Source code summarization aims to generate natural language descriptions of code snippets. Many existing studies learn the syntactic and semantic knowledge of code snippets from their token sequences and Abstract Syntax Trees (ASTs). They use the learned code representations as input to code summarization models, which can accordingly generate summaries describing source code. Traditional models traverse ASTs as sequences or split ASTs into paths as input. However, the former loses the structural properties of ASTs, and the latter destroys the overall structure of ASTs. Therefore, comprehensively capturing the structural features of ASTs in learning code representations for source code summarization remains a challenging problem to be solved. In this paper, we propose M2TS, a Multi-scale Multi-modal approach based on Transformer for source code Summarization. M2TS uses a multi-scale AST feature extraction method, which can extract the structures of ASTs more completely and accurately at multiple local and global levels. To complement missing semantic information in ASTs, we also obtain code token features, and further combine them with the extracted AST features using a cross modality fusion method that not only fuses the syntactic and contextual semantic information of source code, but also highlights the key features of each modality. We conduct experiments on two Java and one Python datasets, and the experimental results demonstrate that M2TS outperforms current state-of-the-art methods. We release our code at https://github.com/TranSMS/M2TS.
While there has been a recent burgeoning of applications at the intersection of natural and programming languages, such as code generation and code summarization, these applications are usually English-centric. This creates a barrier for program developers who are not proficient in English. To mitigate this gap in technology development across languages, we propose a multilingual dataset, MCoNaLa, to benchmark code generation from natural language commands extending beyond English. Modeled off of the methodology from the English Code/Natural Language Challenge (CoNaLa) dataset, we annotated a total of 896 NL-code pairs in three languages: Spanish, Japanese, and Russian. We present a quantitative evaluation of performance on the MCoNaLa dataset by testing with state-of-the-art code generation systems. While the difficulties vary across these three languages, all systems lag significantly behind their English counterparts, revealing the challenges in adapting code generation to new languages.
Source code summarization aims at generating concise and clear natural language descriptions for programming languages. Well-written code summaries are beneficial for programmers to participate in the software development and maintenance process. To learn the semantic representations of source code, recent efforts focus on incorporating the syntax structure of code into neural networks such as Transformer. Such Transformer-based approaches can better capture the long-range dependencies than other neural networks including Recurrent Neural Networks (RNNs), however, most of them do not consider the structural relative correlations between tokens, e.g., relative positions in Abstract Syntax Trees (ASTs), which is beneficial for code semantics learning. To model the structural dependency, we propose a Structural Relative Position guided Transformer, named SCRIPT. SCRIPT first obtains the structural relative positions between tokens via parsing the ASTs of source code, and then passes them into two types of Transformer encoders. One Transformer directly adjusts the input according to the structural relative distance; and the other Transformer encodes the structural relative positions during computing the self-attention scores. Finally, we stack these two types of Transformer encoders to learn representations of source code. Experimental results show that the proposed SCRIPT outperforms the state-of-the-art methods by at least 1.6%, 1.4% and 2.8% with respect to BLEU, ROUGE-L and METEOR on benchmark datasets, respectively. We further show that how the proposed SCRIPT captures the structural relative dependencies.
Recently, many pre-trained language models for source code have been proposed to model the context of code and serve as a basis for downstream code intelligence tasks such as code completion, code search, and code summarization. These models leverage masked pre-training and Transformer and have achieved promising results. However, currently there is still little progress regarding interpretability of existing pre-trained code models. It is not clear why these models work and what feature correlations they can capture. In this paper, we conduct a thorough structural analysis aiming to provide an interpretation of pre-trained language models for source code (e.g., CodeBERT, and GraphCodeBERT) from three distinctive perspectives: (1) attention analysis, (2) probing on the word embedding, and (3) syntax tree induction. Through comprehensive analysis, this paper reveals several insightful findings that may inspire future studies: (1) Attention aligns strongly with the syntax structure of code. (2) Pre-training language models of code can preserve the syntax structure of code in the intermediate representations of each Transformer layer. (3) The pre-trained models of code have the ability of inducing syntax trees of code. Theses findings suggest that it may be helpful to incorporate the syntax structure of code into the process of pre-training for better code representations.
Graphs are a highly expressive data structure, but it is often difficult for humans to find patterns from a complex graph. Hence, generating human-interpretable sequences from graphs have gained interest, called graph2seq learning. It is expected that the compositionality in a graph can be associated to the compositionality in the output sequence in many graph2seq tasks. Therefore, applying compositionality-aware GNN architecture would improve the model performance. In this study, we adopt the multi-level attention pooling (MLAP) architecture, that can aggregate graph representations from multiple levels of information localities. As a real-world example, we take up the extreme source code summarization task, where a model estimate the name of a program function from its source code. We demonstrate that the model having the MLAP architecture outperform the previous state-of-the-art model with more than seven times fewer parameters than it.
