Work

Contextual Graph Grammars for Software Evolution

Year: 2026
Type: preprint
Abstract: This paper introduces a novel approach to software evolution utilizing Contextual Graph Grammars (CGGs). Traditional software evolution techniques often rely on ad-hoc code mutation, lacking a formal understanding of the system's semantics and context-awareness. CGGs offer a solution by providing a rigorous, context-sensitive framework for modifying software architectures represented as graphs. The core claim is that automatically evolving software systems is achievable through the application of CGGs that capture the underlying semantic relationships. The proposed mechanism involves defining a grammar that represents the original software's architecture, followed by dynamic modification of the graph structure based on runtime context, feedback, and semantic constraints. This approach represents a significant departure from conventional mutation techniques, providing a formal and context-aware methodology for software evolution, ultimately leading to more robust and adaptable systems. The theoretical framework presented here outlines a process for evolving software based on contextual understanding, enabling more intelligent and targeted modifications.
Language: English
Subfield: Software
Open Access status: green
OpenAlex ID: W7204864510
Cites:
Cited by:
Related to: 10