classification
Title: Implement tzinfo.dst() method in timezone
Type: enhancement Stage: test needed
Components: Extension Modules Versions: Python 3.2
process
Status: closed Resolution: rejected
Dependencies: 5094 Superseder:
Assigned To: belopolsky Nosy List: belopolsky, brett.cannon, mark.dickinson
Priority: normal Keywords:

Created on 2010-06-16 17:58 by belopolsky, last changed 2010-10-14 18:22 by belopolsky. This issue is now closed.

Messages (2)
msg107940 - (view) Author: Alexander Belopolsky (belopolsky) * (Python committer) Date: 2010-06-16 17:58
This idea was brought in the original fixed offset timezone proposal (see issue 5094), but was met with some opposition.  See msg106914, point 2. As a result, the timezone implementation is in conflict with tzinfo documentation that says "If utcoffset() does not return None, dst() should not return None either." http://docs.python.org/dev/py3k/library/datetime.html#datetime.tzinfo.utcoffset


Note that dst() is not needed for time calculations involving fixed offset timezones because DST shift is included in utcoffset().  It is however needed for interoperability with timetuple time representation. See issue9004.

If the long term goal is to move users from using timetuples to to datetime objects, it is important for time objects to be able to store DST flag.
msg118707 - (view) Author: Alexander Belopolsky (belopolsky) * (Python committer) Date: 2010-10-14 18:22
After thinking some more about this issue, I am going to withdraw this proposal.  If we want to support storing dst flag in datetime instances, it should be stored in the datetime object itself, not in tzinfo. 

See

http://mail.python.org/pipermail/python-dev/2010-August/102842.html
History
Date User Action Args
2010-10-14 18:22:35belopolskysetstatus: open -> closed
resolution: rejected
messages: + msg118707
2010-06-16 18:00:56belopolskysetnosy: + brett.cannon, mark.dickinson
dependencies: + datetime lacks concrete tzinfo implementation for UTC
2010-06-16 17:58:30belopolskycreate