Source code summarization is the task of generating a high-level natural language description for a segment of programming language code. Current neural models for the task differ in their architecture and the aspects of code they consider. In this paper, we show that three SOTA models for code summarization work well on largely disjoint subsets of a large code-base. This complementarity motivates model combination: We propose three meta-models that select the best candidate summary for a given code segment. The two neural models improve significantly over the performance of the best individual model, obtaining an improvement of 2.1 BLEU points on a dataset of code segments where at least one of the individual models obtains a non-zero BLEU.
Automatic code summarization is beneficial to daily software development since it could help reduce the requirement of manual writing. Currently, artificial intelligence is undergoing a paradigm shift. The foundation models pretrained on massive data and finetuned to downstream tasks surpass specially customized models. This trend inspired us to consider reusing foundation models instead of learning from scratch. Thereby, we propose a flexible and robust approach for automatic code summarization, based on neural models. We assemble available foundation models, such as CodeBERT and GPT-2, into a single neural model named AdaMo. Moreover, we utilize Gaussian noise as the simulation of contextual information to optimize the latent representation. Furthermore, we introduce two adaptive schemes from the perspective of knowledge transfer, namely continuous pretraining and intermediate finetuning, and design intermediate stage tasks for general sequence-to-sequence learning. Finally, we evaluate AdaMo against a benchmark dataset for code summarization, by comparing it with state-of-the-art models.
Well-trained machine-learning models, which leverage large amounts of open-source software data, have now become an interesting approach to automating many software engineering tasks. Several SE tasks have all been subject to this approach, with performance gradually improving over the past several years with better models and training methods. More, and more diverse, clean, labeled data is better for training; but constructing good-quality datasets is time-consuming and challenging. Ways of augmenting the volume and diversity of clean, labeled data generally have wide applicability. For some languages (e.g., Ruby) labeled data is less abundant; in others (e.g., JavaScript) the available data maybe more focused on some application domains, and thus less diverse. As a way around such data bottlenecks, we present evidence suggesting that human-written code in different languages (which performs the same function), is rather similar, and particularly preserving of identifier naming patterns; we further present evidence suggesting that identifiers are a very important element of training data for software engineering tasks. We leverage this rather fortuitous phenomenon to find evidence that available multilingual training data (across different languages) can be used to amplify performance. We study this for 3 different tasks: code summarization, code retrieval, and function naming. We note that this data-augmenting approach is broadly compatible with different tasks, languages, and machine-learning models.
Code summarization aims to generate brief natural language descriptions for source code. As source code is highly structured and follows strict programming language grammars, its Abstract Syntax Tree (AST) is often leveraged to inform the encoder about the structural information. However, ASTs are usually much longer than the source code. Current approaches ignore the size limit and simply feed the whole linearized AST into the encoder. To address this problem, we propose AST-Transformer to efficiently encode tree-structured ASTs. Experiments show that AST-Transformer outperforms the state-of-arts by a substantial margin while being able to reduce $90\sim95\%$ of the computational complexity in the encoding process.
Transformer architectures have been successfully used in learning source code representations. The fusion between a graph representation like Abstract Syntax Tree (AST) and a source code sequence makes the use of current approaches computationally intractable for large input sequence lengths. Source code can have long-range dependencies that require larger sequence lengths to model effectively. Current approaches have a quadratic growth in computational and memory costs with respect to the sequence length. Using such models in practical scenarios is difficult. In this work, we propose the conditioning of a source code snippet with its graph modality by using the graph adjacency matrix as an attention mask for a sparse self-attention mechanism and the use of a graph diffusion mechanism to model longer-range token dependencies. Our model reaches state-of-the-art results in BLEU, METEOR, and ROUGE-L metrics for the code summarization task and near state-of-the-art accuracy in the variable misuse task. The memory use and inference time of our model have linear growth with respect to the input sequence length as compared to the quadratic growth of previous works.
Automated source code summarization is a task that generates summarized information about the purpose, usage, and--or implementation of methods and classes to support understanding of these code entities. Multiple approaches and techniques have been proposed for supervised and unsupervised learning in code summarization, however, they were mostly focused on generating a summary for a piece of code. In addition, very few works have leveraged unofficial documentation. This paper proposes an automatic and novel approach for summarizing Android API methods discussed in Stack Overflow that we consider as unofficial documentation in this research. Our approach takes the API method's name as an input and generates a natural language summary based on Stack Overflow discussions of that API method. We have conducted a survey that involves 16 Android developers to evaluate the quality of our automatically generated summaries and compare them with the official Android documentation. Our results demonstrate that while developers find the official documentation more useful in general, the generated summaries are also competitive, in particular for offering implementation details, and can be used as a complementary source for guiding developers in software development and maintenance tasks.
