Deprecate isdst argument in email.utils.localtime

[abalkin]

BPO	28159
Nosy	@abalkin, @bitdancer, @aw1231, @Mariatta

^{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 = None
created_at = <Date 2016-09-14.21:55:13.766>
labels = ['easy', 'library', '3.11']
title = 'Deprecate isdst argument in email.utils.localtime'
updated_at = <Date 2022-02-08.16:46:38.005>
user = 'https://github.com/abalkin'

bugs.python.org fields:

activity = <Date 2022-02-08.16:46:38.005>
actor = 'Alan.Williams'
assignee = 'none'
closed = False
closed_date = None
closer = None
components = ['Library (Lib)']
creation = <Date 2016-09-14.21:55:13.766>
creator = 'belopolsky'
dependencies = []
files = []
hgrepos = []
issue_num = 28159
keywords = ['easy']
message_count = 7.0
messages = ['276492', '281078', '281125', '281127', '281139', '281161', '412846']
nosy_count = 4.0
nosy_names = ['belopolsky', 'r.david.murray', 'Alan.Williams', 'Mariatta']
pr_nums = []
priority = 'normal'
resolution = None
stage = None
status = 'open'
superseder = None
type = None
url = 'https://bugs.python.org/issue28159'
versions = ['Python 3.11']

[abalkin]

With the implementation of PEP-495, email.utils.localtime() does not need the isdst argument to disambiguate naive time. In fact the utility method itself is redundant given that astimezone() now works for naive instances.

[Mariatta]

Hi Alexander,

Is the idea here to raise deprecation warning starting in 3.6 and remove it by 3.8?

Thanks :)

[bitdancer]

At this point I think the deprecation warning will go into 3.7, but we should be able to remove it in 3.8. I think I'll probably keep the 'localtime' method and have it delegate it to astimezone, because frankly I would never be able to remember that that was the datetime method I needed to get localtime.

[Mariatta]

Thanks David :)
So, I found that isdst was never added as method signature in the docs, but it is mentioned in the paragraph.
https://docs.python.org/3.7/library/email.util.html?#email.utils.localtime

The intention is still just to deprecate isdst argument, right? not the localtime().

I can add a sentence in the docs indicating that we favor datetime.astimezone() instead of email.utils.localtime(), if that works for you. What do you think about this?

[bitdancer]

No, localtime does something subtly different, in that it will accept a naive datetime add the local timezone to it. Calling localtime() is also more convenient/readable than calling datetime.now().astimezone().

[abalkin]

Keeping localtime as a convenience function in email.util is fine, but isdst argument should be eliminated at some point. There is a fundamental problem with disambiguating fold times with isdst: some folds do not involve the change in dst or happen in the regions that do not have a dst regime at all. PEP-495 covers this in all the gory details.

[aw1231]

Hi, I'd like to work on this issue. Based on the discussion, the main thing to do here is to raise a deprecation warning when isdst is used?