Replies: 1 comment
-
Hi there, thanks for the interest! Nice work on that SIMD matcher. Just ran a couple of benches locally and it indeed seems consistently faster by a factor >2. The only concern I have is that the high-level All I'm saying is if we're heading that way, we might also have to replicate these features (if it's even possible with SIMD) to achieve similar performance. |
Beta Was this translation helpful? Give feedback.
-
Hey there, love the project! I recently wrote a fuzzy matching library in SIMD that outperforms nucleo and includes many of the scoring bonuses from FZY. It also optionally allows for typos by disabling the prefiltering step in nucleo/fzf. If you're interested in testing it, I'd be more than happy to throw up a PR integrating it!
https://github.com/Saghen/frizbee
Beta Was this translation helpful? Give feedback.
All reactions