Unicode collation algorithm

The Unicode collation algorithm, often shortened to UCA, is defined in Unicode Technical Report 10. It is a customizable method for turning strings of text in any writing system and language that Unicode can represent into binary sort keys. Because the keys can be compared byte by byte, text can be collated efficiently according to the rules of a given language, with options to ignore case, accents and similar differences. The algorithm requires canonically equivalent character sequences to sort identically and specifies that input is first normalized to Normalization Form D. The same technical report defines the Default Unicode Collation Element Table, known as DUCET, which sets a default ordering that can be tailored for particular languages, with many such tailorings kept in the Unicode Common Locale Data Repository. An open source implementation is included in International Components for Unicode.