SBIR Phase I: Programmer-Friendly Automatic Code Fixes

SBIR 第一阶段：程序员友好的自动代码修复

• 批准号: 1747219
• 负责人: Cosmin Radoi
• 金额: $ 22.5万
• 依托单位: Iuvo AI, Inc.
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2018
• 资助国家: 美国
• 起止时间: 2018-01-01 至 2018-12-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1747219&HistoricalAwards=false
• 关键词: SBIR; Phase; Programmer; Friendly; Automatic

The broader impact/commercial potential of this Small Business Innovation Research (SBIR) Phase I project is to develop and bring to market tools that improve code quality and help avoid dangerous bugs, while also improving programmer productivity. Poor code quality leads to bugs which often translate into financial loses and can even endanger human life. A study by NIST estimated software bugs cost $59 billion annually. The key aspect of this project is a programmer-friendly language for describing software fixes. Successful adoption of the language by the programmer community can have a far-reaching positive impact upon the software community: programmers will have a common, formal, language to discuss ways to fix bugs, and the result of their collaboration could be used to automatically fix software systems.This Small Business Innovation Research (SBIR) Phase I project will address the research and technical challenges in providing programmers with a language to express their knowledge of error patterns and their fixes, and a tool that uses this information to check and fix software systems. The research objective is to refine the language such that it is easy to use by programmers, and to provide a scalable infrastructure. This will be achieved by iteratively improving both the language and the supporting infrastructure with feedback from early users. The anticipated result is that the technology will be applicable to a wide variety of bugs and to large code bases.

这个小型企业创新研究（SBIR）第一阶段项目的更广泛的影响/商业潜力是开发和推向市场的工具，提高代码质量，帮助避免危险的错误，同时提高程序员的生产力。代码质量差会导致bug，这些bug通常会转化为经济损失，甚至可能危及人类生命。美国国家标准与技术研究院（NIST）的一项研究估计，软件错误每年造成590亿美元的损失。这个项目的关键方面是一个程序员友好的语言来描述软件修复。程序员社区成功采用该语言可以对软件社区产生深远的积极影响：程序员将有一个共同的，正式的，语言来讨论如何修复错误，他们合作的结果可以用来自动修复软件系统。这个小企业创新研究（SBIR）第一阶段项目将解决研究和技术挑战，为程序员提供一种语言来表达他们对错误模式及其修复的知识，以及使用这些信息来检查和修复软件系统的工具。研究的目标是改进语言，使其易于程序员使用，并提供一个可扩展的基础设施。这将通过迭代地改进语言和支持基础设施来实现，并从早期用户那里获得反馈。预期的结果是，该技术将适用于各种各样的错误和大型代码库。