arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2405.06233cs.PL
收藏

无畏之旅 [草稿]

The Fearless Journey [Draft]

Nick Webster, Marco Servetto, Michael Homer

AI总结:

本文提出最小可扩展对象演算 Fearless Heart,并通过引用能力扩展为 R-Fearless,以兼顾真实应用构建、副作用支持与函数式推理。

摘要:

现有的最小面向对象(OO)模型,如 Featherweight Java(FJ),对于程序建模以及设计新的编程语言和工具很有价值。然而,它们在开发真实世界程序方面的实用性受到限制。我们提出“Fearless Heart”,这是一种新颖的对象演算,保留了 FJ 的最小性和可扩展本质,同时更适合构建复杂的真实世界应用。为说明 Fearless Heart 的可扩展性,我们用引用能力(Reference Capabilities,RC)对其进行扩展,创建了 R-Fearless。它支持可变性和其他副作用,同时保留函数式编程的推理优势,并获得对众所周知由 RC 启用的特性的支持,例如自动并行化、缓存和不变量。R-Fearless 仍然足够最小,允许进一步扩展。它是构建实用系统和形式化模型的理想基础。

英文摘要:

Existing minimal Object-Oriented models (OO), like Featherweight Java (FJ), are valuable for modelling programs and designing new programming languages and tools. However, their utility in developing real-world programs is limited. We introduce the 'Fearless Heart', a novel object calculus preserving FJ's minimal and extensible nature while being more suited for constructing complex, real-world applications. To illustrate the extensibility of the Fearless Heart, we extend it with Reference Capabilities (RC), creating R-Fearless. It supports mutability and other side effects while retaining the reasoning advantages of functional programming and gaining support for features that are well-known to be enabled by RC, like automatic parallelism, caching and invariants. R-Fearless is still minimal enough to allow further extensions. It is an ideal foundation for constructing both practical systems and formal models.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