[issue39720] Signature.bind TypeErrors could be more helpful
New submission from Frazer McLean :
Signature.bind does not tell you if a missing argument is keyword-only for
example. I created the following snippet to examine the differences:
import inspect
def run(f):
try:
f()
except TypeError as exc:
print(exc.args[0])
else:
raise RuntimeError('Expected to raise!')
sig = inspect.signature(f)
try:
sig.bind()
except TypeError as exc:
print(exc.args[0])
else:
raise RuntimeError('Expected to raise!')
print()
@run
def f1(pos_only, /): ...
@run
def f2(pos_or_kw): ...
@run
def f3(*, kw_only): ...
Output on current 3.9 master:
f1() missing 1 required positional argument: 'pos_only'
missing a required argument: 'pos_only'
f2() missing 1 required positional argument: 'pos_or_kw'
missing a required argument: 'pos_or_kw'
f3() missing 1 required keyword-only argument: 'kw_only'
missing a required argument: 'kw_only'
I am willing to create a PR so that the TypeError for f3 says "missing a
required keyword-only argument: 'kw_only'", if this would be accepted.
--
components: Library (Lib)
messages: 362439
nosy: RazerM
priority: normal
severity: normal
status: open
title: Signature.bind TypeErrors could be more helpful
type: enhancement
___
Python tracker
<https://bugs.python.org/issue39720>
___
___
Python-bugs-list mailing list
Unsubscribe:
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue30814] Import dotted name as alias breaks with concurrency
Changes by Frazer McLean <fra...@frazermclean.co.uk>: -- nosy: +RazerM ___ Python tracker <rep...@bugs.python.org> <http://bugs.python.org/issue30814> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue25538] Traceback from __exit__ method is misleading
Changes by Frazer McLean <fra...@frazermclean.co.uk>: -- nosy: +RazerM ___ Python tracker <rep...@bugs.python.org> <http://bugs.python.org/issue25538> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue29922] error message when __aexit__ is not async
Changes by Frazer McLean <fra...@frazermclean.co.uk>: -- nosy: +RazerM ___ Python tracker <rep...@bugs.python.org> <http://bugs.python.org/issue29922> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue29392] msvcrt.locking crashes python
Frazer McLean added the comment: I ran it on AppVeyor to double check, the crash appears from 3.5 onwards as I saw on my computer. https://ci.appveyor.com/project/RazerM/issue29392/build/3 -- ___ Python tracker <rep...@bugs.python.org> <http://bugs.python.org/issue29392> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue29392] msvcrt.locking crashes python
New submission from Frazer McLean:
On 3.5.2 and 3.5.3, but not 3.4.3 or 2.7.10 (Windows 64-bit), the third line
causes Python to crash with exit code 0xC417.
import msvcrt
fp = open('file.txt', 'w')
msvcrt.locking(fp.fileno(), msvcrt.LK_NBLCK, -1)
print('End')
I came across this in the portalocker module. I don't think -1 is a valid third
argument, but it doesn't raise an exception on Python 3.4 or 2.7.
--
components: Windows
messages: 286495
nosy: RazerM, paul.moore, steve.dower, tim.golden, zach.ware
priority: normal
severity: normal
status: open
title: msvcrt.locking crashes python
type: crash
versions: Python 3.5
___
Python tracker <rep...@bugs.python.org>
<http://bugs.python.org/issue29392>
___
___
Python-bugs-list mailing list
Unsubscribe:
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue7434] general pprint rewrite
Changes by Frazer McLean <fra...@frazermclean.co.uk>: -- nosy: +RazerM ___ Python tracker <rep...@bugs.python.org> <http://bugs.python.org/issue7434> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue26423] __len__() returns 32 bit int on windows leading to overflows
Changes by Frazer McLean <fra...@frazermclean.co.uk>: -- nosy: +RazerM ___ Python tracker <rep...@bugs.python.org> <http://bugs.python.org/issue26423> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue25322] contextlib.suppress not tested for nested usage
New submission from Frazer McLean: In Lib/test/test_contextlib.py, there is a bug in the nested usage part of the following function: def test_cm_is_reentrant(self): ignore_exceptions = suppress(Exception) with ignore_exceptions: pass with ignore_exceptions: len(5) with ignore_exceptions: 1/0 with ignore_exceptions: # Check nested usage len(5) Specifically, the final 2 lines aren't reached since the exception raised by 1/0 exits the context manager. -- components: Tests messages: 252382 nosy: RazerM, ncoghlan, yselivanov priority: normal severity: normal status: open title: contextlib.suppress not tested for nested usage type: behavior versions: Python 3.4, Python 3.5 ___ Python tracker <rep...@bugs.python.org> <http://bugs.python.org/issue25322> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue25322] contextlib.suppress not tested for nested usage
Frazer McLean added the comment: I may not have been clear—I understand how context managers work with exceptions. I'm just pointing out that the nested context manager isn't called, so those final two lines aren't testing anything, right? -- ___ Python tracker <rep...@bugs.python.org> <http://bugs.python.org/issue25322> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
