Remove zstandard dependency in favor of numcodecs

pyproject.toml

@@ -16,6 +16,8 @@ dependencies = [
     'numpy>=1.24',
     'fasteners',
     'numcodecs>=0.10.0',
+    'crc32c',
+    'typing_extensions',
 ]
 dynamic = [
   "version",
@@ -52,17 +54,11 @@ docs = [
     'pydata-sphinx-theme',
     'numpydoc',
     'numcodecs[msgpack]',
-    "msgpack",
-    "lmdb",
-    "zstandard",
-    "crc32c",
+    'msgpack',
+    'lmdb',
 ]
 extra = [
-    'attrs',
-    'cattrs',
     'msgpack',
-    'crc32c',
-    'zstandard'
 ]
 optional = [
     'lmdb',
@@ -101,10 +97,7 @@ extra-dependencies = [
     "pytest-cov",
     "msgpack",
     "lmdb",
-    "zstandard",
-    "crc32c",
     "pytest-asyncio",
-    "typing_extensions",
     "mypy"
 ]
 features = ["extra"]

src/zarr/codecs/zstd.py

@@ -1,9 +1,10 @@
 from __future__ import annotations
+from functools import cached_property
 from typing import TYPE_CHECKING
 from dataclasses import dataclass
 
 
-from zstandard import ZstdCompressor, ZstdDecompressor
+from numcodecs.zstd import Zstd
 
 from zarr.abc.codec import BytesBytesCodec
 from zarr.codecs.registry import register_codec
@@ -15,6 +16,8 @@
     from zarr.config import RuntimeConfiguration
     from zarr.common import BytesLike, JSON, ArraySpec
 
+DEFAULT_ZSTD_LEVEL = 3
+
 
 def parse_zstd_level(data: JSON) -> int:
     if isinstance(data, int):
@@ -34,10 +37,10 @@ def parse_checksum(data: JSON) -> bool:
 class ZstdCodec(BytesBytesCodec):
     is_fixed_size = True
 
-    level: int = 0
+    level: int = DEFAULT_ZSTD_LEVEL
     checksum: bool = False
 
-    def __init__(self, *, level: int = 0, checksum: bool = False) -> None:
+    def __init__(self, *, level: int = DEFAULT_ZSTD_LEVEL, checksum: bool = False) -> None:
         level_parsed = parse_zstd_level(level)
         checksum_parsed = parse_checksum(checksum)
 
@@ -52,29 +55,26 @@ def from_dict(cls, data: Dict[str, JSON]) -> Self:
     def to_dict(self) -> Dict[str, JSON]:
         return {"name": "zstd", "configuration": {"level": self.level, "checksum": self.checksum}}
 
-    def _compress(self, data: bytes) -> bytes:
-        ctx = ZstdCompressor(level=self.level, write_checksum=self.checksum)
-        return ctx.compress(data)
-
-    def _decompress(self, data: bytes) -> bytes:
-        ctx = ZstdDecompressor()
-        return ctx.decompress(data)
+    @cached_property
+    def _zstd_codec(self) -> Zstd:
+        config_dict = {"level": self.level, "checksum": self.checksum}
+        return Zstd.from_config(config_dict)
 
     async def decode(
         self,
         chunk_bytes: bytes,
         _chunk_spec: ArraySpec,
         _runtime_configuration: RuntimeConfiguration,
     ) -> BytesLike:
-        return await to_thread(self._decompress, chunk_bytes)
+        return await to_thread(self._zstd_codec.decode, chunk_bytes)
 
     async def encode(
         self,
         chunk_bytes: bytes,
         _chunk_spec: ArraySpec,
         _runtime_configuration: RuntimeConfiguration,
     ) -> Optional[BytesLike]:
-        return await to_thread(self._compress, chunk_bytes)
+        return await to_thread(self._zstd_codec.encode, chunk_bytes)
 
     def compute_encoded_size(self, _input_byte_length: int, _chunk_spec: ArraySpec) -> int:
         raise NotImplementedError