How to check if a datetime object is localized with pytz?

I want to store a datetime object with a localized UTC timezone. The method that stores the datetime object can be given a non-localized datetime (naive) object or an object that already has been localized. How do I determine if localization is needed?

Code with missing if condition:

class MyClass:
  def set_date(self, d):
    # what do i check here?
    # if(d.tzinfo):
      self.date = d.astimezone(pytz.utc)
    # else:
      self.date = pytz.utc.localize(d)

Solution 1:

How do I determine if localization is needed?

From datetime docs:

• a datetime object d is aware iff:

  d.tzinfo is not None and d.tzinfo.utcoffset(d) is not None
  

• d is naive iff:

  d.tzinfo is None or d.tzinfo.utcoffset(d) is None
  

Though if d is a datetime object representing time in UTC timezone then you could use in both cases:

self.date = d.replace(tzinfo=pytz.utc)

It works regardless d is timezone-aware or naive.

Note: don't use datetime.replace() method with a timezone with a non-fixed utc offset (it is ok to use it with UTC timezone but otherwise you should use tz.localize() method).

Solution 2:

if you want to check if a datetime object 'd' is localized, check the d.tzinfo, if it is None, no localization.

Solution 3:

Here is a function wrapping up the top answer.

def tz_aware(dt):
    return dt.tzinfo is not None and dt.tzinfo.utcoffset(dt) is not None

Solution 4:

Here's a more complete function to convert or coerce a timestamp obj to utc. If it reaches the exception this means the timestamp is not localized. Since it's good practice to always work in UTC within the code, this function is very useful at the entry level from persistence.

def convert_or_coerce_timestamp_to_utc(timeobj):
        out = timeobj
        try:
            out = timeobj.astimezone(pytz.utc) # aware object can be in any timezone
        except (ValueError,TypeError) as exc: # naive
            out = timeobj.replace(tzinfo=pytz.utc)
        return out

The small addition from the 'try catch' in the answer by J.F. Sebastian is the additional catch condition, without which not all naive cases will be caught by the function.