Question

I’m looping through an array of C# keywords, and applying a regex for each one. (I’m working on syntax highlighting.)

I only want to match if:

• The keyword is the first word on the line OR the character before is
  a space, period or an open parenthesis

  AND

• The character after the keyword is a space, period or open
  parenthesis.

Here’s what I came up with:

    foreach (string keyword in keyWords)
    {
        regex = new Regex(@"(?<=[\s\.(])" + keyword + @"(?=[\s\.(])");
        foreach (Match match in regex.Matches(code))
        {
            code = code.Replace(match.Value, "<span class='keyword'>" + match.Value + "</span>");
        }
    }

So, in the case of the following text:

“foreach(string s in ss){}”

The word “foreach” matches, BUT also the keyword “in” nested int he word “string” matches – but that’s not good because the characters before and after don’t match the criteria.

Interestingly enough, in the case of the following text:

“xforeachx(string s in ss){}”

The word “foreach” doesn’t match.

So why does the “in” in the word “string” match but not the “foreach” in the second example? What am I missing?

Thank you!

Answer 1

Here is a very simple demo of what I referencing in a comment:

StringBuilder sb = new StringBuilder();
sb.AppendLine("using System;");
sb.AppendLine();
sb.AppendLine("namespace Foo.Bar");
sb.AppendLine("{");
sb.AppendLine("\tpublic class Baz");
sb.AppendLine("\t{");
sb.AppendLine("\t\tpublic static void Main()");
sb.AppendLine("\t\t{");
sb.AppendLine("\t\t\tString[] a = new[]{\"foo\",\"bar\",\"baz\"};");
sb.AppendLine("\t\t\tforeach (var b in a) Console.WriteLine(b);");
sb.AppendLine("\t\t}");
sb.AppendLine("\t}");
sb.AppendLine("}");

Console.Write(sb.ToString());
Console.WriteLine();

String[] keywords = new String[]{
  "using", "namespace", "for", "if", "else", "foreach"
};
String code = sb.ToString();
foreach (String keyword in keywords){
  String pattern = @"(?<=^|[\s\.\(])(" + Regex.Escape(keyword) + @")(?=[\s\.\)])";
  String replacement = "<span class='keyword'>$1</span>";
  code = Regex.Replace(code, pattern, replacement);
}
Console.WriteLine(code);

Resulting in:

<span class='keyword'>using</span> System;

<span class='keyword'>namespace</span> Foo.Bar
{
  public class Baz
  {
    public static void Main()
    {
      String[] a = new[]{"foo","bar","baz"};
      <span class='keyword'>foreach</span> (var b in a) Console.WriteLine(b);
    }
  }
}

Which is I think what you’re after. I used static regex methods but you can refactor it how you’d like. Some things I’d like to point out:

• I highly recommend using Regex.Escape when you’re inserting values in the middle of a regex statement that you’re not constructing yourself. Even if the keywords turn out to be only letters, some changes at a later date may break it. Better safe than sorry.
• If you’re going to use Regex to find the keyword, also use it to replace it. This insures that if “for” was found (and was a keyword) that that instance of for (with the look(ahead|behind) verification is replaced and not a stray “for” found in the string (maybe they have a variable named foreshadow–who knows.
• I slightly modified your look-behind to include ^| which means match the beginning of a line or what is found in the class.
• I also slightly modified your pattern to include a capture group so the replace had something to locate.