使用split删除字符串中用双引号引起来的部分

Question

我将以下内容用作C＃中split函数的参数：

    private char[] delimiterComment = { '(', '{', '[', '\u201C' };
    private char[] delimiterEndComment = { ')', '}', ']', '\u201D' };

它适用于所有“括号”，但不适用于“卷曲双引号”。 我不明白为什么。 它是split的错误还是卷曲引号字符的功能？

作为输入字符串，我输入如下内容：“这是一条预评论”，然后是一些重要信息（可能有嵌入的评论），然后是一些更重要的信息（然后是发布评论）。

我希望删除注释，但将其捕获到结构中，并留下干净的信息字符串。 一切都很好地用了括号，直到我尝试添加卷曲的双引号作为附加的定界符...（我知道嵌入的注释是有意收集的，它们是后注释）。我编写的代码如下：

class CommentSplit
{
    public bool split = false;
    public bool error = false;
    public string original = "";
    public string remainder = "";
    public string preComment = "";
    public string postComment = "";

    public CommentSplit(string inString, char[] startComment, char[] endComment, string[] ignoreStrings, string[] addStrings, bool usePre) // creator
    {
        if (inString == null)
            return;
        original = inString;
        string[] starts = inString.Split(startComment);
        if (starts.Length == 1)
        {
            remainder = inString;
            return;
        }
        if (starts[0] != "")
            remainder += starts[0].TrimEnd();
        for (int i = 1; i < starts.Length; i++)
        {
            string[] ends = starts[i].Split(endComment);
            if (ends.Length != 2) // more than one end comment for a start comment - BUT what about one start and one end comment
            {
                error = true;
                return;
            }
            if (addStrings == null)
            {
                if (ignoreStrings == null)
                {
                    if ((remainder == "") && usePre)
                        preComment += ends[0];
                    else
                        postComment += ends[0];
                }
                else
                {
                    bool ignore = false;
                    for (int z = 0; z < ignoreStrings.Length; z++)
                    {
                        if (ends[0].ToLower() == ignoreStrings[z])
                            ignore = true;
                    }
                    if (!ignore) // was a comment but we might want to ignore it
                    {
                        if ((remainder == "") && usePre)
                        {
                            if (preComment != "")
                                preComment += " ";
                            preComment += ends[0];
                        }
                        else
                        {
                            if (postComment != "")
                                postComment += " ";
                            postComment += ends[0];
                        }
                    }
                }
            }
            else
            {
                bool add = false;
                for (int z = 0; z < addStrings.Length; z++)
                {
                    if (ends[0].ToLower() == addStrings[z])
                        add = true;
                }
                if (add) // was a comment but want it in the remainder
                {
                    if (remainder != "")
                        remainder += " ";
                    remainder += ends[0];
                }
                else
                {
                    if (ignoreStrings == null)
                    {
                        if ((remainder == "") && usePre)
                            preComment += ends[0];
                        else
                            postComment += ends[0];
                    }
                    else
                    {
                        bool ignore = false;
                        for (int z = 0; z < ignoreStrings.Length; z++)
                        {
                            if (ends[0].ToLower() == ignoreStrings[z])
                                ignore = true;
                        }
                        if (!ignore) // was a comment but we might want to ignore it
                        {
                            if ((remainder == "") && usePre)
                            {
                                if (preComment != "")
                                    preComment += " ";
                                preComment += ends[0];
                            }
                            else
                            {
                                if (postComment != "")
                                    postComment += " ";
                                postComment += ends[0];
                            }
                        }
                    }
                }

            }
            if (remainder != "")
                remainder += " ";
            remainder += ends[1].Trim();
        }
        split = true;
    } // CommentSplit
}

我应该注意，我是一位C ++的退休C程序员，所以我的风格可能不是OOP有效的。 我本来确实包含直（非卷曲）双引号，但是它们并不重要，并且确实存在一些问题，因为它们没有前后分隔符版本。

Answer 1

这是一个小的可验证示例，在您的代码中还有其他优点：

char[] delimiterComment = { '(', '{', '[', '\u201C', '\u201D', '"', '“', '”', '}', ']', ')' };
string stringWithComment = "this has a “COMMENT” yeah really";
var result = stringWithComment.Split(delimiterComment);
//Output:
//result[0] = "this has a "
//result[1] = "COMMENT"
//result[2] = " yeah really"

Answer 2

因此，您想删除评论 ，例如

 123 (456 [789) abc ] -> 123 abc ]

在这种情况下，您可以尝试一个简单的循环：

//TODO: I suggest combining starts and ends into array of pairs, e.g.
// KeyValuePair<string,string>[]
private static string CutOffComments(string source, char[] starts, char[] ends) {
  if (string.IsNullOrEmpty(source))
    return source;

  StringBuilder sb = new StringBuilder(source.Length);

  int commentIndex = -1;

  foreach (var c in source) {
    if (commentIndex >= 0) { // within a comment, looking for its end
      if (c == ends[commentIndex])
        commentIndex = -1;
    }
    else { // out of comment, do we starting a new one?
      commentIndex = Array.IndexOf(starts, c);

      if (commentIndex < 0)
        sb.Append(c);
    }
  }

  //TODO:
  // if (commentIndex >= 0) // dungling comment, e.g. 123[456

  return sb.ToString(); 
}

用法：

string source = "123[456]789";
// 123789
string result = CutOffComments(source, delimiterComment, delimiterEndComment); 

Answer 3

只需将双引号放在单引号之间，而不使用转义符：

private char[] delimiterComment = { '(', '{', '[', '\u201C', '"' };
private char[] delimiterEndComment = { ')', '}', ']', '\u201D', '"' };

输入：

string s = "abc(121), {12}, \" HI \"";
Console.WriteLine(string.Join(Environment.NewLine,(s.Split(delimiterComment)).Select(s=> s)));

输出：

abc
121), 
12}, 
 HI