[issue36918] ValueError warning in test_urllib due to io.IOBase destructor

Change by STINNER Victor : -- resolution: -> fixed stage: patch review -> resolved status: open -> closed versions: +Python 3.9 ___ Python tracker ___

[issue36918] ValueError warning in test_urllib due to io.IOBase destructor

miss-islington added the comment: New changeset 9d37ae0bee25692572c201378cd0692df22fa2ac by Miss Islington (bot) in branch '3.8': bpo-36918: Fix "Exception ignored in" in test_urllib (GH-13996) https://github.com/python/cpython/commit/9d37ae0bee25692572c201378cd0692df22fa2ac --

[issue36918] ValueError warning in test_urllib due to io.IOBase destructor

Change by miss-islington : -- pull_requests: +13861 pull_request: https://github.com/python/cpython/pull/13998 ___ Python tracker ___

[issue36918] ValueError warning in test_urllib due to io.IOBase destructor

STINNER Victor added the comment: New changeset eb976e47e261760330c1bed224019b073b05e994 by Victor Stinner in branch 'master': bpo-36918: Fix "Exception ignored in" in test_urllib (GH-13996) https://github.com/python/cpython/commit/eb976e47e261760330c1bed224019b073b05e994 --

[issue36918] ValueError warning in test_urllib due to io.IOBase destructor

Change by STINNER Victor : -- pull_requests: +13859 pull_request: https://github.com/python/cpython/pull/13996 ___ Python tracker ___

[issue36918] ValueError warning in test_urllib due to io.IOBase destructor

STINNER Victor added the comment: I proposed PR 13955: a fix in 2 lines. -- ___ Python tracker ___ ___ Python-bugs-list mailing

[issue36918] ValueError warning in test_urllib due to io.IOBase destructor

STINNER Victor added the comment: Well, maybe Python GC can be enhanced. In the meanwhile, I would like to make these warnings quiet since they are very annoying when I have to analyze buildbot warnings. Moreover, this issues blocks my bpo-37069. Maybe open a separated issue to propose to

[issue36918] ValueError warning in test_urllib due to io.IOBase destructor

Change by STINNER Victor : -- pull_requests: +13823 pull_request: https://github.com/python/cpython/pull/13955 ___ Python tracker ___

[issue36918] ValueError warning in test_urllib due to io.IOBase destructor

Change by STINNER Victor : -- pull_requests: -13428 ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe:

[issue36918] ValueError warning in test_urllib due to io.IOBase destructor

Change by STINNER Victor : -- pull_requests: +13428 ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe:

[issue36918] ValueError warning in test_urllib due to io.IOBase destructor

Karthikeyan Singaravelan added the comment: My analysis was that close was called on fakesocket which is internally closed when it's counter resets to zero and the destructor was trying to flush on a closed object during destructor call. Comment from

[issue36918] ValueError warning in test_urllib due to io.IOBase destructor

Serhiy Storchaka added the comment: No, BasicIO.close() is correct. This is a bug in the garbage collector: the underlying file is closed before closing HTTPResponse. The HTTPResponse instance has a reference to the file object, the file object does not have a reference to the HTTPResponse

[issue36918] ValueError warning in test_urllib due to io.IOBase destructor

Serhiy Storchaka added the comment: I think this is a bug in BasicIO. close() should be idempotent. Calling it on the closed file should have no effect. -- components: +IO -Tests nosy: +serhiy.storchaka ___ Python tracker

[issue36918] ValueError warning in test_urllib due to io.IOBase destructor

Change by Karthikeyan Singaravelan : -- keywords: +patch pull_requests: +13236 stage: -> patch review ___ Python tracker ___ ___

[issue36918] ValueError warning in test_urllib due to io.IOBase destructor

New submission from Karthikeyan Singaravelan : Issue for https://bugs.python.org/issue18748#msg340059. comment : Is there someone interested to debug remaining "Exception ignored:" logs in test_urllib? test_invalid_redirect (test.test_urllib.urlopen_HttpTests) ... Exception ignored in: