use Numcodec instead of numcodecs.abc.Codec

Author: d-v-b

src/zarr/api/asynchronous.py

@@ -46,8 +46,7 @@
 if TYPE_CHECKING:
     from collections.abc import Iterable
 
-    from zarr.abc.codec import Codec
-    from zarr.codecs._v2 import Numcodec
+    from zarr.abc.codec import Codec, Numcodec
     from zarr.core.buffer import NDArrayLikeOrScalar
     from zarr.core.chunk_key_encodings import ChunkKeyEncoding
     from zarr.storage import StoreLike

src/zarr/api/synchronous.py

@@ -17,9 +17,8 @@
     import numpy as np
     import numpy.typing as npt
 
-    from zarr.abc.codec import Codec
+    from zarr.abc.codec import Codec, Numcodec
     from zarr.api.asynchronous import ArrayLike, PathLike
-    from zarr.codecs._v2 import Numcodec
     from zarr.core.array import (
         CompressorsLike,
         FiltersLike,

src/zarr/codecs/_numcodecs.py

@@ -1,5 +1,4 @@
-from zarr.abc.codec import CodecJSON_V2
-from zarr.abc.codec import Numcodec
+from zarr.abc.codec import CodecJSON_V2, Numcodec
 
 
 def get_numcodec(data: CodecJSON_V2[str]) -> Numcodec:

src/zarr/codecs/_v2.py

@@ -2,39 +2,19 @@
 
 import asyncio
 from dataclasses import dataclass
-from typing import TYPE_CHECKING, ClassVar, Self, TypeGuard
+from typing import TYPE_CHECKING, TypeGuard
 
 import numpy as np
 from numcodecs.compat import ensure_bytes, ensure_ndarray_like
-from typing_extensions import Protocol
 
-from zarr.abc.codec import ArrayBytesCodec, CodecJSON_V2
+from zarr.abc.codec import ArrayBytesCodec, Numcodec
 from zarr.registry import get_ndbuffer_class
 
 if TYPE_CHECKING:
     from zarr.core.array_spec import ArraySpec
     from zarr.core.buffer import Buffer, NDBuffer
 
 
-class Numcodec(Protocol):
-    """
-    A protocol that models the ``numcodecs.abc.Codec`` interface.
-    """
-
-    codec_id: ClassVar[str]
-
-    def encode(self, buf: Buffer | NDBuffer) -> Buffer | NDBuffer: ...
-
-    def decode(
-        self, buf: Buffer | NDBuffer, out: Buffer | NDBuffer | None = None
-    ) -> Buffer | NDBuffer: ...
-
-    def get_config(self) -> CodecJSON_V2[str]: ...
-
-    @classmethod
-    def from_config(cls, config: CodecJSON_V2[str]) -> Self: ...
-
-
 def _is_numcodec(obj: object) -> TypeGuard[Numcodec]:
     """
     Check if the given object implements the Numcodec protocol.

src/zarr/core/_info.py

@@ -4,9 +4,9 @@
 import textwrap
 from typing import TYPE_CHECKING, Literal
 
-if TYPE_CHECKING:
-    import numcodecs.abc
+from zarr.abc.codec import Numcodec
 
+if TYPE_CHECKING:
     from zarr.abc.codec import ArrayArrayCodec, ArrayBytesCodec, BytesBytesCodec
     from zarr.core.common import ZarrFormat
     from zarr.core.dtype.wrapper import TBaseDType, TBaseScalar, ZDType
@@ -88,9 +88,9 @@ class ArrayInfo:
     _order: Literal["C", "F"]
     _read_only: bool
     _store_type: str
-    _filters: tuple[numcodecs.abc.Codec, ...] | tuple[ArrayArrayCodec, ...] = ()
+    _filters: tuple[Numcodec, ...] | tuple[ArrayArrayCodec, ...] = ()
     _serializer: ArrayBytesCodec | None = None
-    _compressors: tuple[numcodecs.abc.Codec, ...] | tuple[BytesBytesCodec, ...] = ()
+    _compressors: tuple[Numcodec, ...] | tuple[BytesBytesCodec, ...] = ()
     _count_bytes: int | None = None
     _count_bytes_stored: int | None = None
     _count_chunks_initialized: int | None = None

