ragrep 🔍

[!WARNING] WIP big time. This codebase is full of broken glass, sharp edges, and dragons.

A semantic code search tool that uses embeddings to find similar code snippets across your codebase. Use it to search for things like:

• "how do we handle http request errors?"
• "where is our data validation?"

Features

• Semantic code search using embeddings
• Fully local, no API keys or dependencies
• Supports multiple programming languages through tree-sitter
• Fast SQLite-based storage for embeddings and code chunks
• Intelligent code chunking based on AST

Installation

# Not actually live yet - still need to publish it 
cargo install ragrep

Building from Source

Prerequisites

• Rust toolchain (1.75.0 or later recommended)
• SQLite 3.x

git clone https://github.com/yourusername/ragrep.git
cd ragrep
cargo build --release

The binary will be available at target/release/ragrep

Usage

[!IMPORTANT] The first time you run ragrep, it will download a model and cache it in is global data directory. This might take a minute and will use about 1.5GB of disk space.

Indexing Your Codebase

Before searching, you need to index your codebase:

# Index the current directory
ragrep index

# Index a specific directory
ragrep index --path /path/to/your/code

Searching Code

# Search for code similar to your query
ragrep "handle http request error"

The search results will show relevant code snippets along with their file locations, formatted in a familiar ripgrep-style output.

Debug Mode

To see similarity scores in the output:

RUST_LOG=debug ragrep "your query"

Supported Languages

• Rust
• Python
• JavaScript
• TypeScript

More languages can be added by including their respective tree-sitter parsers.

How It Works

1. Indexing:

   • Scans your codebase for supported files
   • Uses tree-sitter to parse code into meaningful chunks
   • Generates embeddings for each code chunk
   • Stores chunks and embeddings in a SQLite database

2. Searching:

   • Converts your search query into an embedding
   • Finds code chunks with similar embeddings using vector similarity
   • Ranks and displays the most relevant results

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.