Fix small open_virtual_dataset bugs

earthaccess/dmrpp_zarr.py

@@ -92,8 +92,8 @@ def open_virtual_mfdataset(
     import xarray as xr
 
     if access == "direct":
-        fs = earthaccess.get_s3_filesystem(results=granules[0])
-        fs.storage_options["anon"] = False  # type: ignore
+        fs = earthaccess.get_s3_filesystem(results=granules)  # type: ignore
+        fs.storage_options["anon"] = False
     else:
         fs = earthaccess.get_fsspec_https_session()
     if parallel:
@@ -114,7 +114,7 @@ def open_virtual_mfdataset(
                 filetype="dmrpp",  # type: ignore
                 group=group,
                 indexes={},
-                reader_options={"storage_options": fs.storage_options},  # type: ignore
+                reader_options={"storage_options": fs.storage_options},
             )
         )
     if preprocess is not None:
@@ -127,6 +127,7 @@ def open_virtual_mfdataset(
         vds = xr.combine_nested(vdatasets, **xr_combine_nested_kwargs)
     if load:
         refs = vds.virtualize.to_kerchunk(filepath=None, format="dict")
+        protocol = "s3" if "s3" in fs.protocol else fs.protocol
         return xr.open_dataset(
             "reference://",
             engine="zarr",
@@ -135,8 +136,8 @@ def open_virtual_mfdataset(
                 "consolidated": False,
                 "storage_options": {
                     "fo": refs,  # codespell:ignore
-                    "remote_protocol": fs.protocol,
-                    "remote_options": fs.storage_options,  # type: ignore
+                    "remote_protocol": protocol,
+                    "remote_options": fs.storage_options,
                 },
             },
         )

pyproject.toml

@@ -65,7 +65,10 @@ kerchunk = [
   "xarray",
 ]
 virtualizarr = [
-  "virtualizarr >=1.2.0"
+  "virtualizarr >=1.2.0",
+  "dask",
+  "zarr",
+  "h5py >=3.6.0",
 ]
 dev = [
   "bump-my-version >=0.10.0",

tests/integration/test_virtualizarr.py

@@ -15,41 +15,24 @@
 logger.info(f"earthaccess version: {earthaccess.__version__}")
 
 
-@pytest.fixture(scope="module", params=["MUR25-JPL-L4-GLOB-v04.2"])
+@pytest.fixture(
+    scope="module",
+    params=[
+        "MUR25-JPL-L4-GLOB-v04.2",
+        "AVHRR_OI-NCEI-L4-GLOB-v2.1",
+        "M2T1NXSLV",
+    ],
+)
 def granule(request):
     granules = earthaccess.search_data(
         count=1, temporal=("2024"), short_name=request.param
     )
     return granules[0]
 
 
-def test_dmrpp(granule):
-    from virtualizarr import open_virtual_dataset  # type: ignore
-
-    fs = earthaccess.get_fsspec_https_session()
-    data_path = granule.data_links(access="indirect")[0]
-    dmrpp_path = data_path + ".dmrpp"
-
-    result = open_virtual_dataset(
-        dmrpp_path,
-        filetype="dmrpp",  # type: ignore
-        indexes={},
-        reader_options={"storage_options": fs.storage_options},  # type: ignore
-    )
-
-    expected = open_virtual_dataset(
-        data_path,
-        indexes={},
-        reader_options={"storage_options": fs.storage_options},  # type: ignore
-    )
-
-    # TODO: replace with xr.testing when virtualizarr fill_val is fixed (https://github.com/zarr-developers/VirtualiZarr/issues/287)
-    # and dmrpp deflateLevel (zlib compression level) is always present (https://github.com/OPENDAP/bes/issues/954)
-    for var in result.variables:
-        assert var in expected.variables
-        assert result[var].dims == expected[var].dims
-        assert result[var].shape == expected[var].shape
-        assert result[var].dtype == expected[var].dtype
-        assert result[var].data.manifest == expected[var].data.manifest
-    assert set(result.coords) == set(expected.coords)
-    assert result.attrs == expected.attrs
+def test_open_virtual_dataset(granule):
+    # Simply check that the dmrpp can be found, parsed, and loaded. Actual parser result is checked in virtualizarr
+    vds = earthaccess.open_virtual_dataset(granule, load=False)
+    assert vds is not None
+    vds_load = earthaccess.open_virtual_dataset(granule, load=True)
+    assert vds_load is not None