BUG: Fix na_values dict not working on index column (#57547) (#57965)

BUG: Na_values dict not working on index column (#57547)

* fix base_parser not setting col_na_values when na_values is a dict containing None

* fix python_parser applying na_values in a column None

* add unit test to test_na_values.py;

* update whatsnew.

Author: tomhoq

doc/source/whatsnew/v3.0.0.rst

@@ -349,6 +349,7 @@ Bug fixes
 - Fixed bug in :meth:`Series.diff` allowing non-integer values for the ``periods`` argument. (:issue:`56607`)
 - Fixed bug in :meth:`Series.rank` that doesn't preserve missing values for nullable integers when ``na_option='keep'``. (:issue:`56976`)
 - Fixed bug in :meth:`Series.replace` and :meth:`DataFrame.replace` inconsistently replacing matching instances when ``regex=True`` and missing values are present. (:issue:`56599`)
+- Fixed bug in :meth:`read_csv` raising ``TypeError`` when ``index_col`` is specified and ``na_values`` is a dict containing the key ``None``. (:issue:`57547`)
 
 Categorical
 ^^^^^^^^^^^

pandas/io/parsers/base_parser.py

@@ -487,6 +487,8 @@ def _agg_index(self, index, try_parse_dates: bool = True) -> Index:
                     col_na_values, col_na_fvalues = _get_na_values(
                         col_name, self.na_values, self.na_fvalues, self.keep_default_na
                     )
+                else:
+                    col_na_values, col_na_fvalues = set(), set()
 
             clean_dtypes = self._clean_mapping(self.dtype)
 

pandas/io/parsers/python_parser.py

@@ -354,14 +354,15 @@ def _convert_data(
 
         if isinstance(self.na_values, dict):
             for col in self.na_values:
-                na_value = self.na_values[col]
-                na_fvalue = self.na_fvalues[col]
+                if col is not None:
+                    na_value = self.na_values[col]
+                    na_fvalue = self.na_fvalues[col]
 
-                if isinstance(col, int) and col not in self.orig_names:
-                    col = self.orig_names[col]
+                    if isinstance(col, int) and col not in self.orig_names:
+                        col = self.orig_names[col]
 
-                clean_na_values[col] = na_value
-                clean_na_fvalues[col] = na_fvalue
+                    clean_na_values[col] = na_value
+                    clean_na_fvalues[col] = na_fvalue
         else:
             clean_na_values = self.na_values
             clean_na_fvalues = self.na_fvalues

pandas/tests/io/parser/test_na_values.py

@@ -532,6 +532,47 @@ def test_na_values_dict_aliasing(all_parsers):
     tm.assert_dict_equal(na_values, na_values_copy)
 
 
+def test_na_values_dict_null_column_name(all_parsers):
+    # see gh-57547
+    parser = all_parsers
+    data = ",x,y\n\nMA,1,2\nNA,2,1\nOA,,3"
+    names = [None, "x", "y"]
+    na_values = {name: STR_NA_VALUES for name in names}
+    dtype = {None: "object", "x": "float64", "y": "float64"}
+
+    if parser.engine == "pyarrow":
+        msg = "The pyarrow engine doesn't support passing a dict for na_values"
+        with pytest.raises(ValueError, match=msg):
+            parser.read_csv(
+                StringIO(data),
+                index_col=0,
+                header=0,
+                dtype=dtype,
+                names=names,
+                na_values=na_values,
+                keep_default_na=False,
+            )
+        return
+
+    expected = DataFrame(
+        {None: ["MA", "NA", "OA"], "x": [1.0, 2.0, np.nan], "y": [2.0, 1.0, 3.0]}
+    )
+
+    expected = expected.set_index(None)
+
+    result = parser.read_csv(
+        StringIO(data),
+        index_col=0,
+        header=0,
+        dtype=dtype,
+        names=names,
+        na_values=na_values,
+        keep_default_na=False,
+    )
+
+    tm.assert_frame_equal(result, expected)
+
+
 def test_na_values_dict_col_index(all_parsers):
     # see gh-14203
     data = "a\nfoo\n1"