FFF file search sdk - over 100 times faster than ripgrep and fzf

I've been building my file search in rust for a while as a neovim plugin and it appeared to be so good that have finally I released it in a form for SDK. Also as an official rust crate. 

!it is already adopted by such an actively used applications like opencode!<

In short this is the best in class file name search and content search. That doesn't require a long running indexing and is very fast and extremely accurate (we even support fuzzy content search on the code that actually works)

There is a public demo with a search of 3 different types of repositories (2k, 100k, and 500k files) that you can try yourself - https://fff.dmtrkovalenko.dev/

The biggest flex: on this specific search it is over 500x faster than ripgrep while running on a way slower machine (only 2 cores VPS)

over 500x faster on the same query

FFF is used both for grep and typo-resistant fuzzy file search, the filename fuzzy search is extremely efficient and being able to pin point mostly any query to the file name, for example this
"scripstcocoentetess unsignelessrehangzerococci" will be matched to this file scripts/coccinelle/tests/unsigned_lesser_than_zero.cocci

you can try it yourself

https://fff.dmtrkovalenko.dev/?repo=1&q=unsignelessrehangzerococci

Essentially this is just the fastest file search implementation that doesn't require an index (there is startup time cost but for 100k linux repo it is ~3 seconds) and support both file finding, content search, git status, frecency of file opens, persisting querying history and actually much more

https://github.com/dmtrKovalenko/fff.nvim

P.S. the repo looks like neovim plugin but this is actually a monorepo for everything. Rust crate is here https://crates.io/crates/fff-search