simplify test

python/pyarrow/parquet.py

@@ -214,7 +214,7 @@ class ParquetFile:
         Coalesce and issue file reads in parallel to improve performance on
         high-latency filesystems (e.g. S3). If True, Arrow will use a
         background I/O thread pool.
-    coerce_int96_timestamp_unit: str, default None.
+    coerce_int96_timestamp_unit : str, default None.
         Cast timestamps that are stored in INT96 format to a particular
         resolution (e.g. 'ms'). Setting to None is equivalent to 'ns'
         and therefore INT96 timestamps will be infered as timestamps
@@ -1262,7 +1262,7 @@ class ParquetDataset:
     use_legacy_dataset=False. If using a filesystem layer that itself
     performs readahead (e.g. fsspec's S3FS), disable readahead for best
     results.
-coerce_int96_timestamp_unit: str, default None.
+coerce_int96_timestamp_unit : str, default None.
         Cast timestamps that are stored in INT96 format to a particular
         resolution (e.g. 'ms'). Setting to None is equivalent to 'ns'
         and therefore INT96 timestamps will be infered as timestamps

python/pyarrow/tests/parquet/test_datetime.py

@@ -15,7 +15,6 @@
 # specific language governing permissions and limitations
 # under the License.
 
-import sys
 import datetime
 import io
 
@@ -298,9 +297,6 @@ def test_coerce_int96_timestamp_unit(unit):
 @pytest.mark.parametrize('pq_reader_method', ['ParquetFile', 'read_table'])
 def test_coerce_int96_timestamp_overflow(pq_reader_method, tempdir):
 
-    if sys.platform in ("win32", "cygwin"):
-        pytest.skip("Getting datetime.strftime() error on Windows")
-
     def get_table(pq_reader_method, filename, **kwargs):
         if pq_reader_method == "ParquetFile":
             return pq.ParquetFile(filename, **kwargs).read()
@@ -313,37 +309,24 @@ def get_table(pq_reader_method, filename, **kwargs):
         datetime.datetime(2000, 1, 1),
         datetime.datetime(3000, 1, 1)
     ]
-    oob_dts_str = [
-        x.strftime("%Y-%m-%s %H:%M:%S.%f")
-        for x in oob_dts
-    ]
     df = pd.DataFrame({"a": oob_dts})
-    a_df = pa.Table.from_pandas(df)
+    table = pa.table(df)
 
     filename = tempdir / "test_round_trip_overflow.parquet"
-    pq.write_table(a_df, filename, use_deprecated_int96_timestamps=True,
+    pq.write_table(table, filename, use_deprecated_int96_timestamps=True,
                    version="1.0")
 
+    # with the default resolution of ns, we get wrong values for INT96
+    # that are out of bounds for nanosecond range
     tab_error = get_table(pq_reader_method, filename)
-    df_error = tab_error.to_pandas(timestamp_as_object=True)
-    out_error = [
-        x.strftime("%Y-%m-%s %H:%M:%S.%f")
-        for x in df_error["a"].tolist()
-    ]
-
-    assert out_error != oob_dts_str
+    assert tab_error["a"].to_pylist() != oob_dts
 
+    # avoid this overflow by specifying the resolution to use for INT96 values
     tab_correct = get_table(
-        pq_reader_method,
-        filename,
-        coerce_int96_timestamp_unit="s"
+        pq_reader_method, filename, coerce_int96_timestamp_unit="s"
     )
     df_correct = tab_correct.to_pandas(timestamp_as_object=True)
-    out_correct = [
-        x.strftime("%Y-%m-%s %H:%M:%S.%f")
-        for x in df_correct.a.tolist()
-    ]
-    assert out_correct == oob_dts_str
+    tm.assert_frame_equal(df, df_correct)
 
 
 def test_timestamp_restore_timezone():