1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38
//! FST (Fast Succinct Trie) implementation in Rust. //! //! [Master API Docs](https://laysakura.github.io/fst-rs/fst_rs/) //! | //! [Released API Docs](https://docs.rs/crate/fst-rs) //! | //! [Benchmark Results](https://laysakura.github.io/fst-rs/criterion/report/) //! | //! [Changelog](https://github.com/laysakura/fst-rs/blob/master/CHANGELOG.md) //! //! [](https://travis-ci.com/laysakura/fst-rs) //! [](https://crates.io/crates/fst-rs) //! [](https://crates.io/crates/fst-rs) //! [](https://github.com/laysakura/fst-rs#rust-version-supports) //! [](https://github.com/laysakura/fst-rs/blob/master/LICENSE-MIT) //! [](https://github.com/laysakura/fst-rs/blob/master/LICENSE-APACHE) //! //! FST is a building block of SuRF (Succinct Range Filter). //! SuRF and FST is first introduced in the [SIGMOD 2018 best paper](http://www.pdl.cmu.edu/PDL-FTP/Storage/surf_sigmod18.pdf). //! //! FST is a memory efficient and high performance trie. //! FST is like LOUDS based trie (e.g. [trie-rs](https://github.com/laysakura/trie-rs)) but does not use pure LOUDS. //! It instead uses LOUDS-DS (abbrev of LOUDS-Dense & LOUDS-Sparse), which uses bitmap-based encoding (fast but fat) for upper levels of a trie tree and LOUDS-based encoding (slow but memory-efficient) for lower levels. //! //! # Quickstart //! //! To use fst-rs, add the following to your `Cargo.toml` file: //! //! ```toml //! [dependencies] //! fst-rs = "0.1" # NOTE: Replace to latest minor version. //! ``` //! //! ## Usage Overview //! (TBD) //! //! # Features //! (TBD)