Merge branch 'ig/structured_arrays_v2' of github.com:ilan-gold/zarr-python into ig/structured_arrays_v2

Author: ilan-gold

docs/release-notes.rst

@@ -1,6 +1,20 @@
 Release notes
 =============
 
+Unreleased
+----------
+
+New features
+~~~~~~~~~~~~
+
+Bug fixes
+~~~~~~~~~
+* Fixes ``order`` argument for Zarr format 2 arrays.
+  By :user:`Norman Rzepka <normanrz>` (:issue:`2679`).
+
+Behaviour changes
+~~~~~~~~~~~~~~~~~
+
 .. _release_3.0.0:
 
 3.0.0

src/zarr/core/array.py

@@ -4,7 +4,7 @@
 import warnings
 from asyncio import gather
 from collections.abc import Iterable
-from dataclasses import dataclass, field
+from dataclasses import dataclass, field, replace
 from itertools import starmap
 from logging import getLogger
 from typing import (
@@ -1226,14 +1226,17 @@ async def _get_selection(
                 fill_value=self.metadata.fill_value,
             )
         if product(indexer.shape) > 0:
+            # need to use the order from the metadata for v2
+            _config = self._config
+            if self.metadata.zarr_format == 2:
+                _config = replace(_config, order=self.metadata.order)
+
             # reading chunks and decoding them
             await self.codec_pipeline.read(
                 [
                     (
                         self.store_path / self.metadata.encode_chunk_key(chunk_coords),
-                        self.metadata.get_chunk_spec(
-                            chunk_coords, self._config, prototype=prototype
-                        ),
+                        self.metadata.get_chunk_spec(chunk_coords, _config, prototype=prototype),
                         chunk_selection,
                         out_selection,
                     )
@@ -1350,12 +1353,17 @@ async def _set_selection(
         # Buffer and NDBuffer between components.
         value_buffer = prototype.nd_buffer.from_ndarray_like(value)
 
+        # need to use the order from the metadata for v2
+        _config = self._config
+        if self.metadata.zarr_format == 2:
+            _config = replace(_config, order=self.metadata.order)
+
         # merging with existing data and encoding chunks
         await self.codec_pipeline.write(
             [
                 (
                     self.store_path / self.metadata.encode_chunk_key(chunk_coords),
-                    self.metadata.get_chunk_spec(chunk_coords, self._config, prototype),
+                    self.metadata.get_chunk_spec(chunk_coords, _config, prototype),
                     chunk_selection,
                     out_selection,
                 )

tests/test_v2.py

@@ -12,6 +12,8 @@
 import zarr.core.buffer
 import zarr.storage
 from zarr import config
+from zarr.core.buffer.core import default_buffer_prototype
+from zarr.core.sync import sync
 from zarr.storage import MemoryStore, StorePath
 
 
@@ -166,36 +168,54 @@ def test_v2_filters_codecs(filters: Any, order: Literal["C", "F"]) -> None:
 
 @pytest.mark.parametrize("array_order", ["C", "F"])
 @pytest.mark.parametrize("data_order", ["C", "F"])
-def test_v2_non_contiguous(array_order: Literal["C", "F"], data_order: Literal["C", "F"]) -> None:
+@pytest.mark.parametrize("memory_order", ["C", "F"])
+def test_v2_non_contiguous(
+    array_order: Literal["C", "F"], data_order: Literal["C", "F"], memory_order: Literal["C", "F"]
+) -> None:
+    store = MemoryStore()
     arr = zarr.create_array(
-        MemoryStore({}),
+        store,
         shape=(10, 8),
         chunks=(3, 3),
         fill_value=np.nan,
         dtype="float64",
         zarr_format=2,
+        filters=None,
+        compressors=None,
         overwrite=True,
         order=array_order,
+        config={"order": memory_order},
     )
 
     # Non-contiguous write
     a = np.arange(arr.shape[0] * arr.shape[1]).reshape(arr.shape, order=data_order)
-    arr[slice(6, 9, None), slice(3, 6, None)] = a[
-        slice(6, 9, None), slice(3, 6, None)
-    ]  # The slice on the RHS is important
+    arr[6:9, 3:6] = a[6:9, 3:6]  # The slice on the RHS is important
+    np.testing.assert_array_equal(arr[6:9, 3:6], a[6:9, 3:6])
+
     np.testing.assert_array_equal(
-        arr[slice(6, 9, None), slice(3, 6, None)], a[slice(6, 9, None), slice(3, 6, None)]
+        a[6:9, 3:6],
+        np.frombuffer(
+            sync(store.get("2.1", default_buffer_prototype())).to_bytes(), dtype="float64"
+        ).reshape((3, 3), order=array_order),
     )
+    if memory_order == "F":
+        assert (arr[6:9, 3:6]).flags.f_contiguous
+    else:
+        assert (arr[6:9, 3:6]).flags.c_contiguous
 
+    store = MemoryStore()
     arr = zarr.create_array(
-        MemoryStore({}),
+        store,
         shape=(10, 8),
         chunks=(3, 3),
         fill_value=np.nan,
         dtype="float64",
         zarr_format=2,
+        compressors=None,
+        filters=None,
         overwrite=True,
         order=array_order,
+        config={"order": memory_order},
     )
 
     # Contiguous write
@@ -204,8 +224,8 @@ def test_v2_non_contiguous(array_order: Literal["C", "F"], data_order: Literal["
         assert a.flags.f_contiguous
     else:
         assert a.flags.c_contiguous
-    arr[slice(6, 9, None), slice(3, 6, None)] = a
-    np.testing.assert_array_equal(arr[slice(6, 9, None), slice(3, 6, None)], a)
+    arr[6:9, 3:6] = a
+    np.testing.assert_array_equal(arr[6:9, 3:6], a)
 
 
 def test_default_compressor_deprecation_warning():