随着软件产业的快速发展和软件规模、复杂性的不断增长，软件测试的重要性日益体现，只靠传统的手动测试方式已无法满足当前和未来软件产品生产需求，而软件自动化测试技术的产生很大程度地提高了软件测试工作的效率，使得部分枯燥，乏味，重复性的手动测试可以通过录制回放和脚本编写的方式自动化实现。本文对自动化测试的现状，自动化测试工具实现的基本原理以及自动化测试的应用前景进行了研究，对各种商业自动化测试工具软件和开源测试工具软件的优缺点进行了分析，并针对实际项目需求，设计和实现了一个可扩展的自动化测试框架，并对该自动化测试框架各个模块进行了分析和阐述。该自动化测试框架主体使用Python编程语言实现，在框架主体中实现了灵活的开源测试工具包嵌入，并相应地定义了详细的框架主体与嵌入工具包之间的数据交互和通信流程，在测试结果存储上，采用了Wiki架构，实现了测试结果上传至Wiki所需的接口，使得测试用户对测试结果的查询和处理方式更加灵活便捷，在框架配置上采用XML语言配置方式，用户可以对测试中涉及到的各个参数进行灵活配置，并且在后期开发中可以对配置文件进行扩展，同时，该自动化测试框架还实现了超时控制和测试异常错误自动恢复功能，并提供了对分布式测试和数据驱动测试的支持。

As the rapid development of software industry, the size and complexity of software product is growing swiftly, which significantly increased the demand of a rather high-efficient way of software testing. The traditional way of simply executing manual software testing no longer satisfies the demand, and taking advantage of the software testing automation technology to accomplish part of the boring and repetitive testing process has now become prevalent among Software QA engineers. This paper gives an overview of the currently prevailing commercial and open-source software testing automation tools, in addition with the mechanism how different sorts of tools manage to achieve automated testing of software product. Based upon the analysis and project requirements, a flexible software testing automation framework is proposed and the detailed implementation of each module is illustrated in this paper. The main body of the framework is programmed with python language, plus the embedded open-source tool kit which could be replaced by other kits conveniently because of a flexible pattern adopted in this framework to embed tool kit. Wiki architecture is employed in this framework for result storage, and the framework could be configured flexibly with XML language. Moreover, the distributed testing mode and data-driven testing mode is implemented in this framework.