Problems with X509Store Certificates.Find FindByThumbprint

Solution 1:

I suppose that you have copy-pasted the thumbprint from the Windows certificate information dialog box into your code (or to a config file if this is a simplified example). Annoyingly, the first character in the thumbprint textbox is the invisible Unicode "left-to-right-mark" control character. Try selecting the opening string quote and the first character of the thumbprint, deleting them (which will also get rid of the invisible character in-between), and retyping them by hand.

I was subjected to this odd behaviour myself today, and it took me over an hour to figure it out. The way I finally saw it was by using the debugger to check the lengths and hash codes of findValue and of the Thumbprint of the certificate object, which turned out to be different. This led me to inspect the character arrays of those strings in the debugger, where the invisible character showed up.

Solution 2:

I took some of the answers here and combined them into a static method that takes care of removing special characters and upper cases everything. Hopefully someone else can use it.

public static X509Certificate2 GetCertificate(string thumbprint)
{
    // strip any non-hexadecimal values and make uppercase
    thumbprint = Regex.Replace(thumbprint, @"[^\da-fA-F]", string.Empty).ToUpper();
    var store = new X509Store(StoreName.My, StoreLocation.LocalMachine);

    try
    {
        store.Open(OpenFlags.ReadOnly);

        var certCollection = store.Certificates;
        var signingCert = certCollection.Find(X509FindType.FindByThumbprint, thumbprint, false);
        if (signingCert.Count == 0)
        {
            throw new FileNotFoundException(string.Format("Cert with thumbprint: '{0}' not found in local machine cert store.", thumbprint));
        }

        return signingCert[0];
    }
    finally
    {
        store.Close();
    }
}

Solution 3:

I had the same Problem and solved it:

1. I copied the Fingerprint from mmc directly to VS. I compared the strings and didn't find any difference.

2. Checking the length with hash.length, there was a difference, 41 vs. 40.

There is an invisible Char added to the string by copying it out of mmc.

Solving:

1. copy the Fingerprint from mmc to Notepad.exe
2. copy this string again
3. paste to your code

It's working.

Solution 4:

I fell victim to this. Not only was there a Unicode "left-to-right" character in the Windows console snap-in display of the thumbprint, but it also had lowercase hex characters, with spaces between every two characters. The output of CertUtil also had lowercase characters, and spaces. To get a match, I had to specify the findValue as a string which has been transformed to

1. Remove the leading special character,
2. Remove the whitespace between character clusters,
3. Change all the characters to uppercase.

Solution 5:

This tripped me up too, I wrote this function to clean the thumbprint when copied and pasted from MMC:

public string CleanThumbprint(string mmcThumbprint)
    {
        //replace spaces, non word chars and convert to uppercase
        return Regex.Replace(mmcThumbprint, @"\s|\W", "").ToUpper();
    }

...
        var myThumbprint = CleanThumbprint("‎b3 ab 84 e5 1e e5 e4 75 e7 a5 3e 27 8c 87 9d 2f 05 02 27 56");
        var myCertificate = certificates.Find(X509FindType.FindByThumbprint, myThumbprint, true)[0];