rely even more on zarr-cm

Author: d-v-b

src/eopf_geozarr/data_api/geozarr/common.py

@@ -22,7 +22,6 @@
 from pydantic import AfterValidator, BaseModel, Field, model_validator
 from pydantic.experimental.missing_sentinel import MISSING
 from typing_extensions import Protocol, runtime_checkable
-from zarr_cm import ConventionMetadataObject
 
 from eopf_geozarr.data_api.geozarr.projjson import ProjJSON  # noqa: TC001
 
@@ -40,25 +39,6 @@ class UNSET_TYPE:
 GEO_PROJ_VERSION: Final = "0.1"
 
 
-class ZarrConventionMetadata(BaseModel):
-    uuid: str | MISSING = MISSING
-    schema_url: str | MISSING = MISSING
-    spec_url: str | MISSING = MISSING
-    name: str | MISSING = MISSING
-    description: str | MISSING = MISSING
-
-    @model_validator(mode="after")
-    def ensure_identifiable(self) -> Self:
-        if self.uuid is MISSING and self.schema_url is MISSING and self.spec_url is MISSING:
-            raise ValueError("At least one of uuid, schema_url, or spec_url must be provided.")
-
-        return self
-
-
-# Re-export from zarr_cm for backwards compatibility
-ZarrConventionMetadataJSON = ConventionMetadataObject
-
-
 class ProjAttrs(BaseModel, extra="allow"):
     """
     Zarr attributes for coordinate reference system (CRS) encoding.

src/eopf_geozarr/data_api/geozarr/geoproj.py

@@ -7,7 +7,7 @@
 from pydantic import BaseModel, Field, model_validator
 from zarr_cm import geo_proj
 
-from eopf_geozarr.data_api.geozarr.common import ZarrConventionMetadata, is_none
+from eopf_geozarr.data_api.geozarr.common import is_none
 from eopf_geozarr.data_api.geozarr.projjson import ProjJSON  # noqa: TC001
 
 PROJ_UUID = geo_proj.UUID
@@ -16,14 +16,6 @@
 ProjConvention = geo_proj.GeoProjAttrs
 
 
-class ProjConventionMetadata(ZarrConventionMetadata):
-    uuid: str = geo_proj.CMO["uuid"]
-    name: str = geo_proj.CMO["name"]
-    schema_url: str = geo_proj.CMO["schema_url"]
-    spec_url: str = geo_proj.CMO["spec_url"]
-    description: str = geo_proj.CMO["description"]
-
-
 class Proj(BaseModel):
     # At least one of code, wkt2, or projjson must be provided
     code: str | None = Field(None, alias="proj:code", exclude_if=is_none)

src/eopf_geozarr/data_api/geozarr/multiscales/__init__.py

@@ -1,10 +1,9 @@
 """Zarr multiscales convention support."""
 
 from .geozarr import MultiscaleGroupAttrs, MultiscaleMeta
-from .zcm import MULTISCALE_CONVENTION_METADATA, Multiscales, ScaleLevel, ScaleLevelJSON
+from .zcm import Multiscales, ScaleLevel, ScaleLevelJSON
 
 __all__ = [
-    "MULTISCALE_CONVENTION_METADATA",
     "MultiscaleGroupAttrs",
     "MultiscaleMeta",
     "Multiscales",

src/eopf_geozarr/data_api/geozarr/multiscales/geozarr.py

@@ -1,15 +1,16 @@
 from __future__ import annotations
 
-from typing import NotRequired, Self
+from typing import TYPE_CHECKING, NotRequired, Self
 
 from pydantic import BaseModel, model_validator
 from pydantic.experimental.missing_sentinel import MISSING
 from typing_extensions import TypedDict
 
-from eopf_geozarr.data_api.geozarr.common import ZarrConventionMetadata  # noqa: TC001
-
 from . import tms, zcm
 
+if TYPE_CHECKING:
+    from zarr_cm import ConventionMetadataObject
+
 
 class MultiscaleMeta(BaseModel):
     """
@@ -56,7 +57,7 @@ class MultiscaleGroupAttrs(BaseModel):
     multiscales: MultiscaleAttrs
     """
 
-    zarr_conventions: tuple[ZarrConventionMetadata, ...] | MISSING = MISSING
+    zarr_conventions: tuple[ConventionMetadataObject, ...] | MISSING = MISSING
     multiscales: MultiscaleMeta
 
     _zcm_multiscales: zcm.Multiscales | None = None

src/eopf_geozarr/data_api/geozarr/multiscales/zcm.py

@@ -1,14 +1,10 @@
 from __future__ import annotations
 
-from typing import Final
-
 from pydantic import BaseModel, field_validator
 from pydantic.experimental.missing_sentinel import MISSING
 from zarr_cm import ConventionMetadataObject
 from zarr_cm import multiscales as multiscales_cm
 
