fix bare pytest raises in indexes/datetimes (#32884)

Author: quangngd

pandas/tests/indexes/datetimes/test_constructors.py

@@ -415,7 +415,8 @@ def test_construction_dti_with_mixed_timezones(self):
 
         # tz mismatch affecting to tz-aware raises TypeError/ValueError
 
-        with pytest.raises(ValueError):
+        msg = "cannot be converted to datetime64"
+        with pytest.raises(ValueError, match=msg):
             DatetimeIndex(
                 [
                     Timestamp("2011-01-01 10:00", tz="Asia/Tokyo"),
@@ -424,7 +425,6 @@ def test_construction_dti_with_mixed_timezones(self):
                 name="idx",
             )
 
-        msg = "cannot be converted to datetime64"
         with pytest.raises(ValueError, match=msg):
             DatetimeIndex(
                 [
@@ -435,7 +435,7 @@ def test_construction_dti_with_mixed_timezones(self):
                 name="idx",
             )
 
-        with pytest.raises(ValueError):
+        with pytest.raises(ValueError, match=msg):
             DatetimeIndex(
                 [
                     Timestamp("2011-01-01 10:00", tz="Asia/Tokyo"),
@@ -480,7 +480,8 @@ def test_construction_outofbounds(self):
         # coerces to object
         tm.assert_index_equal(Index(dates), exp)
 
-        with pytest.raises(OutOfBoundsDatetime):
+        msg = "Out of bounds nanosecond timestamp"
+        with pytest.raises(OutOfBoundsDatetime, match=msg):
             # can't create DatetimeIndex
             DatetimeIndex(dates)
 
@@ -516,7 +517,8 @@ def test_constructor_coverage(self):
         with pytest.raises(TypeError, match=msg):
             date_range(start="1/1/2000", periods="foo", freq="D")
 
-        with pytest.raises(TypeError):
+        msg = "DatetimeIndex\\(\\) must be called with a collection"
+        with pytest.raises(TypeError, match=msg):
             DatetimeIndex("1/1/2000")
 
         # generator expression
@@ -664,7 +666,8 @@ def test_constructor_dtype(self):
     @pytest.mark.parametrize("dtype", [object, np.int32, np.int64])
     def test_constructor_invalid_dtype_raises(self, dtype):
         # GH 23986
-        with pytest.raises(ValueError):
+        msg = "Unexpected value for 'dtype'"
+        with pytest.raises(ValueError, match=msg):
             DatetimeIndex([1, 2], dtype=dtype)
 
     def test_constructor_name(self):
@@ -681,7 +684,8 @@ def test_000constructor_resolution(self):
     def test_disallow_setting_tz(self):
         # GH 3746
         dti = DatetimeIndex(["2010"], tz="UTC")
-        with pytest.raises(AttributeError):
+        msg = "Cannot directly set timezone"
+        with pytest.raises(AttributeError, match=msg):
             dti.tz = pytz.timezone("US/Pacific")
 
     @pytest.mark.parametrize(
@@ -770,7 +774,8 @@ def test_construction_from_replaced_timestamps_with_dst(self):
     def test_construction_with_tz_and_tz_aware_dti(self):
         # GH 23579
         dti = date_range("2016-01-01", periods=3, tz="US/Central")
-        with pytest.raises(TypeError):
+        msg = "data is already tz-aware US/Central, unable to set specified tz"
+        with pytest.raises(TypeError, match=msg):
             DatetimeIndex(dti, tz="Asia/Tokyo")
 
     def test_construction_with_nat_and_tzlocal(self):
@@ -790,7 +795,8 @@ def test_constructor_no_precision_raises(self):
             pd.Index(["2000"], dtype="datetime64")
 
     def test_constructor_wrong_precision_raises(self):
-        with pytest.raises(ValueError):
+        msg = "Unexpected value for 'dtype': 'datetime64\\[us\\]'"
+        with pytest.raises(ValueError, match=msg):
             pd.DatetimeIndex(["2000"], dtype="datetime64[us]")
 
     def test_index_constructor_with_numpy_object_array_and_timestamp_tz_with_nan(self):

pandas/tests/indexes/datetimes/test_date_range.py

@@ -153,9 +153,10 @@ def test_date_range_int64_overflow_stride_endpoint_different_signs(self):
 
