TST: Added match argument for most uses of tm.assert_produces_warning (#58396)

* Fix for all FutureWarnings

* Add match for most warnings

* Cleaner code

---------

Co-authored-by: Abdulaziz Aloqeely <[email protected]>

Author: Aloqeely

doc/source/development/contributing_codebase.rst

@@ -557,11 +557,12 @@ is being raised, using ``pytest.raises`` instead.
 Testing a warning
 ^^^^^^^^^^^^^^^^^
 
-Use ``tm.assert_produces_warning`` as a context manager to check that a block of code raises a warning.
+Use ``tm.assert_produces_warning`` as a context manager to check that a block of code raises a warning
+and specify the warning message using the ``match`` argument.
 
 .. code-block:: python
 
-    with tm.assert_produces_warning(DeprecationWarning):
+    with tm.assert_produces_warning(DeprecationWarning, match="the warning message"):
         pd.deprecated_function()
 
 If a warning should specifically not happen in a block of code, pass ``False`` into the context manager.

pandas/tests/arrays/sparse/test_constructors.py

@@ -90,13 +90,13 @@ def test_constructor_warns_when_losing_timezone(self):
         dti = pd.date_range("2016-01-01", periods=3, tz="US/Pacific")
 
         expected = SparseArray(np.asarray(dti, dtype="datetime64[ns]"))
-
-        with tm.assert_produces_warning(UserWarning):
+        msg = "loses timezone information"
+        with tm.assert_produces_warning(UserWarning, match=msg):
             result = SparseArray(dti)
 
         tm.assert_sp_array_equal(result, expected)
 
-        with tm.assert_produces_warning(UserWarning):
+        with tm.assert_produces_warning(UserWarning, match=msg):
             result = SparseArray(pd.Series(dti))
 
         tm.assert_sp_array_equal(result, expected)

pandas/tests/arrays/test_datetimelike.py

@@ -778,7 +778,7 @@ def test_to_period_2d(self, arr1d):
         arr2d = arr1d.reshape(1, -1)
 
         warn = None if arr1d.tz is None else UserWarning
-        with tm.assert_produces_warning(warn):
+        with tm.assert_produces_warning(warn, match="will drop timezone information"):
             result = arr2d.to_period("D")
             expected = arr1d.to_period("D").reshape(1, -1)
         tm.assert_period_array_equal(result, expected)

pandas/tests/computation/test_eval.py

@@ -1014,7 +1014,8 @@ def test_performance_warning_for_poor_alignment(
         else:
             seen = False
 
-        with tm.assert_produces_warning(seen):
+        msg = "Alignment difference on axis 1 is larger than an order of magnitude"
+        with tm.assert_produces_warning(seen, match=msg):
             pd.eval("df + s", engine=engine, parser=parser)
 
         s = Series(np.random.default_rng(2).standard_normal(1000))
@@ -1036,7 +1037,7 @@ def test_performance_warning_for_poor_alignment(
         else:
             wrn = False
 
-        with tm.assert_produces_warning(wrn) as w:
+        with tm.assert_produces_warning(wrn, match=msg) as w:
             pd.eval("df + s", engine=engine, parser=parser)
 
             if not is_python_engine and performance_warning:

pandas/tests/dtypes/test_common.py

@@ -797,5 +797,5 @@ def test_pandas_dtype_numpy_warning():
 
 def test_pandas_dtype_ea_not_instance():
     # GH 31356 GH 54592
-    with tm.assert_produces_warning(UserWarning):
+    with tm.assert_produces_warning(UserWarning, match="without any arguments"):
         assert pandas_dtype(CategoricalDtype) == CategoricalDtype()

pandas/tests/dtypes/test_generic.py

@@ -124,7 +124,7 @@ def test_setattr_warnings():
         #  this should not raise a warning
         df.two.not_an_index = [1, 2]
 
