Riptable

Changes:
* Improve Dataset.from_pandas tz handling
* Fix bad comments for TimeZone.normalize_tz_to_tzdb_name()
* Add Hong Kong timezone support to DateTimeNano
* Fix tests so they write to proper tmpdir
* Add Korea/Japan timezones
* Docstring: rt.merge_asof()
* Doc: Remove riptable.rst file in docs/source
* Docstring: Cat.expand_array() (add note)
* Switch from short tz names to tz db identifiers
* Convert setup.py to pyproject build
* Fix annotations for concat_rows and concat_columns
* Docs: Use groupwise ordering, remove "Module Contents" header
* Make _version.py an optional version source

Changes:
* Add cummin and cummax to groupbyops
* Docstring: DateTimeNano class
* Update to more modern compilers
* rt.Dataset.__setitem__: handle integral row index
* Pin to riptide_cpp-1.14.0a0
* Docstring: Categorical class, plus Categoricals User Guide
* Docstrings: Cat.invalid_set(), Cat.invalid_category
* Docstrings: Cat.isnan(), Cat.isnotnan()
* Doc: Add TOC to API Reference
* Pin to riptide_cpp-1.14.0 final
* Don't auto-update conda
* GH CI: Disable auto-conda-update everywhere

Changes:
* Allow creating multiple columns for a ds with length 1
* Make riptable a noarch package
* Missed GH CI changes for noarch builds
* Update cov for riptable module testing
* Fix rt.Struct._A
* Fix empty Dataset copy()
* Have merge as of sort and unsort
* Improve riptable test package selection

Changes:
* Update README with supported Python.
* Add boa to conda requirements to support mama builds
* Code change: Raise an error when start_date array length isn't 1 or length of arr
* Change deprecated np.bool to np.bool_
* Switch GitHub CI to use mamba
* Add --quote option to gen_requirements.py
* simplify quoting

Changes:
* Fix merge2 w.r.t. empty input datasets
* Follow-up fix on merge2 handling empty input datasets
* Make the dtype of inputs different to test dtype handling
* Code change: For DTN constructor, fix conversion of np.datetime64 arrays.
* Add edge case for ncountgroup when len(ikey) == 0
* Docstring: FA.doc: Add links, remove some print statements and output name
* Docstring: FA.info: Minor edits
* Docstring: FA.inv: Minor edits
* Docstring: fa.save()
* Code change: For DTN constructor, convert list of Matlab datenums to FastArray
* Docstring: ds.merge_lookup()
* Docstring: ds.filter(): Typo fix and other minor fixes
* Docstring: FA.fill_invalid(): Minor edits

Changes:
* Replace py-3.8 with py-3.11
* Docstring: FA.copy(): Fix line breaks and See Alsos
* Docstring: FA.between(): Minor edits
* Docstring: FA.nunique(): Minor edits
* Docstring: FA.shift(): Minor edits
* Docstring: FA.sample(): Fix line breaks
* Docstring: FA._np: Minor edits
* Docstring: FA.duplicated(): Shorten line lengths and fix See Alsos
* Docstring: FA.copy_invalid(): Minor edits
* Docstring: FA.differs(): Line breaks and minor edits
* Docstring: FA.diff(): Minor edits
* Docstrings: Fix other line breaks