example data-type extension

Author: joshmoore

src/zarr/core/metadata/v3.py

@@ -612,6 +612,7 @@ class DataType(Enum):
     complex128 = "complex128"
     string = "string"
     bytes = "bytes"
+    example = "https://example.com"
 
     @property
     def byte_count(self) -> int | None:
@@ -630,6 +631,7 @@ def byte_count(self) -> int | None:
             DataType.float64: 8,
             DataType.complex64: 8,
             DataType.complex128: 16,
+            DataType.example: 32,
         }
         try:
             return data_type_byte_counts[self]
@@ -657,9 +659,11 @@ def to_numpy_shortname(self) -> str:
             DataType.float64: "f8",
             DataType.complex64: "c8",
             DataType.complex128: "c16",
+            DataType.example: np.dtype(object),
         }
         return data_type_to_numpy[self]
 
+
     def to_numpy(self) -> np.dtypes.StringDType | np.dtypes.ObjectDType | np.dtype[Any]:
         # note: it is not possible to round trip DataType <-> np.dtype
         # due to the fact that DataType.string and DataType.bytes both

tests/test_metadata/test_v3.py

@@ -411,3 +411,18 @@ def test_dtypes(dtype_str: str) -> None:
     else:
         # return type for vlen types may vary depending on numpy version
         assert dt.byte_count is None
+
+async def test_ext_uri() -> None:
+    metadata_dict = {
+        "zarr_format": 3,
+        "node_type": "array",
+        "shape": (1,),
+        "chunk_grid": {"name": "regular", "configuration": {"chunk_shape": (1,)}},
+        "data_type": "https://example.com",
+        "chunk_key_encoding": {"name": "default", "separator": "."},
+        "codecs": [{"name": "bytes"}],
+        "fill_value": 0,
+    }
+    m = ArrayV3Metadata.from_dict(metadata_dict)
+    d = json.loads(m.to_buffer_dict(default_buffer_prototype())["zarr.json"].to_bytes())
+    assert m.data_type is DataType.example