identifying file types

[thedougster]

Within a C# program, is there any way to tell if a file is a text file or not? I mean a real way, not just basing your conclusion on a file name extension.

 

[RyanEK]

Hi,

Try this function...

static bool IsTextFile(string fileName)
  {
    byte[] file;
    using (System.IO.FileStream stream = new System.IO.FileStream(fileName, System.IO.FileMode.Open, System.IO.FileAccess.Read))
    {
      file = new byte[stream.Length];
      stream.Read(file, 0, file.Length);
    }

    if (file.Length > 3 && ((file[0] == 0x00 && file[1] == 0x00 && file[2] == 0xFE && file[2] == 0xFF /*UCS-4*/)))
      return true;
    else if (file.Length > 2 && ((file[0] == 0xEF && file[1] == 0xBB && file[2] == 0xBF /*UTF-8*/)))
      return true;
    else if (file.Length > 1 && ((file[0] == 0xFF && file[1] == 0xFE /*Unicode*/)))
      return true;
    else if (file.Length > 1 && (file[0] == 0xFE && file[1] == 0xFF /*Unicode Big Endian*/))
      return true;
    else
    {
      for (int i = 0; i < file.Length; i++)
        if (file[i] > 0x80)
          return false;

      return true;
    }
  }

Ryan

 

[thedougster]

Thanks for your help, RyanEK. Just one question: the last 2 terms in your first "if" statement (for the UCS-4 text file format) are:

file[2] == 0xFE && file[2] == 0xFF

Did you mean that the latter term should be:

file[3] == 0xFF

That is, the index should be 3?

Thanks again.

 

[phinoppix]

Just a few comments on the posted code. Be aware that line 5, will actually read all the contents:

file = new byte[stream.Length];
stream.Read(file, 0, file.Length);

Rather than loading everything in the memory, you can just get the first 3 bytes or so, since that's all you need to test.

If the file is non-unicode, it "may be" a 7-bit ASCII text file. Typically, an ASCII text file contains only all printable characters (char 32-127) and newline chars (char 13 and/or 10).

my 2cents