Python: All possible Timezone Abbreviations for given Timezone Name (and vise versa)

Question

Using pytz, I know how to get a listing a Timezone names, but I would like to get all possible Timezone abbreviations for each Timezone name:

import pytz
list(pytz.common_timezones)
['Africa/Abidjan', 'Africa/Accra', 'Africa/Addis_Ababa',...]

What I am looking for is given any Timezone abbreviation, example PST or PDT, ignoring current datetime (e.g. now), return the all possible Timezone name, in this case a list that would include America/Los_Angeles.

Thanks

Answer 1

Since you wish to ignore the current datetime, it sounds like you want to find any timezone which ever used the given abbreviation at any time in the past. That information is in the Olson database and accessible through pytz. However, pytz stores this information in the private attribute, tzone._transition_info:

import collections
import datetime as DT
import pytz

tzones = collections.defaultdict(set)
abbrevs = collections.defaultdict(set)

for name in pytz.all_timezones:
    tzone = pytz.timezone(name)
    for utcoffset, dstoffset, tzabbrev in getattr(
            tzone, '_transition_info', [[None, None, DT.datetime.now(tzone).tzname()]]):
        tzones[tzabbrev].add(name)
        abbrevs[name].add(tzabbrev)

The reason for the third (default) argument to gettattr is to handle a few timezones, such as Africa/Bujumbura, which never had any transitions. So the abbreviation in these cases is the current abbreviation.

---

In [94]: tzones['PST']
Out[94]: 
{'America/Bahia_Banderas',
 'America/Boise',
 'America/Creston',
 'America/Dawson',
 'America/Dawson_Creek',
 'America/Ensenada',
 'America/Hermosillo',
 'America/Inuvik',
 'America/Juneau',
 'America/Los_Angeles',
 'America/Mazatlan',
 'America/Metlakatla',
 'America/Santa_Isabel',
 'America/Sitka',
 'America/Tijuana',
 'America/Vancouver',
 'America/Whitehorse',
 'Canada/Pacific',
 'Canada/Yukon',
 'Mexico/BajaNorte',
 'Mexico/BajaSur',
 'PST8PDT',
 'Pacific/Pitcairn',
 'US/Pacific',
 'US/Pacific-New'}

---

In [95]: tzones['PDT']
Out[95]: 
{'America/Boise',
 'America/Dawson',
 'America/Dawson_Creek',
 'America/Ensenada',
 'America/Juneau',
 'America/Los_Angeles',
 'America/Metlakatla',
 'America/Santa_Isabel',
 'America/Sitka',
 'America/Tijuana',
 'America/Vancouver',
 'America/Whitehorse',
 'Canada/Pacific',
 'Canada/Yukon',
 'Mexico/BajaNorte',
 'PST8PDT',
 'US/Pacific',
 'US/Pacific-New'}

---

In [97]: abbrevs['America/Los_Angeles']
Out[97]: {'LMT', 'PDT', 'PPT', 'PST', 'PWT'}

---

As Paul points out, note that timezone abbreviations are ambiguous -- they do not necessarily map to timezones with the same utcoffset. For example, both Asia/Shanghai and US/Central use the CST timezone abbreviation.

In [242]: 'Asia/Shanghai' in tzones['CST']
Out[242]: True

In [243]: 'US/Central' in tzones['CST']
Out[243]: True