Backport PR pandas-dev#25058: Reading a HDF5 created in py2 (pandas-d…

…ev#25139)

doc/source/whatsnew/v0.24.2.rst

@@ -51,7 +51,7 @@ Bug Fixes
 
 **I/O**
 
--
+- Bug in reading a HDF5 table-format ``DataFrame`` created in Python 2, in Python 3 (:issue:`24925`)
 -
 -
 

pandas/io/pytables.py

@@ -3288,7 +3288,7 @@ def get_attrs(self):
         self.nan_rep = getattr(self.attrs, 'nan_rep', None)
         self.encoding = _ensure_encoding(
             getattr(self.attrs, 'encoding', None))
-        self.errors = getattr(self.attrs, 'errors', 'strict')
+        self.errors = _ensure_decoded(getattr(self.attrs, 'errors', 'strict'))
         self.levels = getattr(
             self.attrs, 'levels', None) or []
         self.index_axes = [

pandas/tests/io/test_pytables.py

@@ -4540,7 +4540,7 @@ def test_pytables_native2_read(self, datapath):
 
     def test_legacy_table_fixed_format_read_py2(self, datapath):
         # GH 24510
-        # legacy table with fixed format written en Python 2
+        # legacy table with fixed format written in Python 2
         with ensure_clean_store(
                 datapath('io', 'data', 'legacy_hdf',
                          'legacy_table_fixed_py2.h5'),
@@ -4552,6 +4552,21 @@ def test_legacy_table_fixed_format_read_py2(self, datapath):
                                                    name='INDEX_NAME'))
             assert_frame_equal(expected, result)
 
+    def test_legacy_table_read_py2(self, datapath):
+        # issue: 24925
+        # legacy table written in Python 2
+        with ensure_clean_store(
+                datapath('io', 'data', 'legacy_hdf',
+                         'legacy_table_py2.h5'),
+                mode='r') as store:
+            result = store.select('table')
+
+        expected = pd.DataFrame({
+            "a": ["a", "b"],
+            "b": [2, 3]
+        })
+        assert_frame_equal(expected, result)
+
     def test_legacy_table_read(self, datapath):
         # legacy table types
         with ensure_clean_store(