-from eopf_geozarr.data_api.geozarr.common import ZarrConventionMetadata
-
 # Convention constants from zarr-cm
 CONVENTION_ID = multiscales_cm.UUID
 CONVENTION_SCHEMA_URL = multiscales_cm.SCHEMA_URL
@@ -22,20 +18,9 @@
 MultiscalesJSON = multiscales_cm.MultiscalesAttrs
 MultiscalesConventionAttrsJSON = multiscales_cm.MultiscalesConventionAttrs
 
-# A final dict representation of the Multiscales convention metadata
-MULTISCALE_CONVENTION_METADATA: Final[ConventionMetadataObject] = multiscales_cm.CMO
-
-
-class MultiscaleConventionMetadata(ZarrConventionMetadata):
-    uuid: str = multiscales_cm.CMO["uuid"]
-    schema_url: str = multiscales_cm.CMO["schema_url"]
-    name: str = multiscales_cm.CMO["name"]
-    description: str = multiscales_cm.CMO["description"]
-    spec_url: str = multiscales_cm.CMO["spec_url"]
-
 
 class ZarrConventionAttrs(BaseModel):
-    zarr_conventions: tuple[ZarrConventionMetadata, ...]
+    zarr_conventions: tuple[ConventionMetadataObject, ...]
 
     model_config = {"extra": "allow"}
 
@@ -68,20 +53,15 @@ class MultiscalesAttrs(ZarrConventionAttrs):
     @field_validator("zarr_conventions", mode="after")
     @classmethod
     def ensure_multiscales_convention(
-        cls, value: tuple[ZarrConventionMetadata, ...]
-    ) -> tuple[ZarrConventionMetadata, ...]:
+        cls, value: tuple[ConventionMetadataObject, ...]
+    ) -> tuple[ConventionMetadataObject, ...]:
         """
         Iterate over the elements of zarr_conventions and check that at least one of them is
         multiscales
         """
-        success: bool = False
-        errors: dict[int, ValueError] = {}
-        for idx, convention_meta in enumerate(value):
-            try:
-                MultiscaleConventionMetadata(**convention_meta.model_dump())
-                success = True
-            except ValueError as e:
-                errors[idx] = e
-        if not success:
-            raise ValueError("Multiscales convention not found. Errors: " + str(errors))
+        expected_uuid = multiscales_cm.CMO["uuid"]
+        if not any(c["uuid"] == expected_uuid for c in value):
+            raise ValueError(
+                f"Multiscales convention (uuid={expected_uuid}) not found in zarr_conventions"
+            )
         return value

src/eopf_geozarr/data_api/geozarr/spatial.py

@@ -7,22 +7,14 @@
 from pydantic import BaseModel, Field, model_validator
 from zarr_cm import spatial as spatial_cm
 
-from eopf_geozarr.data_api.geozarr.common import ZarrConventionMetadata, is_none
+from eopf_geozarr.data_api.geozarr.common import is_none
 
 SPATIAL_UUID = spatial_cm.UUID
 
 # Re-export the zarr-cm TypedDict for the convention metadata object
 SpatialConvention = spatial_cm.SpatialAttrs
 
 
-class SpatialConventionMetadata(ZarrConventionMetadata):
-    uuid: str = spatial_cm.CMO["uuid"]
-    name: str = spatial_cm.CMO["name"]
-    schema_url: str = spatial_cm.CMO["schema_url"]
-    spec_url: str = spatial_cm.CMO["spec_url"]
-    description: str = spatial_cm.CMO["description"]
-
-
 class Spatial(BaseModel):
     dimensions: list[str] = Field(alias="spatial:dimensions")  # Required field
     bbox: list[float] | None = Field(None, alias="spatial:bbox", exclude_if=is_none)

src/eopf_geozarr/s2_optimization/s2_multiscale.py

@@ -15,19 +15,20 @@
 from dask.array import from_delayed
 from pydantic.experimental.missing_sentinel import MISSING
 from pyproj import CRS
+from zarr_cm import geo_proj
+from zarr_cm import multiscales as multiscales_cm
+from zarr_cm import spatial as spatial_cm
 
 from eopf_geozarr.conversion.fs_utils import sanitize_dataset_attributes
 from eopf_geozarr.conversion.geozarr import (
     _create_tile_matrix_limits,
     create_native_crs_tile_matrix_set,
 )
-from eopf_geozarr.data_api.geozarr.geoproj import ProjConventionMetadata
 from eopf_geozarr.data_api.geozarr.multiscales import tms, zcm
 from eopf_geozarr.data_api.geozarr.multiscales.geozarr import (
     MultiscaleGroupAttrs,
     MultiscaleMeta,
 )
-from eopf_geozarr.data_api.geozarr.spatial import SpatialConventionMetadata
 from eopf_geozarr.data_api.geozarr.types import (
     CF_SCALE_OFFSET_KEYS,
     XARRAY_ENCODING_KEYS,
@@ -625,9 +626,9 @@ def add_multiscales_metadata_to_parent(
     # Create convention metadata for all three conventions
     multiscale_attrs = MultiscaleGroupAttrs(
         zarr_conventions=(
-            zcm.MultiscaleConventionMetadata(),
-            SpatialConventionMetadata(),
-            ProjConventionMetadata(),
+            multiscales_cm.CMO,
+            spatial_cm.CMO,
+            geo_proj.CMO,
         ),
         multiscales=MultiscaleMeta(
             layout=layout,
@@ -1016,8 +1017,8 @@ def write_geo_metadata(
 
