Fstd2nc

* Fixed bug in make_index / open_index when storing scalar parameters.
* Fixed bug in handling large files with masked fields. The mask
addresses were stored in 32-bits, causing an overflow in the address.

* Fixed preferred chunking when creating xarray objects via
open_dataset().
* Reduced the memory footprint of Buffer objects, and some small
speedups in the header decoding.
* Fixed CCC decoder. It was broken after a recent refactor, and there
was bug where the min value wasn't being added to the decoded
values.
* Added xarray backend for CCC files, so they can be opened with
xr.open_dataset(...,engine='ccc').
* Catch some possible runtime issues when fetching data from xarray
interface, such as if the file was removed from disk. Return
missing values if a problem occurs.
* Improved efficiency of --fstd-compat option (less padding used).
* Fixed a bug when using --bounds on surface data that has a
'sfc_type' dimension.
* Silence some warnings from recent vgrid libraries, when querying for
the existence of certain parameters.
* Added an 'accum' axis for accumulated quantities. Also added an
'--accum-vars' option to explicitly set fields that should be
treated as accumulated.
* Explicitly use a spherical globe in the projections returned by
fstd2nc.extra.get_crs().
* New helper methods for working with ridiculously large datasets
with xarray.
- fstd2nc.Buffer.make_index() pre-scans the input files and
generates an 'index' file on disk which contains all the variable
structures.
- fstd2nc.Buffer.open_index() reads the above 'index' file and
returns an xarray Dataset representing the FSTD data. This step
should be *much* faster since the files were already scanned
in the first step.

* Added north_pole_grid_longitude attribute for rotated grids.
* Fixed missing vertical coordinates in series data.
* Fixed bug when using --forecast-axis with series data.
* Fixed bug with --yin and --yang when there are non-YY fields in the
file.
* Improved precision of lat/lon axes for non-rotated 'Z' grids.
* Switch to default of fused=False in .to_xarray(), since fusing the
records can give a performance hit for certain access patterns.
* Added $CMCCONST to search path for --opdict.
* Check if file specified by --metadata-file actually exists, instead
of ignoring the missing file.
* Fixed get_crs() for data generated with no_adjust_rlon=True.
* Register this as a backend entrypoint for xarray, so now it should
work from xarray.open_dataset().
* Fixed interpolation for vector fields like UU, VV.
* Added support for RPN extended date range (for years 0-9999).
* Registered a .to_fstd() method for xarray.Dataset objects, for some
preliminary testing.
* Fixed an array transpose happening in some conditions for
.to_fstpy() and .from_fstpy().
* Debian packaging rules provided by Tom Kralidis.

* From Python interface, allow interp parameter to accept existing
grid ids as input.
* Fixed a crash when converting masked data when some of the mask
records are missing.

* Fixed crop_to_smallest_grid option, which was broken in .to_xarray()
after fused records were implemented.
* Added doc string in .from_fstpy().
* Improved support for mesh grids (grtyp='M'). Removed extra
dimensions from mesh_face_nodes variable that shouldn't be there,
and got rid of warning message about decoding '' record.
* Fixed interp option so it now works in dask clusters.
* Better error message when an empty list of files is passed in.

* Added vertical bounds information to --bounds option.
* Removed "coordinates" from internal encoding of xarray DataArray
objects. This makes it easier for the user to remove it from the
attributes in xarray to_netcdf routine.
* Added CCC file interface to the project (work in progress).
* Refactored internal _decode routines to be class methods, which
means the full Buffer object no longer needs to be included in
dask graphs (greatly reducing the size of dask graphs).
* Automatically "fuse" records together in xarray interface, which
increases the chunk sizes in the dask graphs (making operations work
more efficiently in dask).