Merge branch 'main' of github.com:zarr-developers/zarr-python into feat/v2-v3-codecs

Author: d-v-b

changes/3483.bugfix.md

@@ -0,0 +1 @@
+Fix a bug that prevented `PCodec` from being properly resolved when loading arrays using that compressor.

pyproject.toml

@@ -85,8 +85,6 @@ test = [
     "packaging",
     "tomlkit",
     "uv",
-    "pcodec",
-    "zfpy"
 ]
 remote_tests = [
     'zarr[remote]',

src/zarr/codecs/__init__.py

@@ -97,7 +97,8 @@
 )
 register_codec("jenkins_lookup3", JenkinsLookup3, qualname="zarr.codecs.numcodecs.JenkinsLookup3")
 register_codec("numcodecs.pcodec", PCodec, qualname="zarr.codecs.numcodecs.pcodec")
-register_codec("pcodec", PCodec, qualname="zarr.codecs.numcodecs.pcodec")
+register_codec("pcodec", PCodec, qualname="zarr.codecs.numcodecs.PCodec")
+register_codec("numcodecs.pcodec", PCodec, qualname="zarr.codecs.numcodecs.PCodec")
 register_codec("numcodecs.packbits", PackBits, qualname="zarr.codecs.numcodecs.PackBits")
 register_codec("packbits", PackBits, qualname="zarr.codecs.numcodecs.PackBits")
 register_codec("numcodecs.quantize", Quantize, qualname="zarr.codecs.numcodecs.Quantize")

src/zarr/codecs/sharding.py

@@ -570,7 +570,8 @@ def validate(
             )
         ):
             raise ValueError(
-                "The array's `chunk_shape` needs to be divisible by the shard's inner `chunk_shape`."
+                f"The array's `chunk_shape` (got {chunk_grid.chunk_shape}) "
+                f"needs to be divisible by the shard's inner `chunk_shape` (got {self.chunk_shape})."
             )
 
     async def _decode_single(

tests/test_codecs/test_numcodecs.py

@@ -13,7 +13,7 @@
 from zarr.codecs import numcodecs as _numcodecs
 from zarr.codecs._v2 import codec_json_v2_to_v3
 from zarr.errors import ZarrUserWarning
-from zarr.registry import get_numcodec
+from zarr.registry import get_codec_class, get_numcodec
 
 if TYPE_CHECKING:
     from collections.abc import Iterator
@@ -81,18 +81,23 @@ def test_is_numcodec_cls() -> None:
 
 ALL_CODECS = tuple(
     filter(
-        lambda v: isinstance(v, _numcodecs._NumcodecsCodec),
+        lambda v: issubclass(v, _numcodecs._NumcodecsCodec) and hasattr(v, "codec_name"),
         tuple(getattr(_numcodecs, cls_name) for cls_name in _numcodecs.__all__),
     )
 )
 
 
+@pytest.mark.parametrize("codec_cls", ALL_CODECS)
+def test_get_codec_class(codec_cls: type[_numcodecs._NumcodecsCodec]) -> None:
+    assert get_codec_class(codec_cls.codec_name) == codec_cls  # type: ignore[comparison-overlap]
+
+
 @pytest.mark.parametrize("codec_class", ALL_CODECS)
 def test_docstring(codec_class: type[_numcodecs._NumcodecsCodec]) -> None:
     """
     Test that the docstring for the zarr.numcodecs codecs references the wrapped numcodecs class.
     """
-    assert "See :class:`numcodecs." in codec_class.__doc__  # type: ignore[operator]
+    assert "See [numcodecs." in codec_class.__doc__  # type: ignore[operator]
 
 
 @pytest.mark.parametrize(

tests/test_codecs/test_sharding.py

@@ -1,4 +1,5 @@
 import pickle
+import re
 from typing import TYPE_CHECKING, Any, Literal
 
 import numpy as np
@@ -528,3 +529,20 @@ def test_invalid_metadata(store: Store) -> None:
             dtype=np.dtype("uint8"),
             fill_value=0,
         )
+
+
+def test_invalid_shard_shape() -> None:
+    with pytest.raises(
+        ValueError,
+        match=re.escape(
+            "The array's `chunk_shape` (got (16, 16)) needs to be divisible by the shard's inner `chunk_shape` (got (9,))."
+        ),
+    ):
+        zarr.create_array(
+            {},
+            shape=(16, 16),
+            shards=(16, 16),
+            chunks=(9,),
+            dtype=np.dtype("uint8"),
+            fill_value=0,
+        )