Merge branch 'main' of github.com:zarr-developers/zarr-python into chore/dissolve-async-array

Author: d-v-b

.pre-commit-config.yaml

@@ -11,7 +11,7 @@ default_language_version:
 
 repos:
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.14.10
+    rev: v0.14.14
     hooks:
       - id: ruff-check
         args: ["--fix", "--show-fixes"]

changes/3668.feature.md

@@ -0,0 +1,4 @@
+Exposes the array runtime configuration as an attribute called `config` on the `Array` and
+`AsyncArray` classes. The previous `AsyncArray._config` attribute is now a deprecated alias for `AsyncArray.config`.
+
+Adds a method for creating a new `Array` / `AsyncArray` instance with a new runtime configuration, and fixes inaccurate documentation about the `write_empty_chunks` configuration parameter.

docs/user-guide/arrays.md

@@ -154,6 +154,32 @@ z.append(np.vstack([a, a]), axis=1)
 print(f"Shape after second append: {z.shape}")
 ```
 
+## Runtime configuration
+
+Zarr arrays are parametrized with a configuration that determines certain aspects of array behavior.
+
+We currently support two configuration options for arrays: `write_empty_chunks` and `order`.
+
+| field | type | default | description |
+| - |     - | - | - |
+| `write_empty_chunks` | `bool` | `False` | Controls whether empty chunks are written to storage. See [Empty chunks](performance.md#empty-chunks).
+| `order` | `Literal["C", "F"]` | `"C"` | The memory layout of arrays returned when reading data from the store.
+
+You can specify the configuration when you create an array with the `config` keyword argument.
+`config` can be passed as either a `dict` or an `ArrayConfig` object.
+
+```python exec="true" session="arrays" source="above" result="ansi"
+arr = zarr.create_array({}, shape=(10,), dtype='int8', config={"write_empty_chunks": True})
+print(arr.config)
+```
+
+To get an array view with a different config, use the `with_config` method.
+
+```python exec="true" session="arrays" source="above" result="ansi"
+arr_f = arr.with_config({"order": "F"})
+print(arr_f.config)
+```
+
 ## Compressors
 
 A number of different compressors can be used with Zarr. Zarr includes Blosc,

docs/user-guide/performance.md

@@ -125,7 +125,14 @@ This optimization prevents storing redundant objects and can speed up reads, but
 added computation during array writes, since the contents of
 each chunk must be compared to the fill value, and these advantages are contingent on the content of the array.
 If you know that your data will form chunks that are almost always non-empty, then there is no advantage to the optimization described above.
-In this case, creating an array with `write_empty_chunks=True` (the default) will instruct Zarr to write every chunk without checking for emptiness.
+In this case, creating an array with `write_empty_chunks=True` will instruct Zarr to write every chunk without checking for emptiness.
+
+The default value of `write_empty_chunks` is `False`:
+
+```python exec="true" session="performance" source="above" result="ansi"
+arr = zarr.create_array(store={}, shape=(1,), dtype='uint8')
+assert arr.config.write_empty_chunks == False
+```
 
 The following example illustrates the effect of the `write_empty_chunks` flag on
 the time required to write an array with different values.:

src/zarr/core/array.py

@@ -144,7 +144,7 @@
     from zarr.codecs.sharding import ShardingCodecIndexLocation
     from zarr.core.dtype.wrapper import TBaseDType, TBaseScalar
     from zarr.storage import StoreLike
-    from zarr.types import AnyArray, AnyAsyncArray, AsyncArrayV2, AsyncArrayV3
+    from zarr.types import AnyArray, AnyAsyncArray, ArrayV2, ArrayV3, AsyncArrayV2, AsyncArrayV3
 
 
 # Array and AsyncArray are defined in the base ``zarr`` namespace
@@ -300,14 +300,14 @@ class AsyncArray(Generic[T_ArrayMetadata]):
         The path to the Zarr store.
     codec_pipeline : CodecPipeline
         The codec pipeline used for encoding and decoding chunks.
-    _config : ArrayConfig
+    config : ArrayConfig
         The runtime configuration of the array.
     """
 
     metadata: T_ArrayMetadata
     store_path: StorePath