As opposed to natural languages, source code understanding is influenced by grammatical relationships between tokens regardless of their identifier name. Graph representations of source code such as Abstract Syntax Tree (AST) can capture relationships between tokens that are not obvious from the source code. We propose a novel method, GN-Transformer to learn end-to-end on a fused sequence and graph modality we call Syntax-Code-Graph (SCG). GN-Transformer expands on Graph Networks (GN) framework using a self-attention mechanism. SCG is the result of the early fusion between a source code snippet and the AST representation. We perform experiments on the structure of SCG, an ablation study on the model design, and the hyper-parameters to conclude that the performance advantage is from the fused representation. The proposed methods achieve state-of-the-art performance in two code summarization datasets and across three automatic code summarization metrics (BLEU, METEOR, ROUGE-L). We further evaluate the human perceived quality of our model and previous work with an expert-user study. Our model outperforms the state-of-the-art in human perceived quality and accuracy.
An effective and efficient encoding of the source code of a computer program is critical to the success of sequence-to-sequence deep neural network models for tasks in computer program comprehension, such as automated code summarization and documentation. A significant challenge is to find a sequential representation that captures the structural/syntactic information in a computer program and facilitates the training of the learning models. In this paper, we propose to use the Pr\"ufer sequence of the Abstract Syntax Tree (AST) of a computer program to design a sequential representation scheme that preserves the structural information in an AST. Our representation makes it possible to develop deep-learning models in which signals carried by lexical tokens in the training examples can be exploited automatically and selectively based on their syntactic role and importance. Unlike other recently-proposed approaches, our representation is concise and lossless in terms of the structural information of the AST. Empirical studies on real-world benchmark datasets, using a sequence-to-sequence learning model we designed for code summarization, show that our Pr\"ufer-sequence-based representation is indeed highly effective and efficient, outperforming significantly all the recently-proposed deep-learning models we used as the baseline models.
Code search aims to retrieve accurate code snippets based on a natural language query to improve software productivity and quality. With the massive amount of available programs such as (on GitHub or Stack Overflow), identifying and localizing the precise code is critical for the software developers. In addition, Deep learning has recently been widely applied to different code-related scenarios, e.g., vulnerability detection, source code summarization. However, automated deep code search is still challenging since it requires a high-level semantic mapping between code and natural language queries. Most existing deep learning-based approaches for code search rely on the sequential text i.e., feeding the program and the query as a flat sequence of tokens to learn the program semantics while the structural information is not fully considered. Furthermore, the widely adopted Graph Neural Networks (GNNs) have proved their effectiveness in learning program semantics, however, they also suffer the problem of capturing the global dependencies in the constructed graph, which limits the model learning capacity. To address these challenges, in this paper, we design a novel neural network framework, named GraphSearchNet, to enable an effective and accurate source code search by jointly learning the rich semantics of both source code and natural language queries. Specifically, we propose to construct graphs for the source code and queries with bidirectional GGNN (BiGGNN) to capture the local structural information of the source code and queries. Furthermore, we enhance BiGGNN by utilizing the multi-head attention module to supplement the global dependencies that BiGGNN missed to improve the model learning capacity. The extensive experiments on Java and Python programming language from the public benchmark CodeSearchNet confirm that GraphSearchNet outperforms current state-of-the-art works.
Statistical language modeling and translation with transformers have found many successful applications in program understanding and generation tasks, setting high benchmarks for tools in modern software development environments. The finite context window of these neural models means, however, that they will be unable to leverage the entire relevant context of large files and packages for any given task. While there are many efforts to extend the context window, we introduce an architecture-independent approach for leveraging the syntactic hierarchies of source code for incorporating entire file-level context into a fixed-length window. Using concrete syntax trees of each source file we extract syntactic hierarchies and integrate them into context window by selectively removing from view more specific, less relevant scopes for a given task. We evaluate this approach on code generation tasks and joint translation of natural language and source code in Python programming language, achieving a new state-of-the-art in code completion and summarization for Python in the CodeXGLUE benchmark. We also introduce new CodeXGLUE benchmarks for user-experience-motivated tasks: code completion with normalized literals, method body completion/code summarization conditioned on file-level context.
