Message196973
| Author |
ethan.furman |
| Recipients |
eli.bendersky, ethan.furman |
| Date |
2013年09月05日.02:51:02 |
| SpamBayes Score |
-1.0 |
| Marked as misclassified |
Yes |
| Message-id |
<1378349462.72.0.930877473496.issue18929@psf.upfronthosting.co.za> |
| In-reply-to |
| Content |
Part of the solution for Issue18693 is to have `inspect.classify_class_attrs()` properly consider the metaclass (or type) of the class when searching for the origination point of class attributes.
The fix is changing line 325:
- for base in (cls,) + mro:
+ for base in (cls,) + mro + (type(cls),):
or line 361:
- return cls.__mro__
+ return cls.__mro__ + (type(cls), )
Should we target previous Pythons with this fix? |
|
History
|
|---|
| Date |
User |
Action |
Args |
| 2013年09月05日 02:51:02 | ethan.furman | set | recipients:
+ ethan.furman, eli.bendersky |
| 2013年09月05日 02:51:02 | ethan.furman | set | messageid: <1378349462.72.0.930877473496.issue18929@psf.upfronthosting.co.za> |
| 2013年09月05日 02:51:02 | ethan.furman | link | issue18929 messages |
| 2013年09月05日 02:51:02 | ethan.furman | create |
|