-    with tm.assert_produces_warning(UserWarning):
+    with tm.assert_produces_warning(UserWarning, match="doesn't allow columns"):
         #  warn when setting column to nonexistent name
         df.four = df.two + 2
         assert df.four.sum() > df.two.sum()

pandas/tests/frame/indexing/test_indexing.py

@@ -145,7 +145,7 @@ def test_getitem_boolean(self, mixed_float_frame, mixed_int_frame, datetime_fram
         # we are producing a warning that since the passed boolean
         # key is not the same as the given index, we will reindex
         # not sure this is really necessary
-        with tm.assert_produces_warning(UserWarning):
+        with tm.assert_produces_warning(UserWarning, match="will be reindexed"):
             indexer_obj = indexer_obj.reindex(datetime_frame.index[::-1])
             subframe_obj = datetime_frame[indexer_obj]
             tm.assert_frame_equal(subframe_obj, subframe)

pandas/tests/frame/indexing/test_setitem.py

@@ -711,7 +711,10 @@ def test_setitem_npmatrix_2d(self):
         df["np-array"] = a
 
         # Instantiation of `np.matrix` gives PendingDeprecationWarning
-        with tm.assert_produces_warning(PendingDeprecationWarning):
+        with tm.assert_produces_warning(
+            PendingDeprecationWarning,
+            match="matrix subclass is not the recommended way to represent matrices",
+        ):
             df["np-matrix"] = np.matrix(a)
 
         tm.assert_frame_equal(df, expected)

pandas/tests/frame/methods/test_to_dict.py

@@ -166,7 +166,7 @@ def test_to_dict_not_unique_warning(self):
         # GH#16927: When converting to a dict, if a column has a non-unique name
         # it will be dropped, throwing a warning.
         df = DataFrame([[1, 2, 3]], columns=["a", "a", "b"])
-        with tm.assert_produces_warning(UserWarning):
+        with tm.assert_produces_warning(UserWarning, match="columns will be omitted"):
             df.to_dict()
 
     @pytest.mark.filterwarnings("ignore::UserWarning")

pandas/tests/frame/test_arithmetic.py

@@ -1097,25 +1097,24 @@ def test_binop_other(self, op, value, dtype, switch_numexpr_min_elements):
                     and expr.USE_NUMEXPR
                     and switch_numexpr_min_elements == 0
                 ):
-                    warn = UserWarning  # "evaluating in Python space because ..."
+                    warn = UserWarning
             else:
                 msg = (
                     f"cannot perform __{op.__name__}__ with this "
                     "index type: (DatetimeArray|TimedeltaArray)"
                 )
 
             with pytest.raises(TypeError, match=msg):
-                with tm.assert_produces_warning(warn):
+                with tm.assert_produces_warning(warn, match="evaluating in Python"):
                     op(df, elem.value)
 
         elif (op, dtype) in skip:
             if op in [operator.add, operator.mul]:
                 if expr.USE_NUMEXPR and switch_numexpr_min_elements == 0:
-                    # "evaluating in Python space because ..."
                     warn = UserWarning
                 else:
                     warn = None
-                with tm.assert_produces_warning(warn):
+                with tm.assert_produces_warning(warn, match="evaluating in Python"):
                     op(df, elem.value)
 
             else:

pandas/tests/frame/test_reductions.py

@@ -699,7 +699,7 @@ def test_mode_sortwarning(self, using_infer_string):
         expected = DataFrame({"A": ["a", np.nan]})
 
         warning = None if using_infer_string else UserWarning
-        with tm.assert_produces_warning(warning):
+        with tm.assert_produces_warning(warning, match="Unable to sort modes"):
             result = df.mode(dropna=False)
             result = result.sort_values(by="A").reset_index(drop=True)
 

pandas/tests/indexes/base_class/test_setops.py

@@ -84,13 +84,13 @@ def test_union_sort_other_incomparable(self):
         # https://github.com/pandas-dev/pandas/issues/24959
         idx = Index([1, pd.Timestamp("2000")])
         # default (sort=None)
-        with tm.assert_produces_warning(RuntimeWarning):
+        with tm.assert_produces_warning(RuntimeWarning, match="not supported between"):
             result = idx.union(idx[:1])
 
         tm.assert_index_equal(result, idx)
 
