linux - Python Timezone Timestamp Wierdness

Question

I am finding a very strange behavior for different timezones in Python. The following code:

import datetime as dtm
import pytz


def test_tz():
    america_ny_tz: dtm.tzinfo = pytz.timezone("America/New_York")
    est_tz: dtm.tzinfo = pytz.timezone("EST5EDT")

    today = dtm.date.today()

    ny_dtm = dtm.datetime(
        year=today.year, month=today.month, day=today.day, tzinfo=america_ny_tz
    )
    est_dtm = dtm.datetime(
        year=today.year, month=today.month, day=today.day, tzinfo=est_tz
    )

    print(f"New York: {ny_dtm.timestamp()}, EST: {est_dtm.timestamp()}")


if __name__ == "__main__":
    test_tz()

outputs:

New York: 1609995360.0, EST: 1609995600.0

As you may notice the difference is about 4 minutes where one would have to assume the time should be the same.

Am I accessing Time Zone information incorrectly or is my understanding that the Time Zone information should be the same?

Running on Linux, Ubuntu 20.04 but the behavior on 18.04 was the same.

P.S. I haven't tried other languages or different OS' to see if the behavior will be the same.

Answer 1

You need to localize correctly - with pytz's timezone objects, setting tzinfo directly when creating the datetime object is not the right way. It's in the docs:

This library only supports two ways of building a localized time. The first is to use the localize() method provided by the pytz library.

and

The second way of building a localized time is by converting an existing localized time using the standard astimezone() method

---

On the other hand, if you happen to use Python 3.9+, you can use the zoneinfo module from the standard lib to make you life easier - see the docs / using-zoneinfo or example usage here.