修改Itertools.cycle（）

Question

I'm currently using the itertools.cycle() object, and I was wondering if there was anyway to modify the cycle after it's creation. The following:

my_cycle = itertools.cycle([1,2,3])
print my_cycle.next()
my_cycle.delete()    #function doesn't exist
print my_cycle.next()

would have an output of:

1
3

Is there any way to achieve this through itertools? Or perhaps another object? Or do I need to implement my own object to do this.

Answer 1

Itertools doesn't provide such an option. You can build it with a deque :

from collections import deque

class ModifiableCycle(object):
    def __init__(self, items=()):
        self.deque = deque(items)
    def __iter__(self):
        return self
    def __next__(self):
        if not self.deque:
            raise StopIteration
        item = self.deque.popleft()
        self.deque.append(item)
        return item
    next = __next__
    def delete_next(self):
        self.deque.popleft()
    def delete_prev(self):
        # Deletes the item just returned.
        # I suspect this will be more useful than the other method.
        self.deque.pop()

Answer 2

If you want to skip over a value you can just call next once.

If you want to permanently remove one of the values from the cycle, by value, you can create a generator.

def skip_matching(gen, value):
    for v in gen:
        if v == value:
            continue
        yield v

Then you can wrap a cycle like so:

>>> a = skip_matching(itertools.cycle([1,2,3]), 2)
>>> a.next()
1
>>> a.next()
3
>>> a.next()
1

If you want to remove the next value by position (without removing any others that match it), you have to know the cycle length n. Then you can skip one and collect n-1 to create a new cycle.

c.next()
c = itertools.cycle(c.next() for i in range(n-1))