Regex Not Matching Unicode

Question

How would I go about using Regex to match Unicode strings? I'm loading in a couple keywords from a text file and using them with Regex on another file. The keywords both contain unicode (such as á, etc). I'm not sure where the problem is. Is there some option I have to set?

Code:

foreach (string currWord in _keywordList)
{
    MatchCollection mCount = Regex.Matches(
        nSearch.InnerHtml, "\\b" + @currWord + "\\b", RegexOptions.IgnoreCase);

    if (mCount.Count > 0)
    {
        wordFound.Add(currWord);
        MessageBox.Show(@currWord, mCount.ToString());
    }
}

And reading the keywords to a list:

var rdComp = new StreamReader(opnDiag.FileName);
string compSplit = rdComp.ReadToEnd()
                         .Replace("\r\n", "\n")
                         .Replace("\n\r", "\n");
rdComp.Dispose();
string[] compList = compSplit.Split(new[] {'\n'});

Then I change the array to a list.

Can you post a snippet of your code that's using the regex? It could be a charset problem (e.g., nothing to do with regex), or a regex problem, or... — T.J. Crowder, Mar 29 '10 at 13:33
Do the keywords always start and end with word characters (i.e., letters, digits or underscores)? — Alan Moore, Mar 29 '10 at 15:08
It looks like this should work. Can you reduce it to a single regex scenario that fails to match? — Jeremy Stein, Mar 29 '10 at 15:37

score 1 · Answer 1 · answered Mar 29 '10 at 13:56

1

When matching on a specific character, I believe regular expressions only support literals for the ASCII character set. Beyond that, you can use \uxxxx to match on the Unicode code point.

See here.

answered Mar 29 '10 at 13:56

mbeckish

10,485
5
30
55

I'm not sure that's the problem. She/he isn't using character classes but verbatim strings, surrounded by word boundaries. – Tim Pietzcker Mar 29 '10 at 14:08
@Pietzcker - That's the problem. S/he needs to parse the string and add each character as a unicode code point. – mbeckish Mar 29 '10 at 20:15
1

Well, I've just tried using Unicode literals in a regex in C#, and it worked perfectly. `Console.WriteLine(Regex.Replace("It BӦЯӁڀ!", @"\bBӦЯӁڀ\b", "works"));` returns `It works!` – Tim Pietzcker Mar 29 '10 at 20:56

score 0 · Answer 2 · answered Nov 18 '14 at 16:35

0

You can use [\u0000-\uffff]+ to match at least the BMP

answered Nov 18 '14 at 16:35

brighty

406
3
10

Regex Not Matching Unicode

2 Answers2

Linked