Leveraging LLMs to support co-evolution between definitions and instances of textual DSLs: A Systematic Evaluation

TL;DR

This paper systematically evaluates the effectiveness of large language models in supporting the co-evolution of textual domain-specific languages and their instances, highlighting strengths and limitations across different scales and complexities.

Contribution

It provides an empirical assessment of LLMs like Claude and GPT-5.2 for grammar and instance co-evolution in textual DSLs, filling a gap in existing model-driven engineering approaches.

Findings

Strong performance on small-scale cases with high precision and recall.

Performance degradation observed with larger instances and increased complexity.

Response time increases significantly with instance size, indicating scalability limits.

Abstract

Software languages evolve over time for reasons such as feature additions. When grammars evolve, textual instances that originally conformed to them may become outdated. While model-driven engineering provides many techniques for co-evolving models with metamodel changes, these approaches are not designed for textual DSLs and may lose human-relevant information such as layout and comments. This study systematically evaluates the potential of large language models (LLMs) for co-evolving grammars and instances of textual DSLs. Using Claude Sonnet 4.5 and GPT-5.2 across ten case languages with ten runs each, we assess both correctness and preservation of human-oriented information. Results show strong performance on small-scale cases ($\geq$94% precision and recall for instances requiring fewer than 20 modified lines), but performance degraded with scale: Claude maintains 85% recall at 40 lines, while GPT fails on the largest instances. Response time increases substantially with instance size, and grammar evolution complexity and deletion granularity affect performance more than change type. These findings clarify when LLM-based co-evolution is effective and where current limitations remain.