     codec_pipeline: CodecPipeline = field(init=False)
-    _config: ArrayConfig
+    config: ArrayConfig
 
     @overload
     def __init__(
@@ -336,7 +336,7 @@ def __init__(
 
         object.__setattr__(self, "metadata", metadata_parsed)
         object.__setattr__(self, "store_path", store_path)
-        object.__setattr__(self, "_config", config_parsed)
+        object.__setattr__(self, "config", config_parsed)
         object.__setattr__(
             self,
             "codec_pipeline",
@@ -1012,6 +1012,11 @@ async def example():
     def store(self) -> Store:
         return self.store_path.store
 
+    @property
+    @deprecated("Use AsyncArray.config instead.", category=ZarrDeprecationWarning)
+    def _config(self) -> ArrayConfig:
+        return self.config
+
     @property
     def ndim(self) -> int:
         """Returns the number of dimensions in the Array.
@@ -1165,7 +1170,7 @@ def order(self) -> MemoryOrder:
         if self.metadata.zarr_format == 2:
             return self.metadata.order
         else:
-            return self._config.order
+            return self.config.order
 
     @property
     def attrs(self) -> dict[str, JSON]:
@@ -1298,6 +1303,35 @@ def _nshards(self) -> int:
         """
         return product(self._shard_grid_shape)
 
+    @overload
+    def with_config(self: AsyncArrayV2, config: ArrayConfigLike) -> AsyncArrayV2: ...
+
+    @overload
+    def with_config(self: AsyncArrayV3, config: ArrayConfigLike) -> AsyncArrayV3: ...
+
+    def with_config(self, config: ArrayConfigLike) -> Self:
+        """
+        Return a copy of this Array with a new runtime configuration.
+
+        Parameters
+        ----------
+
+        config : ArrayConfigLike
+            The runtime config for the new Array. Any keys not specified will be inherited
+            from the current array's config.
+
+        Returns
+        -------
+        A new Array
+        """
+        if isinstance(config, ArrayConfig):
+            new_config = config
+        else:
+            # Merge new config with existing config, so missing keys are inherited
+            # from the current array rather than from global defaults
+            new_config = ArrayConfig(**{**self.config.to_dict(), **config})  # type: ignore[arg-type]
+        return type(self)(metadata=self.metadata, store_path=self.store_path, config=new_config)
+
     async def nchunks_initialized(self) -> int:
         """
         Calculate the number of chunks that have been initialized in storage.
@@ -1926,6 +1960,19 @@ def async_array(self) -> AsyncArray[T_ArrayMetadata]:
         """
         return self._async_array
 
+    @property
+    def config(self) -> ArrayConfig:
+        """
+        The runtime configuration for this array. This is a read-only property. To modify the
+        runtime configuration, use `Array.with_config` to create a new `Array` with the modified
+        configuration.
+
+        Returns
+        -------
+        An `ArrayConfig` object that defines the runtime configuration for the array.
+        """
+        return self.async_array.config
+
     @classmethod
     @deprecated("Use zarr.create_array instead.", category=ZarrDeprecationWarning)
     def create(
@@ -2387,6 +2434,29 @@ def _nshards(self) -> int:
         """
         return self.async_array._nshards
 
+    @overload
+    def with_config(self: ArrayV2, config: ArrayConfigLike) -> ArrayV2: ...
+
+    @overload
+    def with_config(self: ArrayV3, config: ArrayConfigLike) -> ArrayV3: ...
+
+    def with_config(self, config: ArrayConfigLike) -> Self:
+        """
+        Return a copy of this Array with a new runtime configuration.
+
+        Parameters
+        ----------
+
+        config : ArrayConfigLike
+            The runtime config for the new Array. Any keys not specified will be inherited
+            from the current array's config.
+
+        Returns
+        -------
+        A new Array
+        """
+        return type(self)(self._async_array.with_config(config))
+
     @property
     def nbytes(self) -> int:
         """

src/zarr/core/array_spec.py

