In a message dated 10/22/2007 8:46:59 PM Central Daylight Time, [EMAIL PROTECTED] writes:
> > Still, why do you want None instead of raisng an exception > > (as is the case in other factorial implementations)? > > A null value is as good/bad as raising an exception in my book. > Since you can't do math on a None object, any attempt to do so > will raise an exception: > > >>> 42 + fact(-1) Fair enough. > > I generally prefer my functions to return semi-sensible results > (in this case, None makes sense to me, as there isn't really a > definition of "negative-one factorial"). It also fits in my head > alongside my SQL where NULL values/expressions can be returned > and evaluated without the whole query falling over. Ok, but in SQL, 42 + fact(-1) wouldn't raise an exception, it would simply return a NULL if fact(-1) returned None. I suppose the only reason to not use None is for consistent behavior. If all the functions from the gmpy library raise exceptions on error, I think I would want mine to do the same. > > I suppose if you really wanted to throw an exception using this > lambda craziness, you could wrap the whole result in "0 + > ([body])" which, if the body returned Null, would push up > exception daisies (with slightly misleading exception information). > > -tkc ************************************** See what's new at http://www.aol.com
-- http://mail.python.org/mailman/listinfo/python-list
