October 30, 2025
Fuzzy feels, sharper takes
Show HN: Front End Fuzzy and Substring and Prefix Search
A speedy new name search lands—and the crowd instantly asks, “But how is it different”
TLDR: A new fuzzy search library promises fast, accurate, multilingual results with no dependencies and easy updates. The top community reaction: enthusiasm tempered by demands for head‑to‑head benchmarks—especially against uFuzzy—and real‑world tests to prove the speed claims and show how it handles non‑English names.
A slick new search tool just hit Hacker News promising lightning‑fast, typo‑friendly results that work in the browser and on servers. It says it’s fast, accurate, multilingual, and dependency‑free—basically, a neat little package that finds “Alice King” even when you type “alice kign.” Cue the community drama: the very first vibe was, “Cool… but show us the tradeoffs.” One commenter immediately pointed to the gold standard comparison chart and asked for an interactive showdown like μFuzzy. Translation: screenshots are cute; benchmarks are better.
From there, the peanut gallery wanted receipts: how does it stack up against the dozens of similar JavaScript fuzzy finders? What datasets were used? How does it handle non‑English names and accents? Folks loved the promise of adding, updating, and removing entries on the fly—great for apps with constantly changing lists—but they also dropped the classic HN line: show the numbers or it didn’t happen. The “no dependencies” flex got nods, and the “works in frontend and backend” bit earned a few chef’s kiss emojis in spirit.
Meanwhile, jokesters crowned it the “kign of typos,” and someone quipped that devs collect fuzzy search libraries like Pokémon. The consensus mood: intriguing tool, now give us side‑by‑side tests—preferably with a flashy demo and a spicy bake‑off against uFuzzy and friends.
Key Points
- •@m31coding/fuzzy-search provides fuzzy, substring, and prefix search for object entities by names and features.
- •Queries typically run in under 10 ms and use a suffix array and n-grams with a novel character sorting approach.
- •The library is multilingual, has no dependencies, and works in both browser and Node.js environments.
- •Distributed via microbundle with ESM, CommonJS, UMD, modern builds, and TypeScript definitions; installed via npm.
- •Usage examples cover indexing, querying, removing, and upserting entities, with configurable searchers and quality thresholds.