Commits messages play a crucial role in helping developers efficiently comprehend code modifications. Due to the time pressure of project iteration or poor message-writing practices, many commits suffer from missing messages. To address this issue, researchers have explored the automated generation of commit messages. Because of the truncation mechanism of the learning-based model, most of the current studies focus on code changes appearing at the beginning of a commit into the model for commit message generation. This may not be the best strategy for commit message generation because each code change in a commit contributes unequally to its overall purpose. To better generate commit messages, we propose a novel method that identifies the core code change in a commit for commit message generation. Specifically, we employ a method to predict the relative importance of the classes contained in a commit, and the code change of the class with the highest importance score (i.e., core change) is used to generate the commit message. Incorporating core change information can boost the performance of other existing methods (such as NMT, NNGen, and CoreGen). Building on this insight, we develop CCGen—a Core Change-Based Generation model that integrates a Transformer architecture with CodeBERT-enhanced encoding to leverage code semantics. The experiment demonstrates that the proposed method for commit message generation outperforms the state-of-the-art by 18.47% on average across seven metrics including 19.97 on ROUGE-L.

Pre-trained Vision-Language Models (VLMs) have demonstrated promising performance in remote sensing image-text retrieval tasks. However, the scarcity of high-quality image-text datasets remains a challenge in fine-tuning VLMs for remote sensing. The captions in existing datasets tend to be uniform and lack details. To fully utilize rich detailed information from remote sensing images, we propose a method to fine-tune VLMs. We first construct a new visual-language dataset that balances both Global and Local information for Remote Sensing image-text retrieval (GLRS). Specifically, a Multi-modal Large Language Model (MLLM) is utilized to generate captions for local patches and global captions for the entire image. To effectively utilize local information, we propose a Global and Local image Captioning method (GLCap). With a Large Language Model (LLM), we further obtain higher-quality captions by merging both global and local captions. Finally, we fine-tune the weights of RS-M-CLIP with a progressive global-local fine-tuning strategy on GLRS. Experimental results demonstrate that our method outperforms state-of-the-art approaches on two common remote sensing image-text retrieval downstream tasks. The dataset will be publicly available once the paper is accepted.

Abstract In software development and maintenance, code comments can help developers understand source code and improve communication among developers. However, developers sometimes neglect to update the corresponding comment when changing the code, resulting in outdated comments (i.e., inconsistent codes and comments). Outdated comments are dangerous and harmful and may mislead subsequent developers. More seriously, the outdated comments may lead to a fatal flaw sometime in the future. To automatically identify the outdated comments in source code, we proposed a learning‐based method, called CoCC, to detect the consistency between code and comment. To efficiently identify outdated comments, we extract multiple features from both codes and comments before and after they change. Besides, we also consider the relation between code and comment in our model. Experiment results show that CoCC can effectively detect outdated comments with precision over 90%. In addition, we have identified the 15 most important factors that cause outdated comments and verified the applicability of CoCC in different programming languages. We also used CoCC to find outdated comments in the latest commits of open source projects, which further proves the effectiveness of the proposed method.

Abstract Ethereum, as a leading blockchain platform, has attracted a significant number of practitioners. These practitioners require a platform for communication and collaborative problem‐solving, which led to Ethereum Stack Exchange (ESE), a Q&A site dedicated to Ethereum‐related issues. While the Q&A site facilitates communication among practitioners, it also introduces new challenges. Practitioners adopt code snippets from Q&A sites to address problems encountered. However, the quality of code snippets on ESE remains largely unexplored. Vulnerabilities and gas‐inefficient patterns in ESE may spread to the code in Ethereum and threaten its regular operation. In this article, we conduct an empirical study investigating the distribution of vulnerabilities and gas‐inefficient patterns in ESE. Further, we analyze the potential impact of vulnerabilities and gas‐inefficient patterns from ESE on Ethereum. However, we encounter a problem during the vulnerability and gas‐inefficient pattern detection. Established smart contract analysis tools in the mainstream realm necessitate complete source code files for thorough analysis, while codes on ESE are often incomplete code snippets. To address this, we introduce the AST‐based code clone detection technique to construct detectable files corresponding to code snippets. This enables us to detect vulnerabilities and gas‐inefficient patterns in code snippets. In the end, our findings demonstrate that 11.18% of the contract‐level code snippets and 4.06% of function‐level code snippets in ESE have vulnerabilities. And 27.21% of contract‐level code snippets and 17.89% of function‐level code snippets contain gas‐inefficient patterns. The additional consumption caused by the gas‐inefficient pattern in ESE is approximately $1,695,002. Based on these findings, we provide recommendations for both ESE and its users, aiming to foster collaborative efforts and create a more reliable Q&A site for practitioners.

