Wordnik List in Rust

A quick, local library to query valid English words.

Description

This is a small library that exports useful methods for recognizing valid words from the Wordnik Word List. Definitions, usage, pronunciation, etymology, and examples are omitted. This library uses the official list from wordnik (last synced December 2024). Internally, it uses a static BTreeSet generated at runtime, meaning cross-thread word checking in a small footprint under 10 MB.

Installation

Add to your Rust project with cargo add wordnik-list

Startup runtime

For the first lookup, generating the map takes around 13 milliseconds in release mode on a mid-range desktop in 2024. For debug mode, initial lookup took around 45 milliseconds.

Random lookup time

Runtime is around 300 nanoseconds per lookup in release mode on a mid-range desktop in 2024. For debug mode, runtime per lookup is around 900 nanoseconds.

Stability note

This crate may change implementation at any time to improve memory efficiency or startup time, but main exported functions should still be there.

Example:

use wordnik_list::word_exists;

let try_words = ["list", "rust", "rusty", "ruster", "abroptly", "assertion"];
// Print which words in that list are invalid
for word in try_words {
    if !word_exists(word) {
        println!("\"{}\" is not a valid word", word);
    }
}

Another example:

use wordnik_list::word_iterator;

// Collect list of every 2-letter word
let vec: Vec<&str> = word_iterator().filter(|word| word.len() == 2).collect();
println!("List of every 2-letter word: {:?}", vec);

Changelog

Version 0.2.1

• Add the no_std flag.

Version 0.2.0

• Added new function: word_iterator_by_len
• Removed guarantee that iterators returned by word_iterator and word_range will be in alphabetical order.
• Improved internal implementation: replaced BTreeMap with a combined string per word length.
  • Runtime memory usage reduced by about 50%.
  • Performance penalty at first call removed.
  • Word lookup time improved

Version 0.1.0

• Initial release with 3 functions: word_iterator, word_exists, and word_range.
• Based on BTreeMap implementation that is computed at first call