Fstd2nc

* New .to_xarray_list() method to return list of variables wrapped in
individual Datasets.
* New command-line --bounds flag, to output grid cell boundaries.
* Remove "leadtime", "reftime", "surface_type", and "station" from
list of coordinates, since some netCDF viewers have trouble with them
being listed in there.
* Fix crash in xarray interface when interpolation is enabled.

* Allow fstd2nc.Buffer objects to be serialized / deserialized.
This allows the objects to be used with the Python multiprocessing
library. Note: not yet compatible with dask multiprocessing, which
seems to use a different mechanism for serialization?
* New keyword args add_px and add_gz for computing pressure or
geopotential height fields in Buffer.to_xarray(). Not available for
command-line usage.
* Allow asymmetric cropping with the --crop-to-smallest-grid option,
i.e., when the smaller grid is not centered inside the larger grid.
* Fixed an address overflow bug when using the xarray interface on
files larger than 4GB.

* Improved memory usage of dask wrapper. Don't pre-load entire block.
* Some miscellaneous speedups to dask wrapper and netCDF conversion.
* New command-line option --only-diag-level, to select only the near-
surface level data (opposite of --ignore-diag-level).
* Fixed --opdict / --vardict parsing for Python 3.9 (it was using a
deprecated xml function).
* Moved 'interp' module to 'gridhacks', since it contains other
non-interpolation features now.
* Added --yin and --yang options to select one of the two pieces of a
yin-yang grid.
* Added --crop-to-smallest-grid option to handle case where there is
a grid that's slightly nested inside of another grid (such as with
GEM output for dynamics vs. permanent bus?). This option will trim
all the data to the smallest of such grids.
* Experimental --turbo option to speed up conversions in the case where
it needs to extract data from many (thousands) of files.
* Allow progress bar to be turned on, even if there does not appear to
be a TTY available. Useful for interactive work within a Jupyter
notebook.

* Fixed a bug with xarray interface when used with masked fields.
Sometimes it was identifying a directory page as the mask of a field,
with catastrophic consequences.

* Fixed a bug with rotated grids. They were not properly oriented for
the particular case where the rlon values didn't cross the numerical
dateline for the coordinate.
* Added --quiet option to suppress messages from the converter.

* Fixed regression where some data records were being identified as
(unused) coordinate records, causing them to be omitted from the
netCDF file.
* Metadata can now be applied based on forced-renamed variable names.
E.g., can have sections [SSH_vgrid0] and [SSH_vgrid3] in the metadata
file to control the attributes separately for each vertical
coordinate type of an SSH variable.
* Fixed the handling of strings in --filter option. It should now work
with normal (not byte) strings without any padding.
* Also added 'np' shortcut to filters to allow numpy functions to be
used in the filter criteria (such as np.isin).