CLN: Remove unicode u string prefix in regexes (#26433)

Author: simonjayhawkins

pandas/tests/groupby/test_groupby.py

@@ -730,7 +730,7 @@ def test_omit_nuisance(df):
     grouped = df.groupby({'A': 0, 'C': 0, 'D': 1, 'E': 1}, axis=1)
     msg = (r'\("unsupported operand type\(s\) for \+: '
            "'Timestamp' and 'float'\""
-           r", u?'occurred at index 0'\)")
+           r", 'occurred at index 0'\)")
     with pytest.raises(TypeError, match=msg):
         grouped.agg(lambda x: x.sum(0, numeric_only=False))
 

pandas/tests/indexes/datetimes/test_construction.py

@@ -747,7 +747,7 @@ def test_from_freq_recreate_from_data(self, freq):
 
     def test_datetimeindex_constructor_misc(self):
         arr = ['1/1/2005', '1/2/2005', 'Jn 3, 2005', '2005-01-04']
-        msg = r"(\(u?')?Unknown string format(:', 'Jn 3, 2005'\))?"
+        msg = r"(\(')?Unknown string format(:', 'Jn 3, 2005'\))?"
         with pytest.raises(ValueError, match=msg):
             DatetimeIndex(arr)
 

pandas/tests/indexes/datetimes/test_tools.py

@@ -1203,7 +1203,7 @@ def test_to_datetime_on_datetime64_series(self, cache):
     def test_to_datetime_with_space_in_series(self, cache):
         # GH 6428
         s = Series(['10/18/2006', '10/18/2008', ' '])
-        msg = r"(\(u?')?String does not contain a date(:', ' '\))?"
+        msg = r"(\(')?String does not contain a date(:', ' '\))?"
         with pytest.raises(ValueError, match=msg):
             to_datetime(s, errors='raise', cache=cache)
         result_coerce = to_datetime(s, errors='coerce', cache=cache)

pandas/tests/indexing/test_loc.py

@@ -232,7 +232,7 @@ def test_loc_to_fail(self):
         with pytest.raises(KeyError, match=msg):
             s.loc[[-1, -2]]
 
-        msg = (r"\"None of \[Index\(\[u?'4'\], dtype='object'\)\] are"
+        msg = (r"\"None of \[Index\(\['4'\], dtype='object'\)\] are"
                r" in the \[index\]\"")
         with pytest.raises(KeyError, match=msg):
             s.loc[['4']]

pandas/tests/indexing/test_partial.py

@@ -316,7 +316,7 @@ def test_series_partial_set_with_name(self):
 
         # raises as nothing in in the index
         msg = (r"\"None of \[Int64Index\(\[3, 3, 3\], dtype='int64',"
-               r" name=u?'idx'\)\] are in the \[index\]\"")
+               r" name='idx'\)\] are in the \[index\]\"")
         with pytest.raises(KeyError, match=msg):
             ser.loc[[3, 3, 3]]
 

pandas/tests/io/json/test_pandas.py

@@ -864,7 +864,7 @@ def test_misc_example(self):
         error_msg = """DataFrame\\.index are different
 
 DataFrame\\.index values are different \\(100\\.0 %\\)
-\\[left\\]:  Index\\(\\[u?'a', u?'b'\\], dtype='object'\\)
+\\[left\\]:  Index\\(\\['a', 'b'\\], dtype='object'\\)
 \\[right\\]: RangeIndex\\(start=0, stop=2, step=1\\)"""
         with pytest.raises(AssertionError, match=error_msg):
             assert_frame_equal(result, expected, check_index_type=False)

pandas/tests/io/msgpack/test_pack.py

@@ -64,7 +64,7 @@ def testStrictUnicodeUnpack(self):
             unpackb(packb(b'abc\xeddef'), encoding='utf-8', use_list=1)
 
     def testStrictUnicodePack(self):
-        msg = (r"'ascii' codec can't encode character u*'\\xed' in position 3:"
+        msg = (r"'ascii' codec can't encode character '\\xed' in position 3:"
                r" ordinal not in range\(128\)")
         with pytest.raises(UnicodeEncodeError, match=msg):
             packb("abc\xeddef", encoding='ascii', unicode_errors='strict')

pandas/tests/reshape/merge/test_merge.py

@@ -652,7 +652,7 @@ def test_overlapping_columns_error_message(self):
 
