C＃-如何优化SystemFileWatcher

Question

I have a SystemFileWatcher that allows a user to view live log files. Some of these log files can get quite large, and they are sometimes read through VPN, slowing down the process even more. Is there a way to optimize my Changed Event? I've gotten it so that it works, and displays the information, and doesn't duplicate or show past events, but my buttons on the form lock down as well. So my first question is how can I optimize my code further? My second question would be how do I still interact with my form?

private void fileWatcher_Changed(object sender, FileSystemEventArgs e)
    {
        fileWatcher.EnableRaisingEvents = false;
        var fs = new FileStream(GlobalVar.GlobalString, FileMode.Open, FileAccess.Read, FileShare.ReadWrite);
        using (var sr = new StreamReader(fs))
        {
            var s = "";
            int copyGood = 0;
            bool result = false;
            bool copyModeOn = false;
            var newDateTime = new DateTime();
            List<string> list = new List<string>();
            if (GlobalVar.GlobalLineCount > 0)
            {
                for (var i = 0; i < GlobalVar.GlobalLineCount; i++)
                {
                    sr.ReadLine();
                }
                while ((s = sr.ReadLine()) != null)
                {
                    list.Add(s);
                }
            }
            else
            {
                while ((s = sr.ReadLine()) != null)
                {
                    list.Add(s);
                }
            }
            GlobalVar.GlobalLineCount = list.Count();
            list.Reverse();
            var list2 = list.Take(50).Reverse().ToList();

              foreach (string s1 in list2)
              {
                    if (s1.Length > 23)
                    {
                        var lineTime = s1.Substring(0, 23);
                        result = DateTime.TryParse(lineTime, out newDateTime);
                    }
                    if (newDateTime != null && result)
                    {
                        DateTime lineTime = Convert.ToDateTime(newDateTime.ToString(@"HH:mm:ss.fff"));
                        DateTime dt = Convert.ToDateTime(GlobalVar.GlobalDateTimeString);
                        DateTime globalTimeSet = Convert.ToDateTime(dt.ToString(@"HH:mm:ss.fff"));
                        copyGood = DateTime.Compare(globalTimeSet, lineTime);
                    }
                    if (result && copyGood < 0)
                    {
                        if (richTextBox1.Lines.Length > 0)
                        {
                            foreach (string line in richTextBox1.Lines)
                            {
                                if (line == s1)
                                {
                                    copyModeOn = false;
                                    break;
                                }
                                copyModeOn = true;
                            }
                        }
                        else
                        {
                            copyModeOn = true;
                        }
                        if (copyModeOn)
                        {
                            richTextBox1.AppendText(Environment.NewLine + s1 );
                            richTextBox1.Focus();
                            richTextBox1.Select(richTextBox1.TextLength, 0);
                            richTextBox1.ScrollToCaret();
                        }
                    }
                }
        }
        fileWatcher.EnableRaisingEvents = true;
    }`

Any help you can provide would be greatly appreciated!

Answer 1

Not really. The big gain would be to not have the remote network connection in between - which would mean moving most of the processing to an application server and only sending the lines that are needed over the network.

what you ould do is not read the whole file. Obviously logfile entries are of limited size. There is no need to start at the end over and over again ;) You can assume a max size per line (let's say 512 bytes) which alone would cut out a lot of reading for large files. Or you can actually remember the position of the last read.