Pre-trained models for Natural Languages (NL) like BERT and GPT have been recently shown to transfer well to Programming Languages (PL) and largely benefit a broad set of code-related tasks. Despite their success, most current methods either rely on an encoder-only (or decoder-only) pre-training that is suboptimal for generation (resp. understanding) tasks or process the code snippet in the same way as NL, neglecting the special characteristics of PL such as token types. We present CodeT5, a unified pre-trained encoder-decoder Transformer model that better leverages the code semantics conveyed from the developer-assigned identifiers. Our model employs a unified framework to seamlessly support both code understanding and generation tasks and allows for multi-task learning. Besides, we propose a novel identifier-aware pre-training task that enables the model to distinguish which code tokens are identifiers and to recover them when they are masked. Furthermore, we propose to exploit the user-written code comments with a bimodal dual generation task for better NL-PL alignment. Comprehensive experiments show that CodeT5 significantly outperforms prior methods on understanding tasks such as code defect detection and clone detection, and generation tasks across various directions including PL-NL, NL-PL, and PL-PL. Further analysis reveals that our model can better capture semantic information from code. Our code and pre-trained models are released at https: //github.com/salesforce/CodeT5 .
Software developers write a lot of source code and documentation during software development. Intrinsically, developers often recall parts of source code or code summaries that they had written in the past while implementing software or documenting them. To mimic developers' code or summary generation behavior, we propose a retrieval augmented framework, REDCODER, that retrieves relevant code or summaries from a retrieval database and provides them as a supplement to code generation or summarization models. REDCODER has a couple of uniqueness. First, it extends the state-of-the-art dense retrieval technique to search for relevant code or summaries. Second, it can work with retrieval databases that include unimodal (only code or natural language description) or bimodal instances (code-description pairs). We conduct experiments and extensive analysis on two benchmark datasets of code generation and summarization in Java and Python, and the promising results endorse the effectiveness of our proposed retrieval augmented framework.
There has been a growing interest in developing machine learning (ML) models for code summarization tasks, e.g., comment generation and method naming. Despite substantial increase in the effectiveness of ML models, the evaluation methodologies, i.e., the way people split datasets into training, validation, and test sets, were not well studied. Specifically, no prior work on code summarization considered the timestamps of code and comments during evaluation. This may lead to evaluations that are inconsistent with the intended use cases. In this paper, we introduce the time-segmented evaluation methodology, which is novel to the code summarization research community, and compare it with the mixed-project and cross-project methodologies that have been commonly used. Each methodology can be mapped to some use cases, and the time-segmented methodology should be adopted in the evaluation of ML models for code summarization. To assess the impact of methodologies, we collect a dataset of (code, comment) pairs with timestamps to train and evaluate several recent ML models for code summarization. Our experiments show that different methodologies lead to conflicting evaluation results. We invite the community to expand the set of methodologies used in evaluations.
Recently there have been many advances in research on language modeling of source code. Applications range from code suggestion and completion to code summarization. However, complete program synthesis of industry-grade programming languages remains an open problem. In this work, we introduce and experimentally validate a variational autoencoder model for program synthesis of industry-grade programming languages. This model makes use of the inherent tree structure of code and can be used in conjunction with gradient free optimization techniques like evolutionary methods to generate programs that maximize a given fitness function, for instance, passing a set of test cases. A demonstration is avaliable at https://tree2tree.app
Source code summaries are important for program comprehension and maintenance. However, there are plenty of programs with missing, outdated, or mismatched summaries. Recently, deep learning techniques have been exploited to automatically generate summaries for given code snippets. To achieve a profound understanding of how far we are from solving this problem and provide suggestions to future research, in this paper, we conduct a systematic and in-depth analysis of 5 state-of-the-art neural code summarization models on 6 widely used BLEU variants, 4 pre-processing operations and their combinations, and 3 widely used datasets. The evaluation results show that some important factors have a great influence on the model evaluation, especially on the performance of models and the ranking among the models. However, these factors might be easily overlooked. Specifically, (1) the BLEU metric widely used in existing work of evaluating code summarization models has many variants. Ignoring the differences among these variants could greatly affect the validity of the claimed results. Furthermore, we conduct human evaluations and find that the metric BLEU-DC is most correlated to human perception; (2) code pre-processing choices can have a large (from -18\% to +25\%) impact on the summarization performance and should not be neglected. We also explore the aggregation of pre-processing combinations and boost the performance of models; (3) some important characteristics of datasets (corpus sizes, data splitting methods, and duplication ratios) have a significant impact on model evaluation. Based on the experimental results, we give actionable suggestions for evaluating code summarization and choosing the best method in different scenarios. We also build a shared code summarization toolbox to facilitate future research.