March 27, 2026
Faster JSON, louder comments
A Faster Alternative to Jq
Meet jsongrep: the speed‑obsessed jq rival fans love—and roast
TLDR: jsongrep claims to search huge JSON files faster than jq by pre‑compiling the query, drawing cheers from testers and eye‑rolls from skeptics. The thread devolved into speed debates, a Nushell "use this instead" push, a light‑mode bug joke, and Mac users grumbling about no Apple Silicon build.
A new Rust toy just crash‑landed in Dev World: jsongrep, a command‑line helper that promises faster searches in messy JSON files. The author says it’s like ripgrep for JSON—compile the pattern first, then blaze through data. The demo shows neat tricks like wildcards, “either/or” matches, and a one‑flag “find names anywhere” move, all pitched as speed wins inspired by ripgrep.
But the community? Oh, they came to fight. One camp yawns, saying speed is a fake problem—“is jq really slow?”—and drops the classic meme about shaving milliseconds from terminal launches. Another camp actually tried it and went, “okay wait, this slaps,” even if the syntax isn’t identical. Then a third squad rolls in with the plot twist: “Forget both—Nushell replaces all these tools” with one consistent shell environment, linking to Nushell.
Meanwhile, chaos at the sidelines: a reader spotted the blog’s light‑mode color bug and posted the most developer fix ever—toggle dark mode on and off. And Mac users on Apple Silicon? They’re side‑eyeing the lack of native downloads and complaining about Rosetta emulation. Verdict: jsongrep looks fast, the crowd looks divided, and the drama looks delicious.
Key Points
- •jsongrep is a Rust-based, cross-platform CLI tool that queries JSON by matching path expressions and printing matching values.
- •Its query language supports dot paths, wildcards, alternation, recursive descent, and an optional operator, with -F as a shorthand for recursive descent name searches.
- •The tool treats JSON documents as trees and compiles queries as regular languages into deterministic finite automatons (DFAs) for single-pass, O(1) per-symbol processing.
- •It contrasts with tools like jq, JMESPath, and jsonpath-rust that interpret path expressions, highlighting performance benefits of a compiled approach.
- •The article outlines a benchmarking strategy to compare jsongrep’s performance with other JSON path-like tools, inspired by ripgrep’s philosophy.