Handling bytes

The C code I am interfacing with processes fixes size arrays of type
uint8_t. I've seen suggested how to get python strings into and out of C
character arrays, but I don't really like using strings in python code
to hold raw data bytes. Until python 3 arrives with its bytearray stuff,
I often just put byte range ints into traditional sequences (list,
tuple, or array). To handle this in pyrex I could usecode like:

    def encrypt(data):
        if len(data) != 8:
            raise ValueError("data must be 8 bytes long")

        cdef uint8_t cin[8]
        cdef uint8_t cout[8]
        cdef int i
        for i from 0 <= i < 8:
            cin[i] = data[i]

        do_c_encrypt(cin, cout)

        result = []
        for i from 0 <= i < 8:
            result.append(cout[i])
        return result

This could be called from python code like:

    result = encrypt( [0,1,2,3,4,5,6,7] )

But I'm assuming there is a more efficient approach to converting to and
from python sequences. Maybe I'd be better off using python strings
with  PyString_FromStringAndSize and friends? Or perhaps using
PyObject_AsReadBuffer and PyObject_AsWriteBuffer would work, but that
would limit the input object to python array.array('B') or similar
right? Just wondering if there is a preferred approach since I'd assume
this has been hashed over many times.

Thanks,
-Brad