codebase-embedder

English | 简体中文

A powerful code indexing context retrieval service for AI programming assistants.

Overview

codebase-indexer is the context module of ZGSM (ZhuGe Smart Mind) AI Programming Assistant which running on backend. It provides powerful codebase indexing capabilities to support semantic search for RAG (Retrieval-Augmented Generation) systems.

Key Features

• 🔍 Semantic code search with embeddings
• 🌐 Multi-language support

Requirements

• Go 1.24.3 or higher
• Docker
• PostgreSQL
• Redis
• Weavaite

Quick Start

Installation

# Clone the repository
git clone https://github.com/zgsm-ai/codebase-embedder.git
cd codebase-embedder

# Install dependencies
go mod tidy

Configuration

1. Set up PostgreSQL 、 Redis、vector, etc.

vim etc/config.yaml

3. Update the configuration with your database and Redis credentials

Running

# Build the project
make build

Architecture

The system consists of several key components:

• Parser: Code parsing and AST generation
• Embedding: Code semantic vector generation
• Store: Data storage and indexing
• API: RESTful service interface

License

This project is licensed under the Apache 2.0 License.

Acknowledgments

This project builds upon the excellent work of:

• Tree-sitter - For providing robust parsing capabilities