当前位置： 文江博客话题详情

C# 将包含命令行参数的字符串拆分为 string[]

发布于 2024-07-09 05:40:11 字数 976 浏览 9 评论 0原文

我有一个字符串，其中包含要传递给另一个可执行文件的命令行参数，并且我需要提取包含各个参数的 string[]，其方式与在命令行上指定命令时使用 C# 的方式相同。 通过反射执行另一个程序集入口点时将使用 string[]。

有一个标准函数吗？ 或者是否有正确分割参数的首选方法（正则表达式？）？ 它必须正确处理可能包含空格的 '"' 分隔字符串，因此我不能只在 ' ' 上进行分割。

示例字符串：

string parameterString = @"/src:""C:\tmp\Some Folder\Sub Folder"" /users:""[email protected]"" tasks:""SomeTask,Some Other Task"" -someParam foo";

示例结果：

string[] parameterArray = new string[] { 
  @"/src:C:\tmp\Some Folder\Sub Folder",
  @"/users:[email protected]",
  @"tasks:SomeTask,Some Other Task",
  @"-someParam",
  @"foo"
};

我不需要命令行解析库，只是获取 String[ 的一种方法] 应该生成。

更新：我必须更改预期结果以匹配 C# 实际生成的结果（删除了拆分字符串中的额外“”）

原文

I have a single string that contains the command-line parameters to be passed to another executable and I need to extract the string[] containing the individual parameters in the same way that C# would if the commands had been specified on the command-line. The string[] will be used when executing another assemblies entry-point via reflection.

Is there a standard function for this? Or is there a preferred method (regex?) for splitting the parameters correctly? It must handle '"' delimited strings that may contain spaces correctly, so I can't just split on ' '.

Example string:

string parameterString = @"/src:""C:\tmp\Some Folder\Sub Folder"" /users:""[email protected]"" tasks:""SomeTask,Some Other Task"" -someParam foo";

Example result:

string[] parameterArray = new string[] { 
  @"/src:C:\tmp\Some Folder\Sub Folder",
  @"/users:[email protected]",
  @"tasks:SomeTask,Some Other Task",
  @"-someParam",
  @"foo"
};

I do not need a command-line parsing library, just a way to get the String[] that should be generated.

Update: I had to change the expected result to match what is actually generated by C# (removed the extra "'s in the split strings)

如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。

扫码二维码加入Web技术交流群

发布评论

需要 登录 才能够评论， 你可以免费 注册 一个本站的账号。

评论（27）

っ〆星空下的拥抱 2024-07-16 05:40:12

Windows 命令行解析器的行为正如您所说，按空格分割，除非前面有未闭合的引号。 我建议你自己编写解析器。 也许是这样的：

    static string[] ParseArguments(string commandLine)
    {
        char[] parmChars = commandLine.ToCharArray();
        bool inQuote = false;
        for (int index = 0; index < parmChars.Length; index++)
        {
            if (parmChars[index] == '"')
                inQuote = !inQuote;
            if (!inQuote && parmChars[index] == ' ')
                parmChars[index] = '\n';
        }
        return (new string(parmChars)).Split('\n');
    }

The Windows command-line parser behaves just as you say, split on space unless there's a unclosed quote before it. I would recommend writing the parser yourself. Something like this maybe:

    static string[] ParseArguments(string commandLine)
    {
        char[] parmChars = commandLine.ToCharArray();
        bool inQuote = false;
        for (int index = 0; index < parmChars.Length; index++)
        {
            if (parmChars[index] == '"')
                inQuote = !inQuote;
            if (!inQuote && parmChars[index] == ' ')
                parmChars[index] = '\n';
        }
        return (new string(parmChars)).Split('\n');
    }
回复 原文
锦欢 2024-07-16 05:40:12

我从杰弗里那里得到了 答案L Whitledge 并对其进行了一些增强。

现在它支持单引号和双引号。 您可以通过使用其他类型的引号在参数本身中使用引号。

它还从参数中删除引号，因为它们对参数信息没有贡献。

    public static string[] SplitArguments(string commandLine)
    {
        var parmChars = commandLine.ToCharArray();
        var inSingleQuote = false;
        var inDoubleQuote = false;
        for (var index = 0; index < parmChars.Length; index++)
        {
            if (parmChars[index] == '"' && !inSingleQuote)
            {
                inDoubleQuote = !inDoubleQuote;
                parmChars[index] = '\n';
            }
            if (parmChars[index] == '\'' && !inDoubleQuote)
            {
                inSingleQuote = !inSingleQuote;
                parmChars[index] = '\n';
            }
            if (!inSingleQuote && !inDoubleQuote && parmChars[index] == ' ')
                parmChars[index] = '\n';
        }
        return (new string(parmChars)).Split(new[] { '\n' }, StringSplitOptions.RemoveEmptyEntries);
    }

I took the answer from Jeffrey L Whitledge and enhanced it a little.

It now supports both single and double quotes. You can use quotes in the parameters itself by using other typed quotes.

It also strips the quotes from the arguments since these do not contribute to the argument information.

    public static string[] SplitArguments(string commandLine)
    {
        var parmChars = commandLine.ToCharArray();
        var inSingleQuote = false;
        var inDoubleQuote = false;
        for (var index = 0; index < parmChars.Length; index++)
        {
            if (parmChars[index] == '"' && !inSingleQuote)
            {
                inDoubleQuote = !inDoubleQuote;
                parmChars[index] = '\n';
            }
            if (parmChars[index] == '\'' && !inDoubleQuote)
            {
                inSingleQuote = !inSingleQuote;
                parmChars[index] = '\n';
            }
            if (!inSingleQuote && !inDoubleQuote && parmChars[index] == ' ')
                parmChars[index] = '\n';
        }
        return (new string(parmChars)).Split(new[] { '\n' }, StringSplitOptions.RemoveEmptyEntries);
    }
回复 原文
白芷 2024-07-16 05:40:12

良好且纯粹的托管解决方案 by Earwicker 无法处理这样的参数：

Test("\"He whispered to her \\\"I love you\\\".\"", "He whispered to her \"I love you\".");

它返回了 3 个元素：

"He whispered to her \"I
love
you\"."

所以这里有一个修复来支持“引用” \"escape\" quote"：

public static IEnumerable<string> SplitCommandLine(string commandLine)
{
    bool inQuotes = false;
    bool isEscaping = false;

    return commandLine.Split(c => {
        if (c == '\\' && !isEscaping) { isEscaping = true; return false; }

        if (c == '\"' && !isEscaping)
            inQuotes = !inQuotes;

        isEscaping = false;

        return !inQuotes && Char.IsWhiteSpace(c)/*c == ' '*/;
        })
        .Select(arg => arg.Trim().TrimMatchingQuotes('\"').Replace("\\\"", "\""))
        .Where(arg => !string.IsNullOrEmpty(arg));
}

用另外 2 个案例进行测试：

Test("\"C:\\Program Files\"", "C:\\Program Files");
Test("\"He whispered to her \\\"I love you\\\".\"", "He whispered to her \"I love you\".");

还注意到 接受了 Atif Aziz 的答案，它使用 < a href="http://msdn.microsoft.com/en-us/library/bb776391.aspx" rel="nofollow noreferrer">CommandLineToArgvW 也失败。 它返回了 4 个元素：

He whispered to her \ 
I 
love 
you".

希望这可以帮助将来寻找此类解决方案的人。

The good and pure managed solution by Earwicker failed to handle arguments like this:

Test("\"He whispered to her \\\"I love you\\\".\"", "He whispered to her \"I love you\".");

It returned 3 elements:

"He whispered to her \"I
love
you\"."

So here is a fix to support the "quoted \"escape\" quote":

public static IEnumerable<string> SplitCommandLine(string commandLine)
{
    bool inQuotes = false;
    bool isEscaping = false;

    return commandLine.Split(c => {
        if (c == '\\' && !isEscaping) { isEscaping = true; return false; }

        if (c == '\"' && !isEscaping)
            inQuotes = !inQuotes;

        isEscaping = false;

        return !inQuotes && Char.IsWhiteSpace(c)/*c == ' '*/;
        })
        .Select(arg => arg.Trim().TrimMatchingQuotes('\"').Replace("\\\"", "\""))
        .Where(arg => !string.IsNullOrEmpty(arg));
}

Tested with 2 additional cases:

Test("\"C:\\Program Files\"", "C:\\Program Files");
Test("\"He whispered to her \\\"I love you\\\".\"", "He whispered to her \"I love you\".");

Also noted that the accepted answer by Atif Aziz which uses CommandLineToArgvW also failed. It returned 4 elements:

He whispered to her \ 
I 
love 
you".

Hope this helps someone looking for such a solution in the future.

回复 原文
别闹i 2024-07-16 05:40:12

我喜欢迭代器，现在 LINQ 使 IEnumerable 为可以轻松用作字符串数组，因此我遵循 Jeffrey L Whitledge 的答案 是（作为 string 的扩展方法）：

public static IEnumerable<string> ParseArguments(this string commandLine)
{
    if (string.IsNullOrWhiteSpace(commandLine))
        yield break;

    var sb = new StringBuilder();
    bool inQuote = false;
    foreach (char c in commandLine) {
        if (c == '"' && !inQuote) {
            inQuote = true;
            continue;
        }

        if (c != '"' && !(char.IsWhiteSpace(c) && !inQuote)) {
            sb.Append(c);
            continue;
        }

        if (sb.Length > 0) {
            var result = sb.ToString();
            sb.Clear();
            inQuote = false;
            yield return result;
        }
    }

    if (sb.Length > 0)
        yield return sb.ToString();
}

I like iterators, and nowadays LINQ makes IEnumerable<String> as easily usable as arrays of string, so my take following the spirit of Jeffrey L Whitledge's answer is (as a extension method to string):

public static IEnumerable<string> ParseArguments(this string commandLine)
{
    if (string.IsNullOrWhiteSpace(commandLine))
        yield break;

    var sb = new StringBuilder();
    bool inQuote = false;
    foreach (char c in commandLine) {
        if (c == '"' && !inQuote) {
            inQuote = true;
            continue;
        }

        if (c != '"' && !(char.IsWhiteSpace(c) && !inQuote)) {
            sb.Append(c);
            continue;
        }

        if (sb.Length > 0) {
            var result = sb.ToString();
            sb.Clear();
            inQuote = false;
            yield return result;
        }
    }

    if (sb.Length > 0)
        yield return sb.ToString();
}
回复 原文
欲拥i 2024-07-16 05:40:12

在你的问题中，你要求使用正则表达式，我是它们的忠实粉丝和用户，所以当我需要像你一样进行同样的参数拆分时，我在谷歌搜索后没有找到简单的解决方案后编写了自己的正则表达式。 我喜欢简短的解决方案，所以我做了一个，如下所示：

            var re = @"\G(""((""""|[^""])+)""|(\S+)) *";
            var ms = Regex.Matches(CmdLine, re);
            var list = ms.Cast<Match>()
                         .Select(m => Regex.Replace(
                             m.Groups[2].Success
                                 ? m.Groups[2].Value
                                 : m.Groups[4].Value, @"""""", @"""")).ToArray();

它处理引号内的空格和引号，并将包含的“”转换为“。请随意使用该代码！

In your question you asked for a regex, and I am a big fan and user of them, so when I needed to do this same argument split as you, I wrote my own regex after googling around and not finding a simple solution. I like short solutions, so I made one and here it is:

            var re = @"\G(""((""""|[^""])+)""|(\S+)) *";
            var ms = Regex.Matches(CmdLine, re);
            var list = ms.Cast<Match>()
                         .Select(m => Regex.Replace(
                             m.Groups[2].Success
                                 ? m.Groups[2].Value
                                 : m.Groups[4].Value, @"""""", @"""")).ToArray();

It handles blanks and quotes inside quotation marks, and converts enclosed "" to ". Feel free to use the code!

