ReCover Dataset

ReCover dataset Includes 28 projects with 114 pair of versions of open source software mined from TravisCI and GitHub. The dataset is organized as follows: - Projects: is the folder which stores the 7z archives of all projects in the dataset - Scripts: folder with auxiliary scripts to analyze the dataset and to retrieve coverage information of a builder project - dataset.db: the SQLite file storing the dataset projects' structure - artifactData.xml: XML file containing detailed information on projects, builds and commits in the dataset To use a project, just uncompress the related archive in the Projects folder. Structure of Projects After a project is uncompressed, its inner folder structure is the following: - project@slugname/ # The root folder of the project | |- buildId/ # Folder with the pair of versions included in the build | | |- commitSHA/ # Commit folder for the previous or current version | | | |- sources/ # Folder with the original commit sources | | | |- surefire-reports/ # Folder to readily locate the test eecution reports of the commit | | | |- clover-report.xml # XML report of coverage, at statement and per-test granularity level | | | |- mavenCli.log # Dump of the raw Maven output when the build was executed | | |- commitSHA/ | | |- commitSHA_previous.txt # Empty file telling the SHA of the previous version among the pair | | |- commitSHA_current.txt # Empty file telling the SHA of the current version in the pair | |- buildId/ | ..