Improving the Capabilities and Applicability of Umple

提升Umple的能力和适用性

• 批准号: RGPIN-2016-05087
• 负责人: Lethbridge, Timothy
• 金额: $ 1.89万
• 依托单位: University of Ottawa
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2020
• 资助国家: 加拿大
• 起止时间: 2020-01-01 至 2021-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=709585
• 关键词: Improving; Capabilities; Applicability; Umple

In this work we will enable practitioners and students to develop better software by blending easy-to-use model-driven engineering with software engineering techniques such as agile methods and test-driven development. We will achieve this by continued work on Umple, a textual modeling technology that works seamlessly with existing programming languages and generates fully functional systems, diagrams, documentation and analysis results. Umple is textual because most programmers prefer to develop textually, and we want to be able to blend modeling into their work and thus encourage the use of modeling. Umple generates diagrams dynamically to facilitate analysis and understanding of models. Umple comprehensively incorporates Unified-Modeling Language (UML) modeling elements including class diagrams, composite structure diagrams and state machines. We have conducted empirical studies that have shown that Umple is easy to use. It is being used in both academic and industrial environments. In the proposed work we will 1) improve Umple's real-time code generation, particularly for concurrent, distributed composite structure models; 2) improve its separation-of-concerns mechanisms, particularly traits, variants and their synergies; 3) incorporate the requirements modeling standard GRL; and 4) enhance model analysis and validation. We will also continue to enhance Umple's debugging capability, and integration with other tools such as Papyrus. As case studies, we will apply our work to two domains: 1) Enterprise models for municipalities and universities, 2) Small autonomous model cars based on Arduino processors with CAN networking and various sensors (cameras with scene analysis and Lidar). The Umple-based platform will allow groups of researchers and even undergraduate or high-school students to explore different architectures and algorithms for autonomous driving. As with much of our recent research, we will follow a design-science approach, exploring opportunities that arise as we improve the design of both Umple itself and the software being developed in the above case studies. We will also conduct studies of users actually using Umple. Development of Umple will continue to be performed in a fully agile open-source manner, using model-driven development (Umple is developed in itself), test-driven development and continuous integration. The Umple project itself will continue to serve as a teaching tool for how to run a software engineering project.

在这项工作中，我们将通过将易于使用的模型驱动工程与软件工程技术(如敏捷方法和测试驱动开发)相结合，使从业者和学生能够开发更好的软件。 我们将通过继续研究Umple来实现这一点，Umple是一种文本建模技术，可以与现有编程语言无缝协作，并生成功能齐全的系统、图表、文档和分析结果。 Umple是文本的，因为大多数程序员更喜欢文本开发，而我们希望能够将建模融入到他们的工作中，从而鼓励使用建模。Umple动态生成图表，以便于分析和理解模型。 Umple全面整合了统一建模语言(UML)建模元素，包括类图、复合结构图和状态机。我们进行的实证研究表明，Umple很容易使用。它在学术和工业环境中都被使用。 在提出的工作中，我们将1)改进Umple的实时代码生成，特别是对于并发的、分布式的复合结构模型；2)改进其关注点分离机制，特别是特征、变体及其协同作用；3)纳入需求建模标准GRL；以及4)增强模型分析和验证。 我们还将继续增强Umple的调试能力，并与Papyrus等其他工具集成。 作为案例研究，我们将把我们的工作应用到两个领域：1)市政和大学的企业模型；2)基于带有CAN网络和各种传感器(带有场景分析和激光雷达的摄像头)的Arduino处理器的小型自动驾驶模型车。基于Umple的平台将允许研究人员小组，甚至本科生或高中生探索自动驾驶的不同架构和算法。 与我们最近的许多研究一样，我们将遵循设计科学的方法，探索随着我们改进Umple本身和上述案例研究中正在开发的软件的设计而出现的机会。我们还将对实际使用Umple的用户进行研究。 Umple的开发将继续以完全敏捷的开源方式进行，使用模型驱动的开发(Umple本身就是开发的)、测试驱动的开发和持续集成。Umple项目本身将继续作为如何运行软件工程项目的教学工具。