On Sun, Apr 15, 2018 at 8:05 AM, Tim Peters <tim.peters@gmail.com> wrote: [...]
Then I thought "this is stupid! Python already does reference counting." Voila! Vast swaths of tedious code vanished, giving this remarkably simple implementation:
def mytee(xs, n): last = [None, None]
def gen(it, mylast): nonlocal last while True: mylast = mylast[1] if not mylast: mylast = last[1] = last = [next(it), None] yield mylast[0]
it = iter(xs) return tuple(gen(it, last) for _ in range(n))
There's no need to keep a pointer to the start of the shared list at all - we only need a pointer to the end of the list ("last"), and each derived generator only needs a pointer to its own current position in the list ("mylast").
Things here remind me of my implementation design for PEP 555: the "contexts" present in the process are represented by a singly-linked tree of assignment objects. It's definitely possible to write the above in a more readable way, and FWIW I don't think it involves "assignments as expressions".
What I find kind of hilarious is that it's no help at all as a prototype for a C implementation: Python recycles stale `[next(it), None]` pairs all by itself, when their internal refcounts fall to 0. That's the hardest part.
Why can't the C implementation use Python refcounts? Are you talking about standalone C code? Or perhaps you are thinking about overhead? (In PEP 555 that was not a concern, though). Surely it would make sense to reuse the refcounting code that's already there. There are no cycles here, so it's not particulaly complicated -- just duplication. Anyway, the whole linked list is unnecessary if the iterable can be iterated over multiple times. But "tee" won't know when to do that. *That* is what I call overhead (unless of course all the tee branches are consumed in an interleaved manner).
BTW, I certainly don't suggest adding this to the itertools docs either. While it's short and elegant, it's too subtle to grasp easily - if you think "it's obvious", you haven't yet thought hard enough about the problem ;-) _______________________________________________ Python-ideas mailing list Python-ideas@python.org https://mail.python.org/mailman/listinfo/python-ideas Code of Conduct: http://python.org/psf/codeofconduct/
-- + Koos Zevenhoven + http://twitter.com/k7hoven +