Craig Ringer wrote:
> My first thought would be to express your 'A and B' regex as:
>
> (A.*B)|(B.*A)
>
> with whatever padding, etc, is necessary. You can even substitute in the
> sub-regex for A and B to avoid writing them out twice.
That won't work because of overlaps. Consider
barkeep
with a search for A='bark' and B='keep'.
Neither A.*B nor B.*A will match because the 'k' needs to
be in both A and B.
The OP asked for words, so consecutive letters separated
by non-letters or end of string. With that restriction
this solution will work.
Another possibility is to use positive assertions, as in
(?=A)(?=.*B)|(?=B)(?=.*A)
The best solution is to do a string.find and not worry about
implementing this as a regexp.
Andrew
[EMAIL PROTECTED]
--
http://mail.python.org/mailman/listinfo/python-list
