Abstract:

We present the artifact accompanying our paper, “Empirical Study of the Docker Smells Impact on the Image Size,” accepted at the ICSE main track. The artifact comprises two integral components:

1. Parfum Tool: This innovative tool facilitates the identification and rectification of Docker smells within Dockerfiles. Employed in our empirical study, Parfum evaluates the consequential impact of these smells. Notably, Parfum prioritizes reusability and accessibility, boasting a user-friendly interface available at https://durieux.me/docker-parfum/. This feature-rich tool contributes to the broader research community and industry practitioners alike.

2. Empirical Study Data: The artifact includes dataset derived from our empirical study. This dataset encompasses diverse Dockerfiles, detailed results of smell detection, and the subsequent repair of Dockerfiles.

We believe that this dataset, coupled with the Parfum tool, constitutes invaluable resources for researchers and practitioners.