         # sort=None
-        with tm.assert_produces_warning(RuntimeWarning):
+        with tm.assert_produces_warning(RuntimeWarning, match="not supported between"):
             result = idx.union(idx[:1], sort=None)
         tm.assert_index_equal(result, idx)
 

pandas/tests/indexes/datetimes/methods/test_to_period.py

@@ -117,10 +117,10 @@ def test_to_period_infer(self):
             freq="5min",
         )
 
-        with tm.assert_produces_warning(UserWarning):
+        with tm.assert_produces_warning(UserWarning, match="drop timezone info"):
             pi1 = rng.to_period("5min")
 
-        with tm.assert_produces_warning(UserWarning):
+        with tm.assert_produces_warning(UserWarning, match="drop timezone info"):
             pi2 = rng.to_period()
 
         tm.assert_index_equal(pi1, pi2)
@@ -143,8 +143,7 @@ def test_to_period_millisecond(self):
             ]
         )
 
-        with tm.assert_produces_warning(UserWarning):
-            # warning that timezone info will be lost
+        with tm.assert_produces_warning(UserWarning, match="drop timezone info"):
             period = index.to_period(freq="ms")
         assert 2 == len(period)
         assert period[0] == Period("2007-01-01 10:11:12.123Z", "ms")
@@ -158,8 +157,7 @@ def test_to_period_microsecond(self):
             ]
         )
 
-        with tm.assert_produces_warning(UserWarning):
-            # warning that timezone info will be lost
+        with tm.assert_produces_warning(UserWarning, match="drop timezone info"):
             period = index.to_period(freq="us")
         assert 2 == len(period)
         assert period[0] == Period("2007-01-01 10:11:12.123456Z", "us")
@@ -172,19 +170,15 @@ def test_to_period_microsecond(self):
     def test_to_period_tz(self, tz):
         ts = date_range("1/1/2000", "2/1/2000", tz=tz)
 
-        with tm.assert_produces_warning(UserWarning):
-            # GH#21333 warning that timezone info will be lost
-            # filter warning about freq deprecation
-
+        with tm.assert_produces_warning(UserWarning, match="drop timezone info"):
             result = ts.to_period()[0]
             expected = ts[0].to_period(ts.freq)
 
         assert result == expected
 
         expected = date_range("1/1/2000", "2/1/2000").to_period()
 
-        with tm.assert_produces_warning(UserWarning):
-            # GH#21333 warning that timezone info will be lost
+        with tm.assert_produces_warning(UserWarning, match="drop timezone info"):
             result = ts.to_period(ts.freq)
 
         tm.assert_index_equal(result, expected)
@@ -193,7 +187,7 @@ def test_to_period_tz(self, tz):
     def test_to_period_tz_utc_offset_consistency(self, tz):
         # GH#22905
         ts = date_range("1/1/2000", "2/1/2000", tz="Etc/GMT-1")
-        with tm.assert_produces_warning(UserWarning):
+        with tm.assert_produces_warning(UserWarning, match="drop timezone info"):
             result = ts.to_period()[0]
             expected = ts[0].to_period(ts.freq)
             assert result == expected

pandas/tests/indexes/multi/test_setops.py

@@ -382,7 +382,7 @@ def test_union_sort_other_incomparable():
     idx = MultiIndex.from_product([[1, pd.Timestamp("2000")], ["a", "b"]])
 
     # default, sort=None
-    with tm.assert_produces_warning(RuntimeWarning):
+    with tm.assert_produces_warning(RuntimeWarning, match="are unorderable"):
         result = idx.union(idx[:1])
     tm.assert_index_equal(result, idx)
 

pandas/tests/indexes/test_base.py

@@ -1065,10 +1065,10 @@ def test_outer_join_sort(self):
         left_index = Index(np.random.default_rng(2).permutation(15))
         right_index = date_range("2020-01-01", periods=10)
 
