Enhancing Support for Metaprogramming

鉴于软件的重要性日益增加，为软件开发人员提供技术、理论和工具来帮助支持这些持续的进步是势在必行的。在这方面具有巨大潜力的一种编程方法是元编程，即编写读取、处理和生成其他程序的程序的过程。 元编程具有明显的潜力，可以通过减少开发时间，提高软件质量和增加软件灵活性来显着改进软件。 不幸的是，元编程对于大多数主流软件开发人员来说太复杂了。为了解决这个问题，我的长期研究目标是通过使元程序更容易编写，优化和维护来扩展元程序设计的适用性。 这项研究将推进元编程的理解，方法和技术，并产生高素质的毕业生，他们掌握了最先进的软件开发技术，并准备在这一重要和高影响力的研究领域的研究和应用中不断取得进展。 我和我的学生将通过开发新的分析技术来帮助优化编译器加速元程序，从而使元程序更容易用于高性能软件。 我们还将通过扩展安全检查理论来支持使用可重分配变量的元程序，从而提高元程序的质量。 最后，我们将通过扩展C++语言的开发人员友好的功能，以在编译时生成定制的数据结构，来增加元程序的灵活性和可访问性。 这项研究将有助于更智能的编程实践，邀请更有效的方法来构建无bug的代码，并最终加速正在改变社会的软件革命背后的科学。 三个项目中的两个将产生新的编程工具，通用开发人员可以立即使用这些工具来改进他们的软件开发。加拿大软件开发行业将受益于这项研究所产生的技术，技术和高素质的人才。

Given the ever-increasing importance of software, it is imperative that software developers be provided with techniques, theories, and tools that can help support these continuing advances. One programming methodology with immense potential in this regard is metaprogramming, the process of writing programs that read, process, and generate other programs. Metaprogramming has clear potential to significantly improve software by reducing development time, increasing software quality, and increasing software flexibility. Unfortunately, metaprogramming has simply been too complex for most mainstream software developers to master. To address this, my long term research goal is to expand the applicability of metaprogramming by making metaprograms easier to write, optimize, and maintain. This research will advance the understanding, methodology, and technology of metaprogramming, and produce highly qualified graduates who have mastery of state-of-the-art software development techniques and are poised to lead ongoing advances in research and application of this vital and high-impact research area. My students and I will make metaprogramming easier to use for high-performance software by developing new analysis techniques that help optimizing compilers speed up metaprograms. We will also improve the quality of metaprograms by extending safety checking theories to support metaprograms that use re-assignable variables. Finally, we will increase the flexibility and accessibility of metaprograms by extending the C++ language with developer-friendly features to generate custom-tailored data structures at compile-time. This research will contribute to more intelligent programming practices, invite a more efficient means of building bug-free code, and ultimately accelerate the science behind the software revolution that is transforming society. Two of the three projects will produce new programming tools that generalist developers can immediately use to improve their software development. The Canadian software development industry will benefit from the techniques, technology, and highly-qualified personnel produced by this research.