Conflict resolution for structured merge via version space algebra

Abstract

Resolving conflicts is the main challenge for software merging. The existing merge tools usually rely on the developer to manually resolve conflicts. This is of course inefficient. We propose an interactive approach for resolving merge conflicts. To the best of our knowledge, this is the first attempt for conflict resolution of structured merge. To represent the possibly very large set of candidate programs, we propose an expressive and efficient representation by version space algebra. We also design a simple mechanism for ranking resolutions in the program space, such that the top-ranked resolution is very likely to meet the developer's expectation. We prototype our approach as a merge tool AutoMerge, and evaluate it on 244 real-world conflicts arising from 10 open-source projects. Results show great practicality of our approach. CCS Concepts: • Software and its engineering → Software version control; Context specific languages;

Merging problems in modern version control systems

“…So, the developers must intervention to resolve conflicts manually [27]. Therefore, developers manually rely on existing tools to solve merging conflicts [8]. The current merge conflicts tools still rely on structured and unstructured merge strategies.…”

“…Many software projects use an unstructured merge strategy because it is very simple and relying on purely textual analysis to detect and resolve conflicts [2]. Where source code written in any programming language is regarded as plain text [8] and every software artifact that can be represented as text. So, each tool can handle the software artifacts conflicts successfully [16].…”

“…In Fengmin, Zhu et al (2018) authors proposed an interactive approach for resolving merge conflicts, to represent a very large set of programs proposed an expressive and efficient representation by version space algebra, the study was conducted and evaluated based on 244 real-world conflicts arising from 10 open-source projects. The study found that the AutoMerge detects 244 conflicts spread over 138 files, and successfully resolves as high as 95.1% of the conflicts [8]. McKee et al (2017) authors proposed a study on the factors that impact how practitioners approach merge conflicts and the difficulties they face when resolving conflicts.…”

IntelliMerge: a refactoring-aware software merging technique

“…On one hand, there are proposals that target at modeling artifacts like diagrams and models [Jackson et al 1994;Kolovos et al 2006;Mehra et al 2005]. On the other hand, there are proposals that focus on merging programs; some of them are dedicated to specific programming languages such as Java [Apel et al 2012;Leßenich et al 2015;Zhu and He 2018], C++ [Grass 1992] and XML [Lindholm 2004], while others try to be language-independent by representing program elements in an abstract level [Apel et al 2011;Buffenbarger 1993;Westfechtel 1991]. However, these structured techniques are usually quite expensive and not practical in real software development projects.…”

“…Modern software products are usually developed and maintained by a team of developers, either from a company or from an open-source community. In the development process, the version control system (VCS) plays an indispensable role by providing supports that facilitate the collaborative editing and revising of software artifacts among developers [Zhu and He 2018]. For example, an open-source project hosted on Github 1 usually follows a branch-based development workflow, where new features or bug fixes are developed in separate branches forked from a master branch and then different branches are merged into the master branch.…”