.NET Framework 4.8下支持忽略大小写全词匹配的字符串替换方案

原实现中使用的\b正则单词边界仅匹配\w（字母、数字、下划线）与\W（非\w字符）的交界位置，当待匹配词本身以非\w字符开头/结尾时，\b无法匹配到对应的边界，这就是特殊字符全词匹配失效的核心原因。同时原边界规则也无法处理目标词与标点相邻的全词匹配场景。

仅调整全词匹配的边界逻辑，用自定义环视规则替代原生\b，匹配逻辑完全符合需求：

using System;
using System.Text.RegularExpressions;

public static class StringExtensions
{
    public static string Replace(this string s, string word, string by, bool IgnoreCase, bool WholeWord)
    {
        // 空输入直接返回
        if (string.IsNullOrEmpty(s) || string.IsNullOrEmpty(word))
            return s;

        RegexOptions regexOptions = RegexOptions.None;
        if (IgnoreCase)
            regexOptions |= RegexOptions.IgnoreCase;
        regexOptions |= RegexOptions.CultureInvariant;

        string pattern = Regex.Escape(word);
        if (WholeWord)
        {
            // 自定义全词边界：前后不能是字母/数字，天然支持句首、句尾、与标点相邻的场景
            pattern = @"(?<!\p{L}|\p{N})" + pattern + @"(?!\p{L}|\p{N})";
        }

        return Regex.Replace(s, pattern, by, regexOptions);
    }
}

效果验证 #

• 替换逗号场景：输入"Apple , Mango , Banana."，替换,为" and "，全词匹配开启，输出为"Apple and Mango and Banana."
• 替换单词场景：输入"She is Mari and she is Marijane. Mari is at home."，替换"Mari"为"Mira"，全词匹配开启，输出为"She is Mira and she is Marijane. Mira is at home."，不会误改Marijane

适合高频调用场景，无正则解析开销，逻辑完全可控：

using System;
using System.Collections.Generic;

public static class StringExtensions
{
    public static string Replace(this string s, string word, string by, bool IgnoreCase, bool WholeWord)
    {
        if (string.IsNullOrEmpty(s) || string.IsNullOrEmpty(word) || word.Length > s.Length)
            return s;

        StringComparison comparison = IgnoreCase 
            ? StringComparison.OrdinalIgnoreCase 
            : StringComparison.Ordinal;

        List<int> matchIndexes = new List<int>();
        int wordLength = word.Length;
        int i = 0;

        while (i <= s.Length - wordLength)
        {
            if (string.Compare(s, i, word, 0, wordLength, comparison) == 0)
            {
                bool isMatch = true;
                if (WholeWord)
                {
                    // 左边界校验：左侧有字符时不能是字母/数字
                    if (i > 0 && char.IsLetterOrDigit(s[i - 1]))
                        isMatch = false;
                    // 右边界校验：右侧有字符时不能是字母/数字
                    if (isMatch && i + wordLength < s.Length && char.IsLetterOrDigit(s[i + wordLength]))
                        isMatch = false;
                }

                if (isMatch)
                {
                    matchIndexes.Add(i);
                    i += wordLength; // 跳过已匹配段避免重叠匹配
                    continue;
                }
            }
            i++;
        }

        if (matchIndexes.Count == 0)
            return s;

        // 高效构建结果字符串
        char[] result = new char[s.Length + matchIndexes.Count * (by.Length - wordLength)];
        int resultPos = 0;
        int lastPos = 0;

        foreach (int idx in matchIndexes)
        {
            s.CopyTo(lastPos, result, resultPos, idx - lastPos);
            resultPos += idx - lastPos;
            by.CopyTo(0, result, resultPos, by.Length);
            resultPos += by.Length;
            lastPos = idx + wordLength;
        }
        s.CopyTo(lastPos, result, resultPos, s.Length - lastPos);

        return new string(result);
    }
}

内容的提问来源于stack exchange，提问作者Jaspreet Sidhu