-        with tm.assert_produces_warning(RuntimeWarning):
+        with tm.assert_produces_warning(RuntimeWarning, match="not supported between"):
             result = left_index.join(right_index, how="outer")
 
-        with tm.assert_produces_warning(RuntimeWarning):
+        with tm.assert_produces_warning(RuntimeWarning, match="not supported between"):
             expected = left_index.astype(object).union(right_index.astype(object))
 
         tm.assert_index_equal(result, expected)

pandas/tests/indexes/test_index_new.py

@@ -142,25 +142,18 @@ def test_constructor_infer_nat_dt_like(
         data = [ctor]
         data.insert(pos, nulls_fixture)
 
-        warn = None
         if nulls_fixture is NA:
             expected = Index([NA, NaT])
             mark = pytest.mark.xfail(reason="Broken with np.NaT ctor; see GH 31884")
             request.applymarker(mark)
-            # GH#35942 numpy will emit a DeprecationWarning within the
-            #  assert_index_equal calls.  Since we can't do anything
-            #  about it until GH#31884 is fixed, we suppress that warning.
-            warn = DeprecationWarning
 
         result = Index(data)
 
-        with tm.assert_produces_warning(warn):
-            tm.assert_index_equal(result, expected)
+        tm.assert_index_equal(result, expected)
 
         result = Index(np.array(data, dtype=object))
 
-        with tm.assert_produces_warning(warn):
-            tm.assert_index_equal(result, expected)
+        tm.assert_index_equal(result, expected)
 
     @pytest.mark.parametrize("swap_objs", [True, False])
     def test_constructor_mixed_nat_objs_infers_object(self, swap_objs):

pandas/tests/indexes/test_setops.py

@@ -882,7 +882,7 @@ def test_difference_incomparable(self, opname):
         b = Index([2, Timestamp("1999"), 1])
         op = operator.methodcaller(opname, b)
 
-        with tm.assert_produces_warning(RuntimeWarning):
+        with tm.assert_produces_warning(RuntimeWarning, match="not supported between"):
             # sort=None, the default
             result = op(a)
         expected = Index([3, Timestamp("2000"), 2, Timestamp("1999")])

pandas/tests/internals/test_internals.py

@@ -1280,19 +1280,20 @@ def test_interval_can_hold_element(self, dtype, element):
         # `elem` to not have the same length as `arr`
         ii2 = IntervalIndex.from_breaks(arr[:-1], closed="neither")
         elem = element(ii2)
-        with tm.assert_produces_warning(FutureWarning):
+        msg = "Setting an item of incompatible dtype is deprecated"
+        with tm.assert_produces_warning(FutureWarning, match=msg):
             self.check_series_setitem(elem, ii, False)
         assert not blk._can_hold_element(elem)
 
         ii3 = IntervalIndex.from_breaks([Timestamp(1), Timestamp(3), Timestamp(4)])
         elem = element(ii3)
-        with tm.assert_produces_warning(FutureWarning):
+        with tm.assert_produces_warning(FutureWarning, match=msg):
             self.check_series_setitem(elem, ii, False)
         assert not blk._can_hold_element(elem)
 
         ii4 = IntervalIndex.from_breaks([Timedelta(1), Timedelta(3), Timedelta(4)])
         elem = element(ii4)
-        with tm.assert_produces_warning(FutureWarning):
+        with tm.assert_produces_warning(FutureWarning, match=msg):
             self.check_series_setitem(elem, ii, False)
         assert not blk._can_hold_element(elem)
 
