[FEATURE/IDENTITY] Incorporate byte stream processing for encryption/decryption

[stephen-crawford]

This request is being filed to trim the amount of features being introduced with pull request #4730

As is, Principal Identifier Tokens do not have the functionality to be created via byte stream. This functionality is desired for various uses and should be added. You should be able to create a PIT via byte stream which is identical to a counterpart created via string so long as the byte stream encodes the same characters.

Tests should also be added in the PrincipalIdentifierTokenTests.java file to confirm this functionality. This may require the modification of the existing encryption/decryption code.

[peternied]

@scrawfor99 Would you might updating the roadmap [1] with when you think this should be implemented and provide a link back to this issue?

https://github.com/opensearch-project/OpenSearch/blob/feature/identity/IDENTITY_ROADMAP.md

[stephen-crawford]

[Update 11/30] I am going to start looking at this today and will update further with what I find about the steps for implementing this.

I do have a design question for everyone in the meantime: What do we want to do about decryption? While encryption should be possible I am not sure that decrypting the JWTs as a byte stream will work in the same way. I think we will need to simply store the bytes until we get the entire token and then decrypt it all at once since we will need to have the signature.