Neural Machine Translation for Code Generation

TL;DR

This paper surveys neural machine translation techniques applied to code generation, covering various input/output formats, architectures, and evaluation methods, highlighting current limitations and future research directions.

Contribution

It provides a comprehensive overview of NMT for code generation, categorizing methods and identifying gaps in existing research.

Findings

Diverse input/output representations have been explored in NMT for code.

Current methods face limitations in handling complex code generation tasks.

Future research should focus on improving model robustness and evaluation metrics.

Abstract

Neural machine translation (NMT) methods developed for natural language processing have been shown to be highly successful in automating translation from one natural language to another. Recently, these NMT methods have been adapted to the generation of program code. In NMT for code generation, the task is to generate output source code that satisfies constraints expressed in the input. In the literature, a variety of different input scenarios have been explored, including generating code based on natural language description, lower-level representations such as binary or assembly (neural decompilation), partial representations of source code (code completion and repair), and source code in another language (code translation). In this paper we survey the NMT for code generation literature, cataloging the variety of methods that have been explored according to input and output representations, model architectures, optimization techniques used, data sets, and evaluation methods. We discuss the limitations of existing methods and future research directions