Scalable and Accurate Test Case Prioritization in Continuous Integration Contexts

This dataset is a benchmark of 25 open-source subjects with 21.5k builds and 2.5k failed builds that enables a fair comparison and evaluation of Test Case Prioritization (TCP) techniques. We made our data collection tools available (github.com/Ahmadreza-SY/TCP-CI), which can be used to extend and update the subjects. The description of the structure and files of the dataset can be also found in the documentation of the data collection tool.

Please refer to our academic paper, which can be found on arxiv.org/abs/2109.13168, for details on definitions, experiments, and results. Please cite our paper in any published work that uses resources that are provided in this dataset.

We provide two compressed files:

• TCP-CI-dataset.tar.gz: This file contains the dataset, source code of the subjects, the build logs, and the results of the experiments which were conducted in our research. In other words, this file includes all the required resources to replicate the study, and therefore its size is significantly large.
• TCP-CI-main-dataset.tar.gz: This file only contains the dataset which is described in our GitHub repository (link).