回复 原文
萌无敌 2024-07-16 05:40:12

哦，哎呀。 这都是……呃。 但是这个是合法的官方。 来自 Microsoft，使用 C# for .NET Core，也许仅限 Windows，也许跨平台，但 MIT 许可。

选择花絮、方法声明和值得注意的评论；

internal static unsafe string[] InternalCreateCommandLine(bool includeArg0)
private static unsafe int SegmentCommandLine(char * pCmdLine, string[] argArray, bool includeArg0)
private static unsafe int ScanArgument0(ref char* psrc, char[] arg)
private static unsafe int ScanArgument(ref char* psrc, ref bool inquote, char[] arg)

-

// First, parse the program name (argv[0]). Argv[0] is parsed under special rules. Anything up to 
// the first whitespace outside a quoted subtring is accepted. Backslashes are treated as normal 
// characters.

-

// Rules: 2N backslashes + " ==> N backslashes and begin/end quote
//      2N+1 backslashes + " ==> N backslashes + literal "
//         N backslashes     ==> N backslashes

这是从.NET Framework 移植到.NET Core 的代码，我假设是MSVC C 库或CommandLineToArgvW。

这是我用正则表达式处理一些恶作剧的半心半意的尝试，并忽略了零位参数。 这有点神奇。

private static readonly Regex RxWinArgs
  = new Regex("([^\\s\"]+\"|((?<=\\s|^)(?!\"\"(?!\"))\")+)(\"\"|.*?)*\"[^\\s\"]*|[^\\s]+",
    RegexOptions.Compiled
    | RegexOptions.Singleline
    | RegexOptions.ExplicitCapture
    | RegexOptions.CultureInvariant);

internal static IEnumerable<string> ParseArgumentsWindows(string args) {
  var match = RxWinArgs.Match(args);

  while (match.Success) {
    yield return match.Value;
    match = match.NextMatch();
  }
}

在古怪的生成输出上对其进行了相当多的测试。 它的输出与猴子输入并通过 CommandLineToArgvW 运行的内容有相当一部分相匹配。

Oh heck. It's all ... Eugh. But this is legit official. From Microsoft in C# for .NET Core, maybe windows only, maybe cross-platform, but MIT licensed.

Select tidbits, method declarations and notable comments;

internal static unsafe string[] InternalCreateCommandLine(bool includeArg0)
private static unsafe int SegmentCommandLine(char * pCmdLine, string[] argArray, bool includeArg0)
private static unsafe int ScanArgument0(ref char* psrc, char[] arg)
private static unsafe int ScanArgument(ref char* psrc, ref bool inquote, char[] arg)

-

// First, parse the program name (argv[0]). Argv[0] is parsed under special rules. Anything up to 
// the first whitespace outside a quoted subtring is accepted. Backslashes are treated as normal 
// characters.

-

// Rules: 2N backslashes + " ==> N backslashes and begin/end quote
//      2N+1 backslashes + " ==> N backslashes + literal "
//         N backslashes     ==> N backslashes

This is code ported to .NET Core from .NET Framework from what I assume is either the MSVC C library or CommandLineToArgvW.

Here's my half-hearted attempt at handling some of the shenanigans with Regular Expressions, and ignoring the argument zero bit. It's a little bit wizardy.

private static readonly Regex RxWinArgs
  = new Regex("([^\\s\"]+\"|((?<=\\s|^)(?!\"\"(?!\"))\")+)(\"\"|.*?)*\"[^\\s\"]*|[^\\s]+",
    RegexOptions.Compiled
    | RegexOptions.Singleline
    | RegexOptions.ExplicitCapture
    | RegexOptions.CultureInvariant);

internal static IEnumerable<string> ParseArgumentsWindows(string args) {
  var match = RxWinArgs.Match(args);

  while (match.Success) {
    yield return match.Value;
    match = match.NextMatch();
  }
}

Tested it a fair bit on wacky generated output. It's output matches a fair percentage of what the monkeys typed up and ran through CommandLineToArgvW.

回复 原文
给妤﹃绝世温柔 2024-07-16 05:40:12

有一个 NuGet 包恰好包含您需要的功能：

Microsoft.CodeAnalysis.Common 包含类 CommandLineParser 使用方法 SplitCommandLineIntoArguments 。

你像这样使用它：

using Microsoft.CodeAnalysis;
// [...]
var cli = @"/src:""C:\tmp\Some Folder\Sub Folder"" /users:""[email protected]"" tasks:""SomeTask,Some Other Task"" -someParam foo";
var cliArgs = CommandLineParser.SplitCommandLineIntoArguments(cli, true);

Console.WriteLine(string.Join('\n', cliArgs));
// prints out:
// /src:"C:\tmp\Some Folder\Sub Folder"
// /users:"[email protected]"
// tasks:"SomeTask,Some Other Task"
// -someParam
// foo

There's a NuGet package which contains exactly the functionality you need:

Microsoft.CodeAnalysis.Common contains the class CommandLineParser with the method SplitCommandLineIntoArguments.

You use it like this:

using Microsoft.CodeAnalysis;
// [...]
var cli = @"/src:""C:\tmp\Some Folder\Sub Folder"" /users:""[email protected]"" tasks:""SomeTask,Some Other Task"" -someParam foo";
var cliArgs = CommandLineParser.SplitCommandLineIntoArguments(cli, true);

Console.WriteLine(string.Join('\n', cliArgs));
// prints out:
// /src:"C:\tmp\Some Folder\Sub Folder"
// /users:"[email protected]"
// tasks:"SomeTask,Some Other Task"
// -someParam
// foo
回复 原文
断爱 2024-07-16 05:40:12

这篇代码项目文章是我过去使用过的。 这是一段很好的代码，但它可能会起作用。

这篇 MSDN 文章是我能找到的唯一解释 C# 的内容解析命令行参数。

This The Code Project article is what I've used in the past. It's a good bit of code, but it might work.

This MSDN article is the only thing I could find that explains how C# parses command line arguments.

回复 原文
临风闻羌笛 2024-07-16 05:40:12

纯托管解决方案可能会有所帮助。 WINAPI功能有太多“问题”注释，并且在其他平台上不可用。 这是我的代码，它具有明确定义的行为（如果您愿意，可以更改）。

它应该与 .NET/Windows 在提供 string[] args 参数时执行的操作相同，并且我已将其与许多“有趣”的值进行了比较。

这是一个经典的状态机实现，它从输入字符串中获取每个字符并将其解释为当前状态，生成输出和新状态。 状态在变量 escape、inQuote、hadQuote 和 prevCh 中定义，输出收集在 <代码>currentArg和args。

我通过在真实命令提示符 (Windows 7) 上进行实验发现的一些特性： \\ 生成 \，\" 生成引用范围内的 "、"" 会生成 "。

^ 字符似乎也很神奇：它不加倍时总是消失。否则它对真正的命令行没有影响，因为我还没有找到这种行为的模式

。在这种模式中是以下命令：

cmd /c "argdump.exe "a b c""

cmd 命令似乎捕获了外部引号并逐字记录其余内容，其中一定有一些特殊的魔力，

我没有对我的方法进行基准测试。但认为它相当快。它不使用 Regex ，也不执行任何字符串连接，而是使用 StringBuilder 来收集参数的字符并将它们放入其中。一个列表。

/// <summary>
/// Reads command line arguments from a single string.
/// </summary>
/// <param name="argsString">The string that contains the entire command line.</param>
/// <returns>An array of the parsed arguments.</returns>
public string[] ReadArgs(string argsString)
{
    // Collects the split argument strings
    List<string> args = new List<string>();
    // Builds the current argument
    var currentArg = new StringBuilder();
    // Indicates whether the last character was a backslash escape character
    bool escape = false;
    // Indicates whether we're in a quoted range
    bool inQuote = false;
    // Indicates whether there were quotes in the current arguments
    bool hadQuote = false;
    // Remembers the previous character
    char prevCh = '\0';
    // Iterate all characters from the input string
    for (int i = 0; i < argsString.Length; i++)
    {
        char ch = argsString[i];
        if (ch == '\\' && !escape)
        {
            // Beginning of a backslash-escape sequence
            escape = true;
        }
        else if (ch == '\\' && escape)
        {
            // Double backslash, keep one
            currentArg.Append(ch);
            escape = false;
        }
        else if (ch == '"' && !escape)
        {
            // Toggle quoted range
            inQuote = !inQuote;
            hadQuote = true;
            if (inQuote && prevCh == '"')
            {
                // Doubled quote within a quoted range is like escaping
                currentArg.Append(ch);
            }
        }
        else if (ch == '"' && escape)
        {
            // Backslash-escaped quote, keep it
            currentArg.Append(ch);
            escape = false;
        }
        else if (char.IsWhiteSpace(ch) && !inQuote)
        {
            if (escape)
            {
                // Add pending escape char
                currentArg.Append('\\');
                escape = false;
            }
            // Accept empty arguments only if they are quoted
            if (currentArg.Length > 0 || hadQuote)
            {
                args.Add(currentArg.ToString());
            }
            // Reset for next argument
            currentArg.Clear();
            hadQuote = false;
        }
        else
        {
            if (escape)
            {
                // Add pending escape char
                currentArg.Append('\\');
                escape = false;
            }
            // Copy character from input, no special meaning
            currentArg.Append(ch);
        }
        prevCh = ch;
    }
    // Save last argument
    if (currentArg.Length > 0 || hadQuote)
    {
        args.Add(currentArg.ToString());
    }
    return args.ToArray();
}

A purely managed solution might be helpful. There are too many "problem" comments for the WINAPI function and it's not available on other platforms. Here's my code that has a well-defined behaviour (that you can change if you like).

It should do the same as what .NET/Windows do when providing that string[] args parameter, and I've compared it with a number of "interesting" values.

This is a classic state-machine implementation that takes each single character from the input string and interprets it for the current state, producing output and a new state. The state is defined in the variables escape, inQuote, hadQuote and prevCh, and the output is collected in currentArg and args.

Some of the specialties that I've discovered by experiments on a real command prompt (Windows 7): \\ produces \, \" produces ", "" within a quoted range produces ".

The ^ character seems to be magical, too: it always disappears when not doubling it. Otherwise it has no effect on a real command line. My implementation does not support this, as I haven't found a pattern in this behaviour. Maybe somebody knows more about it.

Something that doesn't fit in this pattern is the following command:

cmd /c "argdump.exe "a b c""

The cmd command seems to catch the outer quotes and take the rest verbatim. There must be some special magic sauce in this.

I've done no benchmarks on my method, but consider it reasonably fast. It doesn't use Regex and doesn't do any string concatenation but instead uses a StringBuilder to collect the characters for an argument and puts them in a list.

