[Handshake] Support an will_next_write_payload_encrypted

[fogti]

The HandshakeState should have some way to determine if the next payload will be encrypted, where false-negatives are acceptable (and will probably happen if !is_my_turn() and read_message wasn't called yet). e.g. A way to only send a payload with the next write_message if it will be encrypted.

[mcginty]

I agree that this certainly feels more useful than a "was_write_payload_encrypted", which was implemented because it's trivial.

To implement will_write_payload_encrypt feels hairier, because it basically involves doing a "dry run" of the write_message operation without changing internal state, which feels kind of dirty to me. That said, snow does now have checkpointing to recover from errors and avoid having an unknown internal state, and this feature could be used to implement this.

It's definitely going to be a more expensive function to call than its past-tense sibling though.