On 1 May 2018 at 21:30, Antoine Pitrou <solipsis@pitrou.net> wrote:
Hi Ken,
On Tue, 1 May 2018 19:22:52 +0800 Ken Hilton <kenlhilton@gmail.com> wrote:
So I'm pretty sure everyone here is familiar with how the "bytes" object works in Python 3. It acts mostly like a string, with the exception that 0-dimensional subscripting (var[idx]) returns an integer, not a bytes object - the integer being the ordinal number of the corresponding character. However, 1-dimensional subscripting (var[idx1:idx2]) returns a bytes object. Example:
>>> a = b'hovercraft' >>> a[0] 104 >>> a[4:8] b'rcra'
Though this isn't exactly unexpected behavior (it's not possible to accidentally do 1-dimensional subscripting and expect an integer it's a different syntax), it's still a shame that it isn't possible to quickly
and
easily subscript an integer out of it. Following up from the previous example, The only way to get 493182234161465432041076 out of b'hovercraft' in a single expression is as follows:
list(__import__('itertools').accumulate((i for i in a), lambda x, y: (x << 8) + y))[-1]
Let's see:
a = b'hovercraft' int.from_bytes(a, 'big') 493182234161465432041076
It's also worth noting that if there's more than one integer of interest in the string, than using the struct module is often going to be better than using multiple slices and int.from_bytes calls: >>> import struct >>> data = b"hovercraft" >>> struct.unpack(">IIH", data) (1752135269, 1919119969, 26228) (The struct module doesn't handle arbitrary length integers, but it handles 8, 16, 32, and 64 bit ones, which is enough for a lot of common use cases) Cheers, Nick. -- Nick Coghlan | ncoghlan@gmail.com | Brisbane, Australia