     def test_date_range_out_of_bounds(self):
         # GH#14187
-        with pytest.raises(OutOfBoundsDatetime):
+        msg = "Cannot generate range"
+        with pytest.raises(OutOfBoundsDatetime, match=msg):
             date_range("2016-01-01", periods=100000, freq="D")
-        with pytest.raises(OutOfBoundsDatetime):
+        with pytest.raises(OutOfBoundsDatetime, match=msg):
             date_range(end="1763-10-12", periods=100000, freq="D")
 
     def test_date_range_gen_error(self):
@@ -736,9 +737,10 @@ def test_precision_finer_than_offset(self):
     )
     def test_mismatching_tz_raises_err(self, start, end):
         # issue 18488
-        with pytest.raises(TypeError):
+        msg = "Start and end cannot both be tz-aware with different timezones"
+        with pytest.raises(TypeError, match=msg):
             pd.date_range(start, end)
-        with pytest.raises(TypeError):
+        with pytest.raises(TypeError, match=msg):
             pd.date_range(start, end, freq=BDay())
 
 
@@ -771,16 +773,17 @@ def test_misc(self):
     def test_date_parse_failure(self):
         badly_formed_date = "2007/100/1"
 
-        with pytest.raises(ValueError):
+        msg = "could not convert string to Timestamp"
+        with pytest.raises(ValueError, match=msg):
             Timestamp(badly_formed_date)
 
-        with pytest.raises(ValueError):
+        with pytest.raises(ValueError, match=msg):
             bdate_range(start=badly_formed_date, periods=10)
 
-        with pytest.raises(ValueError):
+        with pytest.raises(ValueError, match=msg):
             bdate_range(end=badly_formed_date, periods=10)
 
-        with pytest.raises(ValueError):
+        with pytest.raises(ValueError, match=msg):
             bdate_range(badly_formed_date, badly_formed_date)
 
     def test_daterange_bug_456(self):
@@ -813,8 +816,9 @@ def test_bday_near_overflow(self):
 
     def test_bday_overflow_error(self):
         # GH#24252 check that we get OutOfBoundsDatetime and not OverflowError
+        msg = "Out of bounds nanosecond timestamp"
         start = pd.Timestamp.max.floor("D").to_pydatetime()
-        with pytest.raises(OutOfBoundsDatetime):
+        with pytest.raises(OutOfBoundsDatetime, match=msg):
             pd.date_range(start, periods=2, freq="B")
 
 

pandas/tests/indexes/datetimes/test_indexing.py

@@ -312,7 +312,8 @@ def test_take_fill_value(self):
         with pytest.raises(ValueError, match=msg):
             idx.take(np.array([1, 0, -5]), fill_value=True)
 
-        with pytest.raises(IndexError):
+        msg = "out of bounds"
+        with pytest.raises(IndexError, match=msg):
             idx.take(np.array([1, -5]))
 
     def test_take_fill_value_with_timezone(self):
@@ -348,7 +349,8 @@ def test_take_fill_value_with_timezone(self):
         with pytest.raises(ValueError, match=msg):
             idx.take(np.array([1, 0, -5]), fill_value=True)
 
-        with pytest.raises(IndexError):
+        msg = "out of bounds"
+        with pytest.raises(IndexError, match=msg):
             idx.take(np.array([1, -5]))
 
 
@@ -428,7 +430,8 @@ def test_get_loc(self):
         tm.assert_numpy_array_equal(
             idx.get_loc(time(12, 30)), np.array([]), check_dtype=False
         )
-        with pytest.raises(NotImplementedError):
+        msg = "cannot yet lookup inexact labels when key is a time object"
+        with pytest.raises(NotImplementedError, match=msg):
             idx.get_loc(time(12, 30), method="pad")
 
     def test_get_loc_tz_aware(self):
@@ -462,7 +465,8 @@ def test_get_loc_nat(self):
     def test_get_loc_timedelta_invalid_key(self, key):
         # GH#20464
         dti = pd.date_range("1970-01-01", periods=10)
-        with pytest.raises(TypeError):
+        msg = "Cannot index DatetimeIndex with [Tt]imedelta"
+        with pytest.raises(TypeError, match=msg):
             dti.get_loc(key)
 
     def test_get_loc_reasonable_key_error(self):
@@ -571,9 +575,9 @@ def test_insert(self):
             idx.insert(3, pd.Timestamp("2000-01-04"))
         with pytest.raises(TypeError, match="Cannot compare tz-naive and tz-aware"):
             idx.insert(3, datetime(2000, 1, 4))
-        with pytest.raises(ValueError):
+        with pytest.raises(ValueError, match="Timezones don't match"):
             idx.insert(3, pd.Timestamp("2000-01-04", tz="US/Eastern"))
-        with pytest.raises(ValueError):
+        with pytest.raises(ValueError, match="Timezones don't match"):
             idx.insert(3, datetime(2000, 1, 4, tzinfo=pytz.timezone("US/Eastern")))
 
         for tz in ["US/Pacific", "Asia/Singapore"]:
@@ -645,7 +649,7 @@ def test_delete(self):
             assert result.name == expected.name
             assert result.freq == expected.freq
 
