The following code(using commons codec Base64):
byte[] a = Hex.decodeHex("9349c513ed080dab".toCharArray());
System.out.println(Base64.encodeBase64URLSafeString(a));
System.out.println(Base64.encodeBase64String(a));
gives the following output:
k0nFE-0IDas //should be k0nFE-0IDas=
k0nFE+0IDas=
Base64.encodeBase64URLSafeString(a) returns k0nFE-0IDas instead of k0nFE-0IDas=. Why is this happening?
Why is this happening?
Because that's what it's documented to do:
Note: no padding is added.
The = characters at the end of a base64 string are called padding. They're used to make sure that the final string's length is a multiple of 4 characters - but they're not really required, in terms of information theory, so it's reasonable to remove them so long as you then convert the data back to binary using a method which doesn't expect padding. The Apache Codec Base64 class claims it transparently handles both regular and URL-safe base64, so presumably does handle a lack of padding.
