String replace exact match in cyrillic

Question

I want to use regex for string replace with Cyrillic characters. I want to use exact match option. My string replace is working with Latin characters and is looking like that:

'Edin'.replace(/\Edin\b/gi, '');  // Output is ""

The same expression is not working with Cyrillic characters

'Един'.replace(/\Един\b/gi, '');  // Output is still 'Един'

Answer 1

The problem here is \b word boundary chracter, which matches position at a word boundary. Word boundary is defined as (^\w|\w$|\W\w|\w\W). And in its turn word character \w is a set of ASCII characters [A-Za-z0-9_]. Obviously Cyrillic characters don't fall into this set.

For example, for the same reason /\w+/ regular expression will not match Cyrillyc string.

Answer 2

As dfsq wrote the problem is with word boundary. If you remove \b you will get desired output, but it is quite different regex. It will replace Един also in cases where it is a part of word. To avoid that you can use negative lookahead and define which letters shouldn't appear behind, because they could be a part of word.

'Един'.replace(/\Един(?![A-я])/gi, '');