正则表达式替换过载？

Question

so I have the following code which generates a sha256 hash of the file data of the file that was uploaded by the user. This works fine but sometimes it includes illegal characters (for the windows os).

So what I'm trying to implement is a try catch to strip of illegal characters. I have pulled this information off of Microsoft's website itself. However, when implemented with Regex.Replace() I'm told that it only accepts 5 overloads.

Which is confusing because that is what I have and I have triple checked that my hashedfile1name is a string type variable.

The other problem is that for the try catch its telling me that it doesn't know what RegexMatchTimeoutException is. But there are no more imports/using statements in Microsoft's example .

try
{
    FileUpload1.SaveAs("C:\\direct\\uploads\\" + FileUpload1.FileName);
    using (fs = File.OpenRead("C:\\direct\\uploads\\" + FileUpload1.FileName))
    {
        byte[] bytes = new byte[fs.Length];
        fs.Read(bytes, 0, Convert.ToInt32(fs.Length));
        hashedfile1name = Convert.ToBase64String(hash.ComputeHash(bytes));
    }
    try
    {
        Regex.Replace(hashedfile1name, @"[^\w\.@-]", "", RegexOptions.None, TimeSpan.FromSeconds(1.5));
    }
    catch (RegexMatchTimeoutException)
    {
        hashedfile1name = "";
    }
    FileUpload1.SaveAs("C:\\direct\\uploads\\" + hashedfile1name);
    File.Delete("C:\\direct\\uploads\\" + FileUpload1.FileName);

    Label1.Text = "File name: " + FileUpload1.PostedFile.FileName + " - " + hashedfile1name;
}
catch (Exception ex)
{
    Label1.Text = "ERROR: " + ex.Message.ToString();
}

Answer 1

Regex.Replace does not replace in-place; it returns the replacement as a string. The segment:

try
{
    Regex.Replace(hashedfile1name, @"[^\w\.@-]", "", RegexOptions.None, TimeSpan.FromSeconds(1.5));
}
catch (RegexMatchTimeoutException)
{
    hashedfile1name = "";
}

Can be replaced with simply:

hashedfile1name = Regex.Replace(hashedfile1name, @"[^\w\.@-]", "");

Your code should then function as expected.

(If you want to preserve the entire hash, you might want to consider Base-32 encoding, which is similar to base64 but uses only alphanumeric characters. .NET does not include a Base32 methods, but an implementation is provided in Shane's answer here .)