src/zarr/core/array.py

@@ -25,9 +25,9 @@
 from typing_extensions import deprecated
 
 import zarr
-from zarr.abc.codec import ArrayArrayCodec, ArrayBytesCodec, BytesBytesCodec, Codec
+from zarr.abc.codec import ArrayArrayCodec, ArrayBytesCodec, BytesBytesCodec, Codec, Numcodec
 from zarr.abc.store import Store, set_or_delete
-from zarr.codecs._v2 import Numcodec, V2Codec
+from zarr.codecs._v2 import V2Codec
 from zarr.codecs.bytes import BytesCodec
 from zarr.codecs.vlen_utf8 import VLenBytesCodec, VLenUTF8Codec
 from zarr.codecs.zstd import ZstdCodec
@@ -1033,7 +1033,7 @@ def size(self) -> int:
         return np.prod(self.metadata.shape).item()
 
     @property
-    def filters(self) -> tuple[numcodecs.abc.Codec, ...] | tuple[ArrayArrayCodec, ...]:
+    def filters(self) -> tuple[Numcodec, ...] | tuple[ArrayArrayCodec, ...]:
         """
         Filters that are applied to each chunk of the array, in order, before serializing that
         chunk to bytes.
@@ -1062,7 +1062,7 @@ def serializer(self) -> ArrayBytesCodec | None:
 
     @property
     @deprecated("Use AsyncArray.compressors instead.")
-    def compressor(self) -> numcodecs.abc.Codec | None:
+    def compressor(self) -> Numcodec | None:
         """
         Compressor that is applied to each chunk of the array.
 
@@ -1075,7 +1075,7 @@ def compressor(self) -> numcodecs.abc.Codec | None:
         raise TypeError("`compressor` is not available for Zarr format 3 arrays.")
 
     @property
-    def compressors(self) -> tuple[numcodecs.abc.Codec, ...] | tuple[BytesBytesCodec, ...]:
+    def compressors(self) -> tuple[Numcodec, ...] | tuple[BytesBytesCodec, ...]:
         """
         Compressors that are applied to each chunk of the array. Compressors are applied in order, and after any
         filters are applied (if any are specified) and the data is serialized into bytes.
@@ -2227,7 +2227,7 @@ def fill_value(self) -> Any:
         return self.metadata.fill_value
 
     @property
-    def filters(self) -> tuple[numcodecs.abc.Codec, ...] | tuple[ArrayArrayCodec, ...]:
+    def filters(self) -> tuple[Numcodec, ...] | tuple[ArrayArrayCodec, ...]:
         """
         Filters that are applied to each chunk of the array, in order, before serializing that
         chunk to bytes.
@@ -2243,7 +2243,7 @@ def serializer(self) -> None | ArrayBytesCodec:
 
     @property
     @deprecated("Use Array.compressors instead.")
-    def compressor(self) -> numcodecs.abc.Codec | None:
+    def compressor(self) -> Numcodec | None:
         """
         Compressor that is applied to each chunk of the array.
 
@@ -2254,7 +2254,7 @@ def compressor(self) -> numcodecs.abc.Codec | None:
         return self._async_array.compressor
 
     @property
-    def compressors(self) -> tuple[numcodecs.abc.Codec, ...] | tuple[BytesBytesCodec, ...]:
+    def compressors(self) -> tuple[Numcodec, ...] | tuple[BytesBytesCodec, ...]:
         """
         Compressors that are applied to each chunk of the array. Compressors are applied in order, and after any
         filters are applied (if any are specified) and the data is serialized into bytes.
@@ -3900,15 +3900,13 @@ def _build_parents(
 FiltersLike: TypeAlias = (
     Iterable[dict[str, JSON] | ArrayArrayCodec | Numcodec]
     | ArrayArrayCodec
-    | Iterable[numcodecs.abc.Codec]
-    | numcodecs.abc.Codec
+    | Iterable[Numcodec]
+    | Numcodec
     | Literal["auto"]
     | None
 )
 # Union of acceptable types for users to pass in for both v2 and v3 compressors
-CompressorLike: TypeAlias = (
-    dict[str, JSON] | BytesBytesCodec | numcodecs.abc.Codec | Literal["auto"] | None
-)
+CompressorLike: TypeAlias = dict[str, JSON] | BytesBytesCodec | Numcodec | Literal["auto"] | None
 
