#prefix-tree #trie #generic #key-value #tree-search #object-key


Generic trie data structure implementation (prefix tree) with support for different key and value types, and functions to search for common prefixes or postfixes

5 unstable releases

0.7.0 Dec 23, 2023
0.6.0 Dec 22, 2023
0.5.4 Dec 21, 2023
0.5.2 Dec 21, 2023
0.5.0 Dec 21, 2023

#802 in Data structures

Download history 5/week @ 2023-12-15 5/week @ 2023-12-22 6/week @ 2023-12-29 71/week @ 2024-02-23 22/week @ 2024-03-01 138/week @ 2024-03-29

138 downloads per month
Used in curies

Custom license

237 lines

🎄 Prefix Trie

Crates.io Test Release Documentation Codecov status MIT license

PTrie is a generic implementation of the trie data structure with no dependencies, tailored for easy and efficient prefix and postfix search within a collection of objects, such as strings.

The structure is defined as Trie<K, V>, where K represents the type of keys in each node (an iterator of the chain to index), and V is the type of the associated values (any object to which the key points to).

💭 Motivation

The trie is particularly effective for operations involving common prefix identification and retrieval, making it a good choice for applications that require fast and efficient prefix-based search functionalities.

🚀 Usage

Results are sorted in ascending order of their length.

✨ Find prefixes

You can return all prefixes in the trie that matches a given string, or directly retrieve the longest prefix.

use ptrie::Trie;

let mut trie = Trie::new();

trie.insert("a".bytes(), "A");
trie.insert("ab".bytes(), "AB");
trie.insert("abc".bytes(), "ABC");
trie.insert("abcde".bytes(), "ABCDE");

let prefixes = trie.find_prefixes("abcd".bytes());
assert_eq!(prefixes, vec![&"A", &"AB", &"ABC"]);

let longest = trie.find_longest_prefix("abcd".bytes());
assert_eq!(longest, Some("ABC").as_ref());

🔍 Find postfixes

You can also find all postfixes in the trie, e.g. all strings which have the given string as a prefix, and extends it.

use ptrie::Trie;

let mut trie = Trie::new();

trie.insert("app".bytes(), "App");
trie.insert("apple".bytes(), "Apple");
trie.insert("applet".bytes(), "Applet");
trie.insert("apricot".bytes(), "Apricot");

let strings = trie.find_postfixes("app".bytes());
assert_eq!(strings, vec![&"App", &"Apple", &"Applet"]);

🔑 Key-based retrieval functions

The crate provides functions to check for the existence of a key, to retrieve the associated value, or iterate the trie nodes.

use ptrie::Trie;

let mut trie = Trie::new();
trie.insert("app".bytes(), "App");
trie.insert("applet".bytes(), "Applet");

assert_eq!(trie.get("app".bytes()), Some("App").as_ref());
assert_eq!(trie.get("none".bytes()), None.as_ref());

for (k, v) in trie.iter() {
    println!("kv: {:?} {}", k, v);

🏷️ Features

The serde feature adds Serde Serialize and Deserialize traits to the Trie and TrieNode struct.

ptrie = { version = "0.6", features = ["serde"] }

🛠️ Contributing

Contributions are welcome, checkout the CONTRIBUTING.md for instructions to run the project in development.

📜 Changelog

Changelog available in the CHANGELOG.md.

⚖️ License

MIT License