/// <summary>
/// Reads command line arguments from a single string.
/// </summary>
/// <param name="argsString">The string that contains the entire command line.</param>
/// <returns>An array of the parsed arguments.</returns>
public string[] ReadArgs(string argsString)
{
    // Collects the split argument strings
    List<string> args = new List<string>();
    // Builds the current argument
    var currentArg = new StringBuilder();
    // Indicates whether the last character was a backslash escape character
    bool escape = false;
    // Indicates whether we're in a quoted range
    bool inQuote = false;
    // Indicates whether there were quotes in the current arguments
    bool hadQuote = false;
    // Remembers the previous character
    char prevCh = '\0';
    // Iterate all characters from the input string
    for (int i = 0; i < argsString.Length; i++)
    {
        char ch = argsString[i];
        if (ch == '\\' && !escape)
        {
            // Beginning of a backslash-escape sequence
            escape = true;
        }
        else if (ch == '\\' && escape)
        {
            // Double backslash, keep one
            currentArg.Append(ch);
            escape = false;
        }
        else if (ch == '"' && !escape)
        {
            // Toggle quoted range
            inQuote = !inQuote;
            hadQuote = true;
            if (inQuote && prevCh == '"')
            {
                // Doubled quote within a quoted range is like escaping
                currentArg.Append(ch);
            }
        }
        else if (ch == '"' && escape)
        {
            // Backslash-escaped quote, keep it
            currentArg.Append(ch);
            escape = false;
        }
        else if (char.IsWhiteSpace(ch) && !inQuote)
        {
            if (escape)
            {
                // Add pending escape char
                currentArg.Append('\\');
                escape = false;
            }
            // Accept empty arguments only if they are quoted
            if (currentArg.Length > 0 || hadQuote)
            {
                args.Add(currentArg.ToString());
            }
            // Reset for next argument
            currentArg.Clear();
            hadQuote = false;
        }
        else
        {
            if (escape)
            {
                // Add pending escape char
                currentArg.Append('\\');
                escape = false;
            }
            // Copy character from input, no special meaning
            currentArg.Append(ch);
        }
        prevCh = ch;
    }
    // Save last argument
    if (currentArg.Length > 0 || hadQuote)
    {
        args.Add(currentArg.ToString());
    }
    return args.ToArray();
}
回复 原文
ヅ她的身影、若隐若现 2024-07-16 05:40:12

使用：

public static string[] SplitArguments(string args) {
    char[] parmChars = args.ToCharArray();
    bool inSingleQuote = false;
    bool inDoubleQuote = false;
    bool escaped = false;
    bool lastSplitted = false;
    bool justSplitted = false;
    bool lastQuoted = false;
    bool justQuoted = false;

    int i, j;

    for(i=0, j=0; i<parmChars.Length; i++, j++) {
        parmChars[j] = parmChars[i];

        if(!escaped) {
            if(parmChars[i] == '^') {
                escaped = true;
                j--;
            } else if(parmChars[i] == '"' && !inSingleQuote) {
                inDoubleQuote = !inDoubleQuote;
                parmChars[j] = '\n';
                justSplitted = true;
                justQuoted = true;
            } else if(parmChars[i] == '\'' && !inDoubleQuote) {
                inSingleQuote = !inSingleQuote;
                parmChars[j] = '\n';
                justSplitted = true;
                justQuoted = true;
            } else if(!inSingleQuote && !inDoubleQuote && parmChars[i] == ' ') {
                parmChars[j] = '\n';
                justSplitted = true;
            }

            if(justSplitted && lastSplitted && (!lastQuoted || !justQuoted))
                j--;

            lastSplitted = justSplitted;
            justSplitted = false;

            lastQuoted = justQuoted;
            justQuoted = false;
        } else {
            escaped = false;
        }
    }

    if(lastQuoted)
        j--;

    return (new string(parmChars, 0, j)).Split(new[] { '\n' });
}

根据 Vapour in the Alley 的回答，这个也支持 ^ 转义。

示例：

  • 这是一个测试
    • 这个
    • 是
    • 一个
    • 测试
  • 这是一个“是”测试
    • 这个
    • 是一个
    • 测试
  • 这是一个测试 这是一个测试
    • 这个
    • “是
    • 一个“
    • 测试
  • 这个“”“是一个 ^^ 测试”
    • 这个
    • 是一个 ^ 测试

它还支持多个空格（每个空格块仅中断参数一次）。

Use:

public static string[] SplitArguments(string args) {
    char[] parmChars = args.ToCharArray();
    bool inSingleQuote = false;
    bool inDoubleQuote = false;
    bool escaped = false;
    bool lastSplitted = false;
    bool justSplitted = false;
    bool lastQuoted = false;
    bool justQuoted = false;

    int i, j;

    for(i=0, j=0; i<parmChars.Length; i++, j++) {
        parmChars[j] = parmChars[i];

        if(!escaped) {
            if(parmChars[i] == '^') {
                escaped = true;
                j--;
            } else if(parmChars[i] == '"' && !inSingleQuote) {
                inDoubleQuote = !inDoubleQuote;
                parmChars[j] = '\n';
                justSplitted = true;
                justQuoted = true;
            } else if(parmChars[i] == '\'' && !inDoubleQuote) {
                inSingleQuote = !inSingleQuote;
                parmChars[j] = '\n';
                justSplitted = true;
                justQuoted = true;
            } else if(!inSingleQuote && !inDoubleQuote && parmChars[i] == ' ') {
                parmChars[j] = '\n';
                justSplitted = true;
            }

            if(justSplitted && lastSplitted && (!lastQuoted || !justQuoted))
                j--;

            lastSplitted = justSplitted;
            justSplitted = false;

            lastQuoted = justQuoted;
            justQuoted = false;
        } else {
            escaped = false;
        }
    }

    if(lastQuoted)
        j--;

    return (new string(parmChars, 0, j)).Split(new[] { '\n' });
}

Based on Vapour in the Alley's answer, this one also supports ^ escapes.

Examples:

  • this is a test
    • this
    • is
    • a
    • test
  • this "is a" test
    • this
    • is a
    • test
  • this ^"is a^" test
    • this
    • "is
    • a"
    • test
  • this "" "is a ^^ test"
    • this
    • ​
    • is a ^ test

It also supports multiple spaces (breaks arguments just one time per block of spaces).

回复 原文
懒猫 2024-07-16 05:40:12

目前，这是我拥有的代码：

    private String[] SplitCommandLineArgument(String argumentString)
    {
        StringBuilder translatedArguments = new StringBuilder(argumentString);
        bool escaped = false;
        for (int i = 0; i < translatedArguments.Length; i++)
        {
            if (translatedArguments[i] == '"')
            {
                escaped = !escaped;
            }
            if (translatedArguments[i] == ' ' && !escaped)
            {
                translatedArguments[i] = '\n';
            }
        }

        string[] toReturn = translatedArguments.ToString().Split(new char[] { '\n' }, StringSplitOptions.RemoveEmptyEntries);
        for(int i = 0; i < toReturn.Length; i++)
        {
            toReturn[i] = RemoveMatchingQuotes(toReturn[i]);
        }
        return toReturn;
    }

    public static string RemoveMatchingQuotes(string stringToTrim)
    {
        int firstQuoteIndex = stringToTrim.IndexOf('"');
        int lastQuoteIndex = stringToTrim.LastIndexOf('"');
        while (firstQuoteIndex != lastQuoteIndex)
        {
            stringToTrim = stringToTrim.Remove(firstQuoteIndex, 1);
            stringToTrim = stringToTrim.Remove(lastQuoteIndex - 1, 1); //-1 because we've shifted the indicies left by one
            firstQuoteIndex = stringToTrim.IndexOf('"');
            lastQuoteIndex = stringToTrim.LastIndexOf('"');
        }
        return stringToTrim;
    }

它不适用于转义引号，但适用于我迄今为止遇到的情况。

Currently, this is the code that I have:

    private String[] SplitCommandLineArgument(String argumentString)
    {
        StringBuilder translatedArguments = new StringBuilder(argumentString);
        bool escaped = false;
        for (int i = 0; i < translatedArguments.Length; i++)
        {
            if (translatedArguments[i] == '"')
            {
                escaped = !escaped;
            }
            if (translatedArguments[i] == ' ' && !escaped)
            {
                translatedArguments[i] = '\n';
            }
        }

        string[] toReturn = translatedArguments.ToString().Split(new char[] { '\n' }, StringSplitOptions.RemoveEmptyEntries);
        for(int i = 0; i < toReturn.Length; i++)
        {
            toReturn[i] = RemoveMatchingQuotes(toReturn[i]);
        }
        return toReturn;
    }

    public static string RemoveMatchingQuotes(string stringToTrim)
    {
        int firstQuoteIndex = stringToTrim.IndexOf('"');
        int lastQuoteIndex = stringToTrim.LastIndexOf('"');
        while (firstQuoteIndex != lastQuoteIndex)
        {
            stringToTrim = stringToTrim.Remove(firstQuoteIndex, 1);
            stringToTrim = stringToTrim.Remove(lastQuoteIndex - 1, 1); //-1 because we've shifted the indicies left by one
            firstQuoteIndex = stringToTrim.IndexOf('"');
            lastQuoteIndex = stringToTrim.LastIndexOf('"');
        }
        return stringToTrim;
    }

It doesn't work with escaped quotes, but it works for the cases that I've come up against so far.

回复 原文
趁年轻赶紧闹 2024-07-16 05:40:12

这是对 Anton 代码的回复，该代码不适用于转义引号。 我修改了3个地方。

  1. SplitCommandLineArguments 中 StringBuilder 的构造函数，将任何 \" 替换为 \r
  2. SplitCommandLineArguments 中的 for 循环，我现在将 \r 字符替换回 \"。
  3. 将 SplitCommandLineArgument 方法从私有更改为公共静态。
public static string[] SplitCommandLineArgument( String argumentString )
{
    StringBuilder translatedArguments = new StringBuilder( argumentString ).Replace( "\\\"", "\r" );
    bool InsideQuote = false;
    for ( int i = 0; i < translatedArguments.Length; i++ )
    {
        if ( translatedArguments[i] == '"' )
        {
            InsideQuote = !InsideQuote;
        }
        if ( translatedArguments[i] == ' ' && !InsideQuote )
        {
            translatedArguments[i] = '\n';
        }
    }

    string[] toReturn = translatedArguments.ToString().Split( new char[] { '\n' }, StringSplitOptions.RemoveEmptyEntries );
    for ( int i = 0; i < toReturn.Length; i++ )
    {
        toReturn[i] = RemoveMatchingQuotes( toReturn[i] );
        toReturn[i] = toReturn[i].Replace( "\r", "\"" );
    }
    return toReturn;
}

public static string RemoveMatchingQuotes( string stringToTrim )
{
    int firstQuoteIndex = stringToTrim.IndexOf( '"' );
    int lastQuoteIndex = stringToTrim.LastIndexOf( '"' );
    while ( firstQuoteIndex != lastQuoteIndex )
    {
        stringToTrim = stringToTrim.Remove( firstQuoteIndex, 1 );
        stringToTrim = stringToTrim.Remove( lastQuoteIndex - 1, 1 ); //-1 because we've shifted the indicies left by one
        firstQuoteIndex = stringToTrim.IndexOf( '"' );
        lastQuoteIndex = stringToTrim.LastIndexOf( '"' );
    }
    return stringToTrim;
}

This is a reply to Anton's code, which do not work with escaped quotes. I modified 3 places.

