Andrew Gallant (BurntSushi) looked at grep, ack, ag, and every other text search tool and said “I can do better.” Then he actually did it.

# Search for patterns in your entire codebase
rg "function.*login" --type js

# Ignore case, show context, search specific files
rg -i "TODO" -A 3 -B 1 --glob "*.py"

# Search for regex patterns with Unicode support
rg "\p{Emoji}" --pcre2

The Numbers Don’t Lie

  • 13x faster than ack
  • 5x faster than ag (the silver searcher)
  • 2-3x faster than git grep
  • Handles gigabyte files without breaking a sweat

Smart Defaults That Actually Make Sense

Respects .gitignore: Won’t waste time searching files you don’t care about

Binary file detection: Automatically skips binaries, PDFs, images

Unicode support: Proper UTF-8 handling (looking at you, ancient tools)

Regex engine: Uses Rust’s regex crate - fast, safe, and feature-complete

Why Developers Love It

Installation: Single binary, no dependencies, works everywhere

Ergonomics: Sensible flags, colored output, great error messages

Speed: Written in Rust with performance as a first-class concern

Reliability: Memory safe, handles edge cases gracefully

Real Talk: The Migration Story

I switched from ag to rg three years ago and never looked back. The speed difference is immediately noticeable on large codebases. But the real win is the smart defaults - it just does what you expect without tons of flags.

For Vim users: Works perfectly with fzf, telescope, or any fuzzy finder

For VS Code users: Powers the search in most Rust extensions

For CI/CD: Fast enough to use in build scripts and linting pipelines

The Rust Connection

This is showcase Rust code. The performance comes from:

  • Zero-cost abstractions
  • No garbage collector overhead
  • Fearless concurrency (parallel search across CPU cores)
  • Memory safety without runtime penalties

Who Needs This

  • Anyone who searches through code files regularly
  • Rust developers (obviously)
  • Large codebase maintainers where speed matters
  • DevOps teams processing logs and config files
  • Data engineers grepping through text datasets

Bottom Line: If you’re still using basic grep for development work, you’re leaving serious productivity on the table. ripgrep is what search should be in 2025.

Posted by RyanMalloy on Wednesday January 16, @02:45PM from the dept-of-blazing-fast-search dept.