fix handling of empty DF in pandas 1.0 (#2010)

#### Reference Issues/PRs
<!--Example: Fixes #1234. See also #3456.-->
Fix test failure for new merged readd_batch(), due to problem in Pandas
1.x when arctic returns empty DF


#### What does this implement or fix?

#### Any other comments?

#### Checklist

<details>
  <summary>
   Checklist for code changes...
  </summary>
 
- [ ] Have you updated the relevant docstrings, documentation and
copyright notice?
- [ ] Is this contribution tested against [all ArcticDB's
features](../docs/mkdocs/docs/technical/contributing.md)?
- [ ] Do all exceptions introduced raise appropriate [error
messages](https://docs.arcticdb.io/error_messages/)?
 - [ ] Are API changes highlighted in the PR description?
- [ ] Is the PR labelled as enhancement or bug so it appears in
autogenerated release notes?
</details>

<!--
Thanks for contributing a Pull Request to ArcticDB! Please ensure you
have taken a look at:
- ArcticDB's Code of Conduct:
https://github.com/man-group/ArcticDB/blob/master/CODE_OF_CONDUCT.md
- ArcticDB's Contribution Licensing:
https://github.com/man-group/ArcticDB/blob/master/docs/mkdocs/docs/technical/contributing.md#contribution-licensing
-->

Co-authored-by: Georgi Rusev <Georgi Rusev>

Author: grusev

python/tests/integration/arcticdb/test_read_batch_more.py

@@ -312,8 +312,9 @@ def test_read_batch_multiple_symbols_all_types_data_query_metadata(arctic_librar
         # Filter fload and string condition
         assert_frame_equal_rebuild_index_first(dfqapplied, batch[7].data)
     else:
-        with pytest.raises(AssertionError):
-            assert_frame_equal_rebuild_index_first(dfqapplied, batch[7].data)
+        # Special handling
+        assert dfqapplied.shape[0] == batch[7].data.shape[0]
+        assert dfqapplied.columns.to_list() == batch[7].data.columns.to_list()
 
 def test_read_batch_multiple_wrong_things_at_once(arctic_library):
     """
@@ -458,7 +459,11 @@ def q3(q):
     assert_frame_equal_rebuild_index_first(df_filtered, batch[1].data)
     assert metadata == batch[1].metadata
     df_filtered = q2(df_all)[columns2]
-    assert_frame_equal_rebuild_index_first(df_filtered, batch[2].data)
+    ## When we have [] df then the assertion would be different due to
+    ## a problem in pandas 1.x . We affirm that reurned columns are same
+    ## and the size of frame is same
+    assert df_filtered.shape[0] == batch[2].data.shape[0]
+    assert df_filtered.columns.to_list() == batch[2].data.columns.to_list()
     assert metadata == batch[2].metadata
     df_filtered = q3(df_all)[columns_one_1]
     assert_frame_equal_rebuild_index_first(df_filtered, batch[3].data)