Feature: get_source_data

[jmilloy]

Datasource

>>> node = podpac.data.Array(
...             source=np.ones((3, 4)),
...             coordinates=podpac.Coordinates([range(3), range(4)], ["lat", "lon"]),
...         )
>>> 
>>> node.get_source_data()
<xarray.UnitsDataArray (lat: 3, lon: 4)>
array([[1., 1., 1., 1.],
       [1., 1., 1., 1.],
       [1., 1., 1., 1.]])
Coordinates:
  * lat      (lat) float64 0.0 1.0 2.0
  * lon      (lon) float64 0.0 1.0 2.0 3.0
Attributes:
    layer_style:   <podpac.core.style.Style object at 0x7f7c09fcb730>
    crs:           +proj=longlat +datum=WGS84 +no_defs +vunits=m
    geotransform:  (-0.5, 1.0, 0.0, -0.5, 0.0, 1.0)

TileCompositor

>>> a = ArrayRaw(source=np.arange(5) + 100, coordinates=podpac.Coordinates([[0, 1, 2, 3, 4]], dims=["lat"]))
>>> b = ArrayRaw(source=np.arange(5) + 200, coordinates=podpac.Coordinates([[5, 6, 7, 8, 9]], dims=["lat"]))
>>> c = ArrayRaw(source=np.arange(5) + 300, coordinates=podpac.Coordinates([[10, 11, 12, 13, 14]], dims=["lat"]))
>>> node = TileCompositorRaw(sources=[a, b, c])
>>> node.get_source_data()
<xarray.UnitsDataArray (lat: 15)>
array([100., 101., 102., 103., 104., 200., 201., 202., 203., 204., 300.,
       301., 302., 303., 304.])
Coordinates:
  * lat      (lat) float64 0.0 1.0 2.0 3.0 4.0 5.0 ... 10.0 11.0 12.0 13.0 14.0
Attributes:
    layer_style:  <podpac.core.style.Style object at 0x7f7bb627a340>
    crs:          +proj=longlat +datum=WGS84 +no_defs +vunits=m

Bounds

>>> node.get_source_data({"lon": (1.5, 4.5)})
<xarray.UnitsDataArray (lat: 3, lon: 2)>
array([[1., 1.],
       [1., 1.],
       [1., 1.]])
Coordinates:
  * lat      (lat) float64 0.0 1.0 2.0
  * lon      (lon) float64 2.0 3.0
Attributes:
    layer_style:   <podpac.core.style.Style object at 0x7f7bb6566190>
    crs:           +proj=longlat +datum=WGS84 +no_defs +vunits=m
    geotransform:  (1.5, 1.0, 0.0, -0.5, 0.0, 1.0)

closes #468

[jmilloy]

@mpu-creare This is a simple enhancement. It is just a general method implemented in the DataSource class and the TileCompositor class. Neither method calls eval, so it is a true enhancement and not just an API wrapper. However, some child datasource or datalib nodes could probably override get_source_data to be more efficient, such as perhaps the Rasterio node. At this point, I haven't done that (we can do that as needed as optimization), but I want to know if you would prefer to do some of that for this PR.

[mpu-creare]

I'm fine approving this. I did have a few suggestions on wording and potential improvements, but I'll leave it up to you.

The major thing is the OGR node which will NOT work with this implementation. It will throw an exception. Any ideas?