High-quality and appropriate commit messages help developers to quickly understand and track code evolution, which is crucial for the collaborative development and maintenance of software. To relieve developers of the burden of writing commit messages, researchers have proposed various techniques to generate commit messages automatically, among which learning-based techniques have proven to be promising.

Few-shot classification aims to recognize the query sample through a limited amount of support data, where a prototype classifier is commonly applied. However, although the prototype classifier is simple and non-parametric, it does not fully utilize the prior information of samples, leading to prototype bias. To this end, we propose a Feature-weighted Multi-stage Bayesian Prototype Classifier (FMBPC). Specifically, we utilize a feature weighting module to balance the effect of each support sample. Then, features of balanced support samples are utilized as prior information to construct the Bayesian prototype classifier, which can focus more on the important information. Ultimately, a multi-stage inferring strategy is adopted, where the support sample with the greatest distance is filtered in each stage. Prototypes and the corresponding classification score are updated after sample filtering. By integrating the multi-stage classification results, we successfully utilize multi-stage Bayesian inference to enhance the prototype classifier for more accurate few-shot classification results. Experimental results show the efficacy of our method, demonstrating notable advancements in few-shot classification accuracy.

Abstract In software evolution, keeping the test code co‐change with the production code is important, because the outdated test code may not work and is ineffective in revealing faults in the production code. However, due to the tight development time, the production and test code may not be co‐changed immediately by developers. For example, we analysed the top 1003 popular Java projects on GitHub and found that nearly 9.3% of cases (i.e., 464,417) did not update their production and test code at the same time, that is, the production code is updated first, and then the test code is updated at intervals. The result indicates that much test code will not be updated in time. In this paper, we propose a novel approach, Jtup, to remind developers to co‐change the production code and test code in time. Specifically, we first define the co‐changed production and test code as a positive instance, while unchanged test code (i.e., production code changed and test code unchanged) as a negative instance. Then, we extract multidimensional features from the production code to characterize the possibility of their co‐change, including code change features, code complexity features, and code semantic features. Finally, several machine learning‐based methods are employed to identify the co‐changed production and test code. We conduct comprehensive experiments on 20 datasets, and the results show that the Accuracy, Precision, and Recall achieved by Jtup are 76.7%, 78.1%, and 77.4%, which outperforms the state‐of‐the‐art method.

General-purpose foundation models have led to recent breakthroughs in artificial intelligence. In remote sensing, self-supervised learning (SSL) and Masked Image Modeling (MIM) have been adopted to build foundation models. However, these models primarily learn low-level features and require annotated data for fine-tuning. Moreover, they are inapplicable for retrieval and zero-shot applications due to the lack of language understanding. To address these limitations, we propose RemoteCLIP, the first vision-language foundation model for remote sensing that aims to learn robust visual features with rich semantics and aligned text embeddings for seamless downstream application. To address the scarcity of pre-training data, we leverage data scaling which converts heterogeneous annotations into a unified image-caption data format based on Box-to-Caption (B2C) and Mask-to-Box (M2B) conversion. By further incorporating UAV imagery, we produce a 12 × larger pretraining dataset than the combination of all available datasets. RemoteCLIP can be applied to a variety of downstream tasks, including zero-shot image classification, linear probing,k-NN classification, few-shot classification, image-text retrieval, and object counting in remote sensing images. Evaluation on 16 datasets, including a newly introduced RemoteCount benchmark to test the object counting ability, shows that RemoteCLIP consistently outperforms baseline foundation models across different model scales. Impressively, RemoteCLIP beats the state-of-the-art method by 9.14% mean recall on the RSITMD dataset and 8.92% on the RSICD dataset. For zero-shot classification, our RemoteCLIP outperforms the CLIP baseline by up to 6.39% average accuracy on 12 downstream datasets.