  1. The constructor for StringBuilder in SplitCommandLineArguments, replacing any \" with \r
  2. In the for-loop in SplitCommandLineArguments, I now replace the \r character back to \".
  3. Changed the SplitCommandLineArgument method from private to public static.
public static string[] SplitCommandLineArgument( String argumentString )
{
    StringBuilder translatedArguments = new StringBuilder( argumentString ).Replace( "\\\"", "\r" );
    bool InsideQuote = false;
    for ( int i = 0; i < translatedArguments.Length; i++ )
    {
        if ( translatedArguments[i] == '"' )
        {
            InsideQuote = !InsideQuote;
        }
        if ( translatedArguments[i] == ' ' && !InsideQuote )
        {
            translatedArguments[i] = '\n';
        }
    }

    string[] toReturn = translatedArguments.ToString().Split( new char[] { '\n' }, StringSplitOptions.RemoveEmptyEntries );
    for ( int i = 0; i < toReturn.Length; i++ )
    {
        toReturn[i] = RemoveMatchingQuotes( toReturn[i] );
        toReturn[i] = toReturn[i].Replace( "\r", "\"" );
    }
    return toReturn;
}

public static string RemoveMatchingQuotes( string stringToTrim )
{
    int firstQuoteIndex = stringToTrim.IndexOf( '"' );
    int lastQuoteIndex = stringToTrim.LastIndexOf( '"' );
    while ( firstQuoteIndex != lastQuoteIndex )
    {
        stringToTrim = stringToTrim.Remove( firstQuoteIndex, 1 );
        stringToTrim = stringToTrim.Remove( lastQuoteIndex - 1, 1 ); //-1 because we've shifted the indicies left by one
        firstQuoteIndex = stringToTrim.IndexOf( '"' );
        lastQuoteIndex = stringToTrim.LastIndexOf( '"' );
    }
    return stringToTrim;
}
回复 原文
静若繁花 2024-07-16 05:40:12

我认为 C# 应用程序没有单引号或 ^ 引号。
以下功能对我来说工作正常：

public static IEnumerable<String> SplitArguments(string commandLine)
{
    Char quoteChar = '"';
    Char escapeChar = '\\';
    Boolean insideQuote = false;
    Boolean insideEscape = false;

    StringBuilder currentArg = new StringBuilder();

    // needed to keep "" as argument but drop whitespaces between arguments
    Int32 currentArgCharCount = 0;                  

    for (Int32 i = 0; i < commandLine.Length; i++)
    {
        Char c = commandLine[i];
        if (c == quoteChar)
        {
            currentArgCharCount++;

            if (insideEscape)
            {
                currentArg.Append(c);       // found \" -> add " to arg
                insideEscape = false;
            }
            else if (insideQuote)
            {
                insideQuote = false;        // quote ended
            }
            else
            {
                insideQuote = true;         // quote started
            }
        }
        else if (c == escapeChar)
        {
            currentArgCharCount++;

            if (insideEscape)   // found \\ -> add \\ (only \" will be ")
                currentArg.Append(escapeChar + escapeChar);       

            insideEscape = !insideEscape;
        }
        else if (Char.IsWhiteSpace(c))
        {
            if (insideQuote)
            {
                currentArgCharCount++;
                currentArg.Append(c);       // append whitespace inside quote
            }
            else
            {
                if (currentArgCharCount > 0)
                    yield return currentArg.ToString();

                currentArgCharCount = 0;
                currentArg.Clear();
            }
        }
        else
        {
            currentArgCharCount++;
            if (insideEscape)
            {
                // found non-escaping backslash -> add \ (only \" will be ")
                currentArg.Append(escapeChar);                       
                currentArgCharCount = 0;
                insideEscape = false;
            }
            currentArg.Append(c);
        }
    }

    if (currentArgCharCount > 0)
        yield return currentArg.ToString();
}

I don't think there are single quotes or ^ quotes for C# applications.
The following function is working fine for me:

public static IEnumerable<String> SplitArguments(string commandLine)
{
    Char quoteChar = '"';
    Char escapeChar = '\\';
    Boolean insideQuote = false;
    Boolean insideEscape = false;

    StringBuilder currentArg = new StringBuilder();

    // needed to keep "" as argument but drop whitespaces between arguments
    Int32 currentArgCharCount = 0;                  

    for (Int32 i = 0; i < commandLine.Length; i++)
    {
        Char c = commandLine[i];
        if (c == quoteChar)
        {
            currentArgCharCount++;

            if (insideEscape)
            {
                currentArg.Append(c);       // found \" -> add " to arg
                insideEscape = false;
            }
            else if (insideQuote)
            {
                insideQuote = false;        // quote ended
            }
            else
            {
                insideQuote = true;         // quote started
            }
        }
        else if (c == escapeChar)
        {
            currentArgCharCount++;

            if (insideEscape)   // found \\ -> add \\ (only \" will be ")
                currentArg.Append(escapeChar + escapeChar);       

            insideEscape = !insideEscape;
        }
        else if (Char.IsWhiteSpace(c))
        {
            if (insideQuote)
            {
                currentArgCharCount++;
                currentArg.Append(c);       // append whitespace inside quote
            }
            else
            {
                if (currentArgCharCount > 0)
                    yield return currentArg.ToString();

                currentArgCharCount = 0;
                currentArg.Clear();
            }
        }
        else
        {
            currentArgCharCount++;
            if (insideEscape)
            {
                // found non-escaping backslash -> add \ (only \" will be ")
                currentArg.Append(escapeChar);                       
                currentArgCharCount = 0;
                insideEscape = false;
            }
            currentArg.Append(c);
        }
    }

    if (currentArgCharCount > 0)
        yield return currentArg.ToString();
}
回复 原文
沉睡月亮 2024-07-16 05:40:12

您可以查看我昨天发布的代码：

[C#] 路径 & 参数字符串

它将文件名+参数拆分为字符串[]。 处理短路径、环境变量和丢失的文件扩展名。

（最初是针对注册表中的 UninstallString。）

You can have a look at the code I've posted yesterday:

[C#] Path & arguments strings

It splits a filename + arguments into string[]. Short paths, environment variables, and missing file extensions are handled.

(Initially it was for UninstallString in Registry.)

回复 原文
乖乖哒 2024-07-16 05:40:12

试试这个代码：

    string[] str_para_linha_comando(string str, out int argumentos)
    {
        string[] linhaComando = new string[32];
        bool entre_aspas = false;
        int posicao_ponteiro = 0;
        int argc = 0;
        int inicio = 0;
        int fim = 0;
        string sub;

        for(int i = 0; i < str.Length;)
        {
            if (entre_aspas)
            {
                // Está entre aspas
                sub = str.Substring(inicio+1, fim - (inicio+1));
                linhaComando[argc - 1] = sub;

                posicao_ponteiro += ((fim - posicao_ponteiro)+1);
                entre_aspas = false;
                i = posicao_ponteiro;
            }
            else
            {
            tratar_aspas:
                if (str.ElementAt(i) == '\"')
                {
                    inicio = i;
                    fim = str.IndexOf('\"', inicio + 1);
                    entre_aspas = true;
                    argc++;
                }
                else
                {
                    // Se não for aspas, então ler até achar o primeiro espaço em branco
                    if (str.ElementAt(i) == ' ')
                    {
                        if (str.ElementAt(i + 1) == '\"')
                        {
                            i++;
                            goto tratar_aspas;
                        }

                        // Pular os espaços em branco adiconais
                        while(str.ElementAt(i) == ' ') i++;

                        argc++;
                        inicio = i;
                        fim = str.IndexOf(' ', inicio);
                        if (fim == -1) fim = str.Length;
                        sub = str.Substring(inicio, fim - inicio);
                        linhaComando[argc - 1] = sub;
                        posicao_ponteiro += (fim - posicao_ponteiro);

                        i = posicao_ponteiro;
                        if (posicao_ponteiro == str.Length) break;
                    }
                    else
                    {
                        argc++;
                        inicio = i;
                        fim = str.IndexOf(' ', inicio);
                        if (fim == -1) fim = str.Length;

                        sub = str.Substring(inicio, fim - inicio);
                        linhaComando[argc - 1] = sub;
                        posicao_ponteiro += fim - posicao_ponteiro;
                        i = posicao_ponteiro;
                        if (posicao_ponteiro == str.Length) break;
                    }
                }
            }
        }

        argumentos = argc;

        return linhaComando;
    }

它是用葡萄牙语编写的。

Try this code:

    string[] str_para_linha_comando(string str, out int argumentos)
    {
        string[] linhaComando = new string[32];
        bool entre_aspas = false;
        int posicao_ponteiro = 0;
        int argc = 0;
        int inicio = 0;
        int fim = 0;
        string sub;

        for(int i = 0; i < str.Length;)
        {
            if (entre_aspas)
            {
                // Está entre aspas
                sub = str.Substring(inicio+1, fim - (inicio+1));
                linhaComando[argc - 1] = sub;

                posicao_ponteiro += ((fim - posicao_ponteiro)+1);
                entre_aspas = false;
                i = posicao_ponteiro;
            }
            else
            {
            tratar_aspas:
                if (str.ElementAt(i) == '\"')
                {
                    inicio = i;
                    fim = str.IndexOf('\"', inicio + 1);
                    entre_aspas = true;
                    argc++;
                }
                else
                {
                    // Se não for aspas, então ler até achar o primeiro espaço em branco
                    if (str.ElementAt(i) == ' ')
                    {
                        if (str.ElementAt(i + 1) == '\"')
                        {
                            i++;
                            goto tratar_aspas;
                        }

                        // Pular os espaços em branco adiconais
                        while(str.ElementAt(i) == ' ') i++;

                        argc++;
                        inicio = i;
                        fim = str.IndexOf(' ', inicio);
                        if (fim == -1) fim = str.Length;
                        sub = str.Substring(inicio, fim - inicio);
                        linhaComando[argc - 1] = sub;
                        posicao_ponteiro += (fim - posicao_ponteiro);

                        i = posicao_ponteiro;
                        if (posicao_ponteiro == str.Length) break;
                    }
                    else
                    {
                        argc++;
                        inicio = i;
                        fim = str.IndexOf(' ', inicio);
                        if (fim == -1) fim = str.Length;

                        sub = str.Substring(inicio, fim - inicio);
                        linhaComando[argc - 1] = sub;
                        posicao_ponteiro += fim - posicao_ponteiro;
                        i = posicao_ponteiro;
                        if (posicao_ponteiro == str.Length) break;
                    }
                }
            }
        }

        argumentos = argc;

        return linhaComando;
    }

It's written in Portuguese.

