How to Delete a Specific Line in a Text File[Improved_vs2]

What Abhinav suggested is not so practical in terms of performance and may not be suitable for big files.

Here is what you can do:

static void CopyFileWithExcludedLine(string inputFileName, string outputFileName, ulong excludeLine) {
    ulong lineCounter = 0;
    System.IO.StreamWriter writer = null;
    using (System.IO.StreamReader reader = new System.IO.StreamReader(inputFileName, true)) {
        try {
            while (reader.Peek() >= 0) {
                string line = reader.ReadLine();
                if (writer == null)
                    writer =
                       new System.IO.StreamWriter(
                           outputFileName, false,
                           reader.CurrentEncoding);
                if (lineCounter != excludeLine)
                    writer.WriteLine(line);
                lineCounter++;
            } //loop
        } finally {
            if (writer != null)
                writer.Dispose();
        } //exception
    } //using
} //CopyFileWithExcludedLine

Most complex thing here is my solution for encoding. I reproduce encoding of the input file in the output file, but input encoding is accessible only after first read operation. That's why I applied lazy initialization of the writer (when encoding is already known) and added an extra try-finally block (using cannot be used, but the try-catch block is perfectly identical in function.

If you know encoding, the code will be much more simple.

—SA

[Answering a follow-up question about encoding]

Text files comes with different encodings. When you read file in one encoding assuming a different encoding, you may or may not screw up the content of the file. It also depends on the actual set of code points used in the file — for example if all code points lie in the ASCII range, UTF-8 encoding and ANSI encoding will give identical results if UTF-8 without BOM is ised. Encoding issues are covered by the class System.Text.Encoding and derived classes, see http://msdn.microsoft.com/en-us/library/system.text.encoding.aspx[^].

Encodings are roughly subdivided into Unicode UTFs and the encodings used prior to Unicode. .NET supports Unicode. Unicode at its core does not define encodings. It defines one-to-one correspondence between characters as cultural entities, regardless fonts, glyphs, etc., and a set of abstract mathematical numbers (code points), totally abstracted from their computer presentation such as size, little- of big-endian, etc. Instead, a set of different UTF encodings is defined. Unicode is not a 16-bit encoding! It standardize much more code points than 2¹⁶! Even UTF-8 and UTF-16 UTFs support them all (in this way, a character is not one byte or two bytes — it takes 1 to 4 bytes, in fact). They have different convenience in different situations. A text file may or may not contain Byte-Order Marker (BOM) at the beginning of the file, which helps to recognize UTF encoding before reading.

For more information, use:
http://unicode.org/[^],
http://unicode.org/faq/utf_bom.html[^].

—SA

Hope this does something close to what you need to read the lines into a List, so you can sort though them.

public List<string> FileToList(string filePath)
{
    string content = null;
    List<string> lines = new List<string>(){};
    System.IO.StreamReader sr = null;
    // read the file's lines into a List
    try {
        sr = new System.IO.StreamReader(filePath);
        while (sr.Peek() >= 0) {
            lines.Add(sr.ReadLine());
        }
    } finally {
        if ((sr != null))
            sr.Close();
    }
  
    return lines;
}</string></string></string>

Then when you save them back just overwrite the file with the remaining list.

Maybe that will work?

Here is a quick console test that achieves your basic goal, tweak it as you need to fit your needs, but it gives a simple way to remove lines that start with a given ID, if none of the other lines in the file start with "ID like" text, no worries.

List<string> lines = File.ReadAllLines(@"J:\test.txt").ToList();

int removeCount = lines.RemoveAll(item => item.StartsWith("0005"));

Console.WriteLine(String.Format("Removed {0} line(s).", removeCount));

Good luck!