GH-41502: [Python] Fix reading column index with decimal values

python/pyarrow/pandas_compat.py

@@ -23,6 +23,7 @@
 # module bug (ARROW-11983)
 import concurrent.futures.thread  # noqa
 from copy import deepcopy
+import decimal
 from itertools import zip_longest
 import json
 import operator
@@ -1027,6 +1028,7 @@ def _is_generated_index_name(name):
     'string': np.str_,
     'integer': np.int64,
     'floating': np.float64,
+    'decimal': np.object_,
     'empty': np.object_,
 }
 
@@ -1105,6 +1107,9 @@ def _reconstruct_columns_from_metadata(columns, column_indexes):
             tz = pa.lib.string_to_tzinfo(
                 column_indexes[0]['metadata']['timezone'])
             level = pd.to_datetime(level, utc=True).tz_convert(tz)
+        # GH-41503: if the column index was decimal, restore to decimal
+        elif pandas_dtype == "decimal":
+            level = _pandas_api.pd.Index([decimal.Decimal(i) for i in level])
         elif level.dtype != dtype:
             level = level.astype(dtype)
         # ARROW-9096: if original DataFrame was upcast we keep that

python/pyarrow/tests/test_pandas.py

@@ -221,6 +221,17 @@ def test_column_index_names_with_tz(self):
         )
         _check_pandas_roundtrip(df, preserve_index=True)
 
+    def test_column_index_names_with_decimal(self):
+        # GH-41503: Test valid roundtrip with decimal value in column index
+        df = pd.DataFrame(
+            [[decimal.Decimal(5), decimal.Decimal(6)]],
+            columns=pd.MultiIndex.from_product(
+                [[decimal.Decimal(1)], [decimal.Decimal(2), decimal.Decimal(3)]]
+            ),
+            index=[decimal.Decimal(4)],
+        )
+        _check_pandas_roundtrip(df, preserve_index=True)
+
     def test_range_index_shortcut(self):
         # ARROW-1639
         index_name = 'foo'