Message77675
| Author |
mark.dickinson |
| Recipients |
fredrikj, loewis, mark.dickinson, rhettinger, terry.reedy, vstinner |
| Date |
2008年12月12日.16:50:13 |
| SpamBayes Score |
6.91449e-05 |
| Marked as misclassified |
No |
| Message-id |
<1229100615.32.0.735797361223.issue3439@psf.upfronthosting.co.za> |
| In-reply-to |
| Content |
Victor,
This looks good. If it's okay with you, I'll work on the documentation
a little (we need an entry in whatsnew, and a description of the
semantics of numbits for 0 and negative integers.) Then I think this
will be ready.
> I prefer to keep the "fast path" (use _PyLong_NumBits) as *you*
> proposed in another message: Message75767.
Sorry---I wasn't clear. I wasn't suggesting getting rid of the fast
path---I was suggesting inlining the _PyLong_NumBits code in
long_numbits (leaving _PyLong_NumBits itself intact, of course). It
would eliminate all the messing around with exceptions. But I think the
current code is fine. |
|