Abstract The quality of software is closely tied to the effectiveness of the tests it undergoes. Manual test writing, though crucial for bug detection, is time-consuming, which has driven significant research into automated test case generation. However, current methods often struggle to generate relevant inputs, limiting the effectiveness of the tests produced. To address this, we introduce BRMiner , a novel approach that leverages Large Language Models (LLMs) in combination with traditional techniques to extract relevant inputs from bug reports, thereby enhancing automated test generation tools. In this study, we evaluate BRMiner using the Defects4J benchmark and test generation tools such as EvoSuite and Randoop. Our results demonstrate that BRMiner achieves a Relevant Input Rate (RIR) of 60.03% and a Relevant Input Extraction Accuracy Rate (RIEAR) of 31.71%, significantly outperforming methods that rely on LLMs alone. The integration of BRMiner’s input enhances EvoSuite ability to generate more effective test, leading to increased code coverage, with gains observed in branch, instruction, method, and line coverage across multiple projects. Furthermore, BRMiner facilitated the detection of 58 unique bugs, including those that were missed by traditional baseline approaches. Overall, BRMiner ’s combination of LLM filtering with traditional input extraction techniques significantly improves the relevance and effectiveness of automated test generation, advancing the detection of bugs and enhancing code coverage, thereby contributing to higher-quality software development.

Software development requires several stages of code iterations, each one requiring debugging, testing, localizing and fixing bugs. While several tools have been developed to automate one of those tasks individually, integrating and combining those results still requires a huge manual effort from software developers. Additionally, many approaches, e.g., in Automated Program Repair, are based on specifically curated fix templates that fail to generalize to most complex software projects. To address those challenges, I propose a new research agenda to learn the effects of software changes in order to localize and fix bugs without being limited to a specific project or a specific programming language. Additionally, my approach can be used to predict the effects of software changes. My preliminary results indicate the feasibility of successfully training a model on Python software changes and their effects, that is capable of producing 80% of correct patches and predicting the effect of a change with an accuracy of 81%. My results highlight the potential of learning the effects of software changes for better software development.

Recent successes in training word embeddings for Natural Language Processing ( NLP ) tasks have encouraged a wave of research on representation learning for source code, which builds on similar NLP methods. The overall objective is then to produce code embeddings that capture the maximum of program semantics. State-of-the-art approaches invariably rely on a syntactic representation (i.e., raw lexical tokens, abstract syntax trees, or intermediate representation tokens) to generate embeddings, which are criticized in the literature as non-robust or non-generalizable. In this work, we investigate a novel embedding approach based on the intuition that source code has visual patterns of semantics. We further use these patterns to address the outstanding challenge of identifying semantic code clones. We propose the WySiWiM ( ‘ ‘What You See Is What It Means ” ) approach where visual representations of source code are fed into powerful pre-trained image classification neural networks from the field of computer vision to benefit from the practical advantages of transfer learning. We evaluate the proposed embedding approach on the task of vulnerable code prediction in source code and on two variations of the task of semantic code clone identification: code clone detection (a binary classification problem), and code classification (a multi-classification problem). We show with experiments on the BigCloneBench (Java), Open Judge (C) that although simple, our WySiWiM approach performs as effectively as state-of-the-art approaches such as ASTNN or TBCNN. We also showed with data from NVD and SARD that WySiWiM representation can be used to learn a vulnerable code detector with reasonable performance (accuracy ∼90%). We further explore the influence of different steps in our approach, such as the choice of visual representations or the classification algorithm, to eventually discuss the promises and limitations of this research direction.