 CompressorsLike: TypeAlias = (
     Iterable[dict[str, JSON] | BytesBytesCodec | Numcodec]
@@ -4775,7 +4773,7 @@ def default_serializer_v3(dtype: ZDType[Any, Any]) -> ArrayBytesCodec:
     return serializer
 
 
-def default_filters_v2(dtype: ZDType[Any, Any]) -> tuple[numcodecs.abc.Codec] | None:
+def default_filters_v2(dtype: ZDType[Any, Any]) -> tuple[Numcodec] | None:
     """
     Given a data type, return the default filters for that data type.
 
@@ -4813,12 +4811,12 @@ def _parse_chunk_encoding_v2(
     compressor: CompressorsLike,
     filters: FiltersLike,
     dtype: ZDType[TBaseDType, TBaseScalar],
-) -> tuple[tuple[numcodecs.abc.Codec, ...] | None, numcodecs.abc.Codec | None]:
+) -> tuple[tuple[Numcodec, ...] | None, Numcodec | None]:
     """
     Generate chunk encoding classes for Zarr format 2 arrays with optional defaults.
     """
-    _filters: tuple[numcodecs.abc.Codec, ...] | None
-    _compressor: numcodecs.abc.Codec | None
+    _filters: tuple[Numcodec, ...] | None
+    _compressor: Numcodec | None
 
     if compressor is None or compressor == ():
         _compressor = None

src/zarr/core/metadata/v2.py

@@ -7,6 +7,7 @@
 
 import numcodecs.abc
 
+from zarr.abc.codec import Numcodec
 from zarr.abc.metadata import Metadata
 from zarr.core.chunk_grids import RegularChunkGrid
 from zarr.core.dtype import get_data_type_from_json
@@ -56,7 +57,7 @@ class ArrayV2MetadataDict(TypedDict):
 
 
 # Union of acceptable types for v2 compressors
-CompressorLikev2: TypeAlias = dict[str, JSON] | numcodecs.abc.Codec | None
+CompressorLikev2: TypeAlias = dict[str, JSON] | Numcodec | None
 
 
 @dataclass(frozen=True, kw_only=True)
@@ -66,9 +67,9 @@ class ArrayV2Metadata(Metadata):
     dtype: ZDType[TBaseDType, TBaseScalar]
     fill_value: int | float | str | bytes | None = None
     order: MemoryOrder = "C"
-    filters: tuple[numcodecs.abc.Codec, ...] | None = None
+    filters: tuple[Numcodec, ...] | None = None
     dimension_separator: Literal[".", "/"] = "."
-    compressor: numcodecs.abc.Codec | None
+    compressor: Numcodec | None
     attributes: dict[str, JSON] = field(default_factory=dict)
     zarr_format: Literal[2] = field(init=False, default=2)
 
@@ -82,7 +83,7 @@ def __init__(
         order: MemoryOrder,
         dimension_separator: Literal[".", "/"] = ".",
         compressor: CompressorLikev2 = None,
-        filters: Iterable[numcodecs.abc.Codec | dict[str, JSON]] | None = None,
+        filters: Iterable[Numcodec | dict[str, JSON]] | None = None,
         attributes: dict[str, JSON] | None = None,
     ) -> None:
         """
@@ -262,11 +263,11 @@ def parse_zarr_format(data: object) -> Literal[2]:
     raise ValueError(f"Invalid value. Expected 2. Got {data}.")
 