回复 原文
你是暖光i 2024-07-16 05:40:12

这是一个完成工作的一行（请参阅在 BurstCmdLineArgs(...) 方法内完成所有工作的一行）。

这不是我所说的最具可读性的代码行，但您可以为了可读性而将其分解。 它的目的很简单，并且不适用于所有参数情况（例如其中包含分割字符串字符分隔符的文件名参数）。

该解决方案在我使用它的解决方案中运行良好。 就像我说的，它不需要大量代码来处理每个可能的参数格式 n 阶乘，就可以完成工作。

using System;
using System.Collections.Generic;
using System.Linq;

namespace CmdArgProcessor
{
    class Program
    {
        static void Main(string[] args)
        {
            // test switches and switches with values
            // -test1 1 -test2 2 -test3 -test4 -test5 5

            string dummyString = string.Empty;

            var argDict = BurstCmdLineArgs(args);

            Console.WriteLine("Value for switch = -test1: {0}", argDict["test1"]);
            Console.WriteLine("Value for switch = -test2: {0}", argDict["test2"]);
            Console.WriteLine("Switch -test3 is present? {0}", argDict.TryGetValue("test3", out dummyString));
            Console.WriteLine("Switch -test4 is present? {0}", argDict.TryGetValue("test4", out dummyString));
            Console.WriteLine("Value for switch = -test5: {0}", argDict["test5"]);

            // Console output:
            //
            // Value for switch = -test1: 1
            // Value for switch = -test2: 2
            // Switch -test3 is present? True
            // Switch -test4 is present? True
            // Value for switch = -test5: 5
        }

        public static Dictionary<string, string> BurstCmdLineArgs(string[] args)
        {
            var argDict = new Dictionary<string, string>();

            // Flatten the args in to a single string separated by a space.
            // Then split the args on the dash delimiter of a cmd line "switch".
            // E.g. -mySwitch myValue
            //  or -JustMySwitch (no value)
            //  where: all values must follow a switch.
            // Then loop through each string returned by the split operation.
            // If the string can be split again by a space character,
            // then the second string is a value to be paired with a switch,
            // otherwise, only the switch is added as a key with an empty string as the value.
            // Use dictionary indexer to retrieve values for cmd line switches.
            // Use Dictionary::ContainsKey(...) where only a switch is recorded as the key.
            string.Join(" ", args).Split('-').ToList().ForEach(s => argDict.Add(s.Split()[0], (s.Split().Count() > 1 ? s.Split()[1] : "")));

            return argDict;
        }
    }
}

Here's a one liner that gets the job done (see the one line that does all of the work inside the BurstCmdLineArgs(...) method).

Not what I'd call the most readable line of code, but you can break it out for readability's sake. It's simple on purpose and does not work well for all argument cases (like file name arguments that contain the split string character delimiter in them).

This solution has worked well in my solutions that use it. Like I said, it gets the job done without a rat's nest of code to handle every possible argument format n-factorial.

using System;
using System.Collections.Generic;
using System.Linq;

namespace CmdArgProcessor
{
    class Program
    {
        static void Main(string[] args)
        {
            // test switches and switches with values
            // -test1 1 -test2 2 -test3 -test4 -test5 5

            string dummyString = string.Empty;

            var argDict = BurstCmdLineArgs(args);

            Console.WriteLine("Value for switch = -test1: {0}", argDict["test1"]);
            Console.WriteLine("Value for switch = -test2: {0}", argDict["test2"]);
            Console.WriteLine("Switch -test3 is present? {0}", argDict.TryGetValue("test3", out dummyString));
            Console.WriteLine("Switch -test4 is present? {0}", argDict.TryGetValue("test4", out dummyString));
            Console.WriteLine("Value for switch = -test5: {0}", argDict["test5"]);

            // Console output:
            //
            // Value for switch = -test1: 1
            // Value for switch = -test2: 2
            // Switch -test3 is present? True
            // Switch -test4 is present? True
            // Value for switch = -test5: 5
        }

        public static Dictionary<string, string> BurstCmdLineArgs(string[] args)
        {
            var argDict = new Dictionary<string, string>();

            // Flatten the args in to a single string separated by a space.
            // Then split the args on the dash delimiter of a cmd line "switch".
            // E.g. -mySwitch myValue
            //  or -JustMySwitch (no value)
            //  where: all values must follow a switch.
            // Then loop through each string returned by the split operation.
            // If the string can be split again by a space character,
            // then the second string is a value to be paired with a switch,
            // otherwise, only the switch is added as a key with an empty string as the value.
            // Use dictionary indexer to retrieve values for cmd line switches.
            // Use Dictionary::ContainsKey(...) where only a switch is recorded as the key.
            string.Join(" ", args).Split('-').ToList().ForEach(s => argDict.Add(s.Split()[0], (s.Split().Count() > 1 ? s.Split()[1] : "")));

            return argDict;
        }
    }
}
回复 原文
乖乖公主 2024-07-16 05:40:12

在这里找不到我喜欢的东西。 我讨厌用yield magic 来搞乱一个小命令行的堆栈（如果它是一个 TB 的流，那就是另一个故事了）。

这是我的看法，它支持使用双引号进行引号转义，如下所示：

param="15"" 屏幕还不错" param2='15" 屏幕还不错' param3="" param4= /param5

结果：

param="15" 屏幕还不错"

param2='15 英寸屏幕也不错'

参数3=""

参数4=

/参数5

public static string[] SplitArguments(string commandLine)
{
    List<string> args         = new List<string>();
    List<char>   currentArg   = new List<char>();
    char?        quoteSection = null; // Keeps track of a quoted section (and the type of quote that was used to open it)
    char[]       quoteChars   = new[] {'\'', '\"'};
    char         previous     = ' '; // Used for escaping double quotes

    for (var index = 0; index < commandLine.Length; index++)
    {
        char c = commandLine[index];
        if (quoteChars.Contains(c))
        {
            if (previous == c) // Escape sequence detected
            {
                previous = ' '; // Prevent re-escaping
                if (!quoteSection.HasValue)
                {
                    quoteSection = c; // oops, we ended the quoted section prematurely
                    continue;         // don't add the 2nd quote (un-escape)
                }

                if (quoteSection.Value == c)
                    quoteSection = null; // appears to be an empty string (not an escape sequence)
            }
            else if (quoteSection.HasValue)
            {
                if (quoteSection == c)
                    quoteSection = null; // End quoted section
            }
            else
                quoteSection = c; // Start quoted section
        }
        else if (char.IsWhiteSpace(c))
        {
            if (!quoteSection.HasValue)
            {
                args.Add(new string(currentArg.ToArray()));
                currentArg.Clear();
                previous = c;
                continue;
            }
        }

        currentArg.Add(c);
        previous = c;
    }

    if (currentArg.Count > 0)
        args.Add(new string(currentArg.ToArray()));

    return args.ToArray();
}

Couldn't find anything I liked here. I hate to mess up the stack with yield magic for a small command-line (if it were a stream of a terabyte, it would be another story).

Here's my take, it supports quote escapes with double quotes like these:

param="a 15"" screen isn't bad" param2='a 15" screen isn''t bad' param3="" param4= /param5

result:

param="a 15" screen isn't bad"

param2='a 15" screen isn't bad'

param3=""

param4=

/param5

public static string[] SplitArguments(string commandLine)
{
    List<string> args         = new List<string>();
    List<char>   currentArg   = new List<char>();
    char?        quoteSection = null; // Keeps track of a quoted section (and the type of quote that was used to open it)
    char[]       quoteChars   = new[] {'\'', '\"'};
    char         previous     = ' '; // Used for escaping double quotes

    for (var index = 0; index < commandLine.Length; index++)
    {
        char c = commandLine[index];
        if (quoteChars.Contains(c))
        {
            if (previous == c) // Escape sequence detected
            {
                previous = ' '; // Prevent re-escaping
                if (!quoteSection.HasValue)
                {
                    quoteSection = c; // oops, we ended the quoted section prematurely
                    continue;         // don't add the 2nd quote (un-escape)
                }

                if (quoteSection.Value == c)
                    quoteSection = null; // appears to be an empty string (not an escape sequence)
            }
            else if (quoteSection.HasValue)
            {
                if (quoteSection == c)
                    quoteSection = null; // End quoted section
            }
            else
                quoteSection = c; // Start quoted section
        }
        else if (char.IsWhiteSpace(c))
        {
            if (!quoteSection.HasValue)
            {
                args.Add(new string(currentArg.ToArray()));
                currentArg.Clear();
                previous = c;
                continue;
            }
        }

        currentArg.Add(c);
        previous = c;
    }

    if (currentArg.Count > 0)
        args.Add(new string(currentArg.ToArray()));

    return args.ToArray();
}
回复 原文
凉城 2024-07-16 05:40:12

我已经实现了状态机，使其具有相同的解析器结果，就像将 args 传递到 .NET 应用程序并在 static void Main(string[] args) 方法中进行处理一样。

    public static IList<string> ParseCommandLineArgsString(string commandLineArgsString)
    {
        List<string> args = new List<string>();

        commandLineArgsString = commandLineArgsString.Trim();
        if (commandLineArgsString.Length == 0)
            return args;

        int index = 0;
        while (index != commandLineArgsString.Length)
        {
            args.Add(ReadOneArgFromCommandLineArgsString(commandLineArgsString, ref index));
        }

        return args;
    }

    private static string ReadOneArgFromCommandLineArgsString(string line, ref int index)
    {
        if (index >= line.Length)
            return string.Empty;

        var sb = new StringBuilder(512);
        int state = 0;
        while (true)
        {
            char c = line[index];
            index++;
            switch (state)
            {
                case 0: //string outside quotation marks
                    if (c == '\\') //possible escaping character for quotation mark otherwise normal character
                    {
                        state = 1;
                    }
                    else if (c == '"') //opening quotation mark for string between quotation marks
                    {
                        state = 2;
                    }
                    else if (c == ' ') //closing arg
                    {
                        return sb.ToString();
                    }
                    else
                    {
                        sb.Append(c);
                    }

                    break;
                case 1: //possible escaping \ for quotation mark or normal character
                    if (c == '"') //If escaping quotation mark only quotation mark is added into result
                    {
                        state = 0;
                        sb.Append(c);
                    }
                    else // \ works as not-special character
                    {
                        state = 0;
                        sb.Append('\\');
                        index--;
                    }

                    break;
                case 2: //string between quotation marks
                    if (c == '"') //quotation mark in string between quotation marks can be escape mark for following quotation mark or can be ending quotation mark for string between quotation marks
                    {
                        state = 3;
                    }
                    else if (c == '\\') //escaping \ for possible following quotation mark otherwise normal character
                    {
                        state = 4;
                    }
                    else //text in quotation marks
                    {
                        sb.Append(c);
                    }

                    break;
                case 3: //quotation mark in string between quotation marks
                    if (c == '"') //Quotation mark after quotation mark - that means that this one is escaped and can added into result and we will stay in string between quotation marks state
                    {
                        state = 2;
                        sb.Append(c);
                    }
                    else //we had two consecutive quotation marks - this means empty string but the following chars (until space) will be part of same arg result as well
                    {
                        state = 0;
                        index--;
                    }

                    break;
                case 4: //possible escaping \ for quotation mark or normal character in string between quotation marks
                    if (c == '"') //If escaping quotation mark only quotation mark added into result
                    {
                        state = 2;
                        sb.Append(c);
                    }
                    else
                    {
                        state = 2;
                        sb.Append('\\');
                        index--;
                    }

                    break;
            }

            if (index == line.Length)
                return sb.ToString();
        }
    }

I have implemented state machine to have same parser results as if args would be passed into .NET application and processed in static void Main(string[] args) method.

