我如何从c#执行命令行程序,并获得STD OUT结果?具体来说,我想对两个以编程方式选择的文件执行DIFF,并将结果写入一个文本框。


当前回答

这里有一个小例子:

using System;
using System.Diagnostics;

class Program
{
    static void Main(string[] args)
    {
        var p = Process.Start(
            new ProcessStartInfo("git", "branch --show-current")
            {
                CreateNoWindow = true,
                UseShellExecute = false,
                RedirectStandardError = true,
                RedirectStandardOutput = true,
                WorkingDirectory = Environment.CurrentDirectory
            }
        );

        p.WaitForExit();
        string branchName =p.StandardOutput.ReadToEnd().TrimEnd();
        string errorInfoIfAny =p.StandardError.ReadToEnd().TrimEnd();

        if (errorInfoIfAny.Length != 0)
        {
            Console.WriteLine($"error: {errorInfoIfAny}");
        }
        else { 
            Console.WriteLine($"branch: {branchName}");
        }

    }
}

我相信这是最短的形式。

请注意,大多数命令行工具很容易混淆标准输出和标准错误,有时将它们组合成一个字符串是有意义的。

p.e exitcode有时也很有用。

上面的例子用于编写命令行工具,如果你想自己做的话。请注意,对于cli自动化,也可以使用Cake Frosten和Cake Git扩展。

其他回答

The accepted answer on this page has a weakness that is troublesome in rare situations. There are two file handles which programs write to by convention, stdout, and stderr. If you just read a single file handle such as the answer from Ray, and the program you are starting writes enough output to stderr, it will fill up the output stderr buffer and block. Then your two processes are deadlocked. The buffer size may be 4K. This is extremely rare on short-lived programs, but if you have a long running program which repeatedly outputs to stderr, it will happen eventually. This is tricky to debug and track down.

有一些很好的方法来处理这个问题。

One way is to execute cmd.exe instead of your program and use the /c argument to cmd.exe to invoke your program along with the "2>&1" argument to cmd.exe to tell it to merge stdout and stderr. var p = new Process(); p.StartInfo.FileName = "cmd.exe"; p.StartInfo.Arguments = "/c mycmd.exe 2>&1"; Another way is to use a programming model which reads both handles at the same time. var p = new Process(); p.StartInfo.FileName = "cmd.exe"; p.StartInfo.Arguments = @"/c dir \windows"; p.StartInfo.CreateNoWindow = true; p.StartInfo.RedirectStandardError = true; p.StartInfo.RedirectStandardOutput = true; p.StartInfo.RedirectStandardInput = false; p.OutputDataReceived += (a, b) => Console.WriteLine(b.Data); p.ErrorDataReceived += (a, b) => Console.WriteLine(b.Data); p.Start(); p.BeginErrorReadLine(); p.BeginOutputReadLine(); p.WaitForExit();

 System.Diagnostics.ProcessStartInfo psi =
   new System.Diagnostics.ProcessStartInfo(@"program_to_call.exe");
 psi.RedirectStandardOutput = true;
 psi.WindowStyle = System.Diagnostics.ProcessWindowStyle.Hidden;
 psi.UseShellExecute = false;
 System.Diagnostics.Process proc = System.Diagnostics.Process.Start(psi); ////
 System.IO.StreamReader myOutput = proc.StandardOutput;
 proc.WaitForExit(2000);
 if (proc.HasExited)
  {
      string output = myOutput.ReadToEnd();
 }

在PublicDomain开源代码中有一个ProcessHelper类,你可能会感兴趣。

这里有一个小例子:

using System;
using System.Diagnostics;

class Program
{
    static void Main(string[] args)
    {
        var p = Process.Start(
            new ProcessStartInfo("git", "branch --show-current")
            {
                CreateNoWindow = true,
                UseShellExecute = false,
                RedirectStandardError = true,
                RedirectStandardOutput = true,
                WorkingDirectory = Environment.CurrentDirectory
            }
        );

        p.WaitForExit();
        string branchName =p.StandardOutput.ReadToEnd().TrimEnd();
        string errorInfoIfAny =p.StandardError.ReadToEnd().TrimEnd();

        if (errorInfoIfAny.Length != 0)
        {
            Console.WriteLine($"error: {errorInfoIfAny}");
        }
        else { 
            Console.WriteLine($"branch: {branchName}");
        }

    }
}

我相信这是最短的形式。

请注意,大多数命令行工具很容易混淆标准输出和标准错误,有时将它们组合成一个字符串是有意义的。

p.e exitcode有时也很有用。

上面的例子用于编写命令行工具,如果你想自己做的话。请注意,对于cli自动化,也可以使用Cake Frosten和Cake Git扩展。

这可能不是最好/最简单的方法,但可能是一种选择:

当从代码执行时,添加“> output.txt”,然后读入output.txt文件。