@@ -1312,12 +1313,13 @@ def test_period_can_hold_element(self, element):
         # `elem` to not have the same length as `arr`
         pi2 = pi.asfreq("D")[:-1]
         elem = element(pi2)
-        with tm.assert_produces_warning(FutureWarning):
+        msg = "Setting an item of incompatible dtype is deprecated"
+        with tm.assert_produces_warning(FutureWarning, match=msg):
             self.check_series_setitem(elem, pi, False)
 
         dti = pi.to_timestamp("s")[:-1]
         elem = element(dti)
-        with tm.assert_produces_warning(FutureWarning):
+        with tm.assert_produces_warning(FutureWarning, match=msg):
             self.check_series_setitem(elem, pi, False)
 
     def check_can_hold_element(self, obj, elem, inplace: bool):

pandas/tests/io/formats/test_css.py

@@ -38,30 +38,31 @@ def test_css_parse_normalisation(name, norm, abnorm):
 
 
 @pytest.mark.parametrize(
-    "invalid_css,remainder",
+    "invalid_css,remainder,msg",
     [
         # No colon
-        ("hello-world", ""),
-        ("border-style: solid; hello-world", "border-style: solid"),
+        ("hello-world", "", "expected a colon"),
+        ("border-style: solid; hello-world", "border-style: solid", "expected a colon"),
         (
             "border-style: solid; hello-world; font-weight: bold",
             "border-style: solid; font-weight: bold",
+            "expected a colon",
         ),
         # Unclosed string fail
         # Invalid size
-        ("font-size: blah", "font-size: 1em"),
-        ("font-size: 1a2b", "font-size: 1em"),
-        ("font-size: 1e5pt", "font-size: 1em"),
-        ("font-size: 1+6pt", "font-size: 1em"),
-        ("font-size: 1unknownunit", "font-size: 1em"),
-        ("font-size: 10", "font-size: 1em"),
-        ("font-size: 10 pt", "font-size: 1em"),
+        ("font-size: blah", "font-size: 1em", "Unhandled size"),
+        ("font-size: 1a2b", "font-size: 1em", "Unhandled size"),
+        ("font-size: 1e5pt", "font-size: 1em", "Unhandled size"),
+        ("font-size: 1+6pt", "font-size: 1em", "Unhandled size"),
+        ("font-size: 1unknownunit", "font-size: 1em", "Unhandled size"),
+        ("font-size: 10", "font-size: 1em", "Unhandled size"),
+        ("font-size: 10 pt", "font-size: 1em", "Unhandled size"),
         # Too many args
-        ("border-top: 1pt solid red green", "border-top: 1pt solid green"),
+        ("border-top: 1pt solid red green", "border-top: 1pt solid green", "Too many"),
     ],
 )
-def test_css_parse_invalid(invalid_css, remainder):
-    with tm.assert_produces_warning(CSSWarning):
+def test_css_parse_invalid(invalid_css, remainder, msg):
+    with tm.assert_produces_warning(CSSWarning, match=msg):
         assert_same_resolution(invalid_css, remainder)
 
 
@@ -120,7 +121,7 @@ def test_css_side_shorthands(shorthand, expansions):
         {top: "1pt", right: "4pt", bottom: "2pt", left: "0pt"},
     )
 
-    with tm.assert_produces_warning(CSSWarning):
+    with tm.assert_produces_warning(CSSWarning, match="Could not expand"):
         assert_resolves(f"{shorthand}: 1pt 1pt 1pt 1pt 1pt", {})
 
 

pandas/tests/io/formats/test_to_excel.py

@@ -325,7 +325,7 @@ def test_css_to_excel_bad_colors(input_color):
     if input_color is not None:
         expected["fill"] = {"patternType": "solid"}
 
-    with tm.assert_produces_warning(CSSWarning):
+    with tm.assert_produces_warning(CSSWarning, match="Unhandled color format"):
         convert = CSSToExcelConverter()
         assert expected == convert(css)
 

pandas/tests/io/json/test_json_table_schema.py

@@ -639,7 +639,7 @@ def test_warns_non_roundtrippable_names(self, idx):
         # GH 19130
         df = DataFrame(index=idx)
         df.index.name = "index"
-        with tm.assert_produces_warning():
+        with tm.assert_produces_warning(UserWarning, match="not round-trippable"):
             set_default_names(df)
 
     def test_timestamp_in_columns(self):