Python：bytearray对象的长度与派生字节对象的长度不匹配吗？

Question

Got some code like this:

some_buf = build_buffer() # returns a bytearray of ~40 bytes
align_buf(some_buf) # align it to 16 byte boundary
do_something(some_buf[16:32]) # This needs a bytes object, cannot work with bytearray

Alright, let me change it to this:

some_buf = build_buffer() # returns a bytearray of ~40 bytes
align_buf(some_buf) # align it to 16 byte boundary
bytes_buf = bytes(some_buf)[16:32]
do_something(some_buf)

The do_something() complains the bytes object is not 16 byte aligned. To see what's going on, I added a simple condition:

some_buf = build_buffer() # returns a bytearray of ~40 bytes
align_buf(some_buf) # align it to 16 byte boundary
bytes_buf = bytes(some_buf)[16:32]
if len(some_buf) != len(bytes_buf):
    msg = "Length of bytearray ({}) != Length of bytes object ({})".format(len(some_buf), len(bytes_buf))
    raise RuntimeError(msg)
do_something(some_buf)

And this is what I am getting:

RuntimeError: Length of bytearray (144) != Length of bytes object (562)

I am on a Windows 10 x64 system, with Python ( CPython ) interpreter Python 2.7.12 (v2.7.12:d33e0cf91556, Jun 27 2016, 15:19:22) [MSC v.1500 32 bit (Intel)] on win32

Anyone else seen this behavior before? If so, ideas/clues of what I might be missing here?

Answer 1

FWIW, if this behavior is seen, and IntelHex module is in the control flow somewhere, it might be a bug. Or at least, expect this unexpected behavior. Somewhere in the conversion between Intel hex data to binary bytearray and from there to a bytes object, a blackhole probably exists.

In such cases, its probably worth trying with other versions or on other platforms (eg spin up a Linux VM) for the sake of cross-verification, at least.

As mentioned in comments, this behavior was not observed with the same code paths on CPython '3.5.2 (v3.5.2:4def2a2901a5, Jun 25 2016, 22:18:55) [MSC v.1900 64 bit (AMD64)]' interpreter.