RepoCoder: Repository-Level Code Completion Through Iterative Retrieval and Generation

Fengji Zhang; Bei Chen; Yue Zhang; Jacky Keung; Jin Liu; Daoguang Zan; Yi Mao; Jian-Guang Lou; Weizhu Chen

doi:10.18653/v1/2023.emnlp-main.151

RepoCoder: Repository-Level Code Completion Through Iterative Retrieval and Generation

Fengji Zhang, Bei Chen, Yue Zhang, Jacky Keung, Jin Liu, Daoguang Zan, Yi Mao, Jian-Guang Lou, Weizhu Chen

Department of Computer Science

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

68 Citations (Scopus)

90 Downloads (CityUHK Scholars)

Abstract

The task of repository-level code completion is to continue writing the unfinished code based on a broader context of the repository. While for automated code completion tools, it is difficult to utilize the useful information scattered in different files. We propose RepoCoder, a simple, generic, and effective framework to address the challenge. It streamlines the repository-level code completion process by incorporating a similarity-based retriever and a pre-trained code language model in an iterative retrieval-generation pipeline. RepoCoder makes effective utilization of repository-level information for code completion and has the ability to generate code at various levels of granularity. Moreover, we propose a new benchmark RepoBench, which consists of the latest and high-quality real-world repositories covering line, API invocation, and function body completion scenarios. Experimental results indicate that RepoCoder significantly improves the In-File completion baseline by over 10% in all settings and consistently outperforms the vanilla retrieval-augmented code completion approach. Furthermore, we validate the effectiveness of RepoCoder through comprehensive analysis, providing valuable insights for future research. ©2023 Association for Computational Linguistics.

Original language	English
Title of host publication	Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing
Editors	Houda Bouamor, Juan Pino, Kalika Bali
Publisher	Association for Computational Linguistics
Pages	2471–2484
ISBN (Print)	9798891760608
DOIs	https://doi.org/10.18653/v1/2023.emnlp-main.151
Publication status	Published - Dec 2023
Event	2023 Conference on Empirical Methods in Natural Language Processing (EMNLP 2023) - Resorts World Convention Centre (Hybrid), Singapore Duration: 6 Dec 2023 → 10 Dec 2023 https://aclanthology.org/2023.emnlp-main https://2023.emnlp.org/

Publication series

Name	EMNLP - Conference on Empirical Methods in Natural Language Processing, Proceedings

Conference

Conference	2023 Conference on Empirical Methods in Natural Language Processing (EMNLP 2023)
Abbreviated title	EMNLP
Place	Singapore
Period	6/12/23 → 10/12/23
Internet address	https://aclanthology.org/2023.emnlp-main https://2023.emnlp.org/

Bibliographical note

Research Unit(s) information for this publication is provided by the author(s) concerned.

Publisher's Copyright Statement

This full text is made available under CC-BY 4.0. https://creativecommons.org/licenses/by/4.0/

Access Output

10.18653/v1/2023.emnlp-main.151

192460686.pdfFinal Published version, 3.1 MBLicence: CC BY

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

Zhang, F., Chen, B., Zhang, Y., Keung, J., Liu, J., Zan, D., Mao, Y., Lou, J.-G., & Chen, W. (2023). RepoCoder: Repository-Level Code Completion Through Iterative Retrieval and Generation. In H. Bouamor, J. Pino, & K. Bali (Eds.), Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing (pp. 2471–2484). (EMNLP - Conference on Empirical Methods in Natural Language Processing, Proceedings). Association for Computational Linguistics. https://doi.org/10.18653/v1/2023.emnlp-main.151

Zhang, Fengji ; Chen, Bei ; Zhang, Yue et al. / RepoCoder : Repository-Level Code Completion Through Iterative Retrieval and Generation. Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing. editor / Houda Bouamor ; Juan Pino ; Kalika Bali. Association for Computational Linguistics, 2023. pp. 2471–2484 (EMNLP - Conference on Empirical Methods in Natural Language Processing, Proceedings).

@inproceedings{21fe0aa4fb9f478a84c864eb8e91520f,

title = "RepoCoder: Repository-Level Code Completion Through Iterative Retrieval and Generation",

abstract = "The task of repository-level code completion is to continue writing the unfinished code based on a broader context of the repository. While for automated code completion tools, it is difficult to utilize the useful information scattered in different files. We propose RepoCoder, a simple, generic, and effective framework to address the challenge. It streamlines the repository-level code completion process by incorporating a similarity-based retriever and a pre-trained code language model in an iterative retrieval-generation pipeline. RepoCoder makes effective utilization of repository-level information for code completion and has the ability to generate code at various levels of granularity. Moreover, we propose a new benchmark RepoBench, which consists of the latest and high-quality real-world repositories covering line, API invocation, and function body completion scenarios. Experimental results indicate that RepoCoder significantly improves the In-File completion baseline by over 10% in all settings and consistently outperforms the vanilla retrieval-augmented code completion approach. Furthermore, we validate the effectiveness of RepoCoder through comprehensive analysis, providing valuable insights for future research. {\textcopyright}2023 Association for Computational Linguistics.",

author = "Fengji Zhang and Bei Chen and Yue Zhang and Jacky Keung and Jin Liu and Daoguang Zan and Yi Mao and Jian-Guang Lou and Weizhu Chen",

