如何检查richTextBox1中的行在文本文件中是否已经存在？

Question

In form1 top:

private List<string> entriesRlines = new List<string>();
private List<string> entriesLines = new List<string>();
private string[] lines;
private List<string> RedAlerts;
private StreamWriter WriteAlerts;
private string AlertsDirectory;
private string AlertsFilename;
private string combineAlert;

In form1 constructor:

AlertsDirectory = Path.GetDirectoryName(Application.LocalUserAppDataPath) + @"\alerts";
if (!Directory.Exists(AlertsDirectory))
{
  Directory.CreateDirectory(AlertsDirectory);
}
AlertsFilename = "Alerts.txt";
combineAlert = Path.Combine(AlertsDirectory, AlertsFilename);
if (File.Exists(combineAlert))
            {
                lines = File.ReadAllLines(combineAlert);
                label5.Text = lines.Length.ToString();
                RedAlerts = new List<string>(lines);
            }
            RedColorAlert();

Then the RedColorAlert method:

private void RedColorAlert()
        {
            WriteAlerts = new StreamWriter(combineAlert,true);
            string[] rlines = richTextBox1.Lines;
            int linespos;
            if (lines != null)
            {
                for (int x = 0; x < lines.Length; x++)
                {
                    linespos = lines[x].IndexOf("===>");
                    string t = lines[x].Substring(linespos + 4);
                    entriesLines.Add(t);
                }
                for (int y = 0; y < rlines.Length; y += 4)
                {
                    entriesRlines.Add(rlines[y]);
                }
                for (int i = 0; i < entriesLines.Count; i++)
                {
                    for (int f = 0; f < entriesRlines.Count; f++)
                    {
                        if (!entriesLines[i].Contains(entriesRlines[f]))
                        {
                            if (entriesRlines[i].Contains("צבע אדום") || entriesRlines[i].Contains("אזעקה"))
                            {
                                timer3.Start();
                                WriteAlerts.WriteLine(DateTime.Now + "   ===>   " + entriesRlines[i]);
                                break;
                            }
                        }
                    }
                }
            }
            else
            {
                for (int i = 0; i < rlines.Length; i++)
                {
                    if (rlines[i].Contains("צבע אדום") || rlines[i].Contains("אזעקה"))
                    {
                        timer3.Start();
                        WriteAlerts.WriteLine(DateTime.Now + "   ===>   " + rlines[i]);
                    }
                }
            }
            WriteAlerts.Close();
        }

If the first time the text file is empty do the second part the ELSE part. If the text file already exist and have one ore more lines inside do the first part:

for (int i = 0; i < entriesLines.Count; i++)
                {
                    for (int f = 0; f < entriesRlines.Count; f++)
                    {
                        if (!entriesLines[i].Contains(entriesRlines[f]))
                        {
                            if (entriesRlines[i].Contains("צבע אדום") || entriesRlines[i].Contains("אזעקה"))
                            {
                                timer3.Start();
                                WriteAlerts.WriteLine(DateTime.Now + "   ===>   " + entriesRlines[i]);
                                break;
                            }
                        }
                    }
                }

This loop in loop is not written good for some reason it's getting inside all the time and start the timer3 and WriteLine even if the line is already in the text file. I used a breakpoint.

This is how the text file look like now:

8/2/2014 8:11:27 PM ===> 18:46 צבע אדום במ״א אשכול-- נפילה. אזרח נפצע, מפונה במסו8/3/2014 4:25:08 AM ===> 04:13 צבע אדום במ״א שדות נגב ושער הנגב 8/3/2014 4:25:51 AM ===> 04:13 צבע אדום במ״א שדות נגב ושער הנגב 8/3/2014 4:26:50 AM ===> 04:13 צבע אדום במ״א שדות נגב ושער הנגב

The second line is should be like the third and fourth lines but the second line got mixed with the first one for some reason. And the third and fourth lines are should be the same as the second line and never be written.

The idea is that if a line that also contains what i said as condition and not exist already in the text file add it a line new line in the text file. If a line is already exist don't start timer3 and dont write it again.

