IR-based TCP Empirical

Raw Resources

File projectinfo.csv contains the information to necessary resources used in our paper.

In the file, every line has the structure: slug,buildID,jobID,prevBuildID,prevJobID,SHA,prevSHA. For example:

apache/incubator-dubbo,478229404,478229405,478226842,478226844,0721544fed8f,a8af5ce028e7

We can use the information above to construct the links to the resources needed:

slug, buildID: The details about a specific Travis build.
- Rich html: https://travis-ci.org/${slug}/builds/${buildID}
- Json file: https://api.travis-ci.org/builds/${buildID}

slug, jobID: The details about a specific Travis job.
- Rich html: https://travis-ci.org/${slug}/jobs/${jobID}
- Json file: https://api.travis-ci.org/jobs/${jobID}
- Job log: https://api.travis-ci.org/jobs/${jobID}/log

prevBuildID: The details of the previous build.
- Rich html: https://travis-ci.org/${slug}/builds/${prevBuildID}
- Json file: https://api.travis-ci.org/builds/${prevBuildID}

prevJobID: The details of the corresponding previous job.
- Rich html: https://travis-ci.org/${slug}/jobs/${prevJobID}
- Json file: https://api.travis-ci.org/jobs/${prevJobID}
- Job log: https://api.travis-ci.org/jobs/${prevJobID}/log

slug, SHA: The details of the GitHub commit at the current version.
- Rich html: https://github.com/${slug}/commit/${SHA}
- Zip file: https://github.com/${slug}/archive/${SHA}.zip

slug, prevSHA: The details of the GitHub commit at the previous version.
- Rich html: https://github.com/${slug}/commit/${prevSHA}
- Zip file: https://github.com/${slug}/archive/${prevSHA}.zip

File builds.zip contains 4 json files for each Travis build in the dataset:

changedfiles.json: the content and name for each changed file between the old and new code version.
filepatches.json: the patch (unified diff) for each changed file between the old and new code version.
morediff.json: 0, 3, 10 lines of the context of change for each changed file between the old and new code version.
morediff2.json: 1, 2, 5 lines of the context of change for each changed file between the old and new code version.

File jobs.zip contains 2 json files for each Travis job in the dataset:

testfiles.json: the content, state, name and execution time for each test class. In this file, failed tests come before passed tests.
dotestfiles.json: the content, state, name and execution time for each test class. In this file, the order of tests preserves the order in the original job log.

File prevjobs.zip contains 1 json file for each Travis job in the dataset:

File covorganized.zip contains several txt files for each job in the dataset that have coverage information collected:

File report_flaky.csv contains whether the job is flaky or not. 1 mean flaky, 0 means not flaky.