Python package for converting finite regex to a text array?

Question

While it's impossible most of the time to derive all possible text that could match regexes, I'm wondering about a specific case where you have I (like|hate) (pizza|ice cream)!

This would result in four possibilities:

I like ice cream!
I like pizza!
I hate ice cream!
I hate pizza!

Now, it's not bad to homebrew something to split this up.

import re
my_string = "I (like|hate) (ice cream|pizza)!"

def recursive_possibilities(my_list):
    print("Recursing on", my_list)
    my_new_list = []
    for m in my_list:
        checklist = re.findall(r"\([a-z \|]+\)", m)
        if len(checklist) == 0:
            my_new_list.append(m)
            continue
        x = checklist[0]
        for y in x[1:-1].split("|"):
            my_new_list.extend(recursive_possibilities([m.replace(x, y)]))
    return my_new_list

for r in recursive_possibilities([my_string]):
    print(r)

I get the feeling

1. my code could be better and
2. someone else has done it better, and before me, and put it (and other stuff I'd like later) in a module.

Any ideas where I can look for this? Thanks!