         # #2649, #10639
         df2.columns = ['key1', 'foo', 'foo']
-        msg = (r"Data columns not unique: Index\(\[u?'foo', u?'foo'\],"
+        msg = (r"Data columns not unique: Index\(\['foo', 'foo'\],"
                r" dtype='object'\)")
         with pytest.raises(MergeError, match=msg):
             merge(df, df2)

pandas/tests/util/test_assert_frame_equal.py

@@ -140,8 +140,8 @@ def test_frame_equal_index_mismatch():
     msg = """DataFrame\\.index are different
 
 DataFrame\\.index values are different \\(33\\.33333 %\\)
-\\[left\\]:  Index\\(\\[u?'a', u?'b', u?'c'\\], dtype='object'\\)
-\\[right\\]: Index\\(\\[u?'a', u?'b', u?'d'\\], dtype='object'\\)"""
+\\[left\\]:  Index\\(\\['a', 'b', 'c'\\], dtype='object'\\)
+\\[right\\]: Index\\(\\['a', 'b', 'd'\\], dtype='object'\\)"""
 
     df1 = DataFrame({"A": [1, 2, 3], "B": [4, 5, 6]},
                     index=["a", "b", "c"])
@@ -156,8 +156,8 @@ def test_frame_equal_columns_mismatch():
     msg = """DataFrame\\.columns are different
 
 DataFrame\\.columns values are different \\(50\\.0 %\\)
-\\[left\\]:  Index\\(\\[u?'A', u?'B'\\], dtype='object'\\)
-\\[right\\]: Index\\(\\[u?'A', u?'b'\\], dtype='object'\\)"""
+\\[left\\]:  Index\\(\\['A', 'B'\\], dtype='object'\\)
+\\[right\\]: Index\\(\\['A', 'b'\\], dtype='object'\\)"""
 
     df1 = DataFrame({"A": [1, 2, 3], "B": [4, 5, 6]},
                     index=["a", "b", "c"])

pandas/tests/util/test_assert_index_equal.py

@@ -10,7 +10,7 @@ def test_index_equal_levels_mismatch():
 
 Index levels are different
 \\[left\\]:  1, Int64Index\\(\\[1, 2, 3\\], dtype='int64'\\)
-\\[right\\]: 2, MultiIndex\\(levels=\\[\\[u?'A', u?'B'\\], \\[1, 2, 3, 4\\]\\],
+\\[right\\]: 2, MultiIndex\\(levels=\\[\\['A', 'B'\\], \\[1, 2, 3, 4\\]\\],
            codes=\\[\\[0, 0, 1, 1\\], \\[0, 1, 2, 3\\]\\]\\)"""
 
     idx1 = Index([1, 2, 3])
@@ -151,8 +151,8 @@ def test_index_equal_names(name1, name2):
     if name1 == name2 or name1 is name2:
         assert_index_equal(idx1, idx2)
     else:
-        name1 = "u?'x'" if name1 == "x" else name1
-        name2 = "u?'x'" if name2 == "x" else name2
+        name1 = "'x'" if name1 == "x" else name1
+        name2 = "'x'" if name2 == "x" else name2
         msg = msg.format(name1=name1, name2=name2)
 
         with pytest.raises(AssertionError, match=msg):
@@ -163,8 +163,8 @@ def test_index_equal_category_mismatch(check_categorical):
     msg = """Index are different
 
 Attribute "dtype" are different
-\\[left\\]:  CategoricalDtype\\(categories=\\[u?'a', u?'b'\\], ordered=False\\)
-\\[right\\]: CategoricalDtype\\(categories=\\[u?'a', u?'b', u?'c'\\], \
+\\[left\\]:  CategoricalDtype\\(categories=\\['a', 'b'\\], ordered=False\\)
+\\[right\\]: CategoricalDtype\\(categories=\\['a', 'b', 'c'\\], \
 ordered=False\\)"""
 
     idx1 = Index(Categorical(["a", "b"]))

pandas/tests/util/test_assert_series_equal.py

@@ -169,8 +169,8 @@ def test_series_equal_categorical_mismatch(check_categorical):
     msg = """Attributes are different
 
 Attribute "dtype" are different
-\\[left\\]:  CategoricalDtype\\(categories=\\[u?'a', u?'b'\\], ordered=False\\)
-\\[right\\]: CategoricalDtype\\(categories=\\[u?'a', u?'b', u?'c'\\], \
+\\[left\\]:  CategoricalDtype\\(categories=\\['a', 'b'\\], ordered=False\\)
+\\[right\\]: CategoricalDtype\\(categories=\\['a', 'b', 'c'\\], \
 ordered=False\\)"""
 
     s1 = Series(Categorical(["a", "b"]))