Code commenting plays an important role in program comprehension. Automatic comment generation helps improve software maintenance efficiency. The code comments to annotate a method mainly include header comments and snippet comments. The header comment aims to describe the functionality of the entire method, thereby providing a general comment at the beginning of the method. The snippet comment appears at multiple code segments in the body of a method, where a code segment is called a code snippet. Both of them help developers quickly understand code semantics, thereby improving code readability and code maintainability. However, existing automatic comment generation models mainly focus more on header comments, because there are public datasets to validate the performance. By contrast, it is challenging to collect datasets for snippet comments, because it is difficult to determine their scope. Even worse, code snippets are often too short to capture complete syntax and semantic information. To address this challenge, we propose a novel S nippet C omment Gen eration approach called SCGen . First, we utilize the context of the code snippet to expand the syntax and semantic information. Specifically, 600,243 snippet code-comment pairs are collected from 959 Java projects. Then, we capture variables from code snippets and extract variable-related statements from the context. After that, we devise an algorithm to parse and traverse abstract syntax tree (AST) information of code snippets and corresponding context. Finally, SCGen generates snippet comments after inputting the source code snippet and corresponding AST information into a sequence-to-sequence-based model. We conducted extensive experiments on the dataset we collected to evaluate our SCGen . Our approach obtains 18.23 in BLEU-4 metrics, 18.83 in METEOR, and 23.65 in ROUGE-L, which outperforms state-of-the-art comment generation models.

A smart contract is a self-executing program that is stored on the blockchain and runs when predetermined conditions are satisfied. Many frequent transactions involving asset transfers rely on smart contracts deployed on the blockchain, making them highly vulnerable to attack, thus it is essential to ensure the security of smart contracts. Since the smart contract is immutable once deployed, developers must try their best to fix existing vulnerabilities in advance to ensure security. Current approaches for automatic program repair on the smart contracts have mainly adopted the heuristic search algorithms or defined patterns to fix several well-defined types of vulnerabilities. They can only provide security code recommendations for developers in specific scenarios. We explore more general automated program repair of smart contracts in software history.To pave the way for studying code changes related to bug fix of smart contracts in software history, we present a labeled public dataset for method-level program repair task, containing over 12 typical insecure code patterns. Unlike bugs in traditional software, the vulnerabilities of smart contracts are more associated with access control and conditional statements as smart contracts pertain to financial assets. For this problem, we devise a novel double-encoder network and use a code representation designed for the smart contract based on syntax information to repair program. By implementing and evaluating our approach on new dataset comprised of over 10,000 program pairs, we demonstrate the superiority of our approach in both qualitative and quantitative aspects.

Abstract Bytecode is a form of instruction set designed for efficient execution by a software interpreter. Unlike human-readable source code, bytecode is even harder to understand for programmers and researchers. Bytecode has been widely used in various software tasks such as malware detection and clone detection. In order to understand the meaning of the bytecode more quickly and accurately and further help programmers in more software activities, we propose a bytecode comment generation method (called BCGen) using neural language model. Specifically, to get the structured information of the bytecode, we first generate the control flow graph (CFG) of the bytecode, and serialize the CFG with bytecode semantic information. Then a transformer model combining gate recurrent unit is proposed to learn the features of bytecode to generate comments. We obtain the bytecode by building the Jar packages of the well-known open-source projects in the Maven repository and construct a bytecode dataset to train and evaluate our model. Experimental results show that the BLEU of BCGen can reach 0.26, which outperforms several baselines and proves the effectiveness and practicability of our method. It is concluded that it is possible to generate natural language comments directly from the bytecode. Meanwhile, it is important to take structured and semantic information into account in generating bytecode comments.

