【作者】 康宇航；

【导师】 李建明；

【作者基本信息】 大连理工大学 ， 计算机技术（专业学位）， 2019， 硕士

【摘要】 移动互联网的发展使得手机终端设备正逐渐替代个人电脑成为人们在工作生活中进行信息发布、社交、数字办公、购物的主要工具。在一些场景下,如营销人员需要在APP上不断发帖宣传产品,大量重复性的操作降低人们的工作效率且容易导致操作失误。同时,大数据时代的到来使APP所承载的数据信息愈发庞大,从APP中爬取有价值的数据成为当前的热点需求。虽然业界已经采用APP网络抓包的方法爬取移动APP数据,但由于数据保护技术的存在,网络抓包方法不具有通用性。为了解决上述问题,本文设计并实现了安卓平台程序自动化控制系统。自动化控制系统可以在一台PC上运行,同时控制多台安卓手机终端的不同APP自动执行系统预定义的操作任务或自动爬取信息流APP的数据信息。本文通过为APP控件元素生成5种不同呈现形式的XPath路径表达式并提出最优XPath路径表达式选取算法,帮助自动化控制系统高效稳定地定位控件元素。同时,本文设计开发控件定位数据获取工具与定位数据管理平台,实现APP控件元素定位数据配置文件的一键生成。自动化控制系统基于数据驱动模式和PO分层架构设计思想让定位数据配置文件与控制命令脚本分离解耦,系统加载定位数据配置文件,结合编写的简单控制命令脚本实现对多台安卓手机APP的自动化控制。对于信息流APP数据信息的爬取,本文设计实现的自动化控制系统抛弃传统的网络抓包方式,构建控件元素自动遍历模型,采用用户行为模拟技术实现信息流APP数据信息的持久化。在系统运行期间,PC端将创建每台手机终端对应的监控器,监控器映射APP自动运行过程,用户可以在一台PC上查看多台手机终端的运行情况。更多还原

【Abstract】 The development of mobile Internet has made mobile devices gradually replace personal computers as the main tools for people to publish information,socialize,digital office and shop in work and life.In some scenarios,such as marketers need to constantly post products on the APP,a large number of repetitive operations reduce people’s work efficiency and easily lead to operational errors.At the same time,the arrival of the era of big data makes the data information carried by APPs more and more huge.Climbing valuable data from APPs becomes the current hot demand.Although the industry has adopted the APP network packet capture method to crawl mobile APP data,due to the existence of data protection technology,the network packet capture method is not universal.In order to solve the above problems,we designed and implemented the automatic control system based on Android platform application.The automatic control system can be run on one PC,and simultaneously control different APPs of multiple Android mobile phone to automatically execute predefined operational tasks of the system or automatically crawl the data information of the information flow APP.This thesis proposes an optimal XPath path expression selection algorithm to generate five different XPath path expressions for an APP control element and select the best XPath path expression from it,so that the system can locate the control elements efficiently and stably.At the same time,we designed and implemented the control positioning data acquisition tool and the positioning data management platform to realize the one-click generation of the positioning data configuration file of the APP control elements.Based on the data-driven model and PO layered architecture design,the automation control system can separate the positioning data configuration file from the control command script.The system loads the positioning data configuration file,and combines the simple control command script to realize the automatic control of multiple Android mobile phone APPs.For the crawling of information flow APP data information,the automatic control system discards the traditional network packet capture mode,constructs the control element automatic traversal model,and uses the user behavior simulation technology to realize the persistence of the information flow APP data information.During the system operation,the PC will create a monitor corresponding to each mobile phone.The monitormaps the APP automatic running process,and the user can view the running status of multiple mobile on one PC.更多还原