DKNet

Base64 and Base64URL encoding

Part of DKNet.Svc.Encryption. Base64StringExtensions exposes ToBase64String/FromBase64String for the standard alphabet and ToBase64UrlString/FromBase64UrlString for the URL-safe RFC 4648 variant, plus IsBase64String for validating the standard form — all as this string extension methods, static calls work too.

using DKNet.Svc.Encryption;

var token = "???>>>subjects".ToBase64UrlString();
var back = token.FromBase64UrlString();
Console.WriteLine(token); // Pz8_Pj4-c3ViamVjdHM
Console.WriteLine(back);  // ???>>>subjects
Console.WriteLine(token.IsBase64String()); // False (- and _ are not in the standard Base64 alphabet)

The URL variants are useful for JWT-style payloads without pulling in a JWT library: they replace +// with -/_ and omit = padding. Whitespace or empty input returns string.Empty from both decode helpers, and IsBase64String returns false for it.

ToBase64UrlString encodes UTF-8 input with the URL-safe alphabet; IsBase64String recognizes a token that stayed within the standard alphabet but returns false for one containing '-' or '_', while FromBase64UrlString decodes either form back to the original string.

The diagram cannot show the standard-alphabet pair (ToBase64String/FromBase64String) on its own — that path never touches IsBase64String’s URL-character check at all, because it never produces -/_ in the first place.

⚠️ Gotchas & limits

IsBase64String validates the standard alphabet only. A token produced by ToBase64UrlString can return false from IsBase64String even though FromBase64UrlString decodes it correctly — there’s no IsBase64UrlString helper to check the URL-safe form directly.
No implicit alphabet detection. Decoding with the wrong helper (FromBase64String on a Base64URL token containing -/_) throws FormatException rather than falling back.
String-only surface. No byte[] overloads — a caller already holding bytes converts to a string first.

This site is open source. Improve this page.