This is how a line look like in the text file and also in entriesLines:

8/3/2014 4:26:50 AM ===> 04:13 צבע אדום במ״א שדות נגב ושער הנגב

And this is a line that is not in the text file or maybe it does look like in the richTextFile1 and also in entriesRlines:

04:36 צבע אדום במ״א שדות נגב ושער הנגב

entriesRlines is List contain lines of richTextBox1 after psre them. entriesLines is List contain the lines from the richTextBox1.

Answer 1

Some of your Hebrew strings should be displayed right-to-left, for instance

א

is right-to-left: http://www.fileformat.info/info/unicode/char/5d0/index.htm

(And crikey Firefox isn't handling it correctly at all! That's why I had to put it on its own line.)

Maybe the strings displayed in the GUI aren't what you think they are, because they are displayed right-to-left but are actually stored left-to-right? Or vice versa? And who knows what Visual Studio does when displaying string literals with right-to-left characters? I certainly don't... And gosh! VS 2008 seems to be displaying them right-to-left sometimes! If I position my cursor at the beginning of one of your string literals, and press the left arrow, it jumps to the end of the string, and goes backwards!

What I would recommend doing to debug your problem is to make sure your strings contain the characters you expect in the order you expect , with one of the following utilities:

    public static IEnumerable<int> Utf32CodePoints(string s, int index)
    {
        for (int length = s.Length; index < length; index++)
        {
            yield return char.ConvertToUtf32(s, index);
            if (char.IsSurrogatePair(s, index))
                index++;
        }
    }

    public static IEnumerable<KeyValuePair<int, int>> Utf32IndexedCodePoints(string s, int index)
    {
        for (int length = s.Length; index < length; index++)
        {
            yield return new KeyValuePair<int, int>(index, char.ConvertToUtf32(s, index));
            if (char.IsSurrogatePair(s, index))
                index++;
        }
    }

Next remove all those inline string literals (a string literal should never appear more than once in your code anyway) and make then constants, eg:

    const string AlertType1 = "צבע אדום";
    const string AlertType2 = "אזעקה";

Then, using these utilities, examine the constant strings, the TextBox strings, and the file strings, to see what order the characters are really in.

By the way, I recommend you always declare your fields and variables with the smallest possible scope. Eg rather than a field "WriteAlerts" you should define a variable only in the scope required:

        using (var WriteAlerts = new StreamWriter(AlertsDirectoryAndName, true))
        {
            // your code
        }

Update

It looks like Visual Studio 2008 is displaying Hebrew strings right-to-left - even in my English-US locale! I wrote the following code:

    const string Hebrew1 = "צבע אדום";
    const string Hebrew1ApparentFirstChar = "ם";

    static void TestHebrewConstantStringOrder() 
    {
        List<int> Hebrew1Utf32 = new List<int>(TextHelper.Utf32CodePoints(Hebrew1, 0));
        foreach (var codePoint in TextHelper.Utf32CodePoints(Hebrew1ApparentFirstChar, 0))
        {
            int index = Hebrew1Utf32.IndexOf(codePoint);
            Debug.WriteLine(string.Format("Index of U+{0:X8}:  {1}", codePoint, index));
        }
        Debug.WriteLine("Codepoints of " + Hebrew1);
        foreach (var pair in TextHelper.Utf32IndexedCodePoints(Hebrew1, 0))
        {
            string msg = string.Format("{0,7}: U+{1:X8}", pair.Key, pair.Value);
            Debug.WriteLine(msg);
        }
    }

And got the following output:

Index of U+000005DD:  7
Codepoints of צבע אדום
      0: U+000005E6
      1: U+000005D1
      2: U+000005E2
      3: U+00000020
      4: U+000005D0
      5: U+000005D3
      6: U+000005D5
      7: U+000005DD
Codepoints of ם
      0: U+000005DD

And, in case your browser is rendering things differently than mine (a real possibility!) here are screen shots:

As you can see, the "ם" character that looks to be first in the string is actually last.