Function encode_with_padding 

pub(crate) fn encode_with_padding<E: Engine + ?Sized>(
    input: &[u8],
    output: &mut [u8],
    engine: &E,
    expected_encoded_size: usize,
)

B64-encode and pad (if configured).

This helper exists to avoid recalculating encoded_size, which is relatively expensive on short inputs.

encoded_size is the encoded size calculated for input.

output must be of size encoded_size.

All bytes in output will be written to since it is exactly the size of the output.