Dynamically Evaluate Regex Matches-CodePudding

I was hit with a predicament that I can't quite figure out. I am evaluating many large text strings containing large numbers of spaces between words. I have figured out the text to display properly I need to replace about half of the spaces in each segment with non breaking space characters. This varies if the number of spaces are even or odd. I have the replace boiled down to:

if (numberOfSpaces > 3) {
    double mathresult = (numberOfSpaces / 2);
    int numberNBSP = Math.Ceiling(mathresult);
    int numberSpace = Math.Floor(mathresult);
    string replaceText;
    for(numberNBSP > 0, numberNBSP--)
        replaceText ="&nbsp;";
    for(numberSpace > 0, numberSpace--)
        replaceText =" ";

My issue now is calling this code for each instance of space segments. Each segment needs to be be evaluated individually and I feel like I have a blind spot in RegEx of how to do so. I hope this makes sense, thank you for taking the time to read this!

CodePudding user response：

It's just a matter of passing a callback to Replace which will execute for every match that is made.

For example:

using System;
using System.Text.RegularExpressions;

public class Example
{
    public static void Main()
    {
        string input = "I  hope    this makes   sense,   thank you   for    taking  the     time to   read   this!";
        Console.WriteLine("input:  "   input);
        Regex rx = new Regex(@"  ");
        string output = rx.Replace(input, Evaluator);
        Console.WriteLine("output: "   output);
    }

    static string Evaluator(Match match)
    {
        string replaceText;
        int numberOfSpaces = match.Value.Length;          
        if (numberOfSpaces > 3) {   
            double mathresult = (numberOfSpaces / 2);
            int numberNBSP = (int) Math.Ceiling(mathresult);
            int numberSpace = (int) Math.Floor(mathresult);
            replaceText = "";
            for (; numberNBSP > 0; numberNBSP--) replaceText  = "&nbsp;";
            for (; numberSpace > 0; numberSpace--) replaceText  = " ";
        } else {
            replaceText = match.Value;
        }     
        return replaceText;       
    }
}

Obviously, the logic of replacing the spaces is your own, and I didn't look into that.

Alternatively, you could use the regex string " {4,}" which matches 4 or more space characters, then you could get rid of the if (numberOfSpaces > 3) test etc.

If you want to be able to match all whitespace, such as tabs and newlines, then use \s rather than a single space character.