c# Regex non letter characters from a string

Question

My terminology may be a little out here, but i am trying to strip out non letters from a string in C#, so remove dashes ampersands etc, but retain things like accented characters and Chinese characters. All the C# examples i have seen on SO have a regex like this new Regex("[^a-zA-Z0-9 -]");, but my needs are beyond ascii characters.

string input = "I- +AM. 相关 AZURÉE& /30%";

string output = "I AM 相关 AZURÉE 30";

Answer 1

A good starting point would be to remove characters according to their Unicode character class. For example, this code removes everything that is characterized as punctuation, symbol or a control character:

string input = "I- +AM. 相关 AZURÉE& /30%";
var output = Regex.Replace(input, "[\\p{S}\\p{C}\\p{P}]", "");

You could also try the whitelisting approach, by only allowing certain classes. For example, this keeps only characters that are letters, diacritics, digits and spacing:

var output = Regex.Replace(input, "[^\\p{L}\\p{M}\\p{N}\\p{Z}]", "");

See it in action.

Answer 2

string result = string.Concat(input.Where(c => Char.IsLetterOrDigit(c)));