libucd

This library extends the inbuilt char type with the Codepoint trait, which implements 100 properties of the UCD (Unicode Character Database). It aims to be fast and compact, and to have minimal dependencies (it does not require the rust standard library so only needs rust's core crate).

extern crate ucd;
use ucd::Codepoint;

fn main() {
    let salawat: char = 'ﷺ';
    let decomp: String = salawat.decomposition_map().collect();
    println!("{} -> {}", salawat, decomp);
    // ﷺ -> صلى الله عليه وسلم
}

Though the library is fairly extensive, it is not complete. The properties it lacks are:

• Character names and aliases
• Unihan properties
• Tangut source data
• Named sequences
• Standardised variants
• CJK Radicals
• Emoji source data

Data was compressed as arrays of ranges of codepoints having each value for each property, excluding the most common value. Lookup is then implemented by a binary search which should provide O(1) access. A very unscientific test suggests each lookup takes around 100ns on a core i5 processor.

Disclaimer

Please note that this data has been derived from the flat XML version of the UCD. Though this is published by Unicode, it is not the official version of the database. As such, it cannot be guaranteed that this library is entirely consistent with the official database. The unit tests do however suggest consistency with the XML database at least.

Reference

Note, in most cases enum values are simply the full name of the property value converted to camelcase. Perhaps I will eventually create a rustdoc version that will be more helpful.

In the tables below, the first column is the name of the property, and the second column is the method name. The methods are implemented via the ucd::Codepoint trait on char.