Feature: Add flag to disable pyarrow DataTypes conversion

awswrangler/s3/_read_parquet.py

@@ -238,6 +238,7 @@ def _arrowtable2df(
     table: pa.Table,
     categories: Optional[List[str]],
     safe: bool,
+    map_types: bool,
     use_threads: bool,
     dataset: bool,
     path: str,
@@ -257,7 +258,7 @@ def _arrowtable2df(
             strings_to_categorical=False,
             safe=safe,
             categories=categories,
-            types_mapper=_data_types.pyarrow2pandas_extension,
+            types_mapper=_data_types.pyarrow2pandas_extension if map_types else None,
         ),
         dataset=dataset,
         path=path,
@@ -279,6 +280,7 @@ def _read_parquet_chunked(
     columns: Optional[List[str]],
     categories: Optional[List[str]],
     safe: bool,
+    map_types: bool,
     boto3_session: boto3.Session,
     dataset: bool,
     path_root: Optional[str],
@@ -325,6 +327,7 @@ def _read_parquet_chunked(
                     ),
                     categories=categories,
                     safe=safe,
+                    map_types=map_types,
                     use_threads=use_threads,
                     dataset=dataset,
                     path=path,
@@ -404,6 +407,7 @@ def _read_parquet(
     columns: Optional[List[str]],
     categories: Optional[List[str]],
     safe: bool,
+    map_types: bool,
     boto3_session: boto3.Session,
     dataset: bool,
     path_root: Optional[str],
@@ -421,6 +425,7 @@ def _read_parquet(
         ),
         categories=categories,
         safe=safe,
+        map_types=map_types,
         use_threads=use_threads,
         dataset=dataset,
         path=path,
@@ -441,6 +446,7 @@ def read_parquet(
     dataset: bool = False,
     categories: Optional[List[str]] = None,
     safe: bool = True,
+    map_types: bool = True,
     use_threads: bool = True,
     last_modified_begin: Optional[datetime.datetime] = None,
     last_modified_end: Optional[datetime.datetime] = None,
@@ -524,6 +530,10 @@ def read_parquet(
         data in a pandas DataFrame or Series (e.g. timestamps are always
         stored as nanoseconds in pandas). This option controls whether it
         is a safe cast or not.
+    map_types : bool, default True
+        True to convert pyarrow DataTypes to pandas ExtensionDtypes. It is
+        used to override the default pandas type for conversion of built-in
+        pyarrow types or in absence of pandas_metadata in the Table schema.
     use_threads : bool
         True to enable concurrent requests, False to disable multiple threads.
         If enabled os.cpu_count() will be used as the max number of threads.
@@ -597,6 +607,7 @@ def read_parquet(
         "columns": columns,
         "categories": categories,
         "safe": safe,
+        "map_types": map_types,
         "boto3_session": session,
         "dataset": dataset,
         "path_root": path_root,
@@ -633,6 +644,7 @@ def read_parquet_table(
     validate_schema: bool = True,
     categories: Optional[List[str]] = None,
     safe: bool = True,
+    map_types: bool = True,
     chunked: Union[bool, int] = False,
     use_threads: bool = True,
     boto3_session: Optional[boto3.Session] = None,
@@ -699,6 +711,10 @@ def read_parquet_table(
         data in a pandas DataFrame or Series (e.g. timestamps are always
         stored as nanoseconds in pandas). This option controls whether it
         is a safe cast or not.
+    map_types : bool, default True
+        True to convert pyarrow DataTypes to pandas ExtensionDtypes. It is
+        used to override the default pandas type for conversion of built-in
+        pyarrow types or in absence of pandas_metadata in the Table schema.
     chunked : bool
         If True will break the data in smaller DataFrames (Non deterministic number of lines).
         Otherwise return a single DataFrame with the whole data.
@@ -767,6 +783,7 @@ def read_parquet_table(
             validate_schema=validate_schema,
             categories=categories,
             safe=safe,
+            map_types=map_types,
             chunked=chunked,
             dataset=True,
             use_threads=use_threads,

tests/test_s3_parquet.py

@@ -192,6 +192,16 @@ def test_to_parquet_file_dtype(path, use_threads):
     assert str(df2.c1.dtype) == "string"
 
 
+def test_read_parquet_map_types(path):
+    df = pd.DataFrame({"c0": [0, 1, 1, 2]}, dtype=np.int8)
+    file_path = f"{path}0.parquet"
+    wr.s3.to_parquet(df, file_path)
+    df2 = wr.s3.read_parquet(file_path)
+    assert str(df2.c0.dtype) == "Int8"
+    df3 = wr.s3.read_parquet(file_path, map_types=False)
+    assert str(df3.c0.dtype) == "int8"
+
+
 @pytest.mark.parametrize("use_threads", [True, False])
 @pytest.mark.parametrize("max_rows_by_file", [None, 0, 40, 250, 1000])
 def test_parquet_with_size(path, use_threads, max_rows_by_file):