traceback (and threading) drops exception message

[vadmium]

BPO	27348
Nosy	@vstinner, @rbtcollins, @vadmium
PRs	[Do Not Merge] Convert Misc/NEWS so that it is managed by towncrier #552
Files	none-message.patch

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = None
closed_at = <Date 2016-09-22.13:13:08.679>
created_at = <Date 2016-06-18.14:57:36.562>
labels = ['3.7', 'type-bug', 'library']
title = 'traceback (and threading) drops exception message'
updated_at = <Date 2017-03-31.16:36:33.428>
user = 'https://github.com/vadmium'

bugs.python.org fields:

activity = <Date 2017-03-31.16:36:33.428>
actor = 'dstufft'
assignee = 'none'
closed = True
closed_date = <Date 2016-09-22.13:13:08.679>
closer = 'martin.panter'
components = ['Library (Lib)']
creation = <Date 2016-06-18.14:57:36.562>
creator = 'martin.panter'
dependencies = []
files = ['43577']
hgrepos = []
issue_num = 27348
keywords = ['patch', '3.5regression']
message_count = 11.0
messages = ['268814', '269465', '269469', '269471', '269472', '269473', '272494', '276664', '276679', '276683', '277216']
nosy_count = 4.0
nosy_names = ['vstinner', 'rbcollins', 'python-dev', 'martin.panter']
pr_nums = ['552']
priority = 'normal'
resolution = 'fixed'
stage = 'resolved'
status = 'closed'
superseder = None
type = 'behavior'
url = 'https://bugs.python.org/issue27348'
versions = ['Python 3.5', 'Python 3.6', 'Python 3.7']

[vadmium]

If the exception argument is None or repr(None), it is omitted from the report when a thread raises an unhandled exception:

>>> def raise_exception(e):
...     raise e
... 
>>> t = Thread(target=raise_exception, args=(Exception(None),)); t.start(); t.join()
Exception in thread Thread-1:
Traceback (most recent call last):
  File "/home/proj/python/cpython/Lib/threading.py", line 916, in _bootstrap_inner
    self.run()
  File "/home/proj/python/cpython/Lib/threading.py", line 864, in run
    self._target(*self._args, **self._kwargs)
  File "<stdin>", line 2, in raise_exception
Exception

>>> t = Thread(target=raise_exception, args=(Exception("None"),)); t.start(); t.join()
Exception in thread Thread-2:
Traceback (most recent call last):
  File "/home/proj/python/cpython/Lib/threading.py", line 916, in _bootstrap_inner
    self.run()
  File "/home/proj/python/cpython/Lib/threading.py", line 864, in run
    self._target(*self._args, **self._kwargs)
  File "<stdin>", line 2, in raise_exception
Exception

Compare the result with other exception messages, the normal sys.excepthook() report, and Python 2:

>>> t = Thread(target=raise_exception, args=(Exception("NONE"),)); t.start(); t.join()
Exception in thread Thread-3:
Traceback (most recent call last):
  File "/home/proj/python/cpython/Lib/threading.py", line 916, in _bootstrap_inner
    self.run()
  File "/home/proj/python/cpython/Lib/threading.py", line 864, in run
    self._target(*self._args, **self._kwargs)
  File "<stdin>", line 2, in raise_exception
Exception: NONE

>>> raise Exception(None)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
Exception: None

[vadmium]

I traced this back to revision 73afda5a4e4c (bpo-17911), which includes this change to traceback._format_final_exc_line():

-if value is None or not valuestr:
+if value == 'None' or value is None or not valuestr:

[vadmium]

If I remove the value == 'None' check, there are two failures in the test suite:

1. test_decimal fails, but only because the test was changed in revision 5f3dd0a2b1ab to accommodate the very bug I am complaining about. IMO this was a case of “fixing” the test instead of fixing the bug, so it is okay to restore the test.

2. The following test, added in revision ecaafc32c500:

def test_without_exception(self):
    err = traceback.format_exception_only(None, None)
    self.assertEqual(err, ['None\n'])

It was apparently added so that print_exc() would output “None” when called with no exception set. This is the case for Python 2. However in the Python 3 versions I have tried, print_exc() either raises AttributeError or prints “NoneType”. In any case, the test does not seem to be testing anything valid according to the documentation, so I propose to remove it.

[rbtcollins]

hmm, can you give me a change to page this in? I'm pretty sure I saw breakage in external libraries prompting me to add the test and fix. I'd rather not recause that.

[vstinner]

I agree that ignoring the exception message if str(exc) == 'None' is wrong.

[vadmium]

Yeah sure I can give you a chance to consider this (I assume you meant “chance” :). But neither of the tests I mentioned were added by you as far as I know. Maybe you are thinking of some other test.

[vadmium]

Have you had any luck reviewing this Robert?

[vadmium]

I plan to commit this soon, in time for the next release.

[vstinner]

"Issue bpo-27348: In the traceback module, restore the formatting of exception messages like "Exception: None". This fixes a regression introduced in 3.5a2."

Humn, if it is described as a regression, it means that it's a bug no? You plan to push the change into Python 3.5, 3.6 and 3.7, right?

none-message.patch: LGTM.

[vadmium]

Yes, a bug fix for 3.5+.

[python-dev]

New changeset 5859a9e8b214 by Martin Panter in branch '3.5':
Issue bpo-27348: Restore “Exception: None” formatting in traceback module
https://hg.python.org/cpython/rev/5859a9e8b214

New changeset d1455d14accd by Martin Panter in branch '3.6':
Issue bpo-27348: Merge exception formatting fix from 3.5 into 3.6
https://hg.python.org/cpython/rev/d1455d14accd

New changeset 4261ae29d3e2 by Martin Panter in branch 'default':
Issue bpo-27348: Merge exception formatting fix from 3.6
https://hg.python.org/cpython/rev/4261ae29d3e2