This website provides the supplementary stuff of the paper "To Preserve or Not to Preserve Invalid Solutions in Search-Based Software Engineering: A Case Study in Software Product Lines", accepted by ICSE'18: the 40th International Conference on Software Engineering, May 27 - 3 June 2018, Gothenburg, Sweden
We evaluated seven subjects in software produce lines: eCos, FreeBSD, Fiasco, uClinux, Linux, Drupal, and AmazonEC2. The data of these subjects, including feature models, attribute values, and attribute constraints, are publicly available. Here are the sources:
- eCos, FreeBSD, Fiasco, uClinux, Linux : [1]
- Drupal : [2]
- AmazonEC2 : [3]
We used SATIBEA [1] as a baseline and designed five algorithm variants (SATIBEAv1, SATIBEAv2, SATIBEAv3, SATIBEAv4, and SATIBEAv5) that incorporate SAT solving into the initial population generation and the mutation operator of IBEA in different ways. Moreover, one algorithm (SATIBEAv5) designed to preserve valid solutions all along the way further incorporates a subroutine that resolves non-Boolean constraints over integer or real-number variables together with arithmetic or relational operators, which cannot be straightforwardly addressed by the approaches relying on SAT solving. Note that the subroutine (src/lib/amazonMapping.jar) is enabled only for AmazonEC2, the only one containing non-Boolean attribute constraints in our subjects.
- src/lib/ includes the core jar packages used by all algorithms.
- src/common/ includes the common Jave code of all algorithms. The common code must be combined with the specific code (if it exists) to run a particular algorithm for a certain subject.
- results/ includes the experimental results for each algorithm and for each subject.
[1] Christopher Henard, Mike Papadakis, Mark Harman, and Yves Le Traon. 2015. Combining multi-objective search and constraint solving for configuring large software product lines. In Proceedings of 37th IEEE/ACM International Conference on Software Engineering (ICSE). 517–528.
[2] Ana B. Sánchez, Sergio Segura, José Antonio Parejo, and Antonio Ruiz Cortés. 2017. Variability testing in the wild: the Drupal case study. Software & Systems Modeling 16, 1 (2017), 173–194.
[3] Jesús García-Galán, Pablo Trinidad, Omer F. Rana, and Antonio Ruiz Cortés. 2016. Automated configuration support for infrastructure migration to the cloud. Future Generation Comp. Syst. 55 (2016), 200–212.