Encoding dataarrays

Make sure that dataarrays are encoded as well to hopefully solve IndexError issues when datetimes are not coordinates.

Works on #59

.github/workflows/tests.yml

@@ -10,7 +10,7 @@ jobs:
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
-        python-version: ["3.9", "3.10", "3.11"]
+        python-version: ["3.9", "3.10", "3.11", "3.12"]
         os: [windows-latest, ubuntu-latest, macos-latest]
         pydantic-version: ["<2", ">=2"]
 

noxfile.py

@@ -9,6 +9,8 @@
 python_versions = workflow["jobs"]["run"]["strategy"]["matrix"]["python-version"]
 pydantic_versions = workflow["jobs"]["run"]["strategy"]["matrix"]["pydantic-version"]
 
+nox.options.default_venv_backend = "uv"
+
 
 @nox.session(python=python_versions)
 @nox.parametrize("pydantic", pydantic_versions)

requirements-dev.txt

@@ -1,4 +1,5 @@
 black
+cf_xarray
 check-manifest
 doctr
 h5netcdf

tests/conftest.py

@@ -8,7 +8,7 @@
 server_path = Path(__file__).parent / "server.py"
 
 
-@pytest.fixture
+@pytest.fixture(scope="session")
 def xpublish_server(xprocess):
     """Launch an Xpublish server in the background.
 

tests/server.py

@@ -1,5 +1,7 @@
 """Test OpenDAP server with air temperature dataset."""
 
+from pathlib import Path
+
 import numpy as np
 import xarray.tutorial
 import xpublish
@@ -14,8 +16,17 @@
 ds_attrs_cast.attrs["npint"] = np.int16(16)
 ds_attrs_cast.attrs["npintthirtytwo"] = np.int32(32)
 
+ds_coordinates_encoding = xarray.open_dataset(
+    Path(__file__).parent / "min_coordinates_encoding.nc",
+)
+
 rest = xpublish.Rest(
-    {"air": ds, "attrs_quote": ds_attrs_quote, "attrs_cast": ds_attrs_cast},
+    {
+        "air": ds,
+        "attrs_quote": ds_attrs_quote,
+        "attrs_cast": ds_attrs_cast,
+        "coords_encoding": ds_coordinates_encoding,
+    },
     plugins={"opendap": OpenDapPlugin()},
 )
 

tests/test_server.py

@@ -7,6 +7,7 @@
 """
 
 import sys
+from pathlib import Path
 
 import netCDF4
 import pytest
@@ -31,7 +32,9 @@ def test_default_xarray_engine(xpublish_server, dataset):
     """Test opening OpenDAP air dataset with default Xarray engine."""
     url = f"{xpublish_server}/datasets/air/opendap"
     ds = xr.open_dataset(url)
-    assert ds == dataset
+    # assert ds == dataset
+    xr.testing.assert_equal(ds, dataset)
+    # xr.testing.assert_identical(ds, dataset)
 
 
 @pytest.mark.skipif(
@@ -76,3 +79,22 @@ def test_attrs_types(xpublish_server):
 
     assert ds.attrs["npint"] == 16
     assert ds.attrs["npintthirtytwo"] == 32
+
+
+@pytest.mark.skipif(
+    sys.platform == "win32",
+    reason="NetCDF4 is failing on Windows Github Actions workers",
+)
+def test_coordinates_persist_correctly(xpublish_server):
+    """Test that encoded coordinate data makes it through OpenDAP correctly.
+
+    xref: https://github.com/xpublish-community/xpublish/discussions/246
+    """
+    dataset = xr.open_dataset(Path(__file__).parent / "min_coordinates_encoding.nc")
+
+    url = f"{xpublish_server}/datasets/coords_encoding/opendap"
+    ds = xr.open_dataset(url)
+    assert ds.dims == dataset.dims
+    assert ds.coords == dataset.coords
+    assert ds.variables == dataset.variables
+    assert ds == dataset

xpublish_opendap/dap_xarray.py

@@ -90,10 +90,12 @@ def dap_dimension(da: xr.DataArray) -> dap.Array:
 
 def dap_grid(da: xr.DataArray, dims: dict[str, dap.Array]) -> dap.Grid:
     """Transform an xarray DataArray into a DAP Grid."""
+    encoded_da = xr.conventions.encode_cf_variable(da.variable)
+
     data_grid = dap.Grid(
         name=da.name,
-        data=da.astype(da.dtype).data,
-        dtype=dap_dtype(da),
+        data=encoded_da.astype(encoded_da.dtype).data,
+        dtype=dap_dtype(encoded_da),
         dimensions=[dims[dim] for dim in da.dims],
     )