    public static IList<string> ParseCommandLineArgsString(string commandLineArgsString)
    {
        List<string> args = new List<string>();

        commandLineArgsString = commandLineArgsString.Trim();
        if (commandLineArgsString.Length == 0)
            return args;

        int index = 0;
        while (index != commandLineArgsString.Length)
        {
            args.Add(ReadOneArgFromCommandLineArgsString(commandLineArgsString, ref index));
        }

        return args;
    }

    private static string ReadOneArgFromCommandLineArgsString(string line, ref int index)
    {
        if (index >= line.Length)
            return string.Empty;

        var sb = new StringBuilder(512);
        int state = 0;
        while (true)
        {
            char c = line[index];
            index++;
            switch (state)
            {
                case 0: //string outside quotation marks
                    if (c == '\\') //possible escaping character for quotation mark otherwise normal character
                    {
                        state = 1;
                    }
                    else if (c == '"') //opening quotation mark for string between quotation marks
                    {
                        state = 2;
                    }
                    else if (c == ' ') //closing arg
                    {
                        return sb.ToString();
                    }
                    else
                    {
                        sb.Append(c);
                    }

                    break;
                case 1: //possible escaping \ for quotation mark or normal character
                    if (c == '"') //If escaping quotation mark only quotation mark is added into result
                    {
                        state = 0;
                        sb.Append(c);
                    }
                    else // \ works as not-special character
                    {
                        state = 0;
                        sb.Append('\\');
                        index--;
                    }

                    break;
                case 2: //string between quotation marks
                    if (c == '"') //quotation mark in string between quotation marks can be escape mark for following quotation mark or can be ending quotation mark for string between quotation marks
                    {
                        state = 3;
                    }
                    else if (c == '\\') //escaping \ for possible following quotation mark otherwise normal character
                    {
                        state = 4;
                    }
                    else //text in quotation marks
                    {
                        sb.Append(c);
                    }

                    break;
                case 3: //quotation mark in string between quotation marks
                    if (c == '"') //Quotation mark after quotation mark - that means that this one is escaped and can added into result and we will stay in string between quotation marks state
                    {
                        state = 2;
                        sb.Append(c);
                    }
                    else //we had two consecutive quotation marks - this means empty string but the following chars (until space) will be part of same arg result as well
                    {
                        state = 0;
                        index--;
                    }

                    break;
                case 4: //possible escaping \ for quotation mark or normal character in string between quotation marks
                    if (c == '"') //If escaping quotation mark only quotation mark added into result
                    {
                        state = 2;
                        sb.Append(c);
                    }
                    else
                    {
                        state = 2;
                        sb.Append('\\');
                        index--;
                    }

                    break;
            }

            if (index == line.Length)
                return sb.ToString();
        }
    }
回复 原文
节枝 2024-07-16 05:40:12

以下是将空格（单个或多个空格）视为命令行参数分隔符并返回真正的命令行参数的解决方案：

static string[] ParseMultiSpacedArguments(string commandLine)
{
    var isLastCharSpace = false;
    char[] parmChars = commandLine.ToCharArray();
    bool inQuote = false;
    for (int index = 0; index < parmChars.Length; index++)
    {
        if (parmChars[index] == '"')
            inQuote = !inQuote;
        if (!inQuote && parmChars[index] == ' ' && !isLastCharSpace)
            parmChars[index] = '\n';

        isLastCharSpace = parmChars[index] == '\n' || parmChars[index] == ' ';
    }

    return (new string(parmChars)).Split('\n');
}

Here is the solution which treats space(s) (single or multiple spaces) as command line parameter separator and returns the real command line arguments:

static string[] ParseMultiSpacedArguments(string commandLine)
{
    var isLastCharSpace = false;
    char[] parmChars = commandLine.ToCharArray();
    bool inQuote = false;
    for (int index = 0; index < parmChars.Length; index++)
    {
        if (parmChars[index] == '"')
            inQuote = !inQuote;
        if (!inQuote && parmChars[index] == ' ' && !isLastCharSpace)
            parmChars[index] = '\n';

        isLastCharSpace = parmChars[index] == '\n' || parmChars[index] == ' ';
    }

    return (new string(parmChars)).Split('\n');
}
回复 原文
_失温 2024-07-16 05:40:12

我编写了一个方法来将文件名与其参数分开，以便与需要分隔文件名和参数字符串的 ProcessStartInfo 一起使用。

例如 "C:\Users\Me\Something.exe" -a
将给出 { "C:\Users\Me\Something.exe", "-a" } 作为结果

代码如下：

    public static string[] SplitCommandFromArgs(string commandLine)
    {
        commandLine = commandLine.Trim();
        if (commandLine[0] == '"')
        {
            bool isEscaped = false;
            for (int c = 1; c < commandLine.Length; c++)
            {
                if (commandLine[c] == '"' && !isEscaped)
                {
                    return new string[] { commandLine.Substring(1, c - 1), commandLine.Substring(c + 1).Trim() };
                }
                isEscaped = commandLine[c] == '\\';
            }
        }
        else
        {
            for (int c = 1; c < commandLine.Length; c++) {
                if (commandLine[c] == ' ')
                {
                    return new string[] { commandLine.Substring(0, c), commandLine.Substring(c).Trim() };
                }
            }
        }
        return new string[] { commandLine, "" };
    }

I wrote a method to separate a file name from its arguments, for use with ProcessStartInfo which requires separating the file name and argument string.

For instance "C:\Users\Me\Something.exe" -a
would give { "C:\Users\Me\Something.exe", "-a" } as a result

Code below:

    public static string[] SplitCommandFromArgs(string commandLine)
    {
        commandLine = commandLine.Trim();
        if (commandLine[0] == '"')
        {
            bool isEscaped = false;
            for (int c = 1; c < commandLine.Length; c++)
            {
                if (commandLine[c] == '"' && !isEscaped)
                {
                    return new string[] { commandLine.Substring(1, c - 1), commandLine.Substring(c + 1).Trim() };
                }
                isEscaped = commandLine[c] == '\\';
            }
        }
        else
        {
            for (int c = 1; c < commandLine.Length; c++) {
                if (commandLine[c] == ' ')
                {
                    return new string[] { commandLine.Substring(0, c), commandLine.Substring(c).Trim() };
                }
            }
        }
        return new string[] { commandLine, "" };
    }
回复 原文
醉生梦死 2024-07-16 05:40:12

这包含在 System.CommandLine.Parsing 包中

无需导入大型代码分析包。

using System.CommandLine.Parsing;

var cliArgs = CommandLineStringSplitter.Instance.Split(s).ToArray();

https:// learn.microsoft.com/en-us/dotnet/api/system.commandline.parsing.commandlinestringsplitter.split?view=system-commandline

This is included in the System.CommandLine.Parsing package

No need to import the large code analysis package.

using System.CommandLine.Parsing;

var cliArgs = CommandLineStringSplitter.Instance.Split(s).ToArray();

https://learn.microsoft.com/en-us/dotnet/api/system.commandline.parsing.commandlinestringsplitter.split?view=system-commandline

回复 原文
迷鸟归林 2024-07-16 05:40:12

我不确定我是否理解你的意思，但是用作分隔符的字符的问题是否也可以在文本中找到？ （除了它是用双“？”转义的）

如果是这样，我将创建一个 for 循环，并替换 <"> 的所有实例。 与 <|> 一起出现 （或另一个“安全”字符，但请确保它仅替换 <">，而不是 <"">

迭代字符串后，我会像之前发布的那样，分割字符串，但现在在字符上<|>。

I am not sure if I understood you, but is the problem that the character used as splitter, is also to be found inside the text? (Except for that it is escaped with double "?)

If so, I would create a for loop, and replace all instances where <"> is present with <|> (or another "safe" character, but make sure that it only replaces <">, and not <"">

After iterating the string, I would do as previously posted, split the string, but now on the character <|>.

回复 原文
秋叶绚丽 2024-07-16 05:40:12

是的，字符串对象有一个名为 Split() 的内置函数，该函数采用单个参数来指定要查找的字符作为分隔符，并返回一个字符串数组 (string[])，其中包含各个字符其中的价值观。

Yes, the string object has a built in function called Split() that takes a single parameter specifying the character to look for as a delimiter, and returns an array of strings (string[]) with the individual values in it.

回复 原文
笔落惊风雨 2024-07-16 05:40:11

令我烦恼的是，没有函数可以根据检查每个字符的函数来分割字符串。 如果有的话可以这样写：

    public static IEnumerable<string> SplitCommandLine(string commandLine)
    {
        bool inQuotes = false;

        return commandLine.Split(c =>
                                 {
                                     if (c == '\"')
                                         inQuotes = !inQuotes;

                                     return !inQuotes && c == ' ';
                                 })
                          .Select(arg => arg.Trim().TrimMatchingQuotes('\"'))
                          .Where(arg => !string.IsNullOrEmpty(arg));
    }

既然这样写了，为什么不写必要的扩展方法呢。 好吧，你说服了我......

首先，我自己的 Split 版本需要一个函数来决定指定的字符是否应该分割字符串：

    public static IEnumerable<string> Split(this string str, 
                                            Func<char, bool> controller)
    {
        int nextPiece = 0;

        for (int c = 0; c < str.Length; c++)
        {
            if (controller(str[c]))
            {
                yield return str.Substring(nextPiece, c - nextPiece);
                nextPiece = c + 1;
            }
        }

        yield return str.Substring(nextPiece);
    }

根据情况，它可能会产生一些空字符串，但也许该信息会在其他情况下很有用，因此我不会删除此函数中的空条目。

其次（更平常的是）一个小助手，它将从字符串的开头和结尾修剪一对匹配的引号。 它比标准的 Trim 方法更加挑剔 - 它只会从每一端修剪一个字符，并且不会仅从一端修剪：

    public static string TrimMatchingQuotes(this string input, char quote)
    {
        if ((input.Length >= 2) && 
            (input[0] == quote) && (input[input.Length - 1] == quote))
            return input.Substring(1, input.Length - 2);

        return input;
    }

我想您也需要一些测试。 好吧，那么好吧。 但这绝对是最后一件事！ 首先是一个辅助函数，它将分割结果与预期的数组内容进行比较：

    public static void Test(string cmdLine, params string[] args)
    {
        string[] split = SplitCommandLine(cmdLine).ToArray();

        Debug.Assert(split.Length == args.Length);

        for (int n = 0; n < split.Length; n++)
            Debug.Assert(split[n] == args[n]);
    }

然后我可以编写这样的测试：这

        Test("");
        Test("a", "a");
        Test(" abc ", "abc");
        Test("a b ", "a", "b");
        Test("a b \"c d\"", "a", "b", "c d");

是满足您要求的测试：

        Test(@"/src:""C:\tmp\Some Folder\Sub Folder"" /users:""[email protected]"" tasks:""SomeTask,Some Other Task"" -someParam",
             @"/src:""C:\tmp\Some Folder\Sub Folder""", @"/users:""[email protected]""", @"tasks:""SomeTask,Some Other Task""", @"-someParam");

请注意，该实现具有额外的功能，它将删除参数周围的引号（如果出现这种情况）有道理（感谢 TrimMatchingQuotes 函数）。 我相信这是正常命令行解释的一部分。

It annoys me that there's no function to split a string based on a function that examines each character. If there was, you could write it like this:

    public static IEnumerable<string> SplitCommandLine(string commandLine)
    {
        bool inQuotes = false;

        return commandLine.Split(c =>
                                 {
                                     if (c == '\"')
                                         inQuotes = !inQuotes;

                                     return !inQuotes && c == ' ';
                                 })
                          .Select(arg => arg.Trim().TrimMatchingQuotes('\"'))
                          .Where(arg => !string.IsNullOrEmpty(arg));
    }

Although having written that, why not write the necessary extension methods. Okay, you talked me into it...

Firstly, my own version of Split that takes a function that has to decide whether the specified character should split the string:

    public static IEnumerable<string> Split(this string str, 
                                            Func<char, bool> controller)
    {
        int nextPiece = 0;

        for (int c = 0; c < str.Length; c++)
        {
            if (controller(str[c]))
            {
                yield return str.Substring(nextPiece, c - nextPiece);
                nextPiece = c + 1;
            }
        }

        yield return str.Substring(nextPiece);
    }

It may yield some empty strings depending on the situation, but maybe that information will be useful in other cases, so I don't remove the empty entries in this function.

