How to find all possibly overlapping matches?

[MRAB]

kj wrote:
> 
> re.findall finds all non-overlapping matches, but what if one wants
> all (maximal) matches, even those that overlap?
> 
> All the solutions I can come up involve calling re.search iteratively,
> each time giving it a pos parameter starting just after the start
> of the previous match.
> 
> Is there a built-in solution to such a task?
> 
Not in the re module.

It has been requested and is in my regex implementation at
http://bugs.python.org/issue2636 if you want to try that.