在C#中用空格分割字符串

Question

我想用白色空格分割一个字符串,除非字符串中的文本是双引号("text")或单引号('text').

我正在使用此功能:

public static string[] ParseKeywordExpression(string keywordExpressionValue, bool isUniqueKeywordReq)
{
    keywordExpressionValue = keywordExpressionValue.Trim();
    if (keywordExpressionValue == null || !(keywordExpressionValue.Length > 0))
        return new string[0];
    int idx = keywordExpressionValue.Trim().IndexOf(" ");
    if (idx == -1)
        return new string[] { keywordExpressionValue };
    //idx = idx + 1;
    int count = keywordExpressionValue.Length;
    ArrayList extractedList = new ArrayList();
    while (count > 0)
    {
        if (keywordExpressionValue[0] == '"')
        {
            int temp = keywordExpressionValue.IndexOf(BACKSLASH, 1, keywordExpressionValue.Length - 1);
            while (keywordExpressionValue[temp - 1] == '\\')
            {
                temp = keywordExpressionValue.IndexOf(BACKSLASH, temp + 1, keywordExpressionValue.Length - temp - 1);
            }
            idx = temp + 1;
        }
        if (keywordExpressionValue[0] == '\'')
        {
            int temp = keywordExpressionValue.IndexOf(BACKSHASH_QUOTE, 1, keywordExpressionValue.Length - 1);
            while (keywordExpressionValue[temp - 1] == '\\')
            {
                temp = keywordExpressionValue.IndexOf(BACKSHASH_QUOTE, temp + 1, keywordExpressionValue.Length - temp - 1);
            }
            idx = temp + 1;
        }
        string s = keywordExpressionValue.Substring(0, idx);
        int left = count - idx;
        keywordExpressionValue = keywordExpressionValue.Substring(idx, left).Trim();
        if (isUniqueKeywordReq)                    
        {
            if (!extractedList.Contains(s.Trim('"')))
            {
                extractedList.Add(s.Trim('"'));
            }
        }
        else
        {
            extractedList.Add(s.Trim('"'));
        }
        count = keywordExpressionValue.Length;
        idx = keywordExpressionValue.IndexOf(SPACE);
        if (idx == -1)
        {
            string add = keywordExpressionValue.Trim('"', ' ');
            if (add.Length > 0)
            {
                if (isUniqueKeywordReq )
                {
                    if (!extractedList.Contains(add))
                    {
                        extractedList.Add(add);
                    }
                }
                else
                {
                    extractedList.Add(add);
                }
            }                   
            break;
        }
    }
    return (string[])extractedList.ToArray(typeof(string));
}

有没有其他方法可以做到这一点,还是可以优化这个功能？

例如,我希望拆分字符串

%ABC %% aasdf%aalasdjjfas"c:\ Document and Setting\Program Files\abc.exe"

至

%ABC %%
aasdf%
aalasdjjfas
"c:\ Document and Setting\Program Files\abc.exe"

Answer 1

最简单的正则表达式,处理单引号和双引号:

("((\\")|([^"]))*")|('((\\')|([^']))*')|(\S+)

var regex = new Regex(@"(""((\\"")|([^""]))*"")|('((\\')|([^']))*')|(\S+)");
var matches = regex.Matches(inputstring);
foreach (Match match in matches) {
    extractedList.Add(match.Value);
}

所以基本上四到五行代码就足够了.

表达,解释说:

Main structure:
("((\\")|([^"]))*")    Double-quoted token
|                      , or
('((\\')|([^']))*')    single-quoted token
|                      , or
(\S+)                  any group of non-space characters

Double-quoted token:
(                      Group starts
    "                  Initial double-quote
    (                  Inner group starts
        (\\")          Either a backslash followed by a double-quote
        |              , or  
        ([^"])         any non-double-quote character
    )*                 The inner group repeats any number of times (or zero)
    "                  Ending double-quote
)

Single-quoted token:
(                      Group starts
    '                  Initial single-quote
    (                  Inner group starts
        (\\')          Either a backslash followed by a single-quote
        |              , or  
        ([^'])         any non-single-quote character
    )*                 The inner group repeats any number of times (or zero)
    '                  Ending single-quote
)

Non-space characters:
(                      Group starts
    \S                 Non-white-space character
    +                  , repeated at least once
)                      Group ends