Code comments are one of the important documents to help developers review and comprehend source code. In recent studies, researchers have proposed many deep learning models to generate the method header comments (i.e., method comment), which have achieved encouraging results. The comments in the method, which is called inline comment, are also important for program comprehension. Unfortunately, they have not received enough attention in automatic generation when comparing with the method comments. In this paper, we compare and analyze the similarities and differences between the method comments and the inline comments. By applying the existing models of generating method comments to the inline comment generation, we find that these existing models perform worse on the task of inline comment generation. We then further explore the possible reasons and obtain a number of new observations. For example, we find that there are a lot of templates (i.e., comments with the same or similar structures) in the method comment dataset, which makes the models perform better. Some terms were thought to be important (e.g., API calls) in the comment generation by previous study does not significantly affect the quality of the generated comments, which seems counter-intuitive. Our findings may give some implications for building the approaches of method comment or inline comment generation in the future.

Blockchain-based cryptocurrencies have developed rapidly in recent years, however, scalability is one of the biggest challenge. Payment channel networks (PCNs) are one of the important solutions to blockchain scalability and routing is the most critical problem in PCN. Routing algorithms in PCNs have evolved fast and achieved high throughput. However, most of these routing algorithms are designed from the perspective of technical feasibility, and few algorithms focus on the incentives of each off-chain participant, especially the economic incentives for intermediate routing nodes. Besides, due to the highly dynamic nature of off-chain channel deposits, existing routing algorithms rely heavily on channel deposit probing in order to ensure high throughput. In this article, we design routing algorithms from an incentive perspective to improve the profit of intermediate nodes and use deep learning to reduce the dependency of off-chain routing on channel deposit probing. Our experiments show that under the same model, MDRL-IR can increase the profit of intermediate nodes by up to 1.87x and increase the throughput by up to 2.0x compared to the state-of-the-art routing algorithm, while ensuring that the user routing cost per unit throughput remains unchanged. Moreover, approximate performance can be achieved when deposit probing is greatly reduced.

As one of the most well-known programmer Q&A websites, Stack Overflow (i.e., SO) is serving tens of thousands of developers every day. Previous work has shown that many developers reuse the code snippets on SO when they find an answer (from SO) that functionally matches the programming problem they encounter in their development activities. To study how programmers reuse code on SO during project development, we conduct a comprehensive empirical study. First, to capture the development activities of programmers, we collect 342,148 modified code snippets in commits from 793 open-source Java projects, and these modified code can reflect the programming problems encountered during development. We also collect the code snippets from 1,355,617 posts on SO. Then, we employ CCFinder to detect the code clone between the modified code from commits and the code from SO, and further analyze the code reuse when programmer solves a programming problem during development. We count the code reuse ratios of the modified code snippets in the commits of each project in different years, the results show that the average code reuse ratio is 6.32%, and the maximum is 8.38%. The code reuse ratio in project commits has increased year by year, and the proportion of code reuse in the newly established project is higher than that of old projects. We also find that some projects reuse the code snippets from many years ago. Additionally, we find that experienced developers seem to be more likely to reuse the knowledge on SO. Moreover, we find that the code reuse ratio in bug-related commits (6.67%) is slightly higher than that of in non-bug-related commits (6.59%). Furthermore, we also find that the code reuse ratio (14.44%) in Java class files that have undergone multiple modifications is more than double the overall code reuse ratio (6.32%).

Developers may not understand the Gas mechanism of Ethereum, so many smart contracts consume a lot of unnecessary Gas. To address this issue, existing studies have proposed several methods to optimize the code of the contracts to reduce Gas consumption. To verify the effectiveness, most of the methods deploy a private chain to make verification. However, a more reasonable way is to employ the real transactions on Ethereum to trigger the contracts before and after optimization, and then compare the Gas consumption. To achieve this goal, we proposed a method, GOV, to estimate the Gas consumption of the optimized contract by using the real transactions on Ethereum. Our method enables the optimized contract to follow the execution path of the contract before optimization, thus solving the problem of inconsistent execution paths before and after optimization. A preliminary evaluation shows that GOV can effectively estimate the Gas consumption of optimized contract.