@@ -69,6 +69,12 @@ def from_dict(cls, data: ArrayConfigParams) -> Self:
                 kwargs_out[field_name] = data[field_name]
         return cls(**kwargs_out)
 
+    def to_dict(self) -> ArrayConfigParams:
+        """
+        Serialize an instance of this class to a dict.
+        """
+        return {"order": self.order, "write_empty_chunks": self.write_empty_chunks}
+
 
 ArrayConfigLike = ArrayConfig | ArrayConfigParams
 

tests/test_api.py

@@ -232,7 +232,7 @@ def test_open_array_respects_write_empty_chunks_config(zarr_format: ZarrFormat)
     arr2 = zarr.open(store=store, path="test_array", config={"write_empty_chunks": True})
     assert isinstance(arr2, zarr.Array)
 
-    assert arr2.async_array._config.write_empty_chunks is True
+    assert arr2.async_array.config.write_empty_chunks is True
 
     arr2[0:5] = np.zeros(5)
     assert arr2.nchunks_initialized == 1

tests/test_array.py

@@ -44,6 +44,7 @@
     default_filters_v2,
     default_serializer_v3,
 )
+from zarr.core.array_spec import ArrayConfig, ArrayConfigParams
 from zarr.core.buffer import NDArrayLike, NDArrayLikeOrScalar, default_buffer_prototype
 from zarr.core.chunk_grids import _auto_partition
 from zarr.core.chunk_key_encodings import ChunkKeyEncodingParams
@@ -889,7 +890,7 @@ def test_write_empty_chunks_behavior(
         config={"write_empty_chunks": write_empty_chunks},
     )
 
-    assert arr.async_array._config.write_empty_chunks == write_empty_chunks
+    assert arr.async_array.config.write_empty_chunks == write_empty_chunks
 
     # initialize the store with some non-fill value chunks
     arr[:] = fill_value + 1
@@ -1562,7 +1563,7 @@ async def test_write_empty_chunks_config(write_empty_chunks: bool, store: Store)
         """
         with zarr.config.set({"array.write_empty_chunks": write_empty_chunks}):
             arr = await create_array(store, shape=(2, 2), dtype="i4")
-            assert arr._config.write_empty_chunks == write_empty_chunks
+            assert arr.config.write_empty_chunks == write_empty_chunks
 
     @staticmethod
     @pytest.mark.parametrize("path", [None, "", "/", "/foo", "foo", "foo/bar"])
@@ -2194,3 +2195,40 @@ def test_create_array_with_data_num_gets(
     # one get for the metadata and one per shard.
     # Note: we don't actually need one get per shard, but this is the current behavior
     assert store.counter["get"] == 1 + num_shards
+
+
+@pytest.mark.parametrize("config", [{}, {"write_empty_chunks": True}, {"order": "C"}])
+def test_with_config(config: ArrayConfigParams) -> None:
+    """
+    Test that `AsyncArray.with_config` and `Array.with_config` create a copy of the source
+    array with a new runtime configuration.
+    """
+    # the config we start with
+    source_config: ArrayConfigParams = {"write_empty_chunks": False, "order": "F"}
+    source_array = zarr.create_array({}, shape=(1,), dtype="uint8", config=source_config)
+
+    new_async_array_config_dict = source_array._async_array.with_config(config).config.to_dict()
+    new_array_config_dict = source_array.with_config(config).config.to_dict()
+
+    for key in source_config:
+        if key in config:
+            assert new_async_array_config_dict[key] == config[key]  # type: ignore[literal-required]
+            assert new_array_config_dict[key] == config[key]  # type: ignore[literal-required]
+        else:
+            assert new_async_array_config_dict[key] == source_config[key]  # type: ignore[literal-required]
+            assert new_array_config_dict[key] == source_config[key]  # type: ignore[literal-required]
+
+
+def test_with_config_polymorphism() -> None:
+    """
+    Test that `AsyncArray.with_config` and `Array.with_config` accept dicts and full array config
+    objects.
+    """
+    source_config: ArrayConfig = ArrayConfig.from_dict({"write_empty_chunks": False, "order": "F"})
+    source_config_dict = source_config.to_dict()
+
+    arr = zarr.create_array({}, shape=(1,), dtype="uint8")
+    arr_source_config = arr.with_config(source_config)
+    arr_source_config_dict = arr.with_config(source_config_dict)
+
+    assert arr_source_config.config == arr_source_config_dict.config