I see, I didn't think about all the documentations to update, and i should have as I've got the same problem in my project :).
I think the reason is that there isn't enough need for it. The special case of s.rsplit(c, 1) can be coded so easily by using rfind() that I don't see the need to add it. Our Swiss Army Knife string type is beginning to be so loaded with features that I am reluctant to add more. The cost of a new feature these days is measured in the number of books that need to be updated, not the number of lines of code needed to implement it.
For your amusement only (! :-), I offer this implementation of rsplit(), which works in Python 2.3:
def rsplit(string, sep, count=-1): L = [part[::-1] for part in string[::-1].split(sep[::-1], count)] L.reverse() return L Didn't thought about this one, tricky and amusing.
-- Boris Boutillier - Boris.Boutillier@arteris.net