Convert UTC datetime string to local datetime with Python
I've never had to convert time to and from UTC. Recently had a request to have my app be timezone aware, and I've been running myself in circles. Lots of information on converting local time to UTC, which I found fairly elementary (maybe I'm doing that wrong as well), but I can not find any information on easily converting the UTC time to the end-users timezone.
In a nutshell, and android app sends me (appengine app) data and within that data is a timestamp. To store that timestamp to utc time I am using:
datetime.utcfromtimestamp(timestamp)
That seems to be working. When my app stores the data, it is being store as 5 hours ahead (I am EST -5)
The data is being stored on appengine's BigTable, and when retrieved it comes out as a string like so:
"2011-01-21 02:37:21"
How do I convert this string to a DateTime in the users correct time zone?
Also, what is the recommended storage for a users timezone information? (How do you typically store tz info ie: "-5:00" or "EST" etc etc ?) I'm sure the answer to my first question might contain a parameter the answers the second.
If you don't want to provide your own tzinfo objects, check out the python-dateutil library. It provides tzinfo implementations on top of a zoneinfo (Olson) database such that you can refer to time zone rules by a somewhat canonical name.
from datetime import datetime
from dateutil import tz
# METHOD 1: Hardcode zones:
from_zone = tz.gettz('UTC')
to_zone = tz.gettz('America/New_York')
# METHOD 2: Auto-detect zones:
from_zone = tz.tzutc()
to_zone = tz.tzlocal()
# utc = datetime.utcnow()
utc = datetime.strptime('2011-01-21 02:37:21', '%Y-%m-%d %H:%M:%S')
# Tell the datetime object that it's in UTC time zone since
# datetime objects are 'naive' by default
utc = utc.replace(tzinfo=from_zone)
# Convert time zone
central = utc.astimezone(to_zone)
Edit Expanded example to show strptime usage
Edit 2 Fixed API usage to show better entry point method
Edit 3 Included auto-detect methods for timezones (Yarin)
Here's a resilient method that doesn't depend on any external libraries:
from datetime import datetime
import time
def datetime_from_utc_to_local(utc_datetime):
now_timestamp = time.time()
offset = datetime.fromtimestamp(now_timestamp) - datetime.utcfromtimestamp(now_timestamp)
return utc_datetime + offset
This avoids the timing issues in DelboyJay's example. And the lesser timing issues in Erik van Oosten's amendment.
As an interesting footnote, the timezone offset computed above can differ from the following seemingly equivalent expression, probably due to daylight savings rule changes:
offset = datetime.fromtimestamp(0) - datetime.utcfromtimestamp(0) # NO!
Update: This snippet has the weakness of using the UTC offset of the present time, which may differ from the UTC offset of the input datetime. See comments on this answer for another solution.
To get around the different times, grab the epoch time from the time passed in. Here's what I do:
def utc2local (utc):
epoch = time.mktime(utc.timetuple())
offset = datetime.fromtimestamp (epoch) - datetime.utcfromtimestamp (epoch)
return utc + offset
See the datetime documentation on tzinfo objects. You have to implement the timezones you want to support yourself. The are examples at the bottom of the documentation.
Here's a simple example:
from datetime import datetime,tzinfo,timedelta
class Zone(tzinfo):
def __init__(self,offset,isdst,name):
self.offset = offset
self.isdst = isdst
self.name = name
def utcoffset(self, dt):
return timedelta(hours=self.offset) + self.dst(dt)
def dst(self, dt):
return timedelta(hours=1) if self.isdst else timedelta(0)
def tzname(self,dt):
return self.name
GMT = Zone(0,False,'GMT')
EST = Zone(-5,False,'EST')
print datetime.utcnow().strftime('%m/%d/%Y %H:%M:%S %Z')
print datetime.now(GMT).strftime('%m/%d/%Y %H:%M:%S %Z')
print datetime.now(EST).strftime('%m/%d/%Y %H:%M:%S %Z')
t = datetime.strptime('2011-01-21 02:37:21','%Y-%m-%d %H:%M:%S')
t = t.replace(tzinfo=GMT)
print t
print t.astimezone(EST)
Output
01/22/2011 21:52:09
01/22/2011 21:52:09 GMT
01/22/2011 16:52:09 EST
2011-01-21 02:37:21+00:00
2011-01-20 21:37:21-05:00a