Abstract Code review is one of the common activities to guarantee the reliability of software, while code review is time-consuming as it requires reviewers to inspect the source code of each patch. A patch may be reviewed more than once before it is eventually merged or abandoned, and then such a patch may tighten the development schedule of the developers and further affect the development progress of a project. Thus, a tool that predicts early on how long a patch will be reviewed can help developers take self-inspection beforehand for the patches that require long-time review. In this paper, we propose a novel method, PMCost, to predict the reviewing rounds of a patch. PMCost uses a number of features, including patch meta-features, code diff features, personal experience features and patch textual features, to better reflect code changes and review process. To examine the benefits of PMCost, we perform experiments on three large open source projects, namely Eclipse, OpenDaylight and OpenStack. The encouraging experimental results demonstrate the feasibility and effectiveness of our approach. Besides, we further study the why the proposed features contribute to the reviewing rounds prediction.

Several approaches are proposed to detect code smells. A set of important approaches are based on machine learning algorithms, which require the code smells have been labeled in source codes as training data firstly. The common labeling approaches are based on manual or tools, but it is difficult for current approaches to get reliable large-scale datasets. In this paper, an approach using the evolution information of source codes is proposed to get large-scale and more reliable training datasets for detecting code smells based on machine learning algorithms. Our approach analyzes the evolving of the source code smells firstly labeled by a tool from the baseline version into the contrastive version of a software system, and then constructs training datasets based on those "changed smells". Experiments conducted on three open source software projects for detecting four types of code smells(which are Data Class, God Class, Brain Class and Brain Method) show that the models obtained by changed smells datasets have better performance on code smell detection than those obtained by unchanged smells datasets (with an average improvement rate of 7.8% and a maximum increase of 30%). The experiments results indicate that using the evolution information of source codes can construct more reliable training datasets for detecting code smells based on machine learning algorithms.

What reasoning rules can be used for the deduction of bisimulation formulas in coalgebraic specifications is problematic because those rules used in algebraic specifications possibly cannot be applied to bisimulation formulas. Although some categorical bisimulation proof methods for coalgebras have been proposed, they are not based on specification languages of coalgebras so that they cannot be used as reasoning rules. In this paper, a specification language based on paths of polynomial functors is proposed to specify polynomial coalgebras. Paths of polynomial functors give detailed observations and transitions on the state space of coalgebras so that the techniques used in transition system specifications can be applied to such a path-based language. In particular, because bisimulations can be characterized by paths, the notions of progressions, respectful functions and faithful contexts can be defined based on paths, and then bisimulation up-to proof techniques, including bisimulation up-to bisimilarities and up-to contexts for transition systems can be transformed into reasoning rules in the language. Several examples illustrate how to reason syntactically about bisimulations in the language by using the rules induced by the bisimulation proof techniques.

Algorithm animation plays an important role in the education of computing and the development of algorithms. But the lack of explicit modeling of visualization specifications makes it difficult to reuse the algorithm implementations and the specifications. A context-aware algorithm animation framework named CA3F is proposed in this paper to solve the problem. In the CA3F framework, the algorithm execution is abstracted as the creation of algorithm contexts, while the algorithm animation is considered as the visualization of algorithms driven by the change of contexts. Algorithm contexts are modeled as visualization specification explicitly and, in consequence, an algorithm implementation becomes independent of the algorithm animation so as to enhance the reusability of the visualization specification. A group of various sort algorithm animations are developed to demonstrate and evaluate the use of CA3F framework.

Quality of service (QoS) plays an important role in the composition of services. A colored Petri net (CPN) based approach is proposed to model and analyze QoS-driven service composition in a service-oriented architecture (SOA), where all services are considered as dynamic joining and quitting resources. Modeling with an extension of the hierarchical CPN, which is named QSC-net, the QoS-driven features are expressed in the service composition model explicitly, along with the dynamic change of service resources and the uncertain execution of composition processes at runtime. QSC-nets support not only the classical property analysis provided by CPN tools, but also new properties concerning the maximal concurrency of services in the composition model. Compared with existing Petri net based modeling approaches, QSCnets are specially suitable for the modeling and analysis of service composition in a dynamic environment, such as a virtual organization (VO) defined by grid computing.

During maintaining the materialized view in the data warehouse, how to efficiently handle the concurrent updates is an important and intractable problem. The paper discusses typical situations that schema changes mix with data updates concurrently. And the reasons why concurrent updates result in view maintenance anomaly are analyzed. Based on the analysis, an enhanced commit agent is designed for dealing with non-order commit problem. Thus, the consistency between data warehouse and data source is guaranteed.