Secondly (and more mundanely) a little helper that will trim a matching pair of quotes from the start and end of a string. It's more fussy than the standard Trim method - it will only trim one character from each end, and it will not trim from just one end:

    public static string TrimMatchingQuotes(this string input, char quote)
    {
        if ((input.Length >= 2) && 
            (input[0] == quote) && (input[input.Length - 1] == quote))
            return input.Substring(1, input.Length - 2);

        return input;
    }

And I suppose you'll want some tests as well. Well, alright then. But this must be absolutely the last thing! First a helper function that compares the result of the split with the expected array contents:

    public static void Test(string cmdLine, params string[] args)
    {
        string[] split = SplitCommandLine(cmdLine).ToArray();

        Debug.Assert(split.Length == args.Length);

        for (int n = 0; n < split.Length; n++)
            Debug.Assert(split[n] == args[n]);
    }

Then I can write tests like this:

        Test("");
        Test("a", "a");
        Test(" abc ", "abc");
        Test("a b ", "a", "b");
        Test("a b \"c d\"", "a", "b", "c d");

Here's the test for your requirements:

        Test(@"/src:""C:\tmp\Some Folder\Sub Folder"" /users:""[email protected]"" tasks:""SomeTask,Some Other Task"" -someParam",
             @"/src:""C:\tmp\Some Folder\Sub Folder""", @"/users:""[email protected]""", @"tasks:""SomeTask,Some Other Task""", @"-someParam");

Note that the implementation has the extra feature that it will remove quotes around an argument if that makes sense (thanks to the TrimMatchingQuotes function). I believe that's part of the normal command-line interpretation.

回复 原文
月朦胧 2024-07-16 05:40:11

除了良好且纯粹的托管Earwicker 的解决方案，为了完整起见，可能值得一提的是，Windows 还提供了CommandLineToArgvW 函数，用于将字符串分解为数组字符串数：

<前><代码>LPWSTR *CommandLineToArgvW(
LPCWSTR lpCmdLine, int *pNumArgs);

解析 Unicode 命令行字符串
并返回一个指针数组
命令行参数，以及
在某种程度上，对此类论点的计数
类似于标准C
运行时 argv 和 argc 值。

从 C# 调用此 API 并在托管代码中解压生成的字符串数组的示例可以在“使用 CommandLineToArgvW() API 将命令行字符串转换为 Args[]。” 下面是相同代码的稍微简单的版本：

[DllImport("shell32.dll", SetLastError = true)]
static extern IntPtr CommandLineToArgvW(
    [MarshalAs(UnmanagedType.LPWStr)] string lpCmdLine, out int pNumArgs);

public static string[] CommandLineToArgs(string commandLine)
{
    int argc;
    var argv = CommandLineToArgvW(commandLine, out argc);        
    if (argv == IntPtr.Zero)
        throw new System.ComponentModel.Win32Exception();
    try
    {
        var args = new string[argc];
        for (var i = 0; i < args.Length; i++)
        {
            var p = Marshal.ReadIntPtr(argv, i * IntPtr.Size);
            args[i] = Marshal.PtrToStringUni(p);
        }

        return args;
    }
    finally
    {
        Marshal.FreeHGlobal(argv);
    }
}

In addition to the good and pure managed solution by Earwicker, it may be worth mentioning, for sake of completeness, that Windows also provides the CommandLineToArgvW function for breaking up a string into an array of strings:

LPWSTR *CommandLineToArgvW(
    LPCWSTR lpCmdLine, int *pNumArgs);

Parses a Unicode command line string
and returns an array of pointers to
the command line arguments, along with
a count of such arguments, in a way
that is similar to the standard C
run-time argv and argc values.

An example of calling this API from C# and unpacking the resulting string array in managed code can be found at, “Converting Command Line String to Args[] using CommandLineToArgvW() API.” Below is a slightly simpler version of the same code:

[DllImport("shell32.dll", SetLastError = true)]
static extern IntPtr CommandLineToArgvW(
    [MarshalAs(UnmanagedType.LPWStr)] string lpCmdLine, out int pNumArgs);

public static string[] CommandLineToArgs(string commandLine)
{
    int argc;
    var argv = CommandLineToArgvW(commandLine, out argc);        
    if (argv == IntPtr.Zero)
        throw new System.ComponentModel.Win32Exception();
    try
    {
        var args = new string[argc];
        for (var i = 0; i < args.Length; i++)
        {
            var p = Marshal.ReadIntPtr(argv, i * IntPtr.Size);
            args[i] = Marshal.PtrToStringUni(p);
        }

        return args;
    }
    finally
    {
        Marshal.FreeHGlobal(argv);
    }
}
回复 原文
抱猫软卧 2024-07-16 05:40:11

因为我想要与 OP 相同的行为（与 Windows cmd 完全相同的方式分割字符串），所以我编写了一堆测试用例并测试了此处发布的答案：

    Test( 0, m, "One",                    new[] { "One" });
    Test( 1, m, "One ",                   new[] { "One" });
    Test( 2, m, " One",                   new[] { "One" });
    Test( 3, m, " One ",                  new[] { "One" });
    Test( 4, m, "One Two",                new[] { "One", "Two" });
    Test( 5, m, "One  Two",               new[] { "One", "Two" });
    Test( 6, m, "One   Two",              new[] { "One", "Two" });
    Test( 7, m, "\"One Two\"",            new[] { "One Two" });
    Test( 8, m, "One \"Two Three\"",      new[] { "One", "Two Three" });
    Test( 9, m, "One \"Two Three\" Four", new[] { "One", "Two Three", "Four" });
    Test(10, m, "One=\"Two Three\" Four", new[] { "One=Two Three", "Four" });
    Test(11, m, "One\"Two Three\" Four",  new[] { "OneTwo Three", "Four" });
    Test(12, m, "One\"Two Three   Four",  new[] { "OneTwo Three   Four" });
    Test(13, m, "\"One Two\"",            new[] { "One Two" });
    Test(14, m, "One\" \"Two",            new[] { "One Two" });
    Test(15, m, "\"One\"  \"Two\"",       new[] { "One", "Two" });
    Test(16, m, "One\\\"  Two",           new[] { "One\"", "Two" });
    Test(17, m, "\\\"One\\\"  Two",       new[] { "\"One\"", "Two" });
    Test(18, m, "One\"",                  new[] { "One" });
    Test(19, m, "\"One",                  new[] { "One" });
    Test(20, m, "One \"\"",               new[] { "One", "" });
    Test(21, m, "One \"",                 new[] { "One", "" });
    Test(22, m, "1 A=\"B C\"=D 2",        new[] { "1", "A=B C=D", "2" });
    Test(23, m, "1 A=\"B \\\" C\"=D 2",   new[] { "1", "A=B \" C=D", "2" });
    Test(24, m, "1 \\A 2",                new[] { "1", "\\A", "2" });
    Test(25, m, "1 \\\" 2",               new[] { "1", "\"", "2" });
    Test(26, m, "1 \\\\\" 2",             new[] { "1", "\\\"", "2" });
    Test(27, m, "\"",                     new[] { "" });
    Test(28, m, "\\\"",                   new[] { "\"" });
    Test(29, m, "'A B'",                  new[] { "'A", "B'" });
    Test(30, m, "^",                      new[] { "^" });
    Test(31, m, "^A",                     new[] { "A" });
    Test(32, m, "^^",                     new[] { "^" });
    Test(33, m, "\\^^",                   new[] { "\\^" });
    Test(34, m, "^\\\\",                  new[] { "\\\\" });
    Test(35, m, "^\"A B\"",               new[] { "A B" });

    // Test cases Anton

    Test(36, m, @"/src:""C:\tmp\Some Folder\Sub Folder"" /users:""[email protected]"" tasks:""SomeTask,Some Other Task"" -someParam foo", new[] { @"/src:C:\tmp\Some Folder\Sub Folder", @"/users:[email protected]", @"tasks:SomeTask,Some Other Task", @"-someParam", @"foo" });

    // Test cases Daniel Earwicker 

    Test(37, m, "",            new string[] { });
    Test(38, m, "a",           new[] { "a" });
    Test(39, m, " abc ",       new[] { "abc" });
    Test(40, m, "a b ",        new[] { "a", "b" });
    Test(41, m, "a b \"c d\"", new[] { "a", "b", "c d" });

    // Test cases Fabio Iotti 

    Test(42, m, "this is a test ",    new[] { "this", "is", "a", "test" });
    Test(43, m, "this \"is a\" test", new[] { "this", "is a", "test" });

    // Test cases Kevin Thach

    Test(44, m, "\"C:\\Program Files\"",                       new[] { "C:\\Program Files" });
    Test(45, m, "\"He whispered to her \\\"I love you\\\".\"", new[] { "He whispered to her \"I love you\"." });

“预期”值来自直接使用 cmd.exe 进行测试在我的机器（Win10 x64）和一个简单的打印程序上：

static void Main(string[] args) => Console.Out.WriteLine($"Count := {args.Length}\n{string.Join("\n", args.Select((v,i) => $"[{i}] => '{v}'"))}");

这些是结果：


Solution                      | Failed Tests
------------------------------|------------------------------------- 
Atif Aziz (749653)            | 2, 3, 10, 11, 12, 14, 16, 17, 18, 26, 28, 31, 32, 33, 34, 35, 36, 37, 39, 45
Jeffrey L Whitledge (298968)  | 1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 25, 26, 27, 28, 31, 32, 33, 34, 35, 36, 37, 39, 40, 41, 42, 43, 44, 45
Daniel Earwicker (298990)     | 10, 11, 12, 14, 16, 17, 18, 19, 20, 21, 22, 23, 25, 26, 27, 28, 31, 32, 33, 34, 35, 36, 45
Anton (299795)                | 12, 16, 17, 18, 19, 21, 23, 25, 26, 27, 28, 31, 32, 33, 34, 35, 45
CS. (467313)                  | 12, 18, 19, 21, 27, 31, 32, 33, 34, 35
Vapour in the Alley (2132004) | 10, 11, 12, 14, 16, 17, 20, 21, 22, 23, 25, 26, 27, 28, 29, 31, 32, 33, 34, 35, 36, 45
Monoman (7774211)             | 14, 16, 17, 20, 21, 22, 23, 25, 26, 27, 28, 31, 32, 33, 34, 35, 45
Thomas Petersson (19091999)   | 2, 3, 10, 11, 12, 14, 16, 17, 18, 19, 20, 21, 22, 23, 25, 26, 27, 28, 31, 32, 33, 34, 35, 36, 39, 45
Fabio Iotti (19725880)        | 1, 2, 3, 7, 10, 11, 12, 13, 14, 15, 16, 17, 19, 21, 22, 23, 25, 26, 28, 29, 30, 35, 36, 37, 39, 40, 42, 44, 45
ygoe (23961658)               | 26, 31, 32, 33, 34, 35
Kevin Thach (24829691)        | 10, 11, 12, 14, 18, 19, 20, 21, 22, 23, 26, 27, 31, 32, 33, 34, 35, 36
Lucas De Jesus (31621370)     | 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45
HarryP (48008872)             | 24, 26, 31, 32, 33, 34, 35
TylerY86 (53290784)           | 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 25, 26, 27, 28, 31, 32, 33, 34, 35, 36, 41, 43, 44, 45
Louis Somers (55903304)       | 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 25, 26, 27, 28, 29, 31, 32, 33, 34, 35, 36, 39, 41, 43, 44, 45
user2126375 (58233585)        | 5, 6, 15, 16, 17, 31, 32, 33, 34, 35
DilipNannaware (59131568)     | 1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 25, 26, 27, 28, 31, 32, 33, 34, 35, 36, 37, 39, 40, 41, 42, 43, 44, 45
Mikescher (this)              | -

