size)          for child in self.dom.getiterator():              if child.tag == 'entry':                  size = float(child.attrib.get('bytes')) / 1000.0                  url = child.attrib.get('url') + '?'                  url_stem = url.split('?')[0]                  doc = url_stem.split('/')[-1]                  if '.' in doc:                      file_extension = doc.split('.')[-1]                  else:                      file_extension = 'unknown'                  try:                      file_extension_map[file_extension][0] += 1                      file_extension_map[file_extension][1] += size                  except KeyError:                      file_extension_map[file_extension] = [1, size]          return file_extension_map          def get_network_times(self):          timings = []          start_times = []          end_times = []          for child in self.dom.getiterator():              if child.tag == 'entry':                  timings.append(child.attrib.get('timeInMillis'))                  start_times.append(child.attrib.get('start'))                  end_times.append(child.attrib.get('end'))          start_times.sort()          end_times.sort()          start_first_request = self.convert_time(start_times[0])          end_first_request = self.convert_time(end_times[0])          end_last_request = self.convert_time(end_times[-1])          return (start_first_request, end_first_request, end_last_request)          def convert_time(self, date_string):          if '-' in date_string: split_char = '-'          else: split_char = '+'          # dt = datetime.strptime(''.join(date_string.split(split_char)[:-1]), '%Y%m%dT%H:%M:%S.%f')          dt = datetime.strptime(''.join(date_string.split(split_char)[:-1]), '%Y%m')          return dt        if __name__ == '__main__':      main()

最后的结果输出了所有的请求,并 对这次的请求进行简单的分析, 包括异步加载的 ajax、图片啥的,和你在浏览器 F12→network 看到的效果类似。

WEB 自动化测试工具 Selenium 简介及其应用

3、Refer:

(1)http://docs.seleniumhq.org/

(2)Selenium私房菜系列--总章

http://www.cnblogs.com/hyddd/archive/2009/05/30/1492536.html

https://code.google.com/p/selenium-profiler/