Repository-level Code Search with Neural Retrieval Methods

TL;DR

This paper introduces a neural reranking system for repository-level code search that combines BM25 retrieval with CodeBERT to significantly improve relevance ranking using commit histories.

Contribution

It presents a novel multi-stage reranking approach leveraging commit messages and source code, achieving substantial performance gains over baseline methods.

Findings

Up to 80% improvement in MAP, MRR, and P@1 metrics.

Effective use of commit histories for code relevance.

Demonstrated on a new dataset from 7 open-source repositories.

Abstract

This paper presents a multi-stage reranking system for repository-level code search, which leverages the vastly available commit histories of large open-source repositories to aid in bug fixing. We define the task of repository-level code search as retrieving the set of files from the current state of a code repository that are most relevant to addressing a user's question or bug. The proposed approach combines BM25-based retrieval over commit messages with neural reranking using CodeBERT to identify the most pertinent files. By learning patterns from diverse repositories and their commit histories, the system can surface relevant files for the task at hand. The system leverages both commit messages and source code for relevance matching, and is evaluated in both normal and oracle settings. Experiments on a new dataset created from 7 popular open-source repositories demonstrate substantial improvements of up to 80% in MAP, MRR and P@1 over the BM25 baseline, across a diverse set of queries, demonstrating the effectiveness this approach. We hope this work aids LLM agents as a tool for better code search and understanding. Our code and results obtained are publicly available.