没有必要加载整个文件先存入内存。您可以逐行处理：

 使用 System.IO; 
使用 System.Text.RegularExpressions; 

命名空间 ConsoleApplication17 
 {
 class  Program 
 {
私有 静态正则表达式=  new 正则表达式（ @ （PR \ * * -\ * * [A-Za] z0-9  - ] +），RegexOptions.IgnoreCase | RegexOptions.CultureInvariant | RegexOptions.Compiled）; 
静态  void  Main（ string  [] args）
 {
 string  myFile =  @  .. \..\TextFile1.txt跨度>; 
 string  writeFile =  @  .. \\ \\..\TextFile1.csv跨度>; 
使用（TextWriter writer =  new  StreamWriter（writeFile））
 {
 foreach （字符串行 文件.ReadLines（myFile））
 {
匹配m = Pattern.Match（line）; 
 if （m.Success）
 {
 writer.Write（m.Groups [ 1 跨度>]值）。 
 writer.Write（ ，）; 
 writer.WriteLine（line）; 
} 
} 
} 
} 
} 
} 

你可以尝试这样的事情：

  string  input = File.ReadAllText（myfile）; 
正则表达式正则表达式= 新正则表达式（ @  ^（小于COL2> *（小于COL1>？？PR\s * -\s * [A-ZA-Z0-9  - ] +）*）

，RegexOptions.IgnoreCase | RegexOptions.CultureInvariant | RegexOptions.Multiline）;
MatchCollection matchCollection = regex.Matches（input）;
使用（TextWriter tw = new StreamWriter（writeFile））
{
foreach （在 matchCollection中匹配m ）
{
string col1 = m.Groups [ col1]。值; // 仅提取的部分
string col2 = m.Groups [ col2]。 // 完整行
tw.WriteLine（ {0}，{1}，col1，col2）;
}
}

结果

 PR-EG4090-28，\ \192.168.8.212\c\Avago.ATF.Common\Clotho_SDI_Data\FBAR_2MS5-1505_N3131-3033_PR-EG4090-28_FULL-1PASS_CHAN1_20160610_104333_IP192.168.0.2_G6AABRH 

var myfile = (@"C:\Log\all.txt");
                string[] lin = File.ReadAllLines(myfile);
                string pat01 = @"PR -([A-Za-z0-9-]+)";
                MatchCollection matches = Regex.Matches((lin[1]), pat01);
                foreach (Match match in matches)
                {
                    string writeFile = @"E:\DataLocate_PR_id.csv";
                    string readFile = @"C:\Log\all.txt";                                   
                    StringBuilder csv = new StringBuilder();
                    csv.AppendLine(match + readFile);
                    File.AppendAllText(writeFile, csv.ToString());                
                }

What I have tried:

This is my 1s try and many of code hard to understand that I read from other forums.

It isn't necessary to load the whole file into memory first. You can process it line-by-line:

using System.IO;
using System.Text.RegularExpressions;

namespace ConsoleApplication17
{
  class Program
  {
    private static Regex Pattern = new Regex(@"(PR\s*-\s*[A-Za-z0-9-]+)", RegexOptions.IgnoreCase | RegexOptions.CultureInvariant | RegexOptions.Compiled);
    static void Main(string[] args)
    {
      string myFile = @"..\..\TextFile1.txt";
      string writeFile = @"..\..\TextFile1.csv";
      using (TextWriter writer = new StreamWriter(writeFile))
      {
        foreach (string line in File.ReadLines(myFile))
        {
          Match m = Pattern.Match(line);
          if (m.Success)
          {
            writer.Write(m.Groups[1].Value);
            writer.Write(",");
            writer.WriteLine(line);
          }
        }
      }
    }
  }
}

You can try something like this:

string input = File.ReadAllText(myfile);
Regex regex = new Regex(@"^(?<col2>.*(?<col1>PR\s*-\s*[A-Za-z0-9-]+).*)

", RegexOptions.IgnoreCase | RegexOptions.CultureInvariant | RegexOptions.Multiline); MatchCollection matchCollection = regex.Matches(input); using (TextWriter tw = new StreamWriter(writeFile)) { foreach (Match m in matchCollection) { string col1 = m.Groups["col1"].Value; // Only the extracted part string col2 = m.Groups["col2"].Value; // The full line tw.WriteLine("{0},{1}", col1, col2); } }

Result

PR-EG4090-28,\\192.168.8.212\c\Avago.ATF.Common\Clotho_SDI_Data\FBAR_2MS5-1505_N3131-3033_PR-EG4090-28_FULL-1PASS_CHAN1_20160610_104333_IP192.168.0.2_G6AABRH