zlib: improve bytes handling

[sobolevn]

compress

>>> import zlib
>>> zlib.compress(bytearray(b'abc'))
b"x\x9cKLJ\x06\x00\x02M\x01'"
>>> zlib.compress(memoryview(b'abc'))
b"x\x9cKLJ\x06\x00\x02M\x01'"

decompress

>>> zlib.decompress(bytearray(b"x\x9cKLJ\x06\x00\x02M\x01'"))
b'abc'
>>> zlib.decompress(memoryview(b"x\x9cKLJ\x06\x00\x02M\x01'"))
b'abc'

adler32

>>> zlib.adler32(bytearray(b''))
1
>>> zlib.adler32(memoryview(b''))
1

crc32

>>> zlib.crc32(bytearray(b'123'))
2286445522
>>> zlib.crc32(memoryview(b'123'))
2286445522

decompressobj and compressobj

>>> zlib.decompressobj(0, bytearray(b''))
<zlib.Decompress object at 0x103b4b520>
>>> zlib.decompressobj(0, memoryview(b''))
<zlib.Decompress object at 0x103b4b5d0>

>>> zlib.compressobj(zdict=bytearray(b''))
<zlib.Compress object at 0x103b4b520>
>>> zlib.compressobj(zdict=memoryview(b''))
<zlib.Compress object at 0x103b4b5d0>

[github-actions]

According to mypy_primer, this change has no effect on the checked open source code. 🤖🎉

[hauntsaninja]

Lgtm

[JelleZijlstra]

Confirmed in the C code.