cAST: Enhancing Code Retrieval-Augmented Generation with Structural Chunking via Abstract Syntax Tree

TL;DR

This paper introduces a structure-aware chunking method using Abstract Syntax Trees to improve code retrieval and generation quality in Retrieval-Augmented Generation systems.

Contribution

It presents a novel AST-based chunking approach that maintains semantic coherence, outperforming line-based heuristics across multiple code tasks.

Findings

Boosted Recall@5 by 4.3 points on RepoEval

Increased Pass@1 by 2.67 points on SWE-bench

Improved semantic coherence of code chunks

Abstract

Retrieval-Augmented Generation (RAG) has become essential for large-scale code generation, grounding predictions in external code corpora to improve actuality. However, a critical yet underexplored aspect of RAG pipelines is chunking -- the process of dividing documents into retrievable units. Existing line-based chunking heuristics often break semantic structures, splitting functions or merging unrelated code, which can degrade generation quality. We propose chunking via Abstract Syntax Trees (\ourwork), a structure-aware method that recursively breaks large AST nodes into smaller chunks and merges sibling nodes while respecting size limits. This approach generates self-contained, semantically coherent units across programming languages and tasks, improving performance on diverse code generation tasks, e.g., boosting Recall@5 by 4.3 points on RepoEval retrieval and Pass@1 by 2.67 points on SWE-bench generation. Our work highlights the importance of structure-aware chunking for scaling retrieval-enhanced code intelligence.