Program Structure-aware Language Models: Targeted Software Testing beyond Textual Semantics
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910146767093760 |
|---|---|
| author | Tran, Khang Nguyen, Khoa Borcea, Cristian Phan, NhatHai |
| author_facet | Tran, Khang Nguyen, Khoa Borcea, Cristian Phan, NhatHai |
| contents | Recent advances in large language models for test case generation have improved branch coverage via prompt-engineered mutations. However, they still lack principled mechanisms for steering models toward specific high-risk execution branches, limiting their effectiveness for discovering subtle bugs and security vulnerabilities. We propose GLMTest, the first program structure-aware LLM framework for targeted test case generation that seamlessly integrates code property graphs and code semantics using a graph neural network and a language model to condition test case generation on execution branches. This structured conditioning enables controllable and branch-targeted test case generation, thereby potentially enhancing bug and security risk discovery. Experiments on real-world projects show that GLMTest built on a Qwen2.5-Coder-7B-Instruct model improves branch accuracy from 27.4% to 50.2% on TestGenEval benchmark compared with state-of-the-art LLMs, i.e., Claude-Sonnet-4.5 and GPT-4o-mini. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_17715 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Program Structure-aware Language Models: Targeted Software Testing beyond Textual Semantics Tran, Khang Nguyen, Khoa Borcea, Cristian Phan, NhatHai Software Engineering Machine Learning Recent advances in large language models for test case generation have improved branch coverage via prompt-engineered mutations. However, they still lack principled mechanisms for steering models toward specific high-risk execution branches, limiting their effectiveness for discovering subtle bugs and security vulnerabilities. We propose GLMTest, the first program structure-aware LLM framework for targeted test case generation that seamlessly integrates code property graphs and code semantics using a graph neural network and a language model to condition test case generation on execution branches. This structured conditioning enables controllable and branch-targeted test case generation, thereby potentially enhancing bug and security risk discovery. Experiments on real-world projects show that GLMTest built on a Qwen2.5-Coder-7B-Instruct model improves branch accuracy from 27.4% to 50.2% on TestGenEval benchmark compared with state-of-the-art LLMs, i.e., Claude-Sonnet-4.5 and GPT-4o-mini. |
| title | Program Structure-aware Language Models: Targeted Software Testing beyond Textual Semantics |
| topic | Software Engineering Machine Learning |
| url | https://arxiv.org/abs/2604.17715 |