 
-def parse_filters(data: object) -> tuple[numcodecs.abc.Codec, ...] | None:
+def parse_filters(data: object) -> tuple[Numcodec, ...] | None:
     """
     Parse a potential tuple of filters
     """
-    out: list[numcodecs.abc.Codec] = []
+    out: list[Numcodec] = []
 
     if data is None:
         return data
@@ -291,7 +292,7 @@ def parse_filters(data: object) -> tuple[numcodecs.abc.Codec, ...] | None:
     raise TypeError(msg)
 
 
-def parse_compressor(data: object) -> numcodecs.abc.Codec | None:
+def parse_compressor(data: object) -> Numcodec | None:
     """
     Parse a potential compressor.
     """

tests/test_config.py

@@ -180,7 +180,18 @@ async def _encode_single(self, chunk_bytes: Buffer, chunk_spec: ArraySpec) -> Bu
             chunks=(10,),
             zarr_format=3,
             dtype="i4",
-            compressors=[{"name": "blosc", "configuration": {"cname": "lz4", "clevel": 1, "shuffle": "noshuffle", "blocksize": 1, "typesize": 1}},],
+            compressors=[
+                {
+                    "name": "blosc",
+                    "configuration": {
+                        "cname": "lz4",
+                        "clevel": 1,
+                        "shuffle": "noshuffle",
+                        "blocksize": 1,
+                        "typesize": 1,
+                    },
+                },
+            ],
         )
         arr[:] = range(100)
         _mock.call.assert_called()

tests/test_group.py

@@ -11,17 +11,16 @@
 
 import numpy as np
 import pytest
-from numcodecs import Blosc
 
 import zarr
 import zarr.api.asynchronous
 import zarr.api.synchronous
-from zarr.core.array import default_compressor_v2, default_compressors_v3, default_serializer_v3
 import zarr.storage
 from zarr import Array, AsyncArray, AsyncGroup, Group
 from zarr.abc.store import Store
 from zarr.core import sync_group
 from zarr.core._info import GroupInfo
+from zarr.core.array import default_compressor_v2, default_compressors_v3, default_serializer_v3
 from zarr.core.buffer import default_buffer_prototype
 from zarr.core.config import config as zarr_config
 from zarr.core.dtype.common import unpack_dtype_json
@@ -524,7 +523,7 @@ def test_group_child_iterators(store: Store, zarr_format: ZarrFormat, consolidat
                     "chunks": (1,),
                     "order": "C",
                     "filters": None,
-                    "compressor":  default_compressor_v2(dtype).to_json(zarr_format=zarr_format),
+                    "compressor": default_compressor_v2(dtype).to_json(zarr_format=zarr_format),
                     "zarr_format": zarr_format,
                 },
                 "subgroup": {

tests/test_regression/test_v2_dtype_regression.py

@@ -4,7 +4,6 @@
 from pathlib import Path
 from typing import TYPE_CHECKING, Literal
 
-import numcodecs
 import numpy as np
 import pytest
 from numcodecs import LZ4, LZMA, Blosc, GZip, VLenBytes, VLenUTF8, Zstd
@@ -13,6 +12,7 @@
 import zarr.abc
 import zarr.abc.codec
 import zarr.codecs as zarrcodecs
+from zarr.abc.codec import Numcodec
 from zarr.core.array import Array
 from zarr.core.chunk_key_encodings import V2ChunkKeyEncoding
 from zarr.core.dtype.npy.bytes import VariableLengthBytes
@@ -40,9 +40,9 @@ def runner_installed() -> bool:
 class ArrayParams:
     values: np.ndarray[tuple[int], np.dtype[np.generic]]
     fill_value: np.generic | str | int | bytes
-    filters: tuple[numcodecs.abc.Codec, ...] = ()
+    filters: tuple[Numcodec, ...] = ()
     serializer: str | None = None
-    compressor: numcodecs.abc.Codec
+    compressor: Numcodec
 
 
 basic_codecs = GZip(), Blosc(), LZ4(), LZMA(), Zstd()