Option to generate strings for case insensitive matches

[rutsky]

Maybe you will find it a nice and simple to implement feature: to be able to iterate over all strings that matches regular expression ignoring letters case.
E.g. "a{1,2}" will generate "aa", "aA", "Aa", "AA".

P.S. Thanks for this library, it helped me a lot!

[asciimoo]

Yes, it would be a great feature.
But it isn't really easy to implement. Need to permute the case of all letters of constant strings in the regex, because all of "abc", "aBc", "abC", "ABc", "AbC", "aBC", "ABC" matched by "abc" case-insensitively
[aA][bB][cC] produces the same result set and works with the current version

[rutsky]

How about following algorithm?

1. translate regular expression to lower case:

[Aab]{1,2} -> [aab]{1,2}

2. remove redundancies (not sure how they treated in exrex):

   [aab]{1,2} -> [ab]{1,2}

3. generate strings using exrex

4. for each generated string s generate all strings s_i with permuted cases, such as that s_i.lower() == s:

def gen_case_strings(s):
   if s:
       head, tail = s[:-1], s[-1]
       for head_cased in gen_case_strings(head):
           yield head_cased + tail.lower()
           yield head_cased + tail.upper()
   else:
       yield s

So that:

>>> list(gen_case_strings("abc"))
['abc', 'abC', 'aBc', 'aBC', 'Abc', 'AbC', 'ABc', 'ABC']