CI/CD-Enhanced Conflict Resolution: Feasibility and Potential for Smarter Merging
- Contact:
- Alexander Boll
- Timo Kehrer
Context
Concurrent editing in cooperative software development often results in merge conflicts. Human effort is needed to resolve such merge conflicts. Studying merge conflicts and how to (semi-)automatically resolve them is an active research area.
Motivation
Recent research results have shown that the majority of merge conflicts is often resolved using simple patterns. However, an open question is how to decide, for a given conflict, whether it can be resolved using a simple pattern and, if so, which one. In this work, we want to explore whether CI/CD pipelines, notably the test and build stage, may help in answering this question.
Goal
The goal of this work to develop a technique that, given a conflicting merge, creates all the possible resolution variants in which these conflicts are resolved according to the aforementioned simple resolution patterns. The suitability of these variants, in turn, shall then be assessed by running the test and build stages.
Requirements
- Excellent knowledge in software development, including technologies used in CI/CD pipelines (test frameworks, build tools, etc.).
- Excellent programming skills to develop the research prototype and its usage in a fully automated experimental setup.
Pointers
- Boll, Alexander, et al. “Towards Semi-Automated Merge Conflict Resolution: Is It Easier Than We Expected?.” Proceedings of the 28th International Conference on Evaluation and Assessment in Software Engineering. 2024. https://dl.acm.org/doi/pdf/10.1145/3661167.3661197