[issue9025] Non-uniformity in randrange for large arguments.

[Mark Dickinson]

Mark Dickinson <dickinsm at gmail.com> added the comment:

Either of these looks good to me.

If the last line of the second is changed from "return int(r) % n" to "return int(r) // (N // n)" then it'll use the high-order bits of random()  instead of the low-order bits.  This doesn't matter for MT, but might matter for subclasses of Random using a different underlying generator.

----------

_______________________________________
Python tracker <report at bugs.python.org>
<http://bugs.python.org/issue9025>
_______________________________________