New submission from Russel Webber :
The STDLIB_LANDMARKS in Modules/getpath,py refers to os.pyc in an old location:
STDLIB_LANDMARKS = [f'{STDLIB_SUBDIR}/os.py', f'{STDLIB_SUBDIR}/os.pyc']
Since https://www.python.org/dev/peps/pep-3147 the .pyc files are in
__pycache__ folders
Change by Russel Webber :
--
keywords: +patch
pull_requests: +29265
stage: needs patch -> patch review
pull_request: https://github.com/python/cpython/pull/31082
___
Python tracker
<https://bugs.python.org/issu
Russel Webber added the comment:
I agree wholeheartedly, an explantation of sys.path initialisation would be
very helpful. I've seen too many hacks in end user Python code due to
misunderstandings around sys.path initialisation and the available ways to
customise sys.path. I'll work
Russel Webber added the comment:
I am happy to take a look at improving the documentation. I see the new
getpath.py has very extensive commenting. Should I simply mention ._pth files
as in the original issue from 5 years ago, or are you looking for documentation
of how sys.path
Russel Webber added the comment:
I have added a test that I think covers this case.
--
keywords: +patch
message_count: 1.0 -> 2.0
nosy: +RusselWebber
nosy_count: 3.0 -> 4.0
pull_requests: +28893
stage: -> patch review
pull_request: https://github.com/python/cpython/p
Change by Russel Webber :
--
keywords: +patch
pull_requests: +28819
stage: -> patch review
pull_request: https://github.com/python/cpython/pull/30616
___
Python tracker
<https://bugs.python.org/issu
New submission from Russel Webber :
A few lines of the functools.total_ordering implementation are not covered by
the tests.
Specifically, coverage.py shows the returns of NotImplemented:
...
if op_result is NotImplemented:
return op_result
...
as not covered in the functions
