- update to 2.3.3:
* feat: add support for expected termination of transform functions
* feat: export error classes
* feat: unpickle arrays made in Awkward v1 (as v2).
* feat: use pyarrow for string functions
* feat: add support for dictionary encoding from Arrow
* feat: add CPU kernel for to_numpy support for strings/bytestrings
* feat: add support for __arrow_array__
* feat: add __dlpack__, from_dlpack support

- add sle15_python_module_pythons
* bump numpy to 1.17.0
- awkward 2.0.7 actually requires awkward-cpp 8.
* perf: improve Array initialisation performance
- skip python 3.6 (numbas requires scipy)
* Added the ability to select fields inside of nested
records with syntax like array[[“pt”, “eta”, “phi”],
* Fixed Apache Arrow → Awkward Array (ak.from_arrow) for
sliced Arrow arrays: it has an offset attribute we had
been ignoring (#625). Thanks to @trickarcher for the
* Ensured that an ak.Array’s behavior always gets passed to
output of any operation (#626), even if the input is a
dict (#627, affected ak.zip, ak.cartesian, and
* Fixed #629, ak.with_field dropping parameters (#631),
and #634, ak.fill_none not accepting a Record as a fill
* Added a roadmap and upvote/thumbs up instructions to the
* @jpivarski revised the ak.concatenate implementation
(#610), fixed a bug related to axis handling (#612), and
made ak.concatenate friendlier to lazy arrays (#613;
* @chrisburr added setuptools as a requirement because it