BUG: pytz.AmbiguousTimeError not caught in hypothesis `test_on_offset_implementations`

[jackzyliu]

• I have checked that this issue has not already been reported.

• I have confirmed this bug exists on the latest version of pandas.

• (optional) I have confirmed this bug exists on the master branch of pandas.

---

Note: Please read this guide detailing how to provide the necessary information for us to reproduce your bug.

Code Sample, a copy-pastable example

(Hypothesis) Test name: test_on_offset_implementations
Test location: pandas/tests/tseries/offsets/test_offsets_properties.py

The following pair of test inputs at some point came up as failed for the hypothesis test in the checks for #41488, with pytz.AmbiguousTimeError.

dt = datetime.datetime(1900, 1, 1, 0, 0, tzinfo=pytz.timezone('Africa/Kinshasa'))
offset = pd.offsets.MonthBegin(66)

I didn't get the blob for @reproduce_failure, but we can similarly reproduce with @example as follows. Please let me know if there is less messy, more copy-and-paste-able way to reproduce the test failure.

from hypothesis import example
import datetime
dt = datetime.datetime(1900, 1, 1, 0, 0, tzinfo=pytz.timezone('Africa/Kinshasa'))
offset = MonthBegin(66)

@example(dt, offset)
# ------ prepend the code snippet above to reproduce test failure ------
@pytest.mark.arm_slow
@given(gen_random_datetime, gen_yqm_offset)
def test_on_offset_implementations(dt, offset):
    assume(not offset.normalize)
    # check that the class-specific implementations of is_on_offset match
    # the general case definition:
    #   (dt + offset) - offset == dt
    try:
        compare = (dt + offset) - offset
    except pytz.NonExistentTimeError:
        # dt + offset does not exist, assume(False) to indicate
        #  to hypothesis that this is not a valid test case
        assume(False)

    assert offset.is_on_offset(dt) == (compare == dt)

Expected Output

dt + offset yields pd.Timestamp('1905-07-01', tzinfo=pytz.timezone('Africa/Kinshasa')), which I think is correctly an DST-ambiguous timestamp given the following pytz specifications. So we should probably catch the exception and directly assume(False).

In[16]: import pytz
   ...: print(pytz.timezone('Africa/Kinshasa')._utc_transition_times[1])
   ...: print(pytz.timezone('Africa/Kinshasa')._transition_info[1])
   ...: 
1905-06-30 23:46:25
(datetime.timedelta(0), datetime.timedelta(0), 'GMT')

Actual Output

============================================== FAILURES ==============================================
___________________________________ test_on_offset_implementations ___________________________________

    @example(dt, offset)
>   # ------ prepend code snippet to reproduce test failure ------
    @pytest.mark.arm_slow
    @given(gen_random_datetime, gen_yqm_offset)

pandas/tests/tseries/offsets/test_offsets_properties.py:99: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
pandas/tests/tseries/offsets/test_offsets_properties.py:108: in test_on_offset_implementations
    compare = (dt + offset) - offset
pandas/_libs/tslibs/offsets.pyx:430: in pandas._libs.tslibs.offsets.BaseOffset.__add__
    return other.__add__(self)
pandas/_libs/tslibs/offsets.pyx:432: in pandas._libs.tslibs.offsets.BaseOffset.__add__
    return self.apply(other)
pandas/_libs/tslibs/offsets.pyx:178: in pandas._libs.tslibs.offsets.apply_wraps.wrapper
    result = result.tz_localize(tz)
pandas/_libs/tslibs/timestamps.pyx:1705: in pandas._libs.tslibs.timestamps.Timestamp.tz_localize
    value = tz_localize_to_utc_single(self.value, tz,
pandas/_libs/tslibs/tzconversion.pyx:74: in pandas._libs.tslibs.tzconversion.tz_localize_to_utc_single
    return tz_localize_to_utc(
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

>   raise pytz.AmbiguousTimeError(
E   pytz.exceptions.AmbiguousTimeError: Cannot infer dst time from 1905-07-01 00:00:00, try using the 'ambiguous' argument

pandas/_libs/tslibs/tzconversion.pyx:284: AmbiguousTimeError
--------------------------------------------- Hypothesis ---------------------------------------------
Falsifying explicit example: test_on_offset_implementations(
    dt=datetime.datetime(1900, 1, 1, 0, 0, tzinfo=<DstTzInfo 'Africa/Kinshasa' LMT+0:14:00 STD>),
    offset=<66 * MonthBegins>,
)

Problem description

See above.

Output of pd.show_versions()

INSTALLED VERSIONS
------------------
commit           : 4d1019487a73d3e1791e1bace774522307ec2bbc
python           : 3.8.8.final.0
python-bits      : 64
OS               : Linux
OS-release       : 5.8.0-55-generic
Version          : #62~20.04.1-Ubuntu SMP Wed Jun 2 08:55:04 UTC 2021
machine          : x86_64
processor        : x86_64
byteorder        : little
LC_ALL           : None
LANG             : en_US.UTF-8
LOCALE           : en_US.UTF-8

pandas           : 1.3.0.dev0+1850.g4d1019487a.dirty
numpy            : 1.20.1
pytz             : 2021.1
dateutil         : 2.8.1
pip              : 21.0.1
setuptools       : 49.6.0.post20210108
Cython           : 0.29.22
pytest           : 6.2.2
hypothesis       : 6.8.1
sphinx           : 3.5.3
blosc            : None
feather          : None
xlsxwriter       : 1.3.7
lxml.etree       : 4.6.2
html5lib         : 1.1
pymysql          : None
psycopg2         : None
jinja2           : 2.11.3
IPython          : 7.21.0
pandas_datareader: None
bs4              : 4.9.3
bottleneck       : 1.3.2
fsspec           : 0.8.7
fastparquet      : 0.5.0
gcsfs            : 0.7.2
matplotlib       : 3.3.4
numexpr          : 2.7.3
odfpy            : None
openpyxl         : 3.0.7
pandas_gbq       : None
pyarrow          : 3.0.0
pyxlsb           : None
s3fs             : 0.5.2
scipy            : 1.6.1
sqlalchemy       : 1.4.2
tables           : 3.6.1
tabulate         : 0.8.9
xarray           : 0.17.0
xlrd             : 2.0.1
xlwt             : 1.3.0
numba            : 0.53.0