Merge branch 'main' into sharding-storage-transformer

Author: jstriebel

docs/release.rst

@@ -14,6 +14,9 @@ Unreleased
     # .. warning::
     #    Pre-release! Use :command:`pip install --pre zarr` to evaluate this release.
 
+Major changes
+~~~~~~~~~~~~~
+
 * Improve Zarr V3 support, adding partial store read/write and storage transformers.
   Add two features of the [v3 spec](https://zarr-specs.readthedocs.io/en/latest/core/v3.0.html):
   * storage transformers
@@ -22,6 +25,13 @@ Unreleased
   * sharding storage transformer
   By :user:`Jonathan Striebel <jstriebel>`; :issue:`1096`, :issue:`1111`.
 
+
+Bug fixes
+~~~~~~~~~
+
+* Ensure contiguous data is give to ``FSStore``. Only copying if needed.
+  By :user:`Mads R. B. Kristensen <madsbk>` :issue:`1285`.
+
 .. _release_2.13.6:
 
 2.13.6

requirements_dev_optional.txt

@@ -18,6 +18,6 @@ pytest-cov==4.0.0
 pytest-doctestplus==0.12.1
 pytest-timeout==2.1.0
 h5py==3.7.0
-fsspec==2022.11.0
-s3fs==2022.11.0
+fsspec==2023.1.0
+s3fs==2023.1.0
 moto[server]>=4.0.8

zarr/storage.py

@@ -55,8 +55,8 @@
 from zarr.util import (buffer_size, json_loads, nolock, normalize_chunks,
                        normalize_dimension_separator,
                        normalize_dtype, normalize_fill_value, normalize_order,
-                       normalize_shape, normalize_storage_path, retry_call
-                       )
+                       normalize_shape, normalize_storage_path, retry_call,
+                       ensure_contiguous_ndarray_or_bytes)
 
 from zarr._storage.absstore import ABSStore  # noqa: F401
 from zarr._storage.store import (_get_hierarchy_metadata,  # noqa: F401
@@ -1395,13 +1395,19 @@ def __getitem__(self, key):
     def setitems(self, values):
         if self.mode == 'r':
             raise ReadOnlyError()
-        values = {self._normalize_key(key): val for key, val in values.items()}
+
+        # Normalize keys and make sure the values are bytes
+        values = {
+            self._normalize_key(key): ensure_contiguous_ndarray_or_bytes(val)
+            for key, val in values.items()
+        }
         self.map.setitems(values)
 
     def __setitem__(self, key, value):
         if self.mode == 'r':
             raise ReadOnlyError()
         key = self._normalize_key(key)
+        value = ensure_contiguous_ndarray_or_bytes(value)
         path = self.dir_path(key)
         try:
             if self.fs.isdir(path):

zarr/tests/test_core.py

@@ -17,6 +17,7 @@
 from numpy.testing import assert_array_almost_equal, assert_array_equal
 from pkg_resources import parse_version
 
+import zarr
 from zarr._storage.store import (
     v3_api_available,
 )
@@ -3520,3 +3521,27 @@ def test_array_mismatched_store_versions():
         Array(store_v3, path='dataset', read_only=False, chunk_store=chunk_store_v2)
     with pytest.raises(ValueError):
         Array(store_v2, path='dataset', read_only=False, chunk_store=chunk_store_v3)
+
+
+@pytest.mark.skipif(have_fsspec is False, reason="needs fsspec")
+def test_issue_1279(tmpdir):
+    """See <https://github.com/zarr-developers/zarr-python/issues/1279>"""
+
+    data = np.arange(25).reshape((5, 5))
+    ds = zarr.create(
+        shape=data.shape,
+        chunks=(5, 5),
+        dtype=data.dtype,
+        compressor=(None),
+        store=FSStore(url=str(tmpdir), mode="a"),
+        order="F",
+    )
+
+    ds[:] = data
+
+    ds_reopened = zarr.open_array(
+        store=FSStore(url=str(tmpdir), mode="r")
+    )
+
+    written_data = ds_reopened[:]
+    assert_array_equal(data, written_data)

zarr/util.py

@@ -5,17 +5,22 @@
 from textwrap import TextWrapper
 import mmap
 import time
+from typing import Any, Callable, Dict, Optional, Tuple, Union
 
 import numpy as np
 from asciitree import BoxStyle, LeftAligned
 from asciitree.traversal import Traversal
 from collections.abc import Iterable
-from numcodecs.compat import ensure_text, ensure_ndarray_like
+from numcodecs.compat import (
+    ensure_text,
+    ensure_ndarray_like,
+    ensure_bytes,
+    ensure_contiguous_ndarray_like
+)
+from numcodecs.ndarray_like import NDArrayLike
 from numcodecs.registry import codec_registry
 from numcodecs.blosc import cbuffer_sizes, cbuffer_metainfo
 
-from typing import Any, Callable, Dict, Optional, Tuple, Union
-
 
 def flatten(arg: Iterable) -> Iterable:
     for element in arg:
@@ -715,3 +720,28 @@ def all_equal(value: Any, array: Any):
             # using == raises warnings from numpy deprecated pattern, but
             # using np.equal() raises type errors for structured dtypes...
             return np.all(value == array)
+
+
+def ensure_contiguous_ndarray_or_bytes(buf) -> Union[NDArrayLike, bytes]:
+    """Convenience function to coerce `buf` to ndarray-like array or bytes.
+
+    First check if `buf` can be zero-copy converted to a contiguous array.
+    If not, `buf` will be copied to a newly allocated `bytes` object.
+
+    Parameters
+    ----------
+    buf : ndarray-like, array-like, or bytes-like
+        A numpy array like object such as numpy.ndarray, cupy.ndarray, or
+        any object exporting a buffer interface.
+
+    Returns
+    -------
+    arr : NDArrayLike or bytes
+        A ndarray-like or bytes object
+    """
+
+    try:
+        return ensure_contiguous_ndarray_like(buf)
+    except TypeError:
+        # An error is raised if `buf` couldn't be zero-copy converted
+        return ensure_bytes(buf)