Confusion about difference/inconsistency between ctx.from_dataset vs xql.read_xarray

[ahuang11]

2 concerns here:

1. why is one called from_dataset instead of from_xarray, or even read_xarray
2. returns different types

import xarray as xr
import xarray_sql as xql

ds = xr.tutorial.open_dataset('air_temperature')
ctx = xql.XarrayContext()
ctx.from_dataset('air', ds, chunks=dict(time=6))

xarray_sql.sql.XarrayContext

vs

import xarray as xr
import xarray_sql as xql

ds = xr.tutorial.open_dataset('air_temperature')
xql.read_xarray(ds)

pyarrow.lib.RecordBatchReader

[alxmrs]

1. Xarray supports three data structures: DataArray, Dataset, and DataTree. This is more specific to what we support.

2. from_dataset modifies the context state. read_xarray probably shouldn't be exposed as a public API. It should probably be "open_dataset" or the more verbose "dataset_to_table". It is the pure function that translates a dataset to a table, and the Context is the part of the engine that processes the table.

[JHON12091986]

/claim #181

I'll work on fixing read_xarray to work with all xarray datasets.
Will include:

• Analysis of datasets where read_xarray fails
• Implementation of robust handling for different dataset structures
• Tests for various dataset types

PR ready in 12-24 hours.