Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Try to converge on internal types consistency. #574

Closed
wants to merge 1 commit into from
Closed
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 5 additions & 0 deletions docs/release.rst
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,11 @@ Release notes
Next release
------------

* For consistency across stores, the `ConsolidatedMetadataStore` will return
bytes instead of objects for metadata. Internally zarr tries to be more type
stable and attempt to always pass bytes-likes object when storing objects on
stores.

* Fix minor bug in `N5Store`.
By :user:`gsakkis`, :issue:`550`.

Expand Down
2 changes: 1 addition & 1 deletion zarr/core.py
Original file line number Diff line number Diff line change
Expand Up @@ -1489,7 +1489,7 @@ def _set_basic_selection_zd(self, selection, value, fields=None):
chunk[selection] = value

# encode and store
cdata = self._encode_chunk(chunk)
cdata = ensure_bytes(self._encode_chunk(chunk))
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

N.B., ensuring bytes may introduce a memory copy, which may be unnecessary. Memory copies can noticeably affect performance. Ensuring we have something that supports the buffer interface would be better.

self.chunk_store[ckey] = cdata

def _set_basic_selection_nd(self, selection, value, fields=None):
Expand Down
10 changes: 8 additions & 2 deletions zarr/storage.py
Original file line number Diff line number Diff line change
Expand Up @@ -44,7 +44,7 @@
from zarr.meta import encode_array_metadata, encode_group_metadata
from zarr.util import (buffer_size, json_loads, nolock, normalize_chunks,
normalize_dtype, normalize_fill_value, normalize_order,
normalize_shape, normalize_storage_path)
normalize_shape, normalize_storage_path, json_dumps)

__doctest_requires__ = {
('RedisStore', 'RedisStore.*'): ['redis'],
Expand Down Expand Up @@ -2479,6 +2479,10 @@ class ConsolidatedMetadataStore(MutableMapping):

.. versionadded:: 2.3

.. versionchanged:: 2.5

__getitem__ will now return bytes for metadata for consistency across stores.

.. note:: This is an experimental feature.

Parameters
Expand Down Expand Up @@ -2507,7 +2511,9 @@ def __init__(self, store, metadata_key='.zmetadata'):
consolidated_format)

# decode metadata
self.meta_store = meta['metadata']
self.meta_store = {}
for k, v in meta["metadata"].items():
self.meta_store[k] = json_dumps(v)
Comment on lines +2514 to +2516
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

FWIW this was originally how this worked, but this introduces an extra unnecessary JSON encode/decode round trip into the processing of consolidated metadata. In practice that may not be a performance issue, but it seemed awkward to have this extra JSON encode/decode if it wasn't necessary. Hence I introduced a special case which avoided the necessity for this, see the parse_metadata function in the original PR.


def __getitem__(self, key):
return self.meta_store[key]
Expand Down
6 changes: 0 additions & 6 deletions zarr/tests/test_convenience.py
Original file line number Diff line number Diff line change
Expand Up @@ -697,12 +697,6 @@ def test_logging(self):
copy(source['foo'], dest, dry_run=True, log=True)


try:
import h5py
except ImportError: # pragma: no cover
h5py = None


def temp_h5f():
h5py = pytest.importorskip("h5py")
fn = tempfile.mktemp()
Expand Down
8 changes: 4 additions & 4 deletions zarr/tests/test_meta.py
Original file line number Diff line number Diff line change
Expand Up @@ -424,15 +424,15 @@ def test_encode_decode_dtype():
def test_decode_group():

# typical
b = '''{
b = b'''{
"zarr_format": %s
}''' % ZARR_FORMAT
}''' % str(ZARR_FORMAT).encode()
meta = decode_group_metadata(b)
assert ZARR_FORMAT == meta['zarr_format']

# unsupported format
b = '''{
b = b'''{
"zarr_format": %s
}''' % (ZARR_FORMAT - 1)
}''' % str(ZARR_FORMAT - 1).encode()
with pytest.raises(MetadataError):
decode_group_metadata(b)
28 changes: 22 additions & 6 deletions zarr/tests/test_storage.py
Original file line number Diff line number Diff line change
Expand Up @@ -30,6 +30,7 @@
attrs_key, default_compressor, getsize,
group_meta_key, init_array, init_group, migrate_1to2)
from zarr.tests.util import CountingDict, skip_test_env_var
from zarr.util import json_dumps


@contextmanager
Expand Down Expand Up @@ -1627,11 +1628,26 @@ def test_read_write(self):
# setup store with consolidated metdata
store = dict()
consolidated = {
'zarr_consolidated_format': 1,
'metadata': {
'foo': 'bar',
'baz': 42,
}
"zarr_consolidated_format": 1,
"metadata": {
".zgroup": {"zarr_format": 2},
"g2/arr/.zarray": {
"chunks": [5, 5],
"compressor": {
"blocksize": 0,
"clevel": 5,
"cname": "lz4",
"id": "blosc",
"shuffle": 1,
},
"dtype": "<f8",
"fill_value": 0.0,
"filters": None,
"order": "C",
"shape": [20, 20],
"zarr_format": 2,
},
},
}
store['.zmetadata'] = json.dumps(consolidated).encode()

Expand All @@ -1641,7 +1657,7 @@ def test_read_write(self):
# test __contains__, __getitem__
for key, value in consolidated['metadata'].items():
assert key in cs
assert value == cs[key]
assert json_dumps(value) == cs[key]

# test __delitem__, __setitem__
with pytest.raises(PermissionError):
Expand Down