note = "Research Unit(s) information for this publication is provided by the author(s) concerned.; 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP 2023), EMNLP ; Conference date: 06-12-2023 Through 10-12-2023",

year = "2023",

month = dec,

doi = "10.18653/v1/2023.emnlp-main.151",

language = "English",

isbn = "9798891760608",

series = "EMNLP - Conference on Empirical Methods in Natural Language Processing, Proceedings",

publisher = "Association for Computational Linguistics",

pages = "2471–2484",

editor = "Houda Bouamor and Juan Pino and Kalika Bali",

booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",

address = "United States",

url = "https://aclanthology.org/2023.emnlp-main, https://2023.emnlp.org/",

}

Zhang, F, Chen, B, Zhang, Y, Keung, J, Liu, J, Zan, D, Mao, Y, Lou, J-G & Chen, W 2023, RepoCoder: Repository-Level Code Completion Through Iterative Retrieval and Generation. in H Bouamor, J Pino & K Bali (eds), Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing. EMNLP - Conference on Empirical Methods in Natural Language Processing, Proceedings, Association for Computational Linguistics, pp. 2471–2484, 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP 2023), Singapore, 6/12/23. https://doi.org/10.18653/v1/2023.emnlp-main.151

RepoCoder: Repository-Level Code Completion Through Iterative Retrieval and Generation. / Zhang, Fengji; Chen, Bei; Zhang, Yue et al.
Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing. ed. / Houda Bouamor; Juan Pino; Kalika Bali. Association for Computational Linguistics, 2023. p. 2471–2484 (EMNLP - Conference on Empirical Methods in Natural Language Processing, Proceedings).

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

TY - GEN

T1 - RepoCoder

T2 - 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP 2023)

AU - Zhang, Fengji

AU - Chen, Bei

AU - Zhang, Yue

AU - Keung, Jacky

AU - Liu, Jin

AU - Zan, Daoguang

AU - Mao, Yi

AU - Lou, Jian-Guang

AU - Chen, Weizhu

N1 - Research Unit(s) information for this publication is provided by the author(s) concerned.

PY - 2023/12

Y1 - 2023/12

N2 - The task of repository-level code completion is to continue writing the unfinished code based on a broader context of the repository. While for automated code completion tools, it is difficult to utilize the useful information scattered in different files. We propose RepoCoder, a simple, generic, and effective framework to address the challenge. It streamlines the repository-level code completion process by incorporating a similarity-based retriever and a pre-trained code language model in an iterative retrieval-generation pipeline. RepoCoder makes effective utilization of repository-level information for code completion and has the ability to generate code at various levels of granularity. Moreover, we propose a new benchmark RepoBench, which consists of the latest and high-quality real-world repositories covering line, API invocation, and function body completion scenarios. Experimental results indicate that RepoCoder significantly improves the In-File completion baseline by over 10% in all settings and consistently outperforms the vanilla retrieval-augmented code completion approach. Furthermore, we validate the effectiveness of RepoCoder through comprehensive analysis, providing valuable insights for future research. ©2023 Association for Computational Linguistics.

AB - The task of repository-level code completion is to continue writing the unfinished code based on a broader context of the repository. While for automated code completion tools, it is difficult to utilize the useful information scattered in different files. We propose RepoCoder, a simple, generic, and effective framework to address the challenge. It streamlines the repository-level code completion process by incorporating a similarity-based retriever and a pre-trained code language model in an iterative retrieval-generation pipeline. RepoCoder makes effective utilization of repository-level information for code completion and has the ability to generate code at various levels of granularity. Moreover, we propose a new benchmark RepoBench, which consists of the latest and high-quality real-world repositories covering line, API invocation, and function body completion scenarios. Experimental results indicate that RepoCoder significantly improves the In-File completion baseline by over 10% in all settings and consistently outperforms the vanilla retrieval-augmented code completion approach. Furthermore, we validate the effectiveness of RepoCoder through comprehensive analysis, providing valuable insights for future research. ©2023 Association for Computational Linguistics.

UR - http://www.scopus.com/inward/record.url?scp=85183361792&partnerID=8YFLogxK

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-85183361792&origin=recordpage

U2 - 10.18653/v1/2023.emnlp-main.151

DO - 10.18653/v1/2023.emnlp-main.151

M3 - RGC 32 - Refereed conference paper (with host publication)

SN - 9798891760608

T3 - EMNLP - Conference on Empirical Methods in Natural Language Processing, Proceedings

SP - 2471

EP - 2484

BT - Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing

A2 - Bouamor, Houda

A2 - Pino, Juan

A2 - Bali, Kalika

PB - Association for Computational Linguistics

Y2 - 6 December 2023 through 10 December 2023

ER -

Zhang F, Chen B, Zhang Y, Keung J, Liu J, Zan D et al. RepoCoder: Repository-Level Code Completion Through Iterative Retrieval and Generation. In Bouamor H, Pino J, Bali K, editors, Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing. Association for Computational Linguistics. 2023. p. 2471–2484. (EMNLP - Conference on Empirical Methods in Natural Language Processing, Proceedings). doi: 10.18653/v1/2023.emnlp-main.151

RepoCoder: Repository-Level Code Completion Through Iterative Retrieval and Generation

Abstract

Publication series

Conference

Bibliographical note

Publisher's Copyright Statement

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this