Compare vobject VTIMEZONEs for pytz and dateutil.tz timezones

compare_vtimezones.py

# Detect differences between pytz and dateutil.tz.gettz,
# by comparing the vobject VTIMEZONEs for each.
#
# (Might actually be finding problems in vobject's
# timezone rule detection.)
#
# Requirements::
#    pip install pytz dateutil vobject

import dateutil.tz
import os
import pytz

from vobject.icalendar import TimezoneComponent, __tzidMap

OUTPUT_DIR = '/tmp/vtimezone'  # set to None to skip writing files
#OUTPUT_DIR = None


def vtimezones_match(tzname, output_dir=None):
    """Returns True if pytz and dateutil have same VTIMEZONE for tzname.

    Set output_dir to write mismatched VTIMEZONE contents to files,
    allowing easy directory diff. (Will create one file in output_dir/pytz,
    and a corresponding file in output_dir/dateutil.)
    """
    pytz_tzinfo = pytz.timezone(tzname)
    dateutil_tzinfo = dateutil.tz.gettz(tzname)

    # ignore differences in TZID field (see TimezoneComponent.pickTzid)
    pytz_tzinfo.tzid = dateutil_tzinfo.tzid = tzname

    __tzidMap.clear()  # reset tzid registry to avoid caching
    pytz_vtimezone = TimezoneComponent(tzinfo=pytz_tzinfo)
    pytz_serialized = pytz_vtimezone.serialize()

    __tzidMap.clear()  # reset tzid registry to avoid caching
    dateutil_vtimezone = TimezoneComponent(tzinfo=dateutil_tzinfo)
    dateutil_serialized = dateutil_vtimezone.serialize()

    matches = (pytz_serialized == dateutil_serialized)
    if output_dir and not matches:
        # Relies on tzname using slash separators
        write_file(
            os.path.join(output_dir, 'pytz', tzname),
            pytz_serialized)
        write_file(
            os.path.join(output_dir, 'dateutil', tzname),
            dateutil_serialized)
    return matches


def write_file(path, text):
    """Writes text to path, creating intermediate dirs if needed"""
    dirname = os.path.dirname(path)
    if not os.path.exists(dirname):
        os.makedirs(dirname)
    with open(path, 'w') as f:
        f.write(text.replace('\r\n', '\n'))


def get_dateutil_tzdb_version():
    """Return the version of the tzdb files used by dateutil.tz.gettz"""
    # From the contents of the +VERSION file in the tzdb directory
    utc = dateutil.tz.gettz('UTC')
    tzdb_dir = os.path.dirname(utc._filename)
    version_file = os.path.join(tzdb_dir, '+VERSION')
    with open(version_file) as f:
        version = f.read()
    return version.strip()


if __name__ == '__main__':
    dateutil_version = "%s; tzdb %s" % (dateutil.__version__,
                                        get_dateutil_tzdb_version())
    print("pytz version: %s" % pytz.__version__)
    print("dateutil version: %s" % dateutil_version)

    if OUTPUT_DIR:
        write_file(os.path.join(OUTPUT_DIR, 'pytz/VERSION'), pytz.__version__)
        write_file(os.path.join(OUTPUT_DIR, 'dateutil/VERSION'), dateutil_version)

    print("\nDiffering VTIMEZONEs:")
    for tzname in pytz.all_timezones:
        if not vtimezones_match(tzname, output_dir=OUTPUT_DIR):
            print(tzname)

output.txt

pytz version: 2016.4
dateutil version: 2.5.3; tzdb 2016d

Differing VTIMEZONEs:
America/Argentina/Catamarca
America/Argentina/ComodRivadavia
America/Argentina/La_Rioja
America/Argentina/Mendoza
America/Argentina/Rio_Gallegos
America/Argentina/San_Juan
America/Argentina/San_Luis
America/Argentina/Tucuman
America/Argentina/Ushuaia
America/Cambridge_Bay
America/Catamarca
America/Indiana/Knox
America/Indiana/Petersburg
America/Indiana/Tell_City
America/Indiana/Vincennes
America/Indiana/Winamac
America/Knox_IN
America/Mendoza
America/Rankin_Inlet
America/Resolute
Asia/Anadyr
Asia/Kamchatka
Asia/Khandyga
Asia/Novokuznetsk
Asia/Tbilisi
Asia/Tomsk
Europe/Samara
US/Indiana-Starke