LayoutRepair Tool, Automated Tool to Repair Layout Issues in Web Pages Using Galen Testing Tool

Abstract

During the implementation or improvement of a web application, developers may cause some issues which affect the appearance of the layout. Therefore, they tend to use automated testing tools to test the web page based on its specifications to ensure that its appearance is rendered correctly on a web browser. Galen is a common automated test framework for the layout of web applications. However, Galen only detects layout issues in the web applications. The developer must manually fix detected problems, based on Galen report. To address that limitation, this dissertation proposes a tool to automatically repair layout issues likely to occur as regression debugging problems in web applications using the Galen testing tool. The scope of this research addresses repairing only the failed element’s position values in the HTML and CSS files. The repair process was completed by developing two different techniques: positionSpec and guided search-based methods. To improve the efficiency of these techniques, a dependability detection algorithm has been designed to find the root cause of a layout failure. When the developed automated tool was applied to 10 real-world web pages obtained using a random URL generator, an evaluation of the results showed that the repairing tool was able to successfully repair 80% of these web pages using the positionSpec technique and 40% using the search-based algorithm. However, as far as can be determined, no successful work has been done to repair the layout issue using the Galen testing tool.