-        with pytest.raises((IndexError, ValueError)):
+        with pytest.raises((IndexError, ValueError), match="out of bounds"):
             # either depending on numpy version
             idx.delete(5)
 
@@ -804,5 +808,5 @@ def test_get_indexer(self):
         ]
         with pytest.raises(ValueError, match="abbreviation w/o a number"):
             idx.get_indexer(target, "nearest", tolerance=tol_bad)
-        with pytest.raises(ValueError):
+        with pytest.raises(ValueError, match="abbreviation w/o a number"):
             idx.get_indexer(idx[[0]], method="nearest", tolerance="foo")

pandas/tests/indexes/datetimes/test_shift.py

@@ -80,7 +80,7 @@ def test_dti_shift_int(self):
     def test_dti_shift_no_freq(self):
         # GH#19147
         dti = pd.DatetimeIndex(["2011-01-01 10:00", "2011-01-01"], freq=None)
-        with pytest.raises(NullFrequencyError):
+        with pytest.raises(NullFrequencyError, match="Cannot shift with no freq"):
             dti.shift(2)
 
     @pytest.mark.parametrize("tzstr", ["US/Eastern", "dateutil/US/Eastern"])

pandas/tests/indexes/datetimes/test_timezones.py

@@ -319,10 +319,10 @@ def test_dti_tz_localize_nonexistent_raise_coerce(self):
         times = ["2015-03-08 01:00", "2015-03-08 02:00", "2015-03-08 03:00"]
         index = DatetimeIndex(times)
         tz = "US/Eastern"
-        with pytest.raises(pytz.NonExistentTimeError):
+        with pytest.raises(pytz.NonExistentTimeError, match="|".join(times)):
             index.tz_localize(tz=tz)
 
-        with pytest.raises(pytz.NonExistentTimeError):
+        with pytest.raises(pytz.NonExistentTimeError, match="|".join(times)):
             index.tz_localize(tz=tz, nonexistent="raise")
 
         result = index.tz_localize(tz=tz, nonexistent="NaT")
@@ -336,7 +336,7 @@ def test_dti_tz_localize_ambiguous_infer(self, tz):
         # November 6, 2011, fall back, repeat 2 AM hour
         # With no repeated hours, we cannot infer the transition
         dr = date_range(datetime(2011, 11, 6, 0), periods=5, freq=pd.offsets.Hour())
-        with pytest.raises(pytz.AmbiguousTimeError):
+        with pytest.raises(pytz.AmbiguousTimeError, match="Cannot infer dst time"):
             dr.tz_localize(tz)
 
         # With repeated hours, we can infer the transition
@@ -365,7 +365,7 @@ def test_dti_tz_localize_ambiguous_infer(self, tz):
     def test_dti_tz_localize_ambiguous_times(self, tz):
         # March 13, 2011, spring forward, skip from 2 AM to 3 AM
         dr = date_range(datetime(2011, 3, 13, 1, 30), periods=3, freq=pd.offsets.Hour())
-        with pytest.raises(pytz.NonExistentTimeError):
+        with pytest.raises(pytz.NonExistentTimeError, match="2011-03-13 02:30:00"):
             dr.tz_localize(tz)
 
         # after dst transition, it works
@@ -375,7 +375,7 @@ def test_dti_tz_localize_ambiguous_times(self, tz):
 
         # November 6, 2011, fall back, repeat 2 AM hour
         dr = date_range(datetime(2011, 11, 6, 1, 30), periods=3, freq=pd.offsets.Hour())
-        with pytest.raises(pytz.AmbiguousTimeError):
+        with pytest.raises(pytz.AmbiguousTimeError, match="Cannot infer dst time"):
             dr.tz_localize(tz)
 
