CodeS: Natural Language to Code Repository via Multi-Layer Sketch

TL;DR

This paper introduces CodeS, a multi-layer framework that generates entire code repositories from natural language requirements by decomposing the task into sub-tasks, demonstrating effectiveness through benchmarks and user studies.

Contribution

Proposes a novel multi-layer sketch-based framework, CodeS, for converting natural language requirements into complete code repositories, including new benchmarks and evaluation metrics.

Findings

CodeS effectively generates code repositories from natural language.

Benchmark and user studies validate the framework's practicality.

CodeS outperforms baseline methods in experiments.

Abstract

The impressive performance of large language models (LLMs) on code-related tasks has shown the potential of fully automated software development. In light of this, we introduce a new software engineering task, namely Natural Language to code Repository (NL2Repo). This task aims to generate an entire code repository from its natural language requirements. To address this task, we propose a simple yet effective framework CodeS, which decomposes NL2Repo into multiple sub-tasks by a multi-layer sketch. Specifically, CodeS includes three modules: RepoSketcher, FileSketcher, and SketchFiller. RepoSketcher first generates a repository's directory structure for given requirements; FileSketcher then generates a file sketch for each file in the generated structure; SketchFiller finally fills in the details for each function in the generated file sketch. To rigorously assess CodeS on the NL2Repo task, we carry out evaluations through both automated benchmarking and manual feedback analysis. For benchmark-based evaluation, we craft a repository-oriented benchmark, SketchEval, and design an evaluation metric, SketchBLEU. For feedback-based evaluation, we develop a VSCode plugin for CodeS and engage 30 participants in conducting empirical studies. Extensive experiments prove the effectiveness and practicality of CodeS on the NL2Repo task.