1# `matchit` 2 3[](https://docs.rs/matchit) 4[](https://crates.io/crates/matchit) 5[](https://crates.io/crates/matchit) 6 7A high performance, zero-copy URL router. 8 9```rust 10use matchit::Router; 11 12fn main() -> Result<(), Box<dyn std::error::Error>> { 13 let mut router = Router::new(); 14 router.insert("/home", "Welcome!")?; 15 router.insert("/users/:id", "A User")?; 16 17 let matched = router.at("/users/978")?; 18 assert_eq!(matched.params.get("id"), Some("978")); 19 assert_eq!(*matched.value, "A User"); 20 21 Ok(()) 22} 23``` 24 25## Parameters 26 27Along with static routes, the router also supports dynamic route segments. These can either be named or catch-all parameters: 28 29### Named Parameters 30 31Named parameters like `/:id` match anything until the next `/` or the end of the path: 32 33```rust,ignore 34let mut m = Router::new(); 35m.insert("/users/:id", true)?; 36 37assert_eq!(m.at("/users/1")?.params.get("id"), Some("1")); 38assert_eq!(m.at("/users/23")?.params.get("id"), Some("23")); 39assert!(m.at("/users").is_err()); 40``` 41 42### Catch-all Parameters 43 44Catch-all parameters start with `*` and match everything after the `/`. They must always be at the **end** of the route: 45 46```rust,ignore 47let mut m = Router::new(); 48m.insert("/*p", true)?; 49 50assert_eq!(m.at("/foo.js")?.params.get("p"), Some("foo.js")); 51assert_eq!(m.at("/c/bar.css")?.params.get("p"), Some("c/bar.css")); 52 53// note that this would not match: 54assert_eq!(m.at("/").is_err()); 55``` 56 57## Routing Priority 58 59Static and dynamic route segments are allowed to overlap. If they do, static segments will be given higher priority: 60 61```rust,ignore 62let mut m = Router::new(); 63m.insert("/", "Welcome!").unwrap(); // priority: 1 64m.insert("/about", "About Me").unwrap(); // priority: 1 65m.insert("/*filepath", "...").unwrap(); // priority: 2 66``` 67 68## How does it work? 69 70The router takes advantage of the fact that URL routes generally follow a hierarchical structure. Routes are stored them in a radix trie that makes heavy use of common prefixes: 71 72```text 73Priority Path Value 749 \ 1 753 ├s None 762 |├earch\ 2 771 |└upport\ 3 782 ├blog\ 4 791 | └:post None 801 | └\ 5 812 ├about-us\ 6 821 | └team\ 7 831 └contact\ 8 84``` 85 86This allows us to reduce the route search to a small number of branches. Child nodes on the same level of the tree are also prioritized 87by the number of children with registered values, increasing the chance of choosing the correct branch of the first try. 88 89# Benchmarks 90 91As it turns out, this method of routing is extremely fast. In a benchmark matching 4 paths against 130 registered routes, `matchit` find the correct routes 92in under 200 nanoseconds, an order of magnitude faster than most other routers. You can view the benchmark code [here](https://github.com/ibraheemdev/matchit/blob/master/benches/bench.rs). 93 94```text 95Compare Routers/matchit 96time: [197.57 ns 198.74 ns 199.83 ns] 97 98Compare Routers/actix 99time: [26.805 us 26.811 us 26.816 us] 100 101Compare Routers/path-tree 102time: [468.95 ns 470.34 ns 471.65 ns] 103 104Compare Routers/regex 105time: [22.539 us 22.584 us 22.639 us] 106 107Compare Routers/route-recognizer 108time: [3.7552 us 3.7732 us 3.8027 us] 109 110Compare Routers/routefinder 111time: [5.7313 us 5.7405 us 5.7514 us] 112``` 113 114# Credits 115 116A lot of the code in this package was based on Julien Schmidt's [`httprouter`](https://github.com/julienschmidt/httprouter). 117