The REquirements TRacing On target (RETRO).NET Dataset

For this dataset, we took the RETRO Requirement Specification (version 1.0, written to document the features in our original RETRO tool) and used RETRO.NET to trace it to the code used to implement RETRO.NET (copyright Jody Larsen). There are 66 requirements (functional requirements only) that have been extracted from the document. RETRO.NET expects all source elements to be in a folder and all target elements to be in a folder. Therefore, in our dataset, each requirement has been stored in its own file with the identifier as the file name and the file containing its text (RETRONET Requirements folder). The original requirements specification as well as the document subset have been provided in the dataset (as .docx and as data.txt, respectively) along with the python script (parser.py, copyright to Jared Payne) that was used to parse the text only version of the document subset. There are 118 code files in the dataset, primarily C# files. Each code file is listed in the code directory (RETRONET Trunk folder). The answer set has been provided in xml format (result.xml) as well as in traditional answer set format of our research group (results.txt).

搜集汇总

背景与挑战

背景概述

该数据集用于需求追踪研究，基于RETRO.NET工具的需求规范与实现代码，包含66个功能需求（每个需求单独存储）和118个C#代码文件。数据集提供了原始文档、解析脚本和答案集，支持需求与代码之间的追踪分析。