ITR: A Domain-Specific Language for Infopipes

ITR：Infopipes 的特定领域语言

• 批准号: 0219686
• 负责人: Andrew Black
• 金额: $ 37.97万
• 依托单位: Oregon Health & Science University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2002
• 资助国家: 美国
• 起止时间: 2002-09-01 至 2005-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0219686&HistoricalAwards=false
• 关键词: ITR; Domain; Specific; Language; Infopipes

This project is creating programming language technology and tools to aid the development of performance-sensitive distributed streaming applications, such as real-time sensor networks. The basic idea is to design programming abstractions that are specific to the domain of streaming applications, and to develop checking and compilation tools for mapping them to the underlying operating system mechanisms and network protocols.The approach is based on the "Infopipe" programming model, a novel approach to building streaming applications, currently under development by the PIs. Experiments with an early prototype have shown Infopipes to be very promising. However, language tools for simplifying the programming task, checking correctness properties and ensuring adequate performance of complex "Infopipelines" do not currently exist. This proposal outlines the research involved in creating these tools and presents some novel ideas on the language primitives and tools for Infopipes, the kinds of Infopipe properties that can be automatically checked, and the conceptual foundation that enables this checking to be rigorous.

该项目正在创建编程语言技术和工具，以帮助开发对性能敏感的分布式流媒体应用程序，如实时传感器网络。其基本思想是设计特定于流应用领域的编程抽象，并开发检查和编译工具，用于将其映射到底层操作系统机制和网络协议。该方法基于“Infopipe”编程模型，这是一种构建流应用的新方法，目前正在由PI开发。早期原型的实验表明，Infopipes非常有前途。然而，用于简化编程任务、检查正确性属性和确保复杂“信息管道”的充分性能的语言工具目前还不存在。该提案概述了创建这些工具所涉及的研究，并提出了一些关于Infopipes的语言原语和工具的新想法，可以自动检查的Infopipe属性的种类，以及使这种检查严格的概念基础。