Automatic code summarization aims to generate concise natural language descriptions (summary) for source code, which can free software developers from the heavy burden of manual commenting and software maintenance. Existing methods focus on learning a direct mapping from pure code to summaries, overlooking the significant heterogeneity gap between code and summary. Moreover, existing methods lack a human-like re-check process to evaluate whether the generated summaries match well with the code. To address these two limitations, we introduce RBCoSum, a novel framework that incorporates the generated Chain Of Comments (COC) as auxiliary intermediate information for the model to bridge the gap between code and summaries. Also, we propose a rethinking process where a learned ranker trained on our constructed ranking dataset scores the extent of matching between the generated summary and the code, selecting the highest-scoring summary to achieve a re-check process. We conduct extensive experiments to evaluate our approach and compare it with other automatic code summarization models as well as multiple code Large Language Models (LLMs). The experimental results show that RBCoSum is effective and outperforms baselines by a large margin. The human evaluation also proves the summaries generated with RBCoSum are more natural, informative, useful, and truthful.

Over the past decade, the application of deep learning in code clone detection has produced remarkable results. However, the current approaches have two limitations: (a) code representation approaches with low information utilization, such as vanilla Abstract Syntax Tree (AST), leading to information redundancy which results in performance degradation; (b) low efficiency of clone detection on evaluation, resulting in excessive time costs during practical use. In this paper, we propose a Multiple Representation Transformer with an Optimized Abstract Syntax Tree (MRT-OAST) to introduce an efficient code representation method while achieving competitive performance. Specifically, MRT-OAST strategically prunes and enhances the AST, utilizing both pre-order and post-order traversals to represent two different representations. To speed up the evaluation process, MRT-OAST utilizes a pure Siamese Network and employs cosine similarity to compare the similarity between codes. Our approach effectively reduces AST sequences to 40 % and 39 % of their original length in Java and C/C++ while preserving structural information. In code clone detection tasks, our model surpasses state-of-the-art approaches on OJClone and Google Code Jam. During the evaluation of BigCloneBench, our model has a 5x speed improvement compared to the state-of-the-art lightweight model and a 563x speed improvement compared to the BERT-based model, with only a 0.3 % and 0.9 % decrease in$F_{1}$-score.

The code summarization task aims to automatically generate natural language descriptions for code snippets. Recently, pre-trained code language models (CLMs) have demonstrated outstanding performance on code summarization. Additionally, researchers have shown that there is a strong correlation between code function names and summaries, and poorly defined function names lead to worse summaries generated by models. To mitigate this issue, in this paper, we propose CLCoSum, a curriculum learning-based code summarization method for CLMs that improves their performance in poorly named function scenarios. CLCoSum helps CLMs avoid overreliance on function names when they are poorly defined. First, CLCoSum employs data augmentation operators on function names to generate semantically equivalent poorly named codes, which are considered harder data and assist in reducing the model's reliance on unclear function names. Subsequently, CLCoSum uses a curriculum learning paradigm to allow the model to learn these harder codes in an organized way during finetuning. This approach enables CLMs to progress from easier to more difficult training data, similar to the human learning process. Extensive experiments on two existing datasets for Java and Python demonstrate that CLCoSum boosts the performance of various CLMs in code summarization. Specifically, the improvements in BLEU-4 score range from approximately 5% to$\mathbf{2 0. 8 \%}$. The fine-tuning speed of CLCoSum on the augmented dataset is also competitive. Our code and data are available at https://github.com/KuiH/CLCoSum.

Code generation aims to generate code in a general-purpose programming language, such as C++, based on natural language intents. Existing efforts primarily focus on relatively simple programming problems and fail to evaluate the thought process involved in complex programming scenarios. In this paper, we introduce “steps-guided code generation,” a task that assesses the quality of both thought steps and code implementation to evaluate the overall management of handling a complex programming problem. To support this task, we construct CodeStepsEval, a real-world scenario dataset of complex programming problems in the C++ programming language with varying levels of difficulty. Comprehensive experiments on this dataset demonstrate the importance of high-quality steps in enhancing code generation performance and the challenges faced by the code LLMs in this task.