Error handling and documentation of Path.home()

[timhoffm]

BPO	41082
Nosy	@ambv, @timhoffm, @akulakov
PRs	[3.9] bpo-41082: Add note on errors that may be raised by home() and expanduser() #28186

^{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 2021-09-08.17:28:49.896>
created_at = <Date 2020-06-22.21:13:41.917>
labels = ['type-bug', 'library', '3.9']
title = 'Error handling and documentation of Path.home()'
updated_at = <Date 2021-09-08.17:28:49.896>
user = 'https://github.com/timhoffm'

bugs.python.org fields:

activity = <Date 2021-09-08.17:28:49.896>
actor = 'lukasz.langa'
assignee = 'none'
closed = True
closed_date = <Date 2021-09-08.17:28:49.896>
closer = 'lukasz.langa'
components = ['Library (Lib)']
creation = <Date 2020-06-22.21:13:41.917>
creator = 'timhoffm'
dependencies = []
files = []
hgrepos = []
issue_num = 41082
keywords = ['patch']
message_count = 6.0
messages = ['372116', '399598', '401166', '401404', '401412', '401414']
nosy_count = 3.0
nosy_names = ['lukasz.langa', 'timhoffm', 'andrei.avk']
pr_nums = ['28186']
priority = 'normal'
resolution = 'wont fix'
stage = 'resolved'
status = 'closed'
superseder = None
type = 'behavior'
url = 'https://bugs.python.org/issue41082'
versions = ['Python 3.9']

[timhoffm]

Path.home() may fail un
(matplotlib/matplotlib#17707 (comment)).

1. I think the raised key error is too low-level, and it should be something else; what exactly t.b.d.
2. The documentation (https://docs.python.org/3/library/pathlib.html#pathlib.Path.home) should specify what happens in case of failure.
3. The documentation links to https://docs.python.org/3/library/os.path.html#os.path.expanduser, but that's not correct. _PosixFlavor.gethomedir() implements it's own user lookup, which is slightly different from os.path.expanduser()

[akulakov]

In current dev version (3.11), _PosixFlavor doesn't have gethomedir(), and instead os.path.expanduser() is used via _NormalAccessor(). So at least for the current dev version, the docs are accurate.

[akulakov]

I looked a bit more into this, it's been fixed in 3.10, in this PR: #18841

It wasn't backported to 3.9. I'm not quite sure if it's best to backport to 3.9 or fix via docs for 3.9. Just in case I've put up a small PR for such a docfix.

[ambv]

Using os.path.expanduser is a functional change so it wasn't backported at the time to 3.9 and sure shouldn't be backported now. Going with a doc update is the right approach.

[ambv]

New changeset dc2e11e by andrei kulakov in branch '3.9':
[3.9] bpo-41082: Add note on errors that may be raised by home() and expanduser() (GH-28186)
dc2e11e

[ambv]

Behavior is fixed in Python 3.10+. For Python 3.9 the docs are updated to list possible exceptions. Since this was reported with 3.9 in mind, I'm marking this one as wontfix.

Thanks for picking this up, Andrei! ✨ 🍰 ✨