因为没有答案似乎是正确的（至少基于我的用例）这是我的解决方案，它目前通过了所有测试用例（但如果有人有额外的（失败）极端情况请评论）：

public static IEnumerable<string> SplitArgs(string commandLine)
{
    var result = new StringBuilder();

    var quoted = false;
    var escaped = false;
    var started = false;
    var allowcaret = false;
    for (int i = 0; i < commandLine.Length; i++)
    {
        var chr = commandLine[i];

        if (chr == '^' && !quoted)
        {
            if (allowcaret)
            {
                result.Append(chr);
                started = true;
                escaped = false;
                allowcaret = false;
            }
            else if (i + 1 < commandLine.Length && commandLine[i + 1] == '^')
            {
                allowcaret = true;
            }
            else if (i + 1 == commandLine.Length)
            {
                result.Append(chr);
                started = true;
                escaped = false;
            }
        }
        else if (escaped)
        {
            result.Append(chr);
            started = true;
            escaped = false;
        }
        else if (chr == '"')
        {
            quoted = !quoted;
            started = true;
        }
        else if (chr == '\\' && i + 1 < commandLine.Length && commandLine[i + 1] == '"')
        {
            escaped = true;
        }
        else if (chr == ' ' && !quoted)
        {
            if (started) yield return result.ToString();
            result.Clear();
            started = false;
        }
        else
        {
            result.Append(chr);
            started = true;
        }
    }

    if (started) yield return result.ToString();
}

我用来生成测试结果的代码可以在此处找到

Because I wanted the same behavior as OP (split a string exactly the same as windows cmd would do it) I wrote a bunch of test cases and tested the here posted answers:

    Test( 0, m, "One",                    new[] { "One" });
    Test( 1, m, "One ",                   new[] { "One" });
    Test( 2, m, " One",                   new[] { "One" });
    Test( 3, m, " One ",                  new[] { "One" });
    Test( 4, m, "One Two",                new[] { "One", "Two" });
    Test( 5, m, "One  Two",               new[] { "One", "Two" });
    Test( 6, m, "One   Two",              new[] { "One", "Two" });
    Test( 7, m, "\"One Two\"",            new[] { "One Two" });
    Test( 8, m, "One \"Two Three\"",      new[] { "One", "Two Three" });
    Test( 9, m, "One \"Two Three\" Four", new[] { "One", "Two Three", "Four" });
    Test(10, m, "One=\"Two Three\" Four", new[] { "One=Two Three", "Four" });
    Test(11, m, "One\"Two Three\" Four",  new[] { "OneTwo Three", "Four" });
    Test(12, m, "One\"Two Three   Four",  new[] { "OneTwo Three   Four" });
    Test(13, m, "\"One Two\"",            new[] { "One Two" });
    Test(14, m, "One\" \"Two",            new[] { "One Two" });
    Test(15, m, "\"One\"  \"Two\"",       new[] { "One", "Two" });
    Test(16, m, "One\\\"  Two",           new[] { "One\"", "Two" });
    Test(17, m, "\\\"One\\\"  Two",       new[] { "\"One\"", "Two" });
    Test(18, m, "One\"",                  new[] { "One" });
    Test(19, m, "\"One",                  new[] { "One" });
    Test(20, m, "One \"\"",               new[] { "One", "" });
    Test(21, m, "One \"",                 new[] { "One", "" });
    Test(22, m, "1 A=\"B C\"=D 2",        new[] { "1", "A=B C=D", "2" });
    Test(23, m, "1 A=\"B \\\" C\"=D 2",   new[] { "1", "A=B \" C=D", "2" });
    Test(24, m, "1 \\A 2",                new[] { "1", "\\A", "2" });
    Test(25, m, "1 \\\" 2",               new[] { "1", "\"", "2" });
    Test(26, m, "1 \\\\\" 2",             new[] { "1", "\\\"", "2" });
    Test(27, m, "\"",                     new[] { "" });
    Test(28, m, "\\\"",                   new[] { "\"" });
    Test(29, m, "'A B'",                  new[] { "'A", "B'" });
    Test(30, m, "^",                      new[] { "^" });
    Test(31, m, "^A",                     new[] { "A" });
    Test(32, m, "^^",                     new[] { "^" });
    Test(33, m, "\\^^",                   new[] { "\\^" });
    Test(34, m, "^\\\\",                  new[] { "\\\\" });
    Test(35, m, "^\"A B\"",               new[] { "A B" });

    // Test cases Anton

    Test(36, m, @"/src:""C:\tmp\Some Folder\Sub Folder"" /users:""[email protected]"" tasks:""SomeTask,Some Other Task"" -someParam foo", new[] { @"/src:C:\tmp\Some Folder\Sub Folder", @"/users:[email protected]", @"tasks:SomeTask,Some Other Task", @"-someParam", @"foo" });

    // Test cases Daniel Earwicker 

    Test(37, m, "",            new string[] { });
    Test(38, m, "a",           new[] { "a" });
    Test(39, m, " abc ",       new[] { "abc" });
    Test(40, m, "a b ",        new[] { "a", "b" });
    Test(41, m, "a b \"c d\"", new[] { "a", "b", "c d" });

    // Test cases Fabio Iotti 

    Test(42, m, "this is a test ",    new[] { "this", "is", "a", "test" });
    Test(43, m, "this \"is a\" test", new[] { "this", "is a", "test" });

    // Test cases Kevin Thach

    Test(44, m, "\"C:\\Program Files\"",                       new[] { "C:\\Program Files" });
    Test(45, m, "\"He whispered to her \\\"I love you\\\".\"", new[] { "He whispered to her \"I love you\"." });

the "expected" value comes from directly testing it with cmd.exe on my machine (Win10 x64) and a simple print program:

static void Main(string[] args) => Console.Out.WriteLine(
quot;Count := {args.Length}\n{string.Join("\n", args.Select((v,i) => 
quot;[{i}] => '{v}'"))}");

These are the results:

Solution                      | Failed Tests
------------------------------|------------------------------------- 
Atif Aziz (749653)            | 2, 3, 10, 11, 12, 14, 16, 17, 18, 26, 28, 31, 32, 33, 34, 35, 36, 37, 39, 45
Jeffrey L Whitledge (298968)  | 1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 25, 26, 27, 28, 31, 32, 33, 34, 35, 36, 37, 39, 40, 41, 42, 43, 44, 45
Daniel Earwicker (298990)     | 10, 11, 12, 14, 16, 17, 18, 19, 20, 21, 22, 23, 25, 26, 27, 28, 31, 32, 33, 34, 35, 36, 45
Anton (299795)                | 12, 16, 17, 18, 19, 21, 23, 25, 26, 27, 28, 31, 32, 33, 34, 35, 45
CS. (467313)                  | 12, 18, 19, 21, 27, 31, 32, 33, 34, 35
Vapour in the Alley (2132004) | 10, 11, 12, 14, 16, 17, 20, 21, 22, 23, 25, 26, 27, 28, 29, 31, 32, 33, 34, 35, 36, 45
Monoman (7774211)             | 14, 16, 17, 20, 21, 22, 23, 25, 26, 27, 28, 31, 32, 33, 34, 35, 45
Thomas Petersson (19091999)   | 2, 3, 10, 11, 12, 14, 16, 17, 18, 19, 20, 21, 22, 23, 25, 26, 27, 28, 31, 32, 33, 34, 35, 36, 39, 45
Fabio Iotti (19725880)        | 1, 2, 3, 7, 10, 11, 12, 13, 14, 15, 16, 17, 19, 21, 22, 23, 25, 26, 28, 29, 30, 35, 36, 37, 39, 40, 42, 44, 45
ygoe (23961658)               | 26, 31, 32, 33, 34, 35
Kevin Thach (24829691)        | 10, 11, 12, 14, 18, 19, 20, 21, 22, 23, 26, 27, 31, 32, 33, 34, 35, 36
Lucas De Jesus (31621370)     | 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45
HarryP (48008872)             | 24, 26, 31, 32, 33, 34, 35
TylerY86 (53290784)           | 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 25, 26, 27, 28, 31, 32, 33, 34, 35, 36, 41, 43, 44, 45
Louis Somers (55903304)       | 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 25, 26, 27, 28, 29, 31, 32, 33, 34, 35, 36, 39, 41, 43, 44, 45
user2126375 (58233585)        | 5, 6, 15, 16, 17, 31, 32, 33, 34, 35
DilipNannaware (59131568)     | 1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 25, 26, 27, 28, 31, 32, 33, 34, 35, 36, 37, 39, 40, 41, 42, 43, 44, 45
Mikescher (this)              | -

Because no answer seemed correct (at least based on my use case) here is my solution, it currently passes all test cases (but if anyone has additional (failing) corner cases please comment):
public static IEnumerable<string> SplitArgs(string commandLine)
{
    var result = new StringBuilder();

    var quoted = false;
    var escaped = false;
    var started = false;
    var allowcaret = false;
    for (int i = 0; i < commandLine.Length; i++)
    {
        var chr = commandLine[i];

        if (chr == '^' && !quoted)
        {
            if (allowcaret)
            {
                result.Append(chr);
                started = true;
                escaped = false;
                allowcaret = false;
            }
            else if (i + 1 < commandLine.Length && commandLine[i + 1] == '^')
            {
                allowcaret = true;
            }
            else if (i + 1 == commandLine.Length)
            {
                result.Append(chr);
                started = true;
                escaped = false;
            }
        }
        else if (escaped)
        {
            result.Append(chr);
            started = true;
            escaped = false;
        }
        else if (chr == '"')
        {
            quoted = !quoted;
            started = true;
        }
        else if (chr == '\\' && i + 1 < commandLine.Length && commandLine[i + 1] == '"')
        {
            escaped = true;
        }
        else if (chr == ' ' && !quoted)
        {
            if (started) yield return result.ToString();
            result.Clear();
            started = false;
        }
        else
        {
            result.Append(chr);
            started = true;
        }
    }

    if (started) yield return result.ToString();
}

The code I used to generate the test results can be found here
 回复 原文 
~没有更多了~
关于作者
 夏の忆
暂无简介
 文章
 评论
26 人气
 发私信
相关话题
更多 
热门标签
操作系统 程序设计 IT运维 Linux系统管理 JavaScript 服务器应用 solaris C/C++ PHP Shell BSD Vue.js aix Oracle Python HTML 系统管理 HTML5 CSS 前端
更多 
推荐作者
Promise
文章 0 评论 0
qq_lbRlsh
文章 0 评论 0
待＂谢繁草
文章 0 评论 0
yy2010hell
文章 0 评论 0
漫无边际
文章 0 评论 0
傲娇萝莉攻
文章 0 评论 0
更多 
友情链接
文江博客
       
    我们使用 Cookies 和其他技术来定制您的体验包括您的登录状态等。通过阅读我们的 隐私政策 了解更多相关信息。 单击 接受 或继续使用网站，即表示您同意使用 Cookies 和您的相关数据。
     
    原文