altf4arnold/datatracker
1
0
Fork 0
Code Issues Pull requests Actions 3 Packages Projects Releases Wiki Activity

fix: use UTC for model_to_timeline_data() (#4763)

Browse source 
This changes the method to operate in UTC instead of server time.
The existing code was mixing UTC and server time, though, giving
timestamps purporting to be UTC but computed from the date in the
server timezone. To the extent that it matters, I think this change
is preferable to restoring the original behavior.
This commit is contained in:
Jennifer Richards 2022-11-15 15:02:09 -04:00 committed by GitHub
parent 8caa4ce0cc
commit 3c2e004ab8
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 1 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
ietf/doc/views_stats.py
View file

@ -49,7 +49,7 @@ def model_to_timeline_data(model, field='time', **kwargs):
# This is needed for sqlite, when we're running tests: # This is needed for sqlite, when we're running tests:
if type(obj_list[0]['date']) != datetime.date: if type(obj_list[0]['date']) != datetime.date:
obj_list = [ {'date': dt(e['date']), 'count': e['count']} for e in obj_list ] obj_list = [ {'date': dt(e['date']), 'count': e['count']} for e in obj_list ]
today = date_today() today = date_today(datetime.timezone.utc)
if not obj_list[-1]['date'] == today: if not obj_list[-1]['date'] == today:
obj_list += [ {'date': today, 'count': 0} ] obj_list += [ {'date': today, 'count': 0} ]
data = [ ((e['date'].toordinal()-epochday)*1000*60*60*24, e['count']) for e in obj_list ] data = [ ((e['date'].toordinal()-epochday)*1000*60*60*24, e['count']) for e in obj_list ]