Research Report on Feasibility Study of Building a QT Gui Testing Tool, AX Program Code Group Computer Science R&D Project

Grover, B T

Lawrence Livermore National Laboratory

关键词: Testing;    Computers;    Automation;    99;    Architecture;   

DOI  :  10.2172/15003918
RP-ID  :  UCRL-ID-153221
RP-ID  :  W-7405-ENG-48
RP-ID  :  15003918

美国|英语

来源: UNT Digital Library

【 摘 要 】

The main goal of this project was to determine if a tool could be built to test Qt. In determining the feasibility of building a tool the following requirements needed to be researched: (1) Determine if the underlying Qt signal/slot architecture could be leveraged. (2) Research how much impact implementing such a tool would have on existing code, i.e. how much extra code would need to be inserted to use the tool. (3) Determine with the above information if a tool could be built. With the above steps completed, the information needed to make a decision on building a tool could be made. Armed with this information I felt I could make a more educated decision on the possibility of building a tool. This project was divided into two main steps. The first step was to understand the underlying Qt source code much better. The second step was to build a small prototype that I could use to test ideas. The first step was actually much shorter than I had originally anticipated. Understanding the underlying architecture of Qt only took about two weeks. After studying the architecture of qt and working with the support people at Trolltech, the company that develops Qt, I found a way to test Qt. This project was very successful. I accomplished everything I intended to do. I learned and understood the inner workings of the Qt library enough that I could build a simple tool that could leverage some of the information in Qt to test the GUI. I was also able to find a tool that was commercially available to test Qt GUI's. These two things were the main goals of this project. Therefore I consider it a success. In fact I was able to progress farther with my prototype testing then I had originally planned.