         # UTC is OK
@@ -411,11 +411,11 @@ def test_dti_tz_localize(self, prefix):
         tm.assert_numpy_array_equal(dti3.values, dti_utc.values)
 
         dti = pd.date_range(start="11/6/2011 1:59", end="11/6/2011 2:00", freq="L")
-        with pytest.raises(pytz.AmbiguousTimeError):
+        with pytest.raises(pytz.AmbiguousTimeError, match="Cannot infer dst time"):
             dti.tz_localize(tzstr)
 
         dti = pd.date_range(start="3/13/2011 1:59", end="3/13/2011 2:00", freq="L")
-        with pytest.raises(pytz.NonExistentTimeError):
+        with pytest.raises(pytz.NonExistentTimeError, match="2011-03-13 02:00:00"):
             dti.tz_localize(tzstr)
 
     @pytest.mark.parametrize(
@@ -441,7 +441,7 @@ def test_dti_tz_localize_utc_conversion(self, tz):
         # DST ambiguity, this should fail
         rng = date_range("3/11/2012", "3/12/2012", freq="30T")
         # Is this really how it should fail??
-        with pytest.raises(pytz.NonExistentTimeError):
+        with pytest.raises(pytz.NonExistentTimeError, match="2012-03-11 02:00:00"):
             rng.tz_localize(tz)
 
     def test_dti_tz_localize_roundtrip(self, tz_aware_fixture):
@@ -452,7 +452,9 @@ def test_dti_tz_localize_roundtrip(self, tz_aware_fixture):
         tz = tz_aware_fixture
         localized = idx.tz_localize(tz)
         # cant localize a tz-aware object
-        with pytest.raises(TypeError):
+        with pytest.raises(
+            TypeError, match="Already tz-aware, use tz_convert to convert"
+        ):
             localized.tz_localize(tz)
         reset = localized.tz_localize(None)
         assert reset.tzinfo is None
@@ -542,7 +544,8 @@ def test_dti_tz_localize_ambiguous_flags(self, tz):
         di = DatetimeIndex(times)
 
         # When the sizes are incompatible, make sure error is raised
-        with pytest.raises(Exception):
+        msg = "Length of ambiguous bool-array must be the same size as vals"
+        with pytest.raises(Exception, match=msg):
             di.tz_localize(tz, ambiguous=is_dst)
 
         # When sizes are compatible and there are repeats ('infer' won't work)
@@ -564,7 +567,7 @@ def test_dti_construction_ambiguous_endpoint(self, tz):
         # construction with an ambiguous end-point
         # GH#11626
 
-        with pytest.raises(pytz.AmbiguousTimeError):
+        with pytest.raises(pytz.AmbiguousTimeError, match="Cannot infer dst time"):
             date_range(
                 "2013-10-26 23:00", "2013-10-27 01:00", tz="Europe/London", freq="H"
             )
@@ -588,7 +591,7 @@ def test_dti_construction_ambiguous_endpoint(self, tz):
     def test_dti_construction_nonexistent_endpoint(self, tz, option, expected):
         # construction with an nonexistent end-point
 
-        with pytest.raises(pytz.NonExistentTimeError):
+        with pytest.raises(pytz.NonExistentTimeError, match="2019-03-10 02:00:00"):
             date_range(
                 "2019-03-10 00:00", "2019-03-10 02:00", tz="US/Pacific", freq="H"
             )
@@ -613,10 +616,15 @@ def test_dti_tz_localize_nonexistent(self, tz, method, exp):
         n = 60
         dti = date_range(start="2015-03-29 02:00:00", periods=n, freq="min")
         if method == "raise":
-            with pytest.raises(pytz.NonExistentTimeError):
+            with pytest.raises(pytz.NonExistentTimeError, match="2015-03-29 02:00:00"):
                 dti.tz_localize(tz, nonexistent=method)
         elif exp == "invalid":
-            with pytest.raises(ValueError):
+            msg = (
+                "The nonexistent argument must be one of "
+                "'raise', 'NaT', 'shift_forward', 'shift_backward' "
+                "or a timedelta object"
+            )
+            with pytest.raises(ValueError, match=msg):
                 dti.tz_localize(tz, nonexistent=method)
         else:
             result = dti.tz_localize(tz, nonexistent=method)
@@ -1082,7 +1090,8 @@ def test_with_tz(self, tz):
         dr = bdate_range(
             datetime(2005, 1, 1, tzinfo=pytz.utc), datetime(2009, 1, 1, tzinfo=pytz.utc)
         )
-        with pytest.raises(Exception):
+        msg = "Start and end cannot both be tz-aware with different timezones"
+        with pytest.raises(Exception, match=msg):
             bdate_range(datetime(2005, 1, 1, tzinfo=pytz.utc), "1/1/2009", tz=tz)
 
     @pytest.mark.parametrize("prefix", ["", "dateutil/"])