如何仅移除字符串中单词内部的标点符号？

需求说明 #

移除字符串中单词内部的所有标点符号，保留单词开头或结尾的标点。示例：we'll 转为 well，We're 转为 Were；而 build,、crazy! 这类首尾的标点需保留。

测试输入示例：

let testString = "What's with this project I'm trying to build, it's cra!zy!"

已尝试的缺陷方案 #

1. 逐个指定标点移除 #

if let resultRange = myString.range(of: "'") {
    let startIndex = resultRange.lowerBound
    let endIndex = resultRange.upperBound
    let range = startIndex..<endIndex
    let result = myString.replacingCharacters(in: range, with: "")
    print(result) // 输出："Whats with this project Im trying to build, its cra!zy"

问题：只能手动指定要移除的标点，遇到新标点需新增判断，扩展性极差，无法覆盖全量标点类型。

2. 移除所有标点 #

extension String {
  func removingCharacters(inCharacterSet forbiddenCharacters:CharacterSet) -> String
{
    var filteredString = self
    while true {
      if var forbiddenCharRange = filteredString.rangeOfCharacter(from: forbiddenCharacters)  {
        filteredString.removeSubrange(forbiddenCharRange)
      }
      else {
        break
      }
    }

    return filteredString
  }
}

var resultMyString = myString.removingCharacters(inCharacterSet: .punctuationCharacters)
print(resultMyString) // 输出："Whats with this project Im trying to build its crazy"

问题：会一并移除单词首尾的标点，比如build,变成build、crazy!变成crazy，不符合需求。

期望输出 #

"Whats with this project Im trying to build, its crazy!"

解决方案 #

使用正则表达式精准匹配「前后都被字母包围的标点」（即单词内部标点），再将其替换为空字符串。Swift实现如下：

extension String {
    func removeInnerPunctuation() -> String {
        // 正则规则：匹配前后均为字母的标点符号
        let pattern = "(?<=\\p{L})\\p{P}(?=\\p{L})"
        return self.replacingOccurrences(of: pattern, with: "", options: .regularExpression)
    }
}

// 测试验证
let testString = "What's with this project I'm trying to build, it's cra!zy!"
let result = testString.removeInnerPunctuation()
print(result) // 输出："Whats with this project Im trying to build, its crazy!"

正则规则说明 #

• (?<=\\p{L})：正向预查，确保标点前是任意Unicode字母
• \\p{P}：匹配任意Unicode标点符号，覆盖所有标点类型
• (?=\\p{L})：正向预查，确保标点后是任意Unicode字母

该方案可精准移除单词内部标点，同时完整保留首尾标点。

内容的提问来源于stack exchange，提问作者Scott W