         # Add zarr convention declarations
         conventions = [
-            SpatialConventionMetadata().model_dump(),
-            ProjConventionMetadata().model_dump(),
+            spatial_cm.CMO,
+            geo_proj.CMO,
         ]
         dataset.attrs["zarr_conventions"] = conventions
 

tests/test_data_api/test_geoproj.py

@@ -5,44 +5,10 @@
 from pydantic import ValidationError
 from pydantic_zarr.core import tuplify_json
 
-from eopf_geozarr.data_api.geozarr.geoproj import GeoProj, Proj, ProjConventionMetadata
+from eopf_geozarr.data_api.geozarr.geoproj import GeoProj, Proj
 from tests.test_data_api.conftest import view_json_diff
 
 
-class TestProjConventionMetadata:
-    """Test the ProjConventionMetadata class."""
-
-    def test_default_values(self) -> None:
-        """Test that default values are correctly set."""
-        metadata = ProjConventionMetadata()
-
-        assert metadata.uuid == "f17cb550-5864-4468-aeb7-f3180cfb622f"
-        assert metadata.name == "proj:"
-        assert (
-            metadata.schema_url
-            == "https://raw.githubusercontent.com/zarr-experimental/geo-proj/refs/tags/v1/schema.json"
-        )
-        assert (
-            metadata.spec_url == "https://github.com/zarr-experimental/geo-proj/blob/v1/README.md"
-        )
-        assert metadata.description == "Coordinate reference system information for geospatial data"
-
-    def test_serialization(self) -> None:
-        """Test that metadata can be serialized correctly."""
-        metadata = ProjConventionMetadata()
-        result = metadata.model_dump()
-
-        expected = {
-            "uuid": "f17cb550-5864-4468-aeb7-f3180cfb622f",
-            "name": "proj:",
-            "schema_url": "https://raw.githubusercontent.com/zarr-experimental/geo-proj/refs/tags/v1/schema.json",
-            "spec_url": "https://github.com/zarr-experimental/geo-proj/blob/v1/README.md",
-            "description": "Coordinate reference system information for geospatial data",
-        }
-
-        assert result == expected
-
-
 class TestProj:
     """Test the Proj model class."""
 

tests/test_data_api/test_geozarr/test_multiscales/test_geozarr.py

@@ -2,6 +2,7 @@
 
 import pytest
 from pydantic.experimental.missing_sentinel import MISSING
+from zarr_cm import multiscales as multiscales_cm
 
 from eopf_geozarr.data_api.geozarr.multiscales import tms, zcm
 from eopf_geozarr.data_api.geozarr.multiscales.geozarr import (
@@ -27,7 +28,7 @@ def test_multiscale_group_attrs(multiscale_flavor: set[Literal["zcm", "tms"]]) -
             ),
         )
         zcm_meta = zcm.Multiscales(layout=layout, resampling_method="nearest").model_dump()
-        zarr_conventions_meta = (zcm.MULTISCALE_CONVENTION_METADATA,)
+        zarr_conventions_meta = (multiscales_cm.CMO,)
     if "tms" in multiscale_flavor:
         tile_matrix_set = tms.TileMatrixSet(
             id="example_tms",

tests/test_data_api/test_spatial.py

@@ -5,39 +5,7 @@
 import pytest
 from pydantic import ValidationError
 
-from eopf_geozarr.data_api.geozarr.spatial import Spatial, SpatialConventionMetadata
-
-
-class TestSpatialConventionMetadata:
-    """Test the SpatialConventionMetadata class."""
-
-    def test_default_values(self) -> None:
-        """Test that default values are correctly set."""
-        metadata = SpatialConventionMetadata()
-
-        assert metadata.uuid == "689b58e2-cf7b-45e0-9fff-9cfc0883d6b4"
-        assert metadata.name == "spatial:"
-        assert (
-            metadata.schema_url
-            == "https://raw.githubusercontent.com/zarr-conventions/spatial/refs/tags/v1/schema.json"
-        )
-        assert metadata.spec_url == "https://github.com/zarr-conventions/spatial/blob/v1/README.md"
-        assert metadata.description == "Spatial coordinate information"
-
-    def test_serialization(self) -> None:
-        """Test that metadata can be serialized correctly."""
-        metadata = SpatialConventionMetadata()
-        result = metadata.model_dump()
-
-        expected = {
-            "uuid": "689b58e2-cf7b-45e0-9fff-9cfc0883d6b4",
-            "name": "spatial:",
-            "schema_url": "https://raw.githubusercontent.com/zarr-conventions/spatial/refs/tags/v1/schema.json",
-            "spec_url": "https://github.com/zarr-conventions/spatial/blob/v1/README.md",
-            "description": "Spatial coordinate information",
-        }
-
-        assert result == expected
+from eopf_geozarr.data_api.geozarr.spatial import Spatial
 
 
 class TestSpatial: