[issue29327] SystemError or crash in sorted(iterable=[])

2017-03-31 Thread Donald Stufft
Changes by Donald Stufft : -- pull_requests: +895 ___ Python tracker ___ ___

[issue29327] SystemError or crash in sorted(iterable=[])

2017-01-20 Thread STINNER Victor
Changes by STINNER Victor : -- title: SystemError or crash in sorted(iterable= -> SystemError or crash in sorted(iterable=[]) ___ Python tracker

[issue29327] SystemError or crash in sorted(iterable=

2017-01-20 Thread Serhiy Storchaka
Serhiy Storchaka added the comment: Thank you for your reviews Victor and Raymond. As for more general Raymond comments, I agreed with many of them in principle, but this isn't directly related to this issue. Victor opened a topic on the python-committers mailing list:

[issue29327] SystemError or crash in sorted(iterable=

2017-01-20 Thread STINNER Victor
STINNER Victor added the comment: Raymond Hettinger: "A few random thoughts that may or may not be helpful: (...)" I replied on the python-committers mailing list: https://mail.python.org/pipermail/python-committers/2017-January/004129.html I'm not sure that this specific issue is the best

[issue29327] SystemError or crash in sorted(iterable=

2017-01-19 Thread STINNER Victor
STINNER Victor added the comment: "The patch uses new feature of 3.6 -- supporting positional-only parameters in PyArg_ParseTupleAndKeywords() (see issue26282)." Oh, I didn't recall that it's a new feature. It's a nice feature by the way, thanks for implementing it :-) In that case, yeah it's

[issue29327] SystemError or crash in sorted(iterable=[])

2017-01-19 Thread Roundup Robot
Roundup Robot added the comment: New changeset 1827b64cfce8 by Serhiy Storchaka in branch '3.6': Issue #29327: Fixed a crash when pass the iterable keyword argument to sorted(). https://hg.python.org/cpython/rev/1827b64cfce8 New changeset f44f44b14dfc by Serhiy Storchaka in branch 'default':

[issue29327] SystemError or crash in sorted(iterable=[])

2017-01-19 Thread Serhiy Storchaka
Serhiy Storchaka added the comment: The title is spoiled when reply by email. I suppose [] has special meaning in email subject. -- title: SystemError or crash in sorted(iterable= -> SystemError or crash in sorted(iterable=[]) ___ Python tracker

[issue29327] SystemError or crash in sorted(iterable=

2017-01-19 Thread Serhiy Storchaka
Serhiy Storchaka added the comment: > Serhiy, this patch passes tests and looks fine. I think you can go ahead > with this patch. Right now I'm building 3.6 with applied patch before final testing and committing. My netbook is very slow. :( -- title: SystemError or crash in

[issue29327] SystemError or crash in sorted(iterable=[])

2017-01-19 Thread Raymond Hettinger
Raymond Hettinger added the comment: Serhiy, this patch passes tests and looks fine. I think you can go ahead with this patch. -- assignee: -> serhiy.storchaka ___ Python tracker

[issue29327] SystemError or crash in sorted(iterable=[])

2017-01-19 Thread Serhiy Storchaka
Changes by Serhiy Storchaka : -- title: SystemError or crash in sorted(iterable= -> SystemError or crash in sorted(iterable=[]) ___ Python tracker

[issue29327] SystemError or crash in sorted(iterable=

2017-01-19 Thread Serhiy Storchaka
Serhiy Storchaka added the comment: > While Python 3.5 doesn't crash, I consider that it has also the bug. So I > added Python 3.5 in Versions. The patch uses new feature of 3.6 -- supporting positional-only parameters in PyArg_ParseTupleAndKeywords() (see issue26282). Since passing an

[issue29327] SystemError or crash in sorted(iterable=[])

2017-01-19 Thread Raymond Hettinger
Raymond Hettinger added the comment: A few random thoughts that may or may not be helpful: * We now have two seasoned developers and one new core developer that collectively are creating many non-trivial patches to core parts of Python at an unprecedented rate of change. The patches are

[issue29327] SystemError or crash in sorted(iterable=[])

2017-01-19 Thread STINNER Victor
STINNER Victor added the comment: Raymond: "Please be careful with all of these AC changes. They need to have tests. They need to not change APIs. They need to not introduce bugs." The bug was not introduced by an Argument Clinic change. I agree that switching to AC must not change the

[issue29327] SystemError or crash in sorted(iterable=[])

2017-01-19 Thread STINNER Victor
STINNER Victor added the comment: While Python 3.5 doesn't crash, I consider that it has also the bug. So I added Python 3.5 in Versions. sorted.diff LGTM. And thank you for the unit test! -- versions: +Python 3.5 ___ Python tracker

[issue29327] SystemError or crash in sorted(iterable=[])

2017-01-19 Thread STINNER Victor
STINNER Victor added the comment: Oh, this issue is very subtle. Since the list.sorted() class method became a builtin sorted() method (Python 2.4.0, change c06b570adf12 by Raymond Hettinger), the sorted() function accepts an iterable as a keyword argument, whereas list.sort() doesn't.

[issue29327] SystemError or crash in sorted(iterable=[])

2017-01-19 Thread Raymond Hettinger
Raymond Hettinger added the comment: Please be careful with all of these AC changes. They need to have tests. They need to not change APIs. They need to not introduce bugs. The whole effort seems to be being pushed forward in a cavalier and aggressive manner. In this case, there was an

[issue29327] SystemError or crash in sorted(iterable=[])

2017-01-19 Thread Serhiy Storchaka
Serhiy Storchaka added the comment: Proposed patch fixes the bug. -- keywords: +patch nosy: +haypo stage: -> patch review title: SystemError in sorted(iterable=[]) -> SystemError or crash in sorted(iterable=[]) Added file: http://bugs.python.org/file46344/sorted.diff