Evan Schalton wrote: > Peter, > > This isn't a ram consideration as much it's a logical consideration. There > are a lot of ways to handle this, I REALLY don't want to use a package > here. Bit masking is incredibly useful for permutations/combinatoric > algorithms. I can create my own class wrapper or functions, and optimize, > but think that the __and__ operator would be a really useful syntactic > tool. It would streamline these types of operations and *hopefully* help > people write more efficient code. > > I'm less strictly interested in the & operator explicitly working with a > bit int, but think it'd be great if the was a built-in filter something > like: > > [1,2,3,4] & [0,0,1,1] => [3,4] OR > [1,2,3,4] & [False, False, True, True] = [3,4]
But, but, but... for numpy arrays this is done regularly, and the established way to spell it is >>> a = np.array([1,2,3,4]) >>> b = np.array([False, False, True, True]) >>> a[b] array([3, 4]) rather than a & b whereas for lists IMO it's a fringe application. I may be influenced by numpy's example, but I'd expect [1, 2, 3, 4] & [0, 0, 1, 1] --> [0, 0, 1, 0] i. e. bitwise per-element